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

Diff of /code/trunk/ChangeLog

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

revision 5 by nigel, Sat Feb 24 21:38:05 2007 UTC revision 431 by ph10, Wed Sep 2 09:23:21 2009 UTC
# Line 1  Line 1 
1  ChangeLog for PCRE  ChangeLog for PCRE
2  ------------------  ------------------
3    
4    Version 8.00 ??-???-??
5    ----------------------
6    
7    1.  The table for translating pcre_compile() error codes into POSIX error codes
8        was out-of-date, and there was no check on the pcre_compile() error code
9        being within the table. This could lead to an OK return being given in
10        error.
11    
12    2.  Changed the call to open a subject file in pcregrep from fopen(pathname,
13        "r") to fopen(pathname, "rb"), which fixed a problem with some of the tests
14        in a Windows environment.
15    
16    3.  The pcregrep --count option prints the count for each file even when it is
17        zero, as does GNU grep. However, pcregrep was also printing all files when
18        --files-with-matches was added. Now, when both options are given, it prints
19        counts only for those files that have at least one match. (GNU grep just
20        prints the file name in this circumstance, but including the count seems
21        more useful - otherwise, why use --count?) Also ensured that the
22        combination -clh just lists non-zero counts, with no names.
23    
24    4.  The long form of the pcregrep -F option was incorrectly implemented as
25        --fixed_strings instead of --fixed-strings. This is an incompatible change,
26        but it seems right to fix it, and I didn't think it was worth preserving
27        the old behaviour.
28    
29    5.  The command line items --regex=pattern and --regexp=pattern were not
30        recognized by pcregrep, which required --regex pattern or --regexp pattern
31        (with a space rather than an '='). The man page documented the '=' forms,
32        which are compatible with GNU grep; these now work.
33    
34    6.  No libpcreposix.pc file was created for pkg-config; there was just
35        libpcre.pc and libpcrecpp.pc. The omission has been rectified.
36    
37    7.  Added #ifndef SUPPORT_UCP into the pcre_ucd.c module, to reduce its size
38        when UCP support is not needed, by modifying the Python script that
39        generates it from Unicode data files. This should not matter if the module
40        is correctly used as a library, but I received one complaint about 50K of
41        unwanted data. My guess is that the person linked everything into his
42        program rather than using a library. Anyway, it does no harm.
43    
44    8.  A pattern such as /\x{123}{2,2}+/8 was incorrectly compiled; the trigger
45        was a minimum greater than 1 for a wide character in a possessive
46        repetition. The same bug could also affect patterns like /(\x{ff}{0,2})*/8
47        which had an unlimited repeat of a nested, fixed maximum repeat of a wide
48        character. Chaos in the form of incorrect output or a compiling loop could
49        result.
50    
51    9.  The restrictions on what a pattern can contain when partial matching is
52        requested for pcre_exec() have been removed. All patterns can now be
53        partially matched by this function. In addition, if there are at least two
54        slots in the offset vector, the offsets of the first-encountered partial
55        match are set in them when PCRE_ERROR_PARTIAL is returned.
56    
57    10. Partial matching has been split into two forms: PCRE_PARTIAL_SOFT, which is
58        synonymous with PCRE_PARTIAL, for backwards compatibility, and
59        PCRE_PARTIAL_HARD, which causes a partial match to supersede a full match,
60        and may be more useful for multi-segment matching, especially with
61        pcre_exec().
62    
63    11. Partial matching with pcre_exec() is now more intuitive. A partial match
64        used to be given if ever the end of the subject was reached; now it is
65        given only if matching could not proceed because another character was
66        needed. This makes a difference in some odd cases such as Z(*FAIL) with the
67        string "Z", which now yields "no match" instead of "partial match". In the
68        case of pcre_dfa_exec(), "no match" is given if every matching path for the
69        final character ended with (*FAIL).
70    
71    12. Restarting a match using pcre_dfa_exec() after a partial match did not work
72        if the pattern had a "must contain" character that was already found in the
73        earlier partial match, unless partial matching was again requested. For
74        example, with the pattern /dog.(body)?/, the "must contain" character is
75        "g". If the first part-match was for the string "dog", restarting with
76        "sbody" failed.
77    
78    13. Added a pcredemo man page, created automatically from the pcredemo.c file,
79        so that the demonstration program is easily available in environments where
80        PCRE has not been installed from source.
81    
82    14. Arranged to add -DPCRE_STATIC to cflags in libpcre.pc, libpcreposix.cp,
83        libpcrecpp.pc and pcre-config when PCRE is not compiled as a shared
84        library.
85    
86    
87    Version 7.9 11-Apr-09
88    ---------------------
89    
90    1.  When building with support for bzlib/zlib (pcregrep) and/or readline
91        (pcretest), all targets were linked against these libraries. This included
92        libpcre, libpcreposix, and libpcrecpp, even though they do not use these
93        libraries. This caused unwanted dependencies to be created. This problem
94        has been fixed, and now only pcregrep is linked with bzlib/zlib and only
95        pcretest is linked with readline.
96    
97    2.  The "typedef int BOOL" in pcre_internal.h that was included inside the
98        "#ifndef FALSE" condition by an earlier change (probably 7.8/18) has been
99        moved outside it again, because FALSE and TRUE are already defined in AIX,
100        but BOOL is not.
101    
102    3.  The pcre_config() function was treating the PCRE_MATCH_LIMIT and
103        PCRE_MATCH_LIMIT_RECURSION values as ints, when they should be long ints.
104    
105    4.  The pcregrep documentation said spaces were inserted as well as colons (or
106        hyphens) following file names and line numbers when outputting matching
107        lines. This is not true; no spaces are inserted. I have also clarified the
108        wording for the --colour (or --color) option.
109    
110    5.  In pcregrep, when --colour was used with -o, the list of matching strings
111        was not coloured; this is different to GNU grep, so I have changed it to be
112        the same.
113    
114    6.  When --colo(u)r was used in pcregrep, only the first matching substring in
115        each matching line was coloured. Now it goes on to look for further matches
116        of any of the test patterns, which is the same behaviour as GNU grep.
117    
118    7.  A pattern that could match an empty string could cause pcregrep to loop; it
119        doesn't make sense to accept an empty string match in pcregrep, so I have
120        locked it out (using PCRE's PCRE_NOTEMPTY option). By experiment, this
121        seems to be how GNU grep behaves.
122    
123    8.  The pattern (?(?=.*b)b|^) was incorrectly compiled as "match must be at
124        start or after a newline", because the conditional assertion was not being
125        correctly handled. The rule now is that both the assertion and what follows
126        in the first alternative must satisfy the test.
127    
128    9.  If auto-callout was enabled in a pattern with a conditional group whose
129        condition was an assertion, PCRE could crash during matching, both with
130        pcre_exec() and pcre_dfa_exec().
131    
132    10. The PCRE_DOLLAR_ENDONLY option was not working when pcre_dfa_exec() was
133        used for matching.
134    
135    11. Unicode property support in character classes was not working for
136        characters (bytes) greater than 127 when not in UTF-8 mode.
137    
138    12. Added the -M command line option to pcretest.
139    
140    14. Added the non-standard REG_NOTEMPTY option to the POSIX interface.
141    
142    15. Added the PCRE_NO_START_OPTIMIZE match-time option.
143    
144    16. Added comments and documentation about mis-use of no_arg in the C++
145        wrapper.
146    
147    17. Implemented support for UTF-8 encoding in EBCDIC environments, a patch
148        from Martin Jerabek that uses macro names for all relevant character and
149        string constants.
150    
151    18. Added to pcre_internal.h two configuration checks: (a) If both EBCDIC and
152        SUPPORT_UTF8 are set, give an error; (b) If SUPPORT_UCP is set without
153        SUPPORT_UTF8, define SUPPORT_UTF8. The "configure" script handles both of
154        these, but not everybody uses configure.
155    
156    19. A conditional group that had only one branch was not being correctly
157        recognized as an item that could match an empty string. This meant that an
158        enclosing group might also not be so recognized, causing infinite looping
159        (and probably a segfault) for patterns such as ^"((?(?=[a])[^"])|b)*"$
160        with the subject "ab", where knowledge that the repeated group can match
161        nothing is needed in order to break the loop.
162    
163    20. If a pattern that was compiled with callouts was matched using pcre_dfa_
164        exec(), but without supplying a callout function, matching went wrong.
165    
166    21. If PCRE_ERROR_MATCHLIMIT occurred during a recursion, there was a memory
167        leak if the size of the offset vector was greater than 30. When the vector
168        is smaller, the saved offsets during recursion go onto a local stack
169        vector, but for larger vectors malloc() is used. It was failing to free
170        when the recursion yielded PCRE_ERROR_MATCH_LIMIT (or any other "abnormal"
171        error, in fact).
172    
173    22. There was a missing #ifdef SUPPORT_UTF8 round one of the variables in the
174        heapframe that is used only when UTF-8 support is enabled. This caused no
175        problem, but was untidy.
176    
177    23. Steven Van Ingelgem's patch to CMakeLists.txt to change the name
178        CMAKE_BINARY_DIR to PROJECT_BINARY_DIR so that it works when PCRE is
179        included within another project.
180    
181    24. Steven Van Ingelgem's patches to add more options to the CMake support,
182        slightly modified by me:
183    
184          (a) PCRE_BUILD_TESTS can be set OFF not to build the tests, including
185              not building pcregrep.
186    
187          (b) PCRE_BUILD_PCREGREP can be see OFF not to build pcregrep, but only
188              if PCRE_BUILD_TESTS is also set OFF, because the tests use pcregrep.
189    
190    25. Forward references, both numeric and by name, in patterns that made use of
191        duplicate group numbers, could behave incorrectly or give incorrect errors,
192        because when scanning forward to find the reference group, PCRE was not
193        taking into account the duplicate group numbers. A pattern such as
194        ^X(?3)(a)(?|(b)|(q))(Y) is an example.
195    
196    26. Changed a few more instances of "const unsigned char *" to USPTR, making
197        the feature of a custom pointer more persuasive (as requested by a user).
198    
199    27. Wrapped the definitions of fileno and isatty for Windows, which appear in
200        pcretest.c, inside #ifndefs, because it seems they are sometimes already
201        pre-defined.
202    
203    28. Added support for (*UTF8) at the start of a pattern.
204    
205    29. Arrange for flags added by the "release type" setting in CMake to be shown
206        in the configuration summary.
207    
208    
209    Version 7.8 05-Sep-08
210    ---------------------
211    
212    1.  Replaced UCP searching code with optimized version as implemented for Ad
213        Muncher (http://www.admuncher.com/) by Peter Kankowski. This uses a two-
214        stage table and inline lookup instead of a function, giving speed ups of 2
215        to 5 times on some simple patterns that I tested. Permission was given to
216        distribute the MultiStage2.py script that generates the tables (it's not in
217        the tarball, but is in the Subversion repository).
218    
219    2.  Updated the Unicode datatables to Unicode 5.1.0. This adds yet more
220        scripts.
221    
222    3.  Change 12 for 7.7 introduced a bug in pcre_study() when a pattern contained
223        a group with a zero qualifier. The result of the study could be incorrect,
224        or the function might crash, depending on the pattern.
225    
226    4.  Caseless matching was not working for non-ASCII characters in back
227        references. For example, /(\x{de})\1/8i was not matching \x{de}\x{fe}.
228        It now works when Unicode Property Support is available.
229    
230    5.  In pcretest, an escape such as \x{de} in the data was always generating
231        a UTF-8 string, even in non-UTF-8 mode. Now it generates a single byte in
232        non-UTF-8 mode. If the value is greater than 255, it gives a warning about
233        truncation.
234    
235    6.  Minor bugfix in pcrecpp.cc (change "" == ... to NULL == ...).
236    
237    7.  Added two (int) casts to pcregrep when printing the difference of two
238        pointers, in case they are 64-bit values.
239    
240    8.  Added comments about Mac OS X stack usage to the pcrestack man page and to
241        test 2 if it fails.
242    
243    9.  Added PCRE_CALL_CONVENTION just before the names of all exported functions,
244        and a #define of that name to empty if it is not externally set. This is to
245        allow users of MSVC to set it if necessary.
246    
247    10. The PCRE_EXP_DEFN macro which precedes exported functions was missing from
248        the convenience functions in the pcre_get.c source file.
249    
250    11. An option change at the start of a pattern that had top-level alternatives
251        could cause overwriting and/or a crash. This command provoked a crash in
252        some environments:
253    
254          printf "/(?i)[\xc3\xa9\xc3\xbd]|[\xc3\xa9\xc3\xbdA]/8\n" | pcretest
255    
256        This potential security problem was recorded as CVE-2008-2371.
257    
258    12. For a pattern where the match had to start at the beginning or immediately
259        after a newline (e.g /.*anything/ without the DOTALL flag), pcre_exec() and
260        pcre_dfa_exec() could read past the end of the passed subject if there was
261        no match. To help with detecting such bugs (e.g. with valgrind), I modified
262        pcretest so that it places the subject at the end of its malloc-ed buffer.
263    
264    13. The change to pcretest in 12 above threw up a couple more cases when pcre_
265        exec() might read past the end of the data buffer in UTF-8 mode.
266    
267    14. A similar bug to 7.3/2 existed when the PCRE_FIRSTLINE option was set and
268        the data contained the byte 0x85 as part of a UTF-8 character within its
269        first line. This applied both to normal and DFA matching.
270    
271    15. Lazy qualifiers were not working in some cases in UTF-8 mode. For example,
272        /^[^d]*?$/8 failed to match "abc".
273    
274    16. Added a missing copyright notice to pcrecpp_internal.h.
275    
276    17. Make it more clear in the documentation that values returned from
277        pcre_exec() in ovector are byte offsets, not character counts.
278    
279    18. Tidied a few places to stop certain compilers from issuing warnings.
280    
281    19. Updated the Virtual Pascal + BCC files to compile the latest v7.7, as
282        supplied by Stefan Weber. I made a further small update for 7.8 because
283        there is a change of source arrangements: the pcre_searchfuncs.c module is
284        replaced by pcre_ucd.c.
285    
286    
287    Version 7.7 07-May-08
288    ---------------------
289    
290    1.  Applied Craig's patch to sort out a long long problem: "If we can't convert
291        a string to a long long, pretend we don't even have a long long." This is
292        done by checking for the strtoq, strtoll, and _strtoi64 functions.
293    
294    2.  Applied Craig's patch to pcrecpp.cc to restore ABI compatibility with
295        pre-7.6 versions, which defined a global no_arg variable instead of putting
296        it in the RE class. (See also #8 below.)
297    
298    3.  Remove a line of dead code, identified by coverity and reported by Nuno
299        Lopes.
300    
301    4.  Fixed two related pcregrep bugs involving -r with --include or --exclude:
302    
303        (1) The include/exclude patterns were being applied to the whole pathnames
304            of files, instead of just to the final components.
305    
306        (2) If there was more than one level of directory, the subdirectories were
307            skipped unless they satisfied the include/exclude conditions. This is
308            inconsistent with GNU grep (and could even be seen as contrary to the
309            pcregrep specification - which I improved to make it absolutely clear).
310            The action now is always to scan all levels of directory, and just
311            apply the include/exclude patterns to regular files.
312    
313    5.  Added the --include_dir and --exclude_dir patterns to pcregrep, and used
314        --exclude_dir in the tests to avoid scanning .svn directories.
315    
316    6.  Applied Craig's patch to the QuoteMeta function so that it escapes the
317        NUL character as backslash + 0 rather than backslash + NUL, because PCRE
318        doesn't support NULs in patterns.
319    
320    7.  Added some missing "const"s to declarations of static tables in
321        pcre_compile.c and pcre_dfa_exec.c.
322    
323    8.  Applied Craig's patch to pcrecpp.cc to fix a problem in OS X that was
324        caused by fix #2  above. (Subsequently also a second patch to fix the
325        first patch. And a third patch - this was a messy problem.)
326    
327    9.  Applied Craig's patch to remove the use of push_back().
328    
329    10. Applied Alan Lehotsky's patch to add REG_STARTEND support to the POSIX
330        matching function regexec().
331    
332    11. Added support for the Oniguruma syntax \g<name>, \g<n>, \g'name', \g'n',
333        which, however, unlike Perl's \g{...}, are subroutine calls, not back
334        references. PCRE supports relative numbers with this syntax (I don't think
335        Oniguruma does).
336    
337    12. Previously, a group with a zero repeat such as (...){0} was completely
338        omitted from the compiled regex. However, this means that if the group
339        was called as a subroutine from elsewhere in the pattern, things went wrong
340        (an internal error was given). Such groups are now left in the compiled
341        pattern, with a new opcode that causes them to be skipped at execution
342        time.
343    
344    13. Added the PCRE_JAVASCRIPT_COMPAT option. This makes the following changes
345        to the way PCRE behaves:
346    
347        (a) A lone ] character is dis-allowed (Perl treats it as data).
348    
349        (b) A back reference to an unmatched subpattern matches an empty string
350            (Perl fails the current match path).
351    
352        (c) A data ] in a character class must be notated as \] because if the
353            first data character in a class is ], it defines an empty class. (In
354            Perl it is not possible to have an empty class.) The empty class []
355            never matches; it forces failure and is equivalent to (*FAIL) or (?!).
356            The negative empty class [^] matches any one character, independently
357            of the DOTALL setting.
358    
359    14. A pattern such as /(?2)[]a()b](abc)/ which had a forward reference to a
360        non-existent subpattern following a character class starting with ']' and
361        containing () gave an internal compiling error instead of "reference to
362        non-existent subpattern". Fortunately, when the pattern did exist, the
363        compiled code was correct. (When scanning forwards to check for the
364        existencd of the subpattern, it was treating the data ']' as terminating
365        the class, so got the count wrong. When actually compiling, the reference
366        was subsequently set up correctly.)
367    
368    15. The "always fail" assertion (?!) is optimzed to (*FAIL) by pcre_compile;
369        it was being rejected as not supported by pcre_dfa_exec(), even though
370        other assertions are supported. I have made pcre_dfa_exec() support
371        (*FAIL).
372    
373    16. The implementation of 13c above involved the invention of a new opcode,
374        OP_ALLANY, which is like OP_ANY but doesn't check the /s flag. Since /s
375        cannot be changed at match time, I realized I could make a small
376        improvement to matching performance by compiling OP_ALLANY instead of
377        OP_ANY for "." when DOTALL was set, and then removing the runtime tests
378        on the OP_ANY path.
379    
380    17. Compiling pcretest on Windows with readline support failed without the
381        following two fixes: (1) Make the unistd.h include conditional on
382        HAVE_UNISTD_H; (2) #define isatty and fileno as _isatty and _fileno.
383    
384    18. Changed CMakeLists.txt and cmake/FindReadline.cmake to arrange for the
385        ncurses library to be included for pcretest when ReadLine support is
386        requested, but also to allow for it to be overridden. This patch came from
387        Daniel Bergström.
388    
389    19. There was a typo in the file ucpinternal.h where f0_rangeflag was defined
390        as 0x00f00000 instead of 0x00800000. Luckily, this would not have caused
391        any errors with the current Unicode tables. Thanks to Peter Kankowski for
392        spotting this.
393    
394    
395    Version 7.6 28-Jan-08
396    ---------------------
397    
398    1.  A character class containing a very large number of characters with
399        codepoints greater than 255 (in UTF-8 mode, of course) caused a buffer
400        overflow.
401    
402    2.  Patch to cut out the "long long" test in pcrecpp_unittest when
403        HAVE_LONG_LONG is not defined.
404    
405    3.  Applied Christian Ehrlicher's patch to update the CMake build files to
406        bring them up to date and include new features. This patch includes:
407    
408        - Fixed PH's badly added libz and libbz2 support.
409        - Fixed a problem with static linking.
410        - Added pcredemo. [But later removed - see 7 below.]
411        - Fixed dftables problem and added an option.
412        - Added a number of HAVE_XXX tests, including HAVE_WINDOWS_H and
413            HAVE_LONG_LONG.
414        - Added readline support for pcretest.
415        - Added an listing of the option settings after cmake has run.
416    
417    4.  A user submitted a patch to Makefile that makes it easy to create
418        "pcre.dll" under mingw when using Configure/Make. I added stuff to
419        Makefile.am that cause it to include this special target, without
420        affecting anything else. Note that the same mingw target plus all
421        the other distribution libraries and programs are now supported
422        when configuring with CMake (see 6 below) instead of with
423        Configure/Make.
424    
425    5.  Applied Craig's patch that moves no_arg into the RE class in the C++ code.
426        This is an attempt to solve the reported problem "pcrecpp::no_arg is not
427        exported in the Windows port". It has not yet been confirmed that the patch
428        solves the problem, but it does no harm.
429    
430    6.  Applied Sheri's patch to CMakeLists.txt to add NON_STANDARD_LIB_PREFIX and
431        NON_STANDARD_LIB_SUFFIX for dll names built with mingw when configured
432        with CMake, and also correct the comment about stack recursion.
433    
434    7.  Remove the automatic building of pcredemo from the ./configure system and
435        from CMakeLists.txt. The whole idea of pcredemo.c is that it is an example
436        of a program that users should build themselves after PCRE is installed, so
437        building it automatically is not really right. What is more, it gave
438        trouble in some build environments.
439    
440    8.  Further tidies to CMakeLists.txt from Sheri and Christian.
441    
442    
443    Version 7.5 10-Jan-08
444    ---------------------
445    
446    1.  Applied a patch from Craig: "This patch makes it possible to 'ignore'
447        values in parens when parsing an RE using the C++ wrapper."
448    
449    2.  Negative specials like \S did not work in character classes in UTF-8 mode.
450        Characters greater than 255 were excluded from the class instead of being
451        included.
452    
453    3.  The same bug as (2) above applied to negated POSIX classes such as
454        [:^space:].
455    
456    4.  PCRECPP_STATIC was referenced in pcrecpp_internal.h, but nowhere was it
457        defined or documented. It seems to have been a typo for PCRE_STATIC, so
458        I have changed it.
459    
460    5.  The construct (?&) was not diagnosed as a syntax error (it referenced the
461        first named subpattern) and a construct such as (?&a) would reference the
462        first named subpattern whose name started with "a" (in other words, the
463        length check was missing). Both these problems are fixed. "Subpattern name
464        expected" is now given for (?&) (a zero-length name), and this patch also
465        makes it give the same error for \k'' (previously it complained that that
466        was a reference to a non-existent subpattern).
467    
468    6.  The erroneous patterns (?+-a) and (?-+a) give different error messages;
469        this is right because (?- can be followed by option settings as well as by
470        digits. I have, however, made the messages clearer.
471    
472    7.  Patterns such as (?(1)a|b) (a pattern that contains fewer subpatterns
473        than the number used in the conditional) now cause a compile-time error.
474        This is actually not compatible with Perl, which accepts such patterns, but
475        treats the conditional as always being FALSE (as PCRE used to), but it
476        seems to me that giving a diagnostic is better.
477    
478    8.  Change "alphameric" to the more common word "alphanumeric" in comments
479        and messages.
480    
481    9.  Fix two occurrences of "backslash" in comments that should have been
482        "backspace".
483    
484    10. Remove two redundant lines of code that can never be obeyed (their function
485        was moved elsewhere).
486    
487    11. The program that makes PCRE's Unicode character property table had a bug
488        which caused it to generate incorrect table entries for sequences of
489        characters that have the same character type, but are in different scripts.
490        It amalgamated them into a single range, with the script of the first of
491        them. In other words, some characters were in the wrong script. There were
492        thirteen such cases, affecting characters in the following ranges:
493    
494          U+002b0 - U+002c1
495          U+0060c - U+0060d
496          U+0061e - U+00612
497          U+0064b - U+0065e
498          U+0074d - U+0076d
499          U+01800 - U+01805
500          U+01d00 - U+01d77
501          U+01d9b - U+01dbf
502          U+0200b - U+0200f
503          U+030fc - U+030fe
504          U+03260 - U+0327f
505          U+0fb46 - U+0fbb1
506          U+10450 - U+1049d
507    
508    12. The -o option (show only the matching part of a line) for pcregrep was not
509        compatible with GNU grep in that, if there was more than one match in a
510        line, it showed only the first of them. It now behaves in the same way as
511        GNU grep.
512    
513    13. If the -o and -v options were combined for pcregrep, it printed a blank
514        line for every non-matching line. GNU grep prints nothing, and pcregrep now
515        does the same. The return code can be used to tell if there were any
516        non-matching lines.
517    
518    14. Added --file-offsets and --line-offsets to pcregrep.
519    
520    15. The pattern (?=something)(?R) was not being diagnosed as a potentially
521        infinitely looping recursion. The bug was that positive lookaheads were not
522        being skipped when checking for a possible empty match (negative lookaheads
523        and both kinds of lookbehind were skipped).
524    
525    16. Fixed two typos in the Windows-only code in pcregrep.c, and moved the
526        inclusion of <windows.h> to before rather than after the definition of
527        INVALID_FILE_ATTRIBUTES (patch from David Byron).
528    
529    17. Specifying a possessive quantifier with a specific limit for a Unicode
530        character property caused pcre_compile() to compile bad code, which led at
531        runtime to PCRE_ERROR_INTERNAL (-14). Examples of patterns that caused this
532        are: /\p{Zl}{2,3}+/8 and /\p{Cc}{2}+/8. It was the possessive "+" that
533        caused the error; without that there was no problem.
534    
535    18. Added --enable-pcregrep-libz and --enable-pcregrep-libbz2.
536    
537    19. Added --enable-pcretest-libreadline.
538    
539    20. In pcrecpp.cc, the variable 'count' was incremented twice in
540        RE::GlobalReplace(). As a result, the number of replacements returned was
541        double what it should be. I removed one of the increments, but Craig sent a
542        later patch that removed the other one (the right fix) and added unit tests
543        that check the return values (which was not done before).
544    
545    21. Several CMake things:
546    
547        (1) Arranged that, when cmake is used on Unix, the libraries end up with
548            the names libpcre and libpcreposix, not just pcre and pcreposix.
549    
550        (2) The above change means that pcretest and pcregrep are now correctly
551            linked with the newly-built libraries, not previously installed ones.
552    
553        (3) Added PCRE_SUPPORT_LIBREADLINE, PCRE_SUPPORT_LIBZ, PCRE_SUPPORT_LIBBZ2.
554    
555    22. In UTF-8 mode, with newline set to "any", a pattern such as .*a.*=.b.*
556        crashed when matching a string such as a\x{2029}b (note that \x{2029} is a
557        UTF-8 newline character). The key issue is that the pattern starts .*;
558        this means that the match must be either at the beginning, or after a
559        newline. The bug was in the code for advancing after a failed match and
560        checking that the new position followed a newline. It was not taking
561        account of UTF-8 characters correctly.
562    
563    23. PCRE was behaving differently from Perl in the way it recognized POSIX
564        character classes. PCRE was not treating the sequence [:...:] as a
565        character class unless the ... were all letters. Perl, however, seems to
566        allow any characters between [: and :], though of course it rejects as
567        unknown any "names" that contain non-letters, because all the known class
568        names consist only of letters. Thus, Perl gives an error for [[:1234:]],
569        for example, whereas PCRE did not - it did not recognize a POSIX character
570        class. This seemed a bit dangerous, so the code has been changed to be
571        closer to Perl. The behaviour is not identical to Perl, because PCRE will
572        diagnose an unknown class for, for example, [[:l\ower:]] where Perl will
573        treat it as [[:lower:]]. However, PCRE does now give "unknown" errors where
574        Perl does, and where it didn't before.
575    
576    24. Rewrite so as to remove the single use of %n from pcregrep because in some
577        Windows environments %n is disabled by default.
578    
579    
580    Version 7.4 21-Sep-07
581    ---------------------
582    
583    1.  Change 7.3/28 was implemented for classes by looking at the bitmap. This
584        means that a class such as [\s] counted as "explicit reference to CR or
585        LF". That isn't really right - the whole point of the change was to try to
586        help when there was an actual mention of one of the two characters. So now
587        the change happens only if \r or \n (or a literal CR or LF) character is
588        encountered.
589    
590    2.  The 32-bit options word was also used for 6 internal flags, but the numbers
591        of both had grown to the point where there were only 3 bits left.
592        Fortunately, there was spare space in the data structure, and so I have
593        moved the internal flags into a new 16-bit field to free up more option
594        bits.
595    
596    3.  The appearance of (?J) at the start of a pattern set the DUPNAMES option,
597        but did not set the internal JCHANGED flag - either of these is enough to
598        control the way the "get" function works - but the PCRE_INFO_JCHANGED
599        facility is supposed to tell if (?J) was ever used, so now (?J) at the
600        start sets both bits.
601    
602    4.  Added options (at build time, compile time, exec time) to change \R from
603        matching any Unicode line ending sequence to just matching CR, LF, or CRLF.
604    
605    5.  doc/pcresyntax.html was missing from the distribution.
606    
607    6.  Put back the definition of PCRE_ERROR_NULLWSLIMIT, for backward
608        compatibility, even though it is no longer used.
609    
610    7.  Added macro for snprintf to pcrecpp_unittest.cc and also for strtoll and
611        strtoull to pcrecpp.cc to select the available functions in WIN32 when the
612        windows.h file is present (where different names are used). [This was
613        reversed later after testing - see 16 below.]
614    
615    8.  Changed all #include <config.h> to #include "config.h". There were also
616        some further <pcre.h> cases that I changed to "pcre.h".
617    
618    9.  When pcregrep was used with the --colour option, it missed the line ending
619        sequence off the lines that it output.
620    
621    10. It was pointed out to me that arrays of string pointers cause lots of
622        relocations when a shared library is dynamically loaded. A technique of
623        using a single long string with a table of offsets can drastically reduce
624        these. I have refactored PCRE in four places to do this. The result is
625        dramatic:
626    
627          Originally:                          290
628          After changing UCP table:            187
629          After changing error message table:   43
630          After changing table of "verbs"       36
631          After changing table of Posix names   22
632    
633        Thanks to the folks working on Gregex for glib for this insight.
634    
635    11. --disable-stack-for-recursion caused compiling to fail unless -enable-
636        unicode-properties was also set.
637    
638    12. Updated the tests so that they work when \R is defaulted to ANYCRLF.
639    
640    13. Added checks for ANY and ANYCRLF to pcrecpp.cc where it previously
641        checked only for CRLF.
642    
643    14. Added casts to pcretest.c to avoid compiler warnings.
644    
645    15. Added Craig's patch to various pcrecpp modules to avoid compiler warnings.
646    
647    16. Added Craig's patch to remove the WINDOWS_H tests, that were not working,
648        and instead check for _strtoi64 explicitly, and avoid the use of snprintf()
649        entirely. This removes changes made in 7 above.
650    
651    17. The CMake files have been updated, and there is now more information about
652        building with CMake in the NON-UNIX-USE document.
653    
654    
655    Version 7.3 28-Aug-07
656    ---------------------
657    
658     1. In the rejigging of the build system that eventually resulted in 7.1, the
659        line "#include <pcre.h>" was included in pcre_internal.h. The use of angle
660        brackets there is not right, since it causes compilers to look for an
661        installed pcre.h, not the version that is in the source that is being
662        compiled (which of course may be different). I have changed it back to:
663    
664          #include "pcre.h"
665    
666        I have a vague recollection that the change was concerned with compiling in
667        different directories, but in the new build system, that is taken care of
668        by the VPATH setting the Makefile.
669    
670     2. The pattern .*$ when run in not-DOTALL UTF-8 mode with newline=any failed
671        when the subject happened to end in the byte 0x85 (e.g. if the last
672        character was \x{1ec5}). *Character* 0x85 is one of the "any" newline
673        characters but of course it shouldn't be taken as a newline when it is part
674        of another character. The bug was that, for an unlimited repeat of . in
675        not-DOTALL UTF-8 mode, PCRE was advancing by bytes rather than by
676        characters when looking for a newline.
677    
678     3. A small performance improvement in the DOTALL UTF-8 mode .* case.
679    
680     4. Debugging: adjusted the names of opcodes for different kinds of parentheses
681        in debug output.
682    
683     5. Arrange to use "%I64d" instead of "%lld" and "%I64u" instead of "%llu" for
684        long printing in the pcrecpp unittest when running under MinGW.
685    
686     6. ESC_K was left out of the EBCDIC table.
687    
688     7. Change 7.0/38 introduced a new limit on the number of nested non-capturing
689        parentheses; I made it 1000, which seemed large enough. Unfortunately, the
690        limit also applies to "virtual nesting" when a pattern is recursive, and in
691        this case 1000 isn't so big. I have been able to remove this limit at the
692        expense of backing off one optimization in certain circumstances. Normally,
693        when pcre_exec() would call its internal match() function recursively and
694        immediately return the result unconditionally, it uses a "tail recursion"
695        feature to save stack. However, when a subpattern that can match an empty
696        string has an unlimited repetition quantifier, it no longer makes this
697        optimization. That gives it a stack frame in which to save the data for
698        checking that an empty string has been matched. Previously this was taken
699        from the 1000-entry workspace that had been reserved. So now there is no
700        explicit limit, but more stack is used.
701    
702     8. Applied Daniel's patches to solve problems with the import/export magic
703        syntax that is required for Windows, and which was going wrong for the
704        pcreposix and pcrecpp parts of the library. These were overlooked when this
705        problem was solved for the main library.
706    
707     9. There were some crude static tests to avoid integer overflow when computing
708        the size of patterns that contain repeated groups with explicit upper
709        limits. As the maximum quantifier is 65535, the maximum group length was
710        set at 30,000 so that the product of these two numbers did not overflow a
711        32-bit integer. However, it turns out that people want to use groups that
712        are longer than 30,000 bytes (though not repeat them that many times).
713        Change 7.0/17 (the refactoring of the way the pattern size is computed) has
714        made it possible to implement the integer overflow checks in a much more
715        dynamic way, which I have now done. The artificial limitation on group
716        length has been removed - we now have only the limit on the total length of
717        the compiled pattern, which depends on the LINK_SIZE setting.
718    
719    10. Fixed a bug in the documentation for get/copy named substring when
720        duplicate names are permitted. If none of the named substrings are set, the
721        functions return PCRE_ERROR_NOSUBSTRING (7); the doc said they returned an
722        empty string.
723    
724    11. Because Perl interprets \Q...\E at a high level, and ignores orphan \E
725        instances, patterns such as [\Q\E] or [\E] or even [^\E] cause an error,
726        because the ] is interpreted as the first data character and the
727        terminating ] is not found. PCRE has been made compatible with Perl in this
728        regard. Previously, it interpreted [\Q\E] as an empty class, and [\E] could
729        cause memory overwriting.
730    
731    10. Like Perl, PCRE automatically breaks an unlimited repeat after an empty
732        string has been matched (to stop an infinite loop). It was not recognizing
733        a conditional subpattern that could match an empty string if that
734        subpattern was within another subpattern. For example, it looped when
735        trying to match  (((?(1)X|))*)  but it was OK with  ((?(1)X|)*)  where the
736        condition was not nested. This bug has been fixed.
737    
738    12. A pattern like \X?\d or \P{L}?\d in non-UTF-8 mode could cause a backtrack
739        past the start of the subject in the presence of bytes with the top bit
740        set, for example "\x8aBCD".
741    
742    13. Added Perl 5.10 experimental backtracking controls (*FAIL), (*F), (*PRUNE),
743        (*SKIP), (*THEN), (*COMMIT), and (*ACCEPT).
744    
745    14. Optimized (?!) to (*FAIL).
746    
747    15. Updated the test for a valid UTF-8 string to conform to the later RFC 3629.
748        This restricts code points to be within the range 0 to 0x10FFFF, excluding
749        the "low surrogate" sequence 0xD800 to 0xDFFF. Previously, PCRE allowed the
750        full range 0 to 0x7FFFFFFF, as defined by RFC 2279. Internally, it still
751        does: it's just the validity check that is more restrictive.
752    
753    16. Inserted checks for integer overflows during escape sequence (backslash)
754        processing, and also fixed erroneous offset values for syntax errors during
755        backslash processing.
756    
757    17. Fixed another case of looking too far back in non-UTF-8 mode (cf 12 above)
758        for patterns like [\PPP\x8a]{1,}\x80 with the subject "A\x80".
759    
760    18. An unterminated class in a pattern like (?1)\c[ with a "forward reference"
761        caused an overrun.
762    
763    19. A pattern like (?:[\PPa*]*){8,} which had an "extended class" (one with
764        something other than just ASCII characters) inside a group that had an
765        unlimited repeat caused a loop at compile time (while checking to see
766        whether the group could match an empty string).
767    
768    20. Debugging a pattern containing \p or \P could cause a crash. For example,
769        [\P{Any}] did so. (Error in the code for printing property names.)
770    
771    21. An orphan \E inside a character class could cause a crash.
772    
773    22. A repeated capturing bracket such as (A)? could cause a wild memory
774        reference during compilation.
775    
776    23. There are several functions in pcre_compile() that scan along a compiled
777        expression for various reasons (e.g. to see if it's fixed length for look
778        behind). There were bugs in these functions when a repeated \p or \P was
779        present in the pattern. These operators have additional parameters compared
780        with \d, etc, and these were not being taken into account when moving along
781        the compiled data. Specifically:
782    
783        (a) A item such as \p{Yi}{3} in a lookbehind was not treated as fixed
784            length.
785    
786        (b) An item such as \pL+ within a repeated group could cause crashes or
787            loops.
788    
789        (c) A pattern such as \p{Yi}+(\P{Yi}+)(?1) could give an incorrect
790            "reference to non-existent subpattern" error.
791    
792        (d) A pattern like (\P{Yi}{2}\277)? could loop at compile time.
793    
794    24. A repeated \S or \W in UTF-8 mode could give wrong answers when multibyte
795        characters were involved (for example /\S{2}/8g with "A\x{a3}BC").
796    
797    25. Using pcregrep in multiline, inverted mode (-Mv) caused it to loop.
798    
799    26. Patterns such as [\P{Yi}A] which include \p or \P and just one other
800        character were causing crashes (broken optimization).
801    
802    27. Patterns such as (\P{Yi}*\277)* (group with possible zero repeat containing
803        \p or \P) caused a compile-time loop.
804    
805    28. More problems have arisen in unanchored patterns when CRLF is a valid line
806        break. For example, the unstudied pattern [\r\n]A does not match the string
807        "\r\nA" because change 7.0/46 below moves the current point on by two
808        characters after failing to match at the start. However, the pattern \nA
809        *does* match, because it doesn't start till \n, and if [\r\n]A is studied,
810        the same is true. There doesn't seem any very clean way out of this, but
811        what I have chosen to do makes the common cases work: PCRE now takes note
812        of whether there can be an explicit match for \r or \n anywhere in the
813        pattern, and if so, 7.0/46 no longer applies. As part of this change,
814        there's a new PCRE_INFO_HASCRORLF option for finding out whether a compiled
815        pattern has explicit CR or LF references.
816    
817    29. Added (*CR) etc for changing newline setting at start of pattern.
818    
819    
820    Version 7.2 19-Jun-07
821    ---------------------
822    
823     1. If the fr_FR locale cannot be found for test 3, try the "french" locale,
824        which is apparently normally available under Windows.
825    
826     2. Re-jig the pcregrep tests with different newline settings in an attempt
827        to make them independent of the local environment's newline setting.
828    
829     3. Add code to configure.ac to remove -g from the CFLAGS default settings.
830    
831     4. Some of the "internals" tests were previously cut out when the link size
832        was not 2, because the output contained actual offsets. The recent new
833        "Z" feature of pcretest means that these can be cut out, making the tests
834        usable with all link sizes.
835    
836     5. Implemented Stan Switzer's goto replacement for longjmp() when not using
837        stack recursion. This gives a massive performance boost under BSD, but just
838        a small improvement under Linux. However, it saves one field in the frame
839        in all cases.
840    
841     6. Added more features from the forthcoming Perl 5.10:
842    
843        (a) (?-n) (where n is a string of digits) is a relative subroutine or
844            recursion call. It refers to the nth most recently opened parentheses.
845    
846        (b) (?+n) is also a relative subroutine call; it refers to the nth next
847            to be opened parentheses.
848    
849        (c) Conditions that refer to capturing parentheses can be specified
850            relatively, for example, (?(-2)... or (?(+3)...
851    
852        (d) \K resets the start of the current match so that everything before
853            is not part of it.
854    
855        (e) \k{name} is synonymous with \k<name> and \k'name' (.NET compatible).
856    
857        (f) \g{name} is another synonym - part of Perl 5.10's unification of
858            reference syntax.
859    
860        (g) (?| introduces a group in which the numbering of parentheses in each
861            alternative starts with the same number.
862    
863        (h) \h, \H, \v, and \V match horizontal and vertical whitespace.
864    
865     7. Added two new calls to pcre_fullinfo(): PCRE_INFO_OKPARTIAL and
866        PCRE_INFO_JCHANGED.
867    
868     8. A pattern such as  (.*(.)?)*  caused pcre_exec() to fail by either not
869        terminating or by crashing. Diagnosed by Viktor Griph; it was in the code
870        for detecting groups that can match an empty string.
871    
872     9. A pattern with a very large number of alternatives (more than several
873        hundred) was running out of internal workspace during the pre-compile
874        phase, where pcre_compile() figures out how much memory will be needed. A
875        bit of new cunning has reduced the workspace needed for groups with
876        alternatives. The 1000-alternative test pattern now uses 12 bytes of
877        workspace instead of running out of the 4096 that are available.
878    
879    10. Inserted some missing (unsigned int) casts to get rid of compiler warnings.
880    
881    11. Applied patch from Google to remove an optimization that didn't quite work.
882        The report of the bug said:
883    
884          pcrecpp::RE("a*").FullMatch("aaa") matches, while
885          pcrecpp::RE("a*?").FullMatch("aaa") does not, and
886          pcrecpp::RE("a*?\\z").FullMatch("aaa") does again.
887    
888    12. If \p or \P was used in non-UTF-8 mode on a character greater than 127
889        it matched the wrong number of bytes.
890    
891    
892    Version 7.1 24-Apr-07
893    ---------------------
894    
895     1. Applied Bob Rossi and Daniel G's patches to convert the build system to one
896        that is more "standard", making use of automake and other Autotools. There
897        is some re-arrangement of the files and adjustment of comments consequent
898        on this.
899    
900     2. Part of the patch fixed a problem with the pcregrep tests. The test of -r
901        for recursive directory scanning broke on some systems because the files
902        are not scanned in any specific order and on different systems the order
903        was different. A call to "sort" has been inserted into RunGrepTest for the
904        approprate test as a short-term fix. In the longer term there may be an
905        alternative.
906    
907     3. I had an email from Eric Raymond about problems translating some of PCRE's
908        man pages to HTML (despite the fact that I distribute HTML pages, some
909        people do their own conversions for various reasons). The problems
910        concerned the use of low-level troff macros .br and .in. I have therefore
911        removed all such uses from the man pages (some were redundant, some could
912        be replaced by .nf/.fi pairs). The 132html script that I use to generate
913        HTML has been updated to handle .nf/.fi and to complain if it encounters
914        .br or .in.
915    
916     4. Updated comments in configure.ac that get placed in config.h.in and also
917        arranged for config.h to be included in the distribution, with the name
918        config.h.generic, for the benefit of those who have to compile without
919        Autotools (compare pcre.h, which is now distributed as pcre.h.generic).
920    
921     5. Updated the support (such as it is) for Virtual Pascal, thanks to Stefan
922        Weber: (1) pcre_internal.h was missing some function renames; (2) updated
923        makevp.bat for the current PCRE, using the additional files
924        makevp_c.txt, makevp_l.txt, and pcregexp.pas.
925    
926     6. A Windows user reported a minor discrepancy with test 2, which turned out
927        to be caused by a trailing space on an input line that had got lost in his
928        copy. The trailing space was an accident, so I've just removed it.
929    
930     7. Add -Wl,-R... flags in pcre-config.in for *BSD* systems, as I'm told
931        that is needed.
932    
933     8. Mark ucp_table (in ucptable.h) and ucp_gentype (in pcre_ucp_searchfuncs.c)
934        as "const" (a) because they are and (b) because it helps the PHP
935        maintainers who have recently made a script to detect big data structures
936        in the php code that should be moved to the .rodata section. I remembered
937        to update Builducptable as well, so it won't revert if ucptable.h is ever
938        re-created.
939    
940     9. Added some extra #ifdef SUPPORT_UTF8 conditionals into pcretest.c,
941        pcre_printint.src, pcre_compile.c, pcre_study.c, and pcre_tables.c, in
942        order to be able to cut out the UTF-8 tables in the latter when UTF-8
943        support is not required. This saves 1.5-2K of code, which is important in
944        some applications.
945    
946        Later: more #ifdefs are needed in pcre_ord2utf8.c and pcre_valid_utf8.c
947        so as not to refer to the tables, even though these functions will never be
948        called when UTF-8 support is disabled. Otherwise there are problems with a
949        shared library.
950    
951    10. Fixed two bugs in the emulated memmove() function in pcre_internal.h:
952    
953        (a) It was defining its arguments as char * instead of void *.
954    
955        (b) It was assuming that all moves were upwards in memory; this was true
956            a long time ago when I wrote it, but is no longer the case.
957    
958        The emulated memove() is provided for those environments that have neither
959        memmove() nor bcopy(). I didn't think anyone used it these days, but that
960        is clearly not the case, as these two bugs were recently reported.
961    
962    11. The script PrepareRelease is now distributed: it calls 132html, CleanTxt,
963        and Detrail to create the HTML documentation, the .txt form of the man
964        pages, and it removes trailing spaces from listed files. It also creates
965        pcre.h.generic and config.h.generic from pcre.h and config.h. In the latter
966        case, it wraps all the #defines with #ifndefs. This script should be run
967        before "make dist".
968    
969    12. Fixed two fairly obscure bugs concerned with quantified caseless matching
970        with Unicode property support.
971    
972        (a) For a maximizing quantifier, if the two different cases of the
973            character were of different lengths in their UTF-8 codings (there are
974            some cases like this - I found 11), and the matching function had to
975            back up over a mixture of the two cases, it incorrectly assumed they
976            were both the same length.
977    
978        (b) When PCRE was configured to use the heap rather than the stack for
979            recursion during matching, it was not correctly preserving the data for
980            the other case of a UTF-8 character when checking ahead for a match
981            while processing a minimizing repeat. If the check also involved
982            matching a wide character, but failed, corruption could cause an
983            erroneous result when trying to check for a repeat of the original
984            character.
985    
986    13. Some tidying changes to the testing mechanism:
987    
988        (a) The RunTest script now detects the internal link size and whether there
989            is UTF-8 and UCP support by running ./pcretest -C instead of relying on
990            values substituted by "configure". (The RunGrepTest script already did
991            this for UTF-8.) The configure.ac script no longer substitutes the
992            relevant variables.
993    
994        (b) The debugging options /B and /D in pcretest show the compiled bytecode
995            with length and offset values. This means that the output is different
996            for different internal link sizes. Test 2 is skipped for link sizes
997            other than 2 because of this, bypassing the problem. Unfortunately,
998            there was also a test in test 3 (the locale tests) that used /B and
999            failed for link sizes other than 2. Rather than cut the whole test out,
1000            I have added a new /Z option to pcretest that replaces the length and
1001            offset values with spaces. This is now used to make test 3 independent
1002            of link size. (Test 2 will be tidied up later.)
1003    
1004    14. If erroroffset was passed as NULL to pcre_compile, it provoked a
1005        segmentation fault instead of returning the appropriate error message.
1006    
1007    15. In multiline mode when the newline sequence was set to "any", the pattern
1008        ^$ would give a match between the \r and \n of a subject such as "A\r\nB".
1009        This doesn't seem right; it now treats the CRLF combination as the line
1010        ending, and so does not match in that case. It's only a pattern such as ^$
1011        that would hit this one: something like ^ABC$ would have failed after \r
1012        and then tried again after \r\n.
1013    
1014    16. Changed the comparison command for RunGrepTest from "diff -u" to "diff -ub"
1015        in an attempt to make files that differ only in their line terminators
1016        compare equal. This works on Linux.
1017    
1018    17. Under certain error circumstances pcregrep might try to free random memory
1019        as it exited. This is now fixed, thanks to valgrind.
1020    
1021    19. In pcretest, if the pattern /(?m)^$/g<any> was matched against the string
1022        "abc\r\n\r\n", it found an unwanted second match after the second \r. This
1023        was because its rules for how to advance for /g after matching an empty
1024        string at the end of a line did not allow for this case. They now check for
1025        it specially.
1026    
1027    20. pcretest is supposed to handle patterns and data of any length, by
1028        extending its buffers when necessary. It was getting this wrong when the
1029        buffer for a data line had to be extended.
1030    
1031    21. Added PCRE_NEWLINE_ANYCRLF which is like ANY, but matches only CR, LF, or
1032        CRLF as a newline sequence.
1033    
1034    22. Code for handling Unicode properties in pcre_dfa_exec() wasn't being cut
1035        out by #ifdef SUPPORT_UCP. This did no harm, as it could never be used, but
1036        I have nevertheless tidied it up.
1037    
1038    23. Added some casts to kill warnings from HP-UX ia64 compiler.
1039    
1040    24. Added a man page for pcre-config.
1041    
1042    
1043    Version 7.0 19-Dec-06
1044    ---------------------
1045    
1046     1. Fixed a signed/unsigned compiler warning in pcre_compile.c, shown up by
1047        moving to gcc 4.1.1.
1048    
1049     2. The -S option for pcretest uses setrlimit(); I had omitted to #include
1050        sys/time.h, which is documented as needed for this function. It doesn't
1051        seem to matter on Linux, but it showed up on some releases of OS X.
1052    
1053     3. It seems that there are systems where bytes whose values are greater than
1054        127 match isprint() in the "C" locale. The "C" locale should be the
1055        default when a C program starts up. In most systems, only ASCII printing
1056        characters match isprint(). This difference caused the output from pcretest
1057        to vary, making some of the tests fail. I have changed pcretest so that:
1058    
1059        (a) When it is outputting text in the compiled version of a pattern, bytes
1060            other than 32-126 are always shown as hex escapes.
1061    
1062        (b) When it is outputting text that is a matched part of a subject string,
1063            it does the same, unless a different locale has been set for the match
1064            (using the /L modifier). In this case, it uses isprint() to decide.
1065    
1066     4. Fixed a major bug that caused incorrect computation of the amount of memory
1067        required for a compiled pattern when options that changed within the
1068        pattern affected the logic of the preliminary scan that determines the
1069        length. The relevant options are -x, and -i in UTF-8 mode. The result was
1070        that the computed length was too small. The symptoms of this bug were
1071        either the PCRE error "internal error: code overflow" from pcre_compile(),
1072        or a glibc crash with a message such as "pcretest: free(): invalid next
1073        size (fast)". Examples of patterns that provoked this bug (shown in
1074        pcretest format) are:
1075    
1076          /(?-x: )/x
1077          /(?x)(?-x: \s*#\s*)/
1078          /((?i)[\x{c0}])/8
1079          /(?i:[\x{c0}])/8
1080    
1081        HOWEVER: Change 17 below makes this fix obsolete as the memory computation
1082        is now done differently.
1083    
1084     5. Applied patches from Google to: (a) add a QuoteMeta function to the C++
1085        wrapper classes; (b) implement a new function in the C++ scanner that is
1086        more efficient than the old way of doing things because it avoids levels of
1087        recursion in the regex matching; (c) add a paragraph to the documentation
1088        for the FullMatch() function.
1089    
1090     6. The escape sequence \n was being treated as whatever was defined as
1091        "newline". Not only was this contrary to the documentation, which states
1092        that \n is character 10 (hex 0A), but it also went horribly wrong when
1093        "newline" was defined as CRLF. This has been fixed.
1094    
1095     7. In pcre_dfa_exec.c the value of an unsigned integer (the variable called c)
1096        was being set to -1 for the "end of line" case (supposedly a value that no
1097        character can have). Though this value is never used (the check for end of
1098        line is "zero bytes in current character"), it caused compiler complaints.
1099        I've changed it to 0xffffffff.
1100    
1101     8. In pcre_version.c, the version string was being built by a sequence of
1102        C macros that, in the event of PCRE_PRERELEASE being defined as an empty
1103        string (as it is for production releases) called a macro with an empty
1104        argument. The C standard says the result of this is undefined. The gcc
1105        compiler treats it as an empty string (which was what was wanted) but it is
1106        reported that Visual C gives an error. The source has been hacked around to
1107        avoid this problem.
1108    
1109     9. On the advice of a Windows user, included <io.h> and <fcntl.h> in Windows
1110        builds of pcretest, and changed the call to _setmode() to use _O_BINARY
1111        instead of 0x8000. Made all the #ifdefs test both _WIN32 and WIN32 (not all
1112        of them did).
1113    
1114    10. Originally, pcretest opened its input and output without "b"; then I was
1115        told that "b" was needed in some environments, so it was added for release
1116        5.0 to both the input and output. (It makes no difference on Unix-like
1117        systems.) Later I was told that it is wrong for the input on Windows. I've
1118        now abstracted the modes into two macros, to make it easier to fiddle with
1119        them, and removed "b" from the input mode under Windows.
1120    
1121    11. Added pkgconfig support for the C++ wrapper library, libpcrecpp.
1122    
1123    12. Added -help and --help to pcretest as an official way of being reminded
1124        of the options.
1125    
1126    13. Removed some redundant semicolons after macro calls in pcrecpparg.h.in
1127        and pcrecpp.cc because they annoy compilers at high warning levels.
1128    
1129    14. A bit of tidying/refactoring in pcre_exec.c in the main bumpalong loop.
1130    
1131    15. Fixed an occurrence of == in configure.ac that should have been = (shell
1132        scripts are not C programs :-) and which was not noticed because it works
1133        on Linux.
1134    
1135    16. pcretest is supposed to handle any length of pattern and data line (as one
1136        line or as a continued sequence of lines) by extending its input buffer if
1137        necessary. This feature was broken for very long pattern lines, leading to
1138        a string of junk being passed to pcre_compile() if the pattern was longer
1139        than about 50K.
1140    
1141    17. I have done a major re-factoring of the way pcre_compile() computes the
1142        amount of memory needed for a compiled pattern. Previously, there was code
1143        that made a preliminary scan of the pattern in order to do this. That was
1144        OK when PCRE was new, but as the facilities have expanded, it has become
1145        harder and harder to keep it in step with the real compile phase, and there
1146        have been a number of bugs (see for example, 4 above). I have now found a
1147        cunning way of running the real compile function in a "fake" mode that
1148        enables it to compute how much memory it would need, while actually only
1149        ever using a few hundred bytes of working memory and without too many
1150        tests of the mode. This should make future maintenance and development
1151        easier. A side effect of this work is that the limit of 200 on the nesting
1152        depth of parentheses has been removed (though this was never a serious
1153        limitation, I suspect). However, there is a downside: pcre_compile() now
1154        runs more slowly than before (30% or more, depending on the pattern). I
1155        hope this isn't a big issue. There is no effect on runtime performance.
1156    
1157    18. Fixed a minor bug in pcretest: if a pattern line was not terminated by a
1158        newline (only possible for the last line of a file) and it was a
1159        pattern that set a locale (followed by /Lsomething), pcretest crashed.
1160    
1161    19. Added additional timing features to pcretest. (1) The -tm option now times
1162        matching only, not compiling. (2) Both -t and -tm can be followed, as a
1163        separate command line item, by a number that specifies the number of
1164        repeats to use when timing. The default is 50000; this gives better
1165        precision, but takes uncomfortably long for very large patterns.
1166    
1167    20. Extended pcre_study() to be more clever in cases where a branch of a
1168        subpattern has no definite first character. For example, (a*|b*)[cd] would
1169        previously give no result from pcre_study(). Now it recognizes that the
1170        first character must be a, b, c, or d.
1171    
1172    21. There was an incorrect error "recursive call could loop indefinitely" if
1173        a subpattern (or the entire pattern) that was being tested for matching an
1174        empty string contained only one non-empty item after a nested subpattern.
1175        For example, the pattern (?>\x{100}*)\d(?R) provoked this error
1176        incorrectly, because the \d was being skipped in the check.
1177    
1178    22. The pcretest program now has a new pattern option /B and a command line
1179        option -b, which is equivalent to adding /B to every pattern. This causes
1180        it to show the compiled bytecode, without the additional information that
1181        -d shows. The effect of -d is now the same as -b with -i (and similarly, /D
1182        is the same as /B/I).
1183    
1184    23. A new optimization is now able automatically to treat some sequences such
1185        as a*b as a*+b. More specifically, if something simple (such as a character
1186        or a simple class like \d) has an unlimited quantifier, and is followed by
1187        something that cannot possibly match the quantified thing, the quantifier
1188        is automatically "possessified".
1189    
1190    24. A recursive reference to a subpattern whose number was greater than 39
1191        went wrong under certain circumstances in UTF-8 mode. This bug could also
1192        have affected the operation of pcre_study().
1193    
1194    25. Realized that a little bit of performance could be had by replacing
1195        (c & 0xc0) == 0xc0 with c >= 0xc0 when processing UTF-8 characters.
1196    
1197    26. Timing data from pcretest is now shown to 4 decimal places instead of 3.
1198    
1199    27. Possessive quantifiers such as a++ were previously implemented by turning
1200        them into atomic groups such as ($>a+). Now they have their own opcodes,
1201        which improves performance. This includes the automatically created ones
1202        from 23 above.
1203    
1204    28. A pattern such as (?=(\w+))\1: which simulates an atomic group using a
1205        lookahead was broken if it was not anchored. PCRE was mistakenly expecting
1206        the first matched character to be a colon. This applied both to named and
1207        numbered groups.
1208    
1209    29. The ucpinternal.h header file was missing its idempotency #ifdef.
1210    
1211    30. I was sent a "project" file called libpcre.a.dev which I understand makes
1212        building PCRE on Windows easier, so I have included it in the distribution.
1213    
1214    31. There is now a check in pcretest against a ridiculously large number being
1215        returned by pcre_exec() or pcre_dfa_exec(). If this happens in a /g or /G
1216        loop, the loop is abandoned.
1217    
1218    32. Forward references to subpatterns in conditions such as (?(2)...) where
1219        subpattern 2 is defined later cause pcre_compile() to search forwards in
1220        the pattern for the relevant set of parentheses. This search went wrong
1221        when there were unescaped parentheses in a character class, parentheses
1222        escaped with \Q...\E, or parentheses in a #-comment in /x mode.
1223    
1224    33. "Subroutine" calls and backreferences were previously restricted to
1225        referencing subpatterns earlier in the regex. This restriction has now
1226        been removed.
1227    
1228    34. Added a number of extra features that are going to be in Perl 5.10. On the
1229        whole, these are just syntactic alternatives for features that PCRE had
1230        previously implemented using the Python syntax or my own invention. The
1231        other formats are all retained for compatibility.
1232    
1233        (a) Named groups can now be defined as (?<name>...) or (?'name'...) as well
1234            as (?P<name>...). The new forms, as well as being in Perl 5.10, are
1235            also .NET compatible.
1236    
1237        (b) A recursion or subroutine call to a named group can now be defined as
1238            (?&name) as well as (?P>name).
1239    
1240        (c) A backreference to a named group can now be defined as \k<name> or
1241            \k'name' as well as (?P=name). The new forms, as well as being in Perl
1242            5.10, are also .NET compatible.
1243    
1244        (d) A conditional reference to a named group can now use the syntax
1245            (?(<name>) or (?('name') as well as (?(name).
1246    
1247        (e) A "conditional group" of the form (?(DEFINE)...) can be used to define
1248            groups (named and numbered) that are never evaluated inline, but can be
1249            called as "subroutines" from elsewhere. In effect, the DEFINE condition
1250            is always false. There may be only one alternative in such a group.
1251    
1252        (f) A test for recursion can be given as (?(R1).. or (?(R&name)... as well
1253            as the simple (?(R). The condition is true only if the most recent
1254            recursion is that of the given number or name. It does not search out
1255            through the entire recursion stack.
1256    
1257        (g) The escape \gN or \g{N} has been added, where N is a positive or
1258            negative number, specifying an absolute or relative reference.
1259    
1260    35. Tidied to get rid of some further signed/unsigned compiler warnings and
1261        some "unreachable code" warnings.
1262    
1263    36. Updated the Unicode property tables to Unicode version 5.0.0. Amongst other
1264        things, this adds five new scripts.
1265    
1266    37. Perl ignores orphaned \E escapes completely. PCRE now does the same.
1267        There were also incompatibilities regarding the handling of \Q..\E inside
1268        character classes, for example with patterns like [\Qa\E-\Qz\E] where the
1269        hyphen was adjacent to \Q or \E. I hope I've cleared all this up now.
1270    
1271    38. Like Perl, PCRE detects when an indefinitely repeated parenthesized group
1272        matches an empty string, and forcibly breaks the loop. There were bugs in
1273        this code in non-simple cases. For a pattern such as  ^(a()*)*  matched
1274        against  aaaa  the result was just "a" rather than "aaaa", for example. Two
1275        separate and independent bugs (that affected different cases) have been
1276        fixed.
1277    
1278    39. Refactored the code to abolish the use of different opcodes for small
1279        capturing bracket numbers. This is a tidy that I avoided doing when I
1280        removed the limit on the number of capturing brackets for 3.5 back in 2001.
1281        The new approach is not only tidier, it makes it possible to reduce the
1282        memory needed to fix the previous bug (38).
1283    
1284    40. Implemented PCRE_NEWLINE_ANY to recognize any of the Unicode newline
1285        sequences (http://unicode.org/unicode/reports/tr18/) as "newline" when
1286        processing dot, circumflex, or dollar metacharacters, or #-comments in /x
1287        mode.
1288    
1289    41. Add \R to match any Unicode newline sequence, as suggested in the Unicode
1290        report.
1291    
1292    42. Applied patch, originally from Ari Pollak, modified by Google, to allow
1293        copy construction and assignment in the C++ wrapper.
1294    
1295    43. Updated pcregrep to support "--newline=any". In the process, I fixed a
1296        couple of bugs that could have given wrong results in the "--newline=crlf"
1297        case.
1298    
1299    44. Added a number of casts and did some reorganization of signed/unsigned int
1300        variables following suggestions from Dair Grant. Also renamed the variable
1301        "this" as "item" because it is a C++ keyword.
1302    
1303    45. Arranged for dftables to add
1304    
1305          #include "pcre_internal.h"
1306    
1307        to pcre_chartables.c because without it, gcc 4.x may remove the array
1308        definition from the final binary if PCRE is built into a static library and
1309        dead code stripping is activated.
1310    
1311    46. For an unanchored pattern, if a match attempt fails at the start of a
1312        newline sequence, and the newline setting is CRLF or ANY, and the next two
1313        characters are CRLF, advance by two characters instead of one.
1314    
1315    
1316    Version 6.7 04-Jul-06
1317    ---------------------
1318    
1319     1. In order to handle tests when input lines are enormously long, pcretest has
1320        been re-factored so that it automatically extends its buffers when
1321        necessary. The code is crude, but this _is_ just a test program. The
1322        default size has been increased from 32K to 50K.
1323    
1324     2. The code in pcre_study() was using the value of the re argument before
1325        testing it for NULL. (Of course, in any sensible call of the function, it
1326        won't be NULL.)
1327    
1328     3. The memmove() emulation function in pcre_internal.h, which is used on
1329        systems that lack both memmove() and bcopy() - that is, hardly ever -
1330        was missing a "static" storage class specifier.
1331    
1332     4. When UTF-8 mode was not set, PCRE looped when compiling certain patterns
1333        containing an extended class (one that cannot be represented by a bitmap
1334        because it contains high-valued characters or Unicode property items, e.g.
1335        [\pZ]). Almost always one would set UTF-8 mode when processing such a
1336        pattern, but PCRE should not loop if you do not (it no longer does).
1337        [Detail: two cases were found: (a) a repeated subpattern containing an
1338        extended class; (b) a recursive reference to a subpattern that followed a
1339        previous extended class. It wasn't skipping over the extended class
1340        correctly when UTF-8 mode was not set.]
1341    
1342     5. A negated single-character class was not being recognized as fixed-length
1343        in lookbehind assertions such as (?<=[^f]), leading to an incorrect
1344        compile error "lookbehind assertion is not fixed length".
1345    
1346     6. The RunPerlTest auxiliary script was showing an unexpected difference
1347        between PCRE and Perl for UTF-8 tests. It turns out that it is hard to
1348        write a Perl script that can interpret lines of an input file either as
1349        byte characters or as UTF-8, which is what "perltest" was being required to
1350        do for the non-UTF-8 and UTF-8 tests, respectively. Essentially what you
1351        can't do is switch easily at run time between having the "use utf8;" pragma
1352        or not. In the end, I fudged it by using the RunPerlTest script to insert
1353        "use utf8;" explicitly for the UTF-8 tests.
1354    
1355     7. In multiline (/m) mode, PCRE was matching ^ after a terminating newline at
1356        the end of the subject string, contrary to the documentation and to what
1357        Perl does. This was true of both matching functions. Now it matches only at
1358        the start of the subject and immediately after *internal* newlines.
1359    
1360     8. A call of pcre_fullinfo() from pcretest to get the option bits was passing
1361        a pointer to an int instead of a pointer to an unsigned long int. This
1362        caused problems on 64-bit systems.
1363    
1364     9. Applied a patch from the folks at Google to pcrecpp.cc, to fix "another
1365        instance of the 'standard' template library not being so standard".
1366    
1367    10. There was no check on the number of named subpatterns nor the maximum
1368        length of a subpattern name. The product of these values is used to compute
1369        the size of the memory block for a compiled pattern. By supplying a very
1370        long subpattern name and a large number of named subpatterns, the size
1371        computation could be caused to overflow. This is now prevented by limiting
1372        the length of names to 32 characters, and the number of named subpatterns
1373        to 10,000.
1374    
1375    11. Subpatterns that are repeated with specific counts have to be replicated in
1376        the compiled pattern. The size of memory for this was computed from the
1377        length of the subpattern and the repeat count. The latter is limited to
1378        65535, but there was no limit on the former, meaning that integer overflow
1379        could in principle occur. The compiled length of a repeated subpattern is
1380        now limited to 30,000 bytes in order to prevent this.
1381    
1382    12. Added the optional facility to have named substrings with the same name.
1383    
1384    13. Added the ability to use a named substring as a condition, using the
1385        Python syntax: (?(name)yes|no). This overloads (?(R)... and names that
1386        are numbers (not recommended). Forward references are permitted.
1387    
1388    14. Added forward references in named backreferences (if you see what I mean).
1389    
1390    15. In UTF-8 mode, with the PCRE_DOTALL option set, a quantified dot in the
1391        pattern could run off the end of the subject. For example, the pattern
1392        "(?s)(.{1,5})"8 did this with the subject "ab".
1393    
1394    16. If PCRE_DOTALL or PCRE_MULTILINE were set, pcre_dfa_exec() behaved as if
1395        PCRE_CASELESS was set when matching characters that were quantified with ?
1396        or *.
1397    
1398    17. A character class other than a single negated character that had a minimum
1399        but no maximum quantifier - for example [ab]{6,} - was not handled
1400        correctly by pce_dfa_exec(). It would match only one character.
1401    
1402    18. A valid (though odd) pattern that looked like a POSIX character
1403        class but used an invalid character after [ (for example [[,abc,]]) caused
1404        pcre_compile() to give the error "Failed: internal error: code overflow" or
1405        in some cases to crash with a glibc free() error. This could even happen if
1406        the pattern terminated after [[ but there just happened to be a sequence of
1407        letters, a binary zero, and a closing ] in the memory that followed.
1408    
1409    19. Perl's treatment of octal escapes in the range \400 to \777 has changed
1410        over the years. Originally (before any Unicode support), just the bottom 8
1411        bits were taken. Thus, for example, \500 really meant \100. Nowadays the
1412        output from "man perlunicode" includes this:
1413    
1414          The regular expression compiler produces polymorphic opcodes.  That
1415          is, the pattern adapts to the data and automatically switches to
1416          the Unicode character scheme when presented with Unicode data--or
1417          instead uses a traditional byte scheme when presented with byte
1418          data.
1419    
1420        Sadly, a wide octal escape does not cause a switch, and in a string with
1421        no other multibyte characters, these octal escapes are treated as before.
1422        Thus, in Perl, the pattern  /\500/ actually matches \100 but the pattern
1423        /\500|\x{1ff}/ matches \500 or \777 because the whole thing is treated as a
1424        Unicode string.
1425    
1426        I have not perpetrated such confusion in PCRE. Up till now, it took just
1427        the bottom 8 bits, as in old Perl. I have now made octal escapes with
1428        values greater than \377 illegal in non-UTF-8 mode. In UTF-8 mode they
1429        translate to the appropriate multibyte character.
1430    
1431    29. Applied some refactoring to reduce the number of warnings from Microsoft
1432        and Borland compilers. This has included removing the fudge introduced
1433        seven years ago for the OS/2 compiler (see 2.02/2 below) because it caused
1434        a warning about an unused variable.
1435    
1436    21. PCRE has not included VT (character 0x0b) in the set of whitespace
1437        characters since release 4.0, because Perl (from release 5.004) does not.
1438        [Or at least, is documented not to: some releases seem to be in conflict
1439        with the documentation.] However, when a pattern was studied with
1440        pcre_study() and all its branches started with \s, PCRE still included VT
1441        as a possible starting character. Of course, this did no harm; it just
1442        caused an unnecessary match attempt.
1443    
1444    22. Removed a now-redundant internal flag bit that recorded the fact that case
1445        dependency changed within the pattern. This was once needed for "required
1446        byte" processing, but is no longer used. This recovers a now-scarce options
1447        bit. Also moved the least significant internal flag bit to the most-
1448        significant bit of the word, which was not previously used (hangover from
1449        the days when it was an int rather than a uint) to free up another bit for
1450        the future.
1451    
1452    23. Added support for CRLF line endings as well as CR and LF. As well as the
1453        default being selectable at build time, it can now be changed at runtime
1454        via the PCRE_NEWLINE_xxx flags. There are now options for pcregrep to
1455        specify that it is scanning data with non-default line endings.
1456    
1457    24. Changed the definition of CXXLINK to make it agree with the definition of
1458        LINK in the Makefile, by replacing LDFLAGS to CXXFLAGS.
1459    
1460    25. Applied Ian Taylor's patches to avoid using another stack frame for tail
1461        recursions. This makes a big different to stack usage for some patterns.
1462    
1463    26. If a subpattern containing a named recursion or subroutine reference such
1464        as (?P>B) was quantified, for example (xxx(?P>B)){3}, the calculation of
1465        the space required for the compiled pattern went wrong and gave too small a
1466        value. Depending on the environment, this could lead to "Failed: internal
1467        error: code overflow at offset 49" or "glibc detected double free or
1468        corruption" errors.
1469    
1470    27. Applied patches from Google (a) to support the new newline modes and (b) to
1471        advance over multibyte UTF-8 characters in GlobalReplace.
1472    
1473    28. Change free() to pcre_free() in pcredemo.c. Apparently this makes a
1474        difference for some implementation of PCRE in some Windows version.
1475    
1476    29. Added some extra testing facilities to pcretest:
1477    
1478        \q<number>   in a data line sets the "match limit" value
1479        \Q<number>   in a data line sets the "match recursion limt" value
1480        -S <number>  sets the stack size, where <number> is in megabytes
1481    
1482        The -S option isn't available for Windows.
1483    
1484    
1485    Version 6.6 06-Feb-06
1486    ---------------------
1487    
1488     1. Change 16(a) for 6.5 broke things, because PCRE_DATA_SCOPE was not defined
1489        in pcreposix.h. I have copied the definition from pcre.h.
1490    
1491     2. Change 25 for 6.5 broke compilation in a build directory out-of-tree
1492        because pcre.h is no longer a built file.
1493    
1494     3. Added Jeff Friedl's additional debugging patches to pcregrep. These are
1495        not normally included in the compiled code.
1496    
1497    
1498    Version 6.5 01-Feb-06
1499    ---------------------
1500    
1501     1. When using the partial match feature with pcre_dfa_exec(), it was not
1502        anchoring the second and subsequent partial matches at the new starting
1503        point. This could lead to incorrect results. For example, with the pattern
1504        /1234/, partially matching against "123" and then "a4" gave a match.
1505    
1506     2. Changes to pcregrep:
1507    
1508        (a) All non-match returns from pcre_exec() were being treated as failures
1509            to match the line. Now, unless the error is PCRE_ERROR_NOMATCH, an
1510            error message is output. Some extra information is given for the
1511            PCRE_ERROR_MATCHLIMIT and PCRE_ERROR_RECURSIONLIMIT errors, which are
1512            probably the only errors that are likely to be caused by users (by
1513            specifying a regex that has nested indefinite repeats, for instance).
1514            If there are more than 20 of these errors, pcregrep is abandoned.
1515    
1516        (b) A binary zero was treated as data while matching, but terminated the
1517            output line if it was written out. This has been fixed: binary zeroes
1518            are now no different to any other data bytes.
1519    
1520        (c) Whichever of the LC_ALL or LC_CTYPE environment variables is set is
1521            used to set a locale for matching. The --locale=xxxx long option has
1522            been added (no short equivalent) to specify a locale explicitly on the
1523            pcregrep command, overriding the environment variables.
1524    
1525        (d) When -B was used with -n, some line numbers in the output were one less
1526            than they should have been.
1527    
1528        (e) Added the -o (--only-matching) option.
1529    
1530        (f) If -A or -C was used with -c (count only), some lines of context were
1531            accidentally printed for the final match.
1532    
1533        (g) Added the -H (--with-filename) option.
1534    
1535        (h) The combination of options -rh failed to suppress file names for files
1536            that were found from directory arguments.
1537    
1538        (i) Added the -D (--devices) and -d (--directories) options.
1539    
1540        (j) Added the -F (--fixed-strings) option.
1541    
1542        (k) Allow "-" to be used as a file name for -f as well as for a data file.
1543    
1544        (l) Added the --colo(u)r option.
1545    
1546        (m) Added Jeffrey Friedl's -S testing option, but within #ifdefs so that it
1547            is not present by default.
1548    
1549     3. A nasty bug was discovered in the handling of recursive patterns, that is,
1550        items such as (?R) or (?1), when the recursion could match a number of
1551        alternatives. If it matched one of the alternatives, but subsequently,
1552        outside the recursion, there was a failure, the code tried to back up into
1553        the recursion. However, because of the way PCRE is implemented, this is not
1554        possible, and the result was an incorrect result from the match.
1555    
1556        In order to prevent this happening, the specification of recursion has
1557        been changed so that all such subpatterns are automatically treated as
1558        atomic groups. Thus, for example, (?R) is treated as if it were (?>(?R)).
1559    
1560     4. I had overlooked the fact that, in some locales, there are characters for
1561        which isalpha() is true but neither isupper() nor islower() are true. In
1562        the fr_FR locale, for instance, the \xAA and \xBA characters (ordmasculine
1563        and ordfeminine) are like this. This affected the treatment of \w and \W
1564        when they appeared in character classes, but not when they appeared outside
1565        a character class. The bit map for "word" characters is now created
1566        separately from the results of isalnum() instead of just taking it from the
1567        upper, lower, and digit maps. (Plus the underscore character, of course.)
1568    
1569     5. The above bug also affected the handling of POSIX character classes such as
1570        [[:alpha:]] and [[:alnum:]]. These do not have their own bit maps in PCRE's
1571        permanent tables. Instead, the bit maps for such a class were previously
1572        created as the appropriate unions of the upper, lower, and digit bitmaps.
1573        Now they are created by subtraction from the [[:word:]] class, which has
1574        its own bitmap.
1575    
1576     6. The [[:blank:]] character class matches horizontal, but not vertical space.
1577        It is created by subtracting the vertical space characters (\x09, \x0a,
1578        \x0b, \x0c) from the [[:space:]] bitmap. Previously, however, the
1579        subtraction was done in the overall bitmap for a character class, meaning
1580        that a class such as [\x0c[:blank:]] was incorrect because \x0c would not
1581        be recognized. This bug has been fixed.
1582    
1583     7. Patches from the folks at Google:
1584    
1585          (a) pcrecpp.cc: "to handle a corner case that may or may not happen in
1586          real life, but is still worth protecting against".
1587    
1588          (b) pcrecpp.cc: "corrects a bug when negative radixes are used with
1589          regular expressions".
1590    
1591          (c) pcre_scanner.cc: avoid use of std::count() because not all systems
1592          have it.
1593    
1594          (d) Split off pcrecpparg.h from pcrecpp.h and had the former built by
1595          "configure" and the latter not, in order to fix a problem somebody had
1596          with compiling the Arg class on HP-UX.
1597    
1598          (e) Improve the error-handling of the C++ wrapper a little bit.
1599    
1600          (f) New tests for checking recursion limiting.
1601    
1602     8. The pcre_memmove() function, which is used only if the environment does not
1603        have a standard memmove() function (and is therefore rarely compiled),
1604        contained two bugs: (a) use of int instead of size_t, and (b) it was not
1605        returning a result (though PCRE never actually uses the result).
1606    
1607     9. In the POSIX regexec() interface, if nmatch is specified as a ridiculously
1608        large number - greater than INT_MAX/(3*sizeof(int)) - REG_ESPACE is
1609        returned instead of calling malloc() with an overflowing number that would
1610        most likely cause subsequent chaos.
1611    
1612    10. The debugging option of pcretest was not showing the NO_AUTO_CAPTURE flag.
1613    
1614    11. The POSIX flag REG_NOSUB is now supported. When a pattern that was compiled
1615        with this option is matched, the nmatch and pmatch options of regexec() are
1616        ignored.
1617    
1618    12. Added REG_UTF8 to the POSIX interface. This is not defined by POSIX, but is
1619        provided in case anyone wants to the the POSIX interface with UTF-8
1620        strings.
1621    
1622    13. Added CXXLDFLAGS to the Makefile parameters to provide settings only on the
1623        C++ linking (needed for some HP-UX environments).
1624    
1625    14. Avoid compiler warnings in get_ucpname() when compiled without UCP support
1626        (unused parameter) and in the pcre_printint() function (omitted "default"
1627        switch label when the default is to do nothing).
1628    
1629    15. Added some code to make it possible, when PCRE is compiled as a C++
1630        library, to replace subject pointers for pcre_exec() with a smart pointer
1631        class, thus making it possible to process discontinuous strings.
1632    
1633    16. The two macros PCRE_EXPORT and PCRE_DATA_SCOPE are confusing, and perform
1634        much the same function. They were added by different people who were trying
1635        to make PCRE easy to compile on non-Unix systems. It has been suggested
1636        that PCRE_EXPORT be abolished now that there is more automatic apparatus
1637        for compiling on Windows systems. I have therefore replaced it with
1638        PCRE_DATA_SCOPE. This is set automatically for Windows; if not set it
1639        defaults to "extern" for C or "extern C" for C++, which works fine on
1640        Unix-like systems. It is now possible to override the value of PCRE_DATA_
1641        SCOPE with something explicit in config.h. In addition:
1642    
1643        (a) pcreposix.h still had just "extern" instead of either of these macros;
1644            I have replaced it with PCRE_DATA_SCOPE.
1645    
1646        (b) Functions such as _pcre_xclass(), which are internal to the library,
1647            but external in the C sense, all had PCRE_EXPORT in their definitions.
1648            This is apparently wrong for the Windows case, so I have removed it.
1649            (It makes no difference on Unix-like systems.)
1650    
1651    17. Added a new limit, MATCH_LIMIT_RECURSION, which limits the depth of nesting
1652        of recursive calls to match(). This is different to MATCH_LIMIT because
1653        that limits the total number of calls to match(), not all of which increase
1654        the depth of recursion. Limiting the recursion depth limits the amount of
1655        stack (or heap if NO_RECURSE is set) that is used. The default can be set
1656        when PCRE is compiled, and changed at run time. A patch from Google adds
1657        this functionality to the C++ interface.
1658    
1659    18. Changes to the handling of Unicode character properties:
1660    
1661        (a) Updated the table to Unicode 4.1.0.
1662    
1663        (b) Recognize characters that are not in the table as "Cn" (undefined).
1664    
1665        (c) I revised the way the table is implemented to a much improved format
1666            which includes recognition of ranges. It now supports the ranges that
1667            are defined in UnicodeData.txt, and it also amalgamates other
1668            characters into ranges. This has reduced the number of entries in the
1669            table from around 16,000 to around 3,000, thus reducing its size
1670            considerably. I realized I did not need to use a tree structure after
1671            all - a binary chop search is just as efficient. Having reduced the
1672            number of entries, I extended their size from 6 bytes to 8 bytes to
1673            allow for more data.
1674    
1675        (d) Added support for Unicode script names via properties such as \p{Han}.
1676    
1677    19. In UTF-8 mode, a backslash followed by a non-Ascii character was not
1678        matching that character.
1679    
1680    20. When matching a repeated Unicode property with a minimum greater than zero,
1681        (for example \pL{2,}), PCRE could look past the end of the subject if it
1682        reached it while seeking the minimum number of characters. This could
1683        happen only if some of the characters were more than one byte long, because
1684        there is a check for at least the minimum number of bytes.
1685    
1686    21. Refactored the implementation of \p and \P so as to be more general, to
1687        allow for more different types of property in future. This has changed the
1688        compiled form incompatibly. Anybody with saved compiled patterns that use
1689        \p or \P will have to recompile them.
1690    
1691    22. Added "Any" and "L&" to the supported property types.
1692    
1693    23. Recognize \x{...} as a code point specifier, even when not in UTF-8 mode,
1694        but give a compile time error if the value is greater than 0xff.
1695    
1696    24. The man pages for pcrepartial, pcreprecompile, and pcre_compile2 were
1697        accidentally not being installed or uninstalled.
1698    
1699    25. The pcre.h file was built from pcre.h.in, but the only changes that were
1700        made were to insert the current release number. This seemed silly, because
1701        it made things harder for people building PCRE on systems that don't run
1702        "configure". I have turned pcre.h into a distributed file, no longer built
1703        by "configure", with the version identification directly included. There is
1704        no longer a pcre.h.in file.
1705    
1706        However, this change necessitated a change to the pcre-config script as
1707        well. It is built from pcre-config.in, and one of the substitutions was the
1708        release number. I have updated configure.ac so that ./configure now finds
1709        the release number by grepping pcre.h.
1710    
1711    26. Added the ability to run the tests under valgrind.
1712    
1713    
1714    Version 6.4 05-Sep-05
1715    ---------------------
1716    
1717     1. Change 6.0/10/(l) to pcregrep introduced a bug that caused separator lines
1718        "--" to be printed when multiple files were scanned, even when none of the
1719        -A, -B, or -C options were used. This is not compatible with Gnu grep, so I
1720        consider it to be a bug, and have restored the previous behaviour.
1721    
1722     2. A couple of code tidies to get rid of compiler warnings.
1723    
1724     3. The pcretest program used to cheat by referring to symbols in the library
1725        whose names begin with _pcre_. These are internal symbols that are not
1726        really supposed to be visible externally, and in some environments it is
1727        possible to suppress them. The cheating is now confined to including
1728        certain files from the library's source, which is a bit cleaner.
1729    
1730     4. Renamed pcre.in as pcre.h.in to go with pcrecpp.h.in; it also makes the
1731        file's purpose clearer.
1732    
1733     5. Reorganized pcre_ucp_findchar().
1734    
1735    
1736    Version 6.3 15-Aug-05
1737    ---------------------
1738    
1739     1. The file libpcre.pc.in did not have general read permission in the tarball.
1740    
1741     2. There were some problems when building without C++ support:
1742    
1743        (a) If C++ support was not built, "make install" and "make test" still
1744            tried to test it.
1745    
1746        (b) There were problems when the value of CXX was explicitly set. Some
1747            changes have been made to try to fix these, and ...
1748    
1749        (c) --disable-cpp can now be used to explicitly disable C++ support.
1750    
1751        (d) The use of @CPP_OBJ@ directly caused a blank line preceded by a
1752            backslash in a target when C++ was disabled. This confuses some
1753            versions of "make", apparently. Using an intermediate variable solves
1754            this. (Same for CPP_LOBJ.)
1755    
1756     3. $(LINK_FOR_BUILD) now includes $(CFLAGS_FOR_BUILD) and $(LINK)
1757        (non-Windows) now includes $(CFLAGS) because these flags are sometimes
1758        necessary on certain architectures.
1759    
1760     4. Added a setting of -export-symbols-regex to the link command to remove
1761        those symbols that are exported in the C sense, but actually are local
1762        within the library, and not documented. Their names all begin with
1763        "_pcre_". This is not a perfect job, because (a) we have to except some
1764        symbols that pcretest ("illegally") uses, and (b) the facility isn't always
1765        available (and never for static libraries). I have made a note to try to
1766        find a way round (a) in the future.
1767    
1768    
1769    Version 6.2 01-Aug-05
1770    ---------------------
1771    
1772     1. There was no test for integer overflow of quantifier values. A construction
1773        such as {1111111111111111} would give undefined results. What is worse, if
1774        a minimum quantifier for a parenthesized subpattern overflowed and became
1775        negative, the calculation of the memory size went wrong. This could have
1776        led to memory overwriting.
1777    
1778     2. Building PCRE using VPATH was broken. Hopefully it is now fixed.
1779    
1780     3. Added "b" to the 2nd argument of fopen() in dftables.c, for non-Unix-like
1781        operating environments where this matters.
1782    
1783     4. Applied Giuseppe Maxia's patch to add additional features for controlling
1784        PCRE options from within the C++ wrapper.
1785    
1786     5. Named capturing subpatterns were not being correctly counted when a pattern
1787        was compiled. This caused two problems: (a) If there were more than 100
1788        such subpatterns, the calculation of the memory needed for the whole
1789        compiled pattern went wrong, leading to an overflow error. (b) Numerical
1790        back references of the form \12, where the number was greater than 9, were
1791        not recognized as back references, even though there were sufficient
1792        previous subpatterns.
1793    
1794     6. Two minor patches to pcrecpp.cc in order to allow it to compile on older
1795        versions of gcc, e.g. 2.95.4.
1796    
1797    
1798    Version 6.1 21-Jun-05
1799    ---------------------
1800    
1801     1. There was one reference to the variable "posix" in pcretest.c that was not
1802        surrounded by "#if !defined NOPOSIX".
1803    
1804     2. Make it possible to compile pcretest without DFA support, UTF8 support, or
1805        the cross-check on the old pcre_info() function, for the benefit of the
1806        cut-down version of PCRE that is currently imported into Exim.
1807    
1808     3. A (silly) pattern starting with (?i)(?-i) caused an internal space
1809        allocation error. I've done the easy fix, which wastes 2 bytes for sensible
1810        patterns that start (?i) but I don't think that matters. The use of (?i) is
1811        just an example; this all applies to the other options as well.
1812    
1813     4. Since libtool seems to echo the compile commands it is issuing, the output
1814        from "make" can be reduced a bit by putting "@" in front of each libtool
1815        compile command.
1816    
1817     5. Patch from the folks at Google for configure.in to be a bit more thorough
1818        in checking for a suitable C++ installation before trying to compile the
1819        C++ stuff. This should fix a reported problem when a compiler was present,
1820        but no suitable headers.
1821    
1822     6. The man pages all had just "PCRE" as their title. I have changed them to
1823        be the relevant file name. I have also arranged that these names are
1824        retained in the file doc/pcre.txt, which is a concatenation in text format
1825        of all the man pages except the little individual ones for each function.
1826    
1827     7. The NON-UNIX-USE file had not been updated for the different set of source
1828        files that come with release 6. I also added a few comments about the C++
1829        wrapper.
1830    
1831    
1832    Version 6.0 07-Jun-05
1833    ---------------------
1834    
1835     1. Some minor internal re-organization to help with my DFA experiments.
1836    
1837     2. Some missing #ifdef SUPPORT_UCP conditionals in pcretest and printint that
1838        didn't matter for the library itself when fully configured, but did matter
1839        when compiling without UCP support, or within Exim, where the ucp files are
1840        not imported.
1841    
1842     3. Refactoring of the library code to split up the various functions into
1843        different source modules. The addition of the new DFA matching code (see
1844        below) to a single monolithic source would have made it really too
1845        unwieldy, quite apart from causing all the code to be include in a
1846        statically linked application, when only some functions are used. This is
1847        relevant even without the DFA addition now that patterns can be compiled in
1848        one application and matched in another.
1849    
1850        The downside of splitting up is that there have to be some external
1851        functions and data tables that are used internally in different modules of
1852        the library but which are not part of the API. These have all had their
1853        names changed to start with "_pcre_" so that they are unlikely to clash
1854        with other external names.
1855    
1856     4. Added an alternate matching function, pcre_dfa_exec(), which matches using
1857        a different (DFA) algorithm. Although it is slower than the original
1858        function, it does have some advantages for certain types of matching
1859        problem.
1860    
1861     5. Upgrades to pcretest in order to test the features of pcre_dfa_exec(),
1862        including restarting after a partial match.
1863    
1864     6. A patch for pcregrep that defines INVALID_FILE_ATTRIBUTES if it is not
1865        defined when compiling for Windows was sent to me. I have put it into the
1866        code, though I have no means of testing or verifying it.
1867    
1868     7. Added the pcre_refcount() auxiliary function.
1869    
1870     8. Added the PCRE_FIRSTLINE option. This constrains an unanchored pattern to
1871        match before or at the first newline in the subject string. In pcretest,
1872        the /f option on a pattern can be used to set this.
1873    
1874     9. A repeated \w when used in UTF-8 mode with characters greater than 256
1875        would behave wrongly. This has been present in PCRE since release 4.0.
1876    
1877    10. A number of changes to the pcregrep command:
1878    
1879        (a) Refactored how -x works; insert ^(...)$ instead of setting
1880            PCRE_ANCHORED and checking the length, in preparation for adding
1881            something similar for -w.
1882    
1883        (b) Added the -w (match as a word) option.
1884    
1885        (c) Refactored the way lines are read and buffered so as to have more
1886            than one at a time available.
1887    
1888        (d) Implemented a pcregrep test script.
1889    
1890        (e) Added the -M (multiline match) option. This allows patterns to match
1891            over several lines of the subject. The buffering ensures that at least
1892            8K, or the rest of the document (whichever is the shorter) is available
1893            for matching (and similarly the previous 8K for lookbehind assertions).
1894    
1895        (f) Changed the --help output so that it now says
1896    
1897              -w, --word-regex(p)
1898    
1899            instead of two lines, one with "regex" and the other with "regexp"
1900            because that confused at least one person since the short forms are the
1901            same. (This required a bit of code, as the output is generated
1902            automatically from a table. It wasn't just a text change.)
1903    
1904        (g) -- can be used to terminate pcregrep options if the next thing isn't an
1905            option but starts with a hyphen. Could be a pattern or a path name
1906            starting with a hyphen, for instance.
1907    
1908        (h) "-" can be given as a file name to represent stdin.
1909    
1910        (i) When file names are being printed, "(standard input)" is used for
1911            the standard input, for compatibility with GNU grep. Previously
1912            "<stdin>" was used.
1913    
1914        (j) The option --label=xxx can be used to supply a name to be used for
1915            stdin when file names are being printed. There is no short form.
1916    
1917        (k) Re-factored the options decoding logic because we are going to add
1918            two more options that take data. Such options can now be given in four
1919            different ways, e.g. "-fname", "-f name", "--file=name", "--file name".
1920    
1921        (l) Added the -A, -B, and -C options for requesting that lines of context
1922            around matches be printed.
1923    
1924        (m) Added the -L option to print the names of files that do not contain
1925            any matching lines, that is, the complement of -l.
1926    
1927        (n) The return code is 2 if any file cannot be opened, but pcregrep does
1928            continue to scan other files.
1929    
1930        (o) The -s option was incorrectly implemented. For compatibility with other
1931            greps, it now suppresses the error message for a non-existent or non-
1932            accessible file (but not the return code). There is a new option called
1933            -q that suppresses the output of matching lines, which was what -s was
1934            previously doing.
1935    
1936        (p) Added --include and --exclude options to specify files for inclusion
1937            and exclusion when recursing.
1938    
1939    11. The Makefile was not using the Autoconf-supported LDFLAGS macro properly.
1940        Hopefully, it now does.
1941    
1942    12. Missing cast in pcre_study().
1943    
1944    13. Added an "uninstall" target to the makefile.
1945    
1946    14. Replaced "extern" in the function prototypes in Makefile.in with
1947        "PCRE_DATA_SCOPE", which defaults to 'extern' or 'extern "C"' in the Unix
1948        world, but is set differently for Windows.
1949    
1950    15. Added a second compiling function called pcre_compile2(). The only
1951        difference is that it has an extra argument, which is a pointer to an
1952        integer error code. When there is a compile-time failure, this is set
1953        non-zero, in addition to the error test pointer being set to point to an
1954        error message. The new argument may be NULL if no error number is required
1955        (but then you may as well call pcre_compile(), which is now just a
1956        wrapper). This facility is provided because some applications need a
1957        numeric error indication, but it has also enabled me to tidy up the way
1958        compile-time errors are handled in the POSIX wrapper.
1959    
1960    16. Added VPATH=.libs to the makefile; this should help when building with one
1961        prefix path and installing with another. (Or so I'm told by someone who
1962        knows more about this stuff than I do.)
1963    
1964    17. Added a new option, REG_DOTALL, to the POSIX function regcomp(). This
1965        passes PCRE_DOTALL to the pcre_compile() function, making the "." character
1966        match everything, including newlines. This is not POSIX-compatible, but
1967        somebody wanted the feature. From pcretest it can be activated by using
1968        both the P and the s flags.
1969    
1970    18. AC_PROG_LIBTOOL appeared twice in Makefile.in. Removed one.
1971    
1972    19. libpcre.pc was being incorrectly installed as executable.
1973    
1974    20. A couple of places in pcretest check for end-of-line by looking for '\n';
1975        it now also looks for '\r' so that it will work unmodified on Windows.
1976    
1977    21. Added Google's contributed C++ wrapper to the distribution.
1978    
1979    22. Added some untidy missing memory free() calls in pcretest, to keep
1980        Electric Fence happy when testing.
1981    
1982    
1983    
1984    Version 5.0 13-Sep-04
1985    ---------------------
1986    
1987     1. Internal change: literal characters are no longer packed up into items
1988        containing multiple characters in a single byte-string. Each character
1989        is now matched using a separate opcode. However, there may be more than one
1990        byte in the character in UTF-8 mode.
1991    
1992     2. The pcre_callout_block structure has two new fields: pattern_position and
1993        next_item_length. These contain the offset in the pattern to the next match
1994        item, and its length, respectively.
1995    
1996     3. The PCRE_AUTO_CALLOUT option for pcre_compile() requests the automatic
1997        insertion of callouts before each pattern item. Added the /C option to
1998        pcretest to make use of this.
1999    
2000     4. On the advice of a Windows user, the lines
2001    
2002          #if defined(_WIN32) || defined(WIN32)
2003          _setmode( _fileno( stdout ), 0x8000 );
2004          #endif  /* defined(_WIN32) || defined(WIN32) */
2005    
2006        have been added to the source of pcretest. This apparently does useful
2007        magic in relation to line terminators.
2008    
2009     5. Changed "r" and "w" in the calls to fopen() in pcretest to "rb" and "wb"
2010        for the benefit of those environments where the "b" makes a difference.
2011    
2012     6. The icc compiler has the same options as gcc, but "configure" doesn't seem
2013        to know about it. I have put a hack into configure.in that adds in code
2014        to set GCC=yes if CC=icc. This seems to end up at a point in the
2015        generated configure script that is early enough to affect the setting of
2016        compiler options, which is what is needed, but I have no means of testing
2017        whether it really works. (The user who reported this had patched the
2018        generated configure script, which of course I cannot do.)
2019    
2020        LATER: After change 22 below (new libtool files), the configure script
2021        seems to know about icc (and also ecc). Therefore, I have commented out
2022        this hack in configure.in.
2023    
2024     7. Added support for pkg-config (2 patches were sent in).
2025    
2026     8. Negated POSIX character classes that used a combination of internal tables
2027        were completely broken. These were [[:^alpha:]], [[:^alnum:]], and
2028        [[:^ascii]]. Typically, they would match almost any characters. The other
2029        POSIX classes were not broken in this way.
2030    
2031     9. Matching the pattern "\b.*?" against "ab cd", starting at offset 1, failed
2032        to find the match, as PCRE was deluded into thinking that the match had to
2033        start at the start point or following a newline. The same bug applied to
2034        patterns with negative forward assertions or any backward assertions
2035        preceding ".*" at the start, unless the pattern required a fixed first
2036        character. This was a failing pattern: "(?!.bcd).*". The bug is now fixed.
2037    
2038    10. In UTF-8 mode, when moving forwards in the subject after a failed match
2039        starting at the last subject character, bytes beyond the end of the subject
2040        string were read.
2041    
2042    11. Renamed the variable "class" as "classbits" to make life easier for C++
2043        users. (Previously there was a macro definition, but it apparently wasn't
2044        enough.)
2045    
2046    12. Added the new field "tables" to the extra data so that tables can be passed
2047        in at exec time, or the internal tables can be re-selected. This allows
2048        a compiled regex to be saved and re-used at a later time by a different
2049        program that might have everything at different addresses.
2050    
2051    13. Modified the pcre-config script so that, when run on Solaris, it shows a
2052        -R library as well as a -L library.
2053    
2054    14. The debugging options of pcretest (-d on the command line or D on a
2055        pattern) showed incorrect output for anything following an extended class
2056        that contained multibyte characters and which was followed by a quantifier.
2057    
2058    15. Added optional support for general category Unicode character properties
2059        via the \p, \P, and \X escapes. Unicode property support implies UTF-8
2060        support. It adds about 90K to the size of the library. The meanings of the
2061        inbuilt class escapes such as \d and \s have NOT been changed.
2062    
2063    16. Updated pcredemo.c to include calls to free() to release the memory for the
2064        compiled pattern.
2065    
2066    17. The generated file chartables.c was being created in the source directory
2067        instead of in the building directory. This caused the build to fail if the
2068        source directory was different from the building directory, and was
2069        read-only.
2070    
2071    18. Added some sample Win commands from Mark Tetrode into the NON-UNIX-USE
2072        file. No doubt somebody will tell me if they don't make sense... Also added
2073        Dan Mooney's comments about building on OpenVMS.
2074    
2075    19. Added support for partial matching via the PCRE_PARTIAL option for
2076        pcre_exec() and the \P data escape in pcretest.
2077    
2078    20. Extended pcretest with 3 new pattern features:
2079    
2080        (i)   A pattern option of the form ">rest-of-line" causes pcretest to
2081              write the compiled pattern to the file whose name is "rest-of-line".
2082              This is a straight binary dump of the data, with the saved pointer to
2083              the character tables forced to be NULL. The study data, if any, is
2084              written too. After writing, pcretest reads a new pattern.
2085    
2086        (ii)  If, instead of a pattern, "<rest-of-line" is given, pcretest reads a
2087              compiled pattern from the given file. There must not be any
2088              occurrences of "<" in the file name (pretty unlikely); if there are,
2089              pcretest will instead treat the initial "<" as a pattern delimiter.
2090              After reading in the pattern, pcretest goes on to read data lines as
2091              usual.
2092    
2093        (iii) The F pattern option causes pcretest to flip the bytes in the 32-bit
2094              and 16-bit fields in a compiled pattern, to simulate a pattern that
2095              was compiled on a host of opposite endianness.
2096    
2097    21. The pcre-exec() function can now cope with patterns that were compiled on
2098        hosts of opposite endianness, with this restriction:
2099    
2100          As for any compiled expression that is saved and used later, the tables
2101          pointer field cannot be preserved; the extra_data field in the arguments
2102          to pcre_exec() should be used to pass in a tables address if a value
2103          other than the default internal tables were used at compile time.
2104    
2105    22. Calling pcre_exec() with a negative value of the "ovecsize" parameter is
2106        now diagnosed as an error. Previously, most of the time, a negative number
2107        would have been treated as zero, but if in addition "ovector" was passed as
2108        NULL, a crash could occur.
2109    
2110    23. Updated the files ltmain.sh, config.sub, config.guess, and aclocal.m4 with
2111        new versions from the libtool 1.5 distribution (the last one is a copy of
2112        a file called libtool.m4). This seems to have fixed the need to patch
2113        "configure" to support Darwin 1.3 (which I used to do). However, I still
2114        had to patch ltmain.sh to ensure that ${SED} is set (it isn't on my
2115        workstation).
2116    
2117    24. Changed the PCRE licence to be the more standard "BSD" licence.
2118    
2119    
2120    Version 4.5 01-Dec-03
2121    ---------------------
2122    
2123     1. There has been some re-arrangement of the code for the match() function so
2124        that it can be compiled in a version that does not call itself recursively.
2125        Instead, it keeps those local variables that need separate instances for
2126        each "recursion" in a frame on the heap, and gets/frees frames whenever it
2127        needs to "recurse". Keeping track of where control must go is done by means
2128        of setjmp/longjmp. The whole thing is implemented by a set of macros that
2129        hide most of the details from the main code, and operates only if
2130        NO_RECURSE is defined while compiling pcre.c. If PCRE is built using the
2131        "configure" mechanism, "--disable-stack-for-recursion" turns on this way of
2132        operating.
2133    
2134        To make it easier for callers to provide specially tailored get/free
2135        functions for this usage, two new functions, pcre_stack_malloc, and
2136        pcre_stack_free, are used. They are always called in strict stacking order,
2137        and the size of block requested is always the same.
2138    
2139        The PCRE_CONFIG_STACKRECURSE info parameter can be used to find out whether
2140        PCRE has been compiled to use the stack or the heap for recursion. The
2141        -C option of pcretest uses this to show which version is compiled.
2142    
2143        A new data escape \S, is added to pcretest; it causes the amounts of store
2144        obtained and freed by both kinds of malloc/free at match time to be added
2145        to the output.
2146    
2147     2. Changed the locale test to use "fr_FR" instead of "fr" because that's
2148        what's available on my current Linux desktop machine.
2149    
2150     3. When matching a UTF-8 string, the test for a valid string at the start has
2151        been extended. If start_offset is not zero, PCRE now checks that it points
2152        to a byte that is the start of a UTF-8 character. If not, it returns
2153        PCRE_ERROR_BADUTF8_OFFSET (-11). Note: the whole string is still checked;
2154        this is necessary because there may be backward assertions in the pattern.
2155        When matching the same subject several times, it may save resources to use
2156        PCRE_NO_UTF8_CHECK on all but the first call if the string is long.
2157    
2158     4. The code for checking the validity of UTF-8 strings has been tightened so
2159        that it rejects (a) strings containing 0xfe or 0xff bytes and (b) strings
2160        containing "overlong sequences".
2161    
2162     5. Fixed a bug (appearing twice) that I could not find any way of exploiting!
2163        I had written "if ((digitab[*p++] && chtab_digit) == 0)" where the "&&"
2164        should have been "&", but it just so happened that all the cases this let
2165        through by mistake were picked up later in the function.
2166    
2167     6. I had used a variable called "isblank" - this is a C99 function, causing
2168        some compilers to warn. To avoid this, I renamed it (as "blankclass").
2169    
2170     7. Cosmetic: (a) only output another newline at the end of pcretest if it is
2171        prompting; (b) run "./pcretest /dev/null" at the start of the test script
2172        so the version is shown; (c) stop "make test" echoing "./RunTest".
2173    
2174     8. Added patches from David Burgess to enable PCRE to run on EBCDIC systems.
2175    
2176     9. The prototype for memmove() for systems that don't have it was using
2177        size_t, but the inclusion of the header that defines size_t was later. I've
2178        moved the #includes for the C headers earlier to avoid this.
2179    
2180    10. Added some adjustments to the code to make it easier to compiler on certain
2181        special systems:
2182    
2183          (a) Some "const" qualifiers were missing.
2184          (b) Added the macro EXPORT before all exported functions; by default this
2185              is defined to be empty.
2186          (c) Changed the dftables auxiliary program (that builds chartables.c) so
2187              that it reads its output file name as an argument instead of writing
2188              to the standard output and assuming this can be redirected.
2189    
2190    11. In UTF-8 mode, if a recursive reference (e.g. (?1)) followed a character
2191        class containing characters with values greater than 255, PCRE compilation
2192        went into a loop.
2193    
2194    12. A recursive reference to a subpattern that was within another subpattern
2195        that had a minimum quantifier of zero caused PCRE to crash. For example,
2196        (x(y(?2))z)? provoked this bug with a subject that got as far as the
2197        recursion. If the recursively-called subpattern itself had a zero repeat,
2198        that was OK.
2199    
2200    13. In pcretest, the buffer for reading a data line was set at 30K, but the
2201        buffer into which it was copied (for escape processing) was still set at
2202        1024, so long lines caused crashes.
2203    
2204    14. A pattern such as /[ab]{1,3}+/ failed to compile, giving the error
2205        "internal error: code overflow...". This applied to any character class
2206        that was followed by a possessive quantifier.
2207    
2208    15. Modified the Makefile to add libpcre.la as a prerequisite for
2209        libpcreposix.la because I was told this is needed for a parallel build to
2210        work.
2211    
2212    16. If a pattern that contained .* following optional items at the start was
2213        studied, the wrong optimizing data was generated, leading to matching
2214        errors. For example, studying /[ab]*.*c/ concluded, erroneously, that any
2215        matching string must start with a or b or c. The correct conclusion for
2216        this pattern is that a match can start with any character.
2217    
2218    
2219    Version 4.4 13-Aug-03
2220    ---------------------
2221    
2222     1. In UTF-8 mode, a character class containing characters with values between
2223        127 and 255 was not handled correctly if the compiled pattern was studied.
2224        In fixing this, I have also improved the studying algorithm for such
2225        classes (slightly).
2226    
2227     2. Three internal functions had redundant arguments passed to them. Removal
2228        might give a very teeny performance improvement.
2229    
2230     3. Documentation bug: the value of the capture_top field in a callout is *one
2231        more than* the number of the hightest numbered captured substring.
2232    
2233     4. The Makefile linked pcretest and pcregrep with -lpcre, which could result
2234        in incorrectly linking with a previously installed version. They now link
2235        explicitly with libpcre.la.
2236    
2237     5. configure.in no longer needs to recognize Cygwin specially.
2238    
2239     6. A problem in pcre.in for Windows platforms is fixed.
2240    
2241     7. If a pattern was successfully studied, and the -d (or /D) flag was given to
2242        pcretest, it used to include the size of the study block as part of its
2243        output. Unfortunately, the structure contains a field that has a different
2244        size on different hardware architectures. This meant that the tests that
2245        showed this size failed. As the block is currently always of a fixed size,
2246        this information isn't actually particularly useful in pcretest output, so
2247        I have just removed it.
2248    
2249     8. Three pre-processor statements accidentally did not start in column 1.
2250        Sadly, there are *still* compilers around that complain, even though
2251        standard C has not required this for well over a decade. Sigh.
2252    
2253     9. In pcretest, the code for checking callouts passed small integers in the
2254        callout_data field, which is a void * field. However, some picky compilers
2255        complained about the casts involved for this on 64-bit systems. Now
2256        pcretest passes the address of the small integer instead, which should get
2257        rid of the warnings.
2258    
2259    10. By default, when in UTF-8 mode, PCRE now checks for valid UTF-8 strings at
2260        both compile and run time, and gives an error if an invalid UTF-8 sequence
2261        is found. There is a option for disabling this check in cases where the
2262        string is known to be correct and/or the maximum performance is wanted.
2263    
2264    11. In response to a bug report, I changed one line in Makefile.in from
2265    
2266            -Wl,--out-implib,.libs/lib@WIN_PREFIX@pcreposix.dll.a \
2267        to
2268            -Wl,--out-implib,.libs/@WIN_PREFIX@libpcreposix.dll.a \
2269    
2270        to look similar to other lines, but I have no way of telling whether this
2271        is the right thing to do, as I do not use Windows. No doubt I'll get told
2272        if it's wrong...
2273    
2274    
2275    Version 4.3 21-May-03
2276    ---------------------
2277    
2278    1. Two instances of @WIN_PREFIX@ omitted from the Windows targets in the
2279       Makefile.
2280    
2281    2. Some refactoring to improve the quality of the code:
2282    
2283       (i)   The utf8_table... variables are now declared "const".
2284    
2285       (ii)  The code for \cx, which used the "case flipping" table to upper case
2286             lower case letters, now just substracts 32. This is ASCII-specific,
2287             but the whole concept of \cx is ASCII-specific, so it seems
2288             reasonable.
2289    
2290       (iii) PCRE was using its character types table to recognize decimal and
2291             hexadecimal digits in the pattern. This is silly, because it handles
2292             only 0-9, a-f, and A-F, but the character types table is locale-
2293             specific, which means strange things might happen. A private
2294             table is now used for this - though it costs 256 bytes, a table is
2295             much faster than multiple explicit tests. Of course, the standard
2296             character types table is still used for matching digits in subject
2297             strings against \d.
2298    
2299       (iv)  Strictly, the identifier ESC_t is reserved by POSIX (all identifiers
2300             ending in _t are). So I've renamed it as ESC_tee.
2301    
2302    3. The first argument for regexec() in the POSIX wrapper should have been
2303       defined as "const".
2304    
2305    4. Changed pcretest to use malloc() for its buffers so that they can be
2306       Electric Fenced for debugging.
2307    
2308    5. There were several places in the code where, in UTF-8 mode, PCRE would try
2309       to read one or more bytes before the start of the subject string. Often this
2310       had no effect on PCRE's behaviour, but in some circumstances it could
2311       provoke a segmentation fault.
2312    
2313    6. A lookbehind at the start of a pattern in UTF-8 mode could also cause PCRE
2314       to try to read one or more bytes before the start of the subject string.
2315    
2316    7. A lookbehind in a pattern matched in non-UTF-8 mode on a PCRE compiled with
2317       UTF-8 support could misbehave in various ways if the subject string
2318       contained bytes with the 0x80 bit set and the 0x40 bit unset in a lookbehind
2319       area. (PCRE was not checking for the UTF-8 mode flag, and trying to move
2320       back over UTF-8 characters.)
2321    
2322    
2323    Version 4.2 14-Apr-03
2324    ---------------------
2325    
2326    1. Typo "#if SUPPORT_UTF8" instead of "#ifdef SUPPORT_UTF8" fixed.
2327    
2328    2. Changes to the building process, supplied by Ronald Landheer-Cieslak
2329         [ON_WINDOWS]: new variable, "#" on non-Windows platforms
2330         [NOT_ON_WINDOWS]: new variable, "#" on Windows platforms
2331         [WIN_PREFIX]: new variable, "cyg" for Cygwin
2332         * Makefile.in: use autoconf substitution for OBJEXT, EXEEXT, BUILD_OBJEXT
2333           and BUILD_EXEEXT
2334         Note: automatic setting of the BUILD variables is not yet working
2335         set CPPFLAGS and BUILD_CPPFLAGS (but don't use yet) - should be used at
2336           compile-time but not at link-time
2337         [LINK]: use for linking executables only
2338         make different versions for Windows and non-Windows
2339         [LINKLIB]: new variable, copy of UNIX-style LINK, used for linking
2340           libraries
2341         [LINK_FOR_BUILD]: new variable
2342         [OBJEXT]: use throughout
2343         [EXEEXT]: use throughout
2344         <winshared>: new target
2345         <wininstall>: new target
2346         <dftables.o>: use native compiler
2347         <dftables>: use native linker
2348         <install>: handle Windows platform correctly
2349         <clean>: ditto
2350         <check>: ditto
2351         copy DLL to top builddir before testing
2352    
2353       As part of these changes, -no-undefined was removed again. This was reported
2354       to give trouble on HP-UX 11.0, so getting rid of it seems like a good idea
2355       in any case.
2356    
2357    3. Some tidies to get rid of compiler warnings:
2358    
2359       . In the match_data structure, match_limit was an unsigned long int, whereas
2360         match_call_count was an int. I've made them both unsigned long ints.
2361    
2362       . In pcretest the fact that a const uschar * doesn't automatically cast to
2363         a void * provoked a warning.
2364    
2365       . Turning on some more compiler warnings threw up some "shadow" variables
2366         and a few more missing casts.
2367    
2368    4. If PCRE was complied with UTF-8 support, but called without the PCRE_UTF8
2369       option, a class that contained a single character with a value between 128
2370       and 255 (e.g. /[\xFF]/) caused PCRE to crash.
2371    
2372    5. If PCRE was compiled with UTF-8 support, but called without the PCRE_UTF8
2373       option, a class that contained several characters, but with at least one
2374       whose value was between 128 and 255 caused PCRE to crash.
2375    
2376    
2377    Version 4.1 12-Mar-03
2378    ---------------------
2379    
2380    1. Compiling with gcc -pedantic found a couple of places where casts were
2381    needed, and a string in dftables.c that was longer than standard compilers are
2382    required to support.
2383    
2384    2. Compiling with Sun's compiler found a few more places where the code could
2385    be tidied up in order to avoid warnings.
2386    
2387    3. The variables for cross-compiling were called HOST_CC and HOST_CFLAGS; the
2388    first of these names is deprecated in the latest Autoconf in favour of the name
2389    CC_FOR_BUILD, because "host" is typically used to mean the system on which the
2390    compiled code will be run. I can't find a reference for HOST_CFLAGS, but by
2391    analogy I have changed it to CFLAGS_FOR_BUILD.
2392    
2393    4. Added -no-undefined to the linking command in the Makefile, because this is
2394    apparently helpful for Windows. To make it work, also added "-L. -lpcre" to the
2395    linking step for the pcreposix library.
2396    
2397    5. PCRE was failing to diagnose the case of two named groups with the same
2398    name.
2399    
2400    6. A problem with one of PCRE's optimizations was discovered. PCRE remembers a
2401    literal character that is needed in the subject for a match, and scans along to
2402    ensure that it is present before embarking on the full matching process. This
2403    saves time in cases of nested unlimited repeats that are never going to match.
2404    Problem: the scan can take a lot of time if the subject is very long (e.g.
2405    megabytes), thus penalizing straightforward matches. It is now done only if the
2406    amount of subject to be scanned is less than 1000 bytes.
2407    
2408    7. A lesser problem with the same optimization is that it was recording the
2409    first character of an anchored pattern as "needed", thus provoking a search
2410    right along the subject, even when the first match of the pattern was going to
2411    fail. The "needed" character is now not set for anchored patterns, unless it
2412    follows something in the pattern that is of non-fixed length. Thus, it still
2413    fulfils its original purpose of finding quick non-matches in cases of nested
2414    unlimited repeats, but isn't used for simple anchored patterns such as /^abc/.
2415    
2416    
2417    Version 4.0 17-Feb-03
2418    ---------------------
2419    
2420    1. If a comment in an extended regex that started immediately after a meta-item
2421    extended to the end of string, PCRE compiled incorrect data. This could lead to
2422    all kinds of weird effects. Example: /#/ was bad; /()#/ was bad; /a#/ was not.
2423    
2424    2. Moved to autoconf 2.53 and libtool 1.4.2.
2425    
2426    3. Perl 5.8 no longer needs "use utf8" for doing UTF-8 things. Consequently,
2427    the special perltest8 script is no longer needed - all the tests can be run
2428    from a single perltest script.
2429    
2430    4. From 5.004, Perl has not included the VT character (0x0b) in the set defined
2431    by \s. It has now been removed in PCRE. This means it isn't recognized as
2432    whitespace in /x regexes too, which is the same as Perl. Note that the POSIX
2433    class [:space:] *does* include VT, thereby creating a mess.
2434    
2435    5. Added the class [:blank:] (a GNU extension from Perl 5.8) to match only
2436    space and tab.
2437    
2438    6. Perl 5.005 was a long time ago. It's time to amalgamate the tests that use
2439    its new features into the main test script, reducing the number of scripts.
2440    
2441    7. Perl 5.8 has changed the meaning of patterns like /a(?i)b/. Earlier versions
2442    were backward compatible, and made the (?i) apply to the whole pattern, as if
2443    /i were given. Now it behaves more logically, and applies the option setting
2444    only to what follows. PCRE has been changed to follow suit. However, if it
2445    finds options settings right at the start of the pattern, it extracts them into
2446    the global options, as before. Thus, they show up in the info data.
2447    
2448    8. Added support for the \Q...\E escape sequence. Characters in between are
2449    treated as literals. This is slightly different from Perl in that $ and @ are
2450    also handled as literals inside the quotes. In Perl, they will cause variable
2451    interpolation. Note the following examples:
2452    
2453        Pattern            PCRE matches      Perl matches
2454    
2455        \Qabc$xyz\E        abc$xyz           abc followed by the contents of $xyz
2456        \Qabc\$xyz\E       abc\$xyz          abc\$xyz
2457        \Qabc\E\$\Qxyz\E   abc$xyz           abc$xyz
2458    
2459    For compatibility with Perl, \Q...\E sequences are recognized inside character
2460    classes as well as outside them.
2461    
2462    9. Re-organized 3 code statements in pcretest to avoid "overflow in
2463    floating-point constant arithmetic" warnings from a Microsoft compiler. Added a
2464    (size_t) cast to one statement in pcretest and one in pcreposix to avoid
2465    signed/unsigned warnings.
2466    
2467    10. SunOS4 doesn't have strtoul(). This was used only for unpicking the -o
2468    option for pcretest, so I've replaced it by a simple function that does just
2469    that job.
2470    
2471    11. pcregrep was ending with code 0 instead of 2 for the commands "pcregrep" or
2472    "pcregrep -".
2473    
2474    12. Added "possessive quantifiers" ?+, *+, ++, and {,}+ which come from Sun's
2475    Java package. This provides some syntactic sugar for simple cases of what my
2476    documentation calls "once-only subpatterns". A pattern such as x*+ is the same
2477    as (?>x*). In other words, if what is inside (?>...) is just a single repeated
2478    item, you can use this simplified notation. Note that only makes sense with
2479    greedy quantifiers. Consequently, the use of the possessive quantifier forces
2480    greediness, whatever the setting of the PCRE_UNGREEDY option.
2481    
2482    13. A change of greediness default within a pattern was not taking effect at
2483    the current level for patterns like /(b+(?U)a+)/. It did apply to parenthesized
2484    subpatterns that followed. Patterns like /b+(?U)a+/ worked because the option
2485    was abstracted outside.
2486    
2487    14. PCRE now supports the \G assertion. It is true when the current matching
2488    position is at the start point of the match. This differs from \A when the
2489    starting offset is non-zero. Used with the /g option of pcretest (or similar
2490    code), it works in the same way as it does for Perl's /g option. If all
2491    alternatives of a regex begin with \G, the expression is anchored to the start
2492    match position, and the "anchored" flag is set in the compiled expression.
2493    
2494    15. Some bugs concerning the handling of certain option changes within patterns
2495    have been fixed. These applied to options other than (?ims). For example,
2496    "a(?x: b c )d" did not match "XabcdY" but did match "Xa b c dY". It should have
2497    been the other way round. Some of this was related to change 7 above.
2498    
2499    16. PCRE now gives errors for /[.x.]/ and /[=x=]/ as unsupported POSIX
2500    features, as Perl does. Previously, PCRE gave the warnings only for /[[.x.]]/
2501    and /[[=x=]]/. PCRE now also gives an error for /[:name:]/ because it supports
2502    POSIX classes only within a class (e.g. /[[:alpha:]]/).
2503    
2504    17. Added support for Perl's \C escape. This matches one byte, even in UTF8
2505    mode. Unlike ".", it always matches newline, whatever the setting of
2506    PCRE_DOTALL. However, PCRE does not permit \C to appear in lookbehind
2507    assertions. Perl allows it, but it doesn't (in general) work because it can't
2508    calculate the length of the lookbehind. At least, that's the case for Perl
2509    5.8.0 - I've been told they are going to document that it doesn't work in
2510    future.
2511    
2512    18. Added an error diagnosis for escapes that PCRE does not support: these are
2513    \L, \l, \N, \P, \p, \U, \u, and \X.
2514    
2515    19. Although correctly diagnosing a missing ']' in a character class, PCRE was
2516    reading past the end of the pattern in cases such as /[abcd/.
2517    
2518    20. PCRE was getting more memory than necessary for patterns with classes that
2519    contained both POSIX named classes and other characters, e.g. /[[:space:]abc/.
2520    
2521    21. Added some code, conditional on #ifdef VPCOMPAT, to make life easier for
2522    compiling PCRE for use with Virtual Pascal.
2523    
2524    22. Small fix to the Makefile to make it work properly if the build is done
2525    outside the source tree.
2526    
2527    23. Added a new extension: a condition to go with recursion. If a conditional
2528    subpattern starts with (?(R) the "true" branch is used if recursion has
2529    happened, whereas the "false" branch is used only at the top level.
2530    
2531    24. When there was a very long string of literal characters (over 255 bytes
2532    without UTF support, over 250 bytes with UTF support), the computation of how
2533    much memory was required could be incorrect, leading to segfaults or other
2534    strange effects.
2535    
2536    25. PCRE was incorrectly assuming anchoring (either to start of subject or to
2537    start of line for a non-DOTALL pattern) when a pattern started with (.*) and
2538    there was a subsequent back reference to those brackets. This meant that, for
2539    example, /(.*)\d+\1/ failed to match "abc123bc". Unfortunately, it isn't
2540    possible to check for precisely this case. All we can do is abandon the
2541    optimization if .* occurs inside capturing brackets when there are any back
2542    references whatsoever. (See below for a better fix that came later.)
2543    
2544    26. The handling of the optimization for finding the first character of a
2545    non-anchored pattern, and for finding a character that is required later in the
2546    match were failing in some cases. This didn't break the matching; it just
2547    failed to optimize when it could. The way this is done has been re-implemented.
2548    
2549    27. Fixed typo in error message for invalid (?R item (it said "(?p").
2550    
2551    28. Added a new feature that provides some of the functionality that Perl
2552    provides with (?{...}). The facility is termed a "callout". The way it is done
2553    in PCRE is for the caller to provide an optional function, by setting
2554    pcre_callout to its entry point. Like pcre_malloc and pcre_free, this is a
2555    global variable. By default it is unset, which disables all calling out. To get
2556    the function called, the regex must include (?C) at appropriate points. This
2557    is, in fact, equivalent to (?C0), and any number <= 255 may be given with (?C).
2558    This provides a means of identifying different callout points. When PCRE
2559    reaches such a point in the regex, if pcre_callout has been set, the external
2560    function is called. It is provided with data in a structure called
2561    pcre_callout_block, which is defined in pcre.h. If the function returns 0,
2562    matching continues; if it returns a non-zero value, the match at the current
2563    point fails. However, backtracking will occur if possible. [This was changed
2564    later and other features added - see item 49 below.]
2565    
2566    29. pcretest is upgraded to test the callout functionality. It provides a
2567    callout function that displays information. By default, it shows the start of
2568    the match and the current position in the text. There are some new data escapes
2569    to vary what happens:
2570    
2571        \C+         in addition, show current contents of captured substrings
2572        \C-         do not supply a callout function
2573        \C!n        return 1 when callout number n is reached
2574        \C!n!m      return 1 when callout number n is reached for the mth time
2575    
2576    30. If pcregrep was called with the -l option and just a single file name, it
2577    output "<stdin>" if a match was found, instead of the file name.
2578    
2579    31. Improve the efficiency of the POSIX API to PCRE. If the number of capturing
2580    slots is less than POSIX_MALLOC_THRESHOLD, use a block on the stack to pass to
2581    pcre_exec(). This saves a malloc/free per call. The default value of
2582    POSIX_MALLOC_THRESHOLD is 10; it can be changed by --with-posix-malloc-threshold
2583    when configuring.
2584    
2585    32. The default maximum size of a compiled pattern is 64K. There have been a
2586    few cases of people hitting this limit. The code now uses macros to handle the
2587    storing of links as offsets within the compiled pattern. It defaults to 2-byte
2588    links, but this can be changed to 3 or 4 bytes by --with-link-size when
2589    configuring. Tests 2 and 5 work only with 2-byte links because they output
2590    debugging information about compiled patterns.
2591    
2592    33. Internal code re-arrangements:
2593    
2594    (a) Moved the debugging function for printing out a compiled regex into
2595        its own source file (printint.c) and used #include to pull it into
2596        pcretest.c and, when DEBUG is defined, into pcre.c, instead of having two
2597        separate copies.
2598    
2599    (b) Defined the list of op-code names for debugging as a macro in
2600        internal.h so that it is next to the definition of the opcodes.
2601    
2602    (c) Defined a table of op-code lengths for simpler skipping along compiled
2603        code. This is again a macro in internal.h so that it is next to the
2604        definition of the opcodes.
2605    
2606    34. Added support for recursive calls to individual subpatterns, along the
2607    lines of Robin Houston's patch (but implemented somewhat differently).
2608    
2609    35. Further mods to the Makefile to help Win32. Also, added code to pcregrep to
2610    allow it to read and process whole directories in Win32. This code was
2611    contributed by Lionel Fourquaux; it has not been tested by me.
2612    
2613    36. Added support for named subpatterns. The Python syntax (?P<name>...) is
2614    used to name a group. Names consist of alphanumerics and underscores, and must
2615    be unique. Back references use the syntax (?P=name) and recursive calls use
2616    (?P>name) which is a PCRE extension to the Python extension. Groups still have
2617    numbers. The function pcre_fullinfo() can be used after compilation to extract
2618    a name/number map. There are three relevant calls:
2619    
2620      PCRE_INFO_NAMEENTRYSIZE        yields the size of each entry in the map
2621      PCRE_INFO_NAMECOUNT            yields the number of entries
2622      PCRE_INFO_NAMETABLE            yields a pointer to the map.
2623    
2624    The map is a vector of fixed-size entries. The size of each entry depends on
2625    the length of the longest name used. The first two bytes of each entry are the
2626    group number, most significant byte first. There follows the corresponding
2627    name, zero terminated. The names are in alphabetical order.
2628    
2629    37. Make the maximum literal string in the compiled code 250 for the non-UTF-8
2630    case instead of 255. Making it the same both with and without UTF-8 support
2631    means that the same test output works with both.
2632    
2633    38. There was a case of malloc(0) in the POSIX testing code in pcretest. Avoid
2634    calling malloc() with a zero argument.
2635    
2636    39. Change 25 above had to resort to a heavy-handed test for the .* anchoring
2637    optimization. I've improved things by keeping a bitmap of backreferences with
2638    numbers 1-31 so that if .* occurs inside capturing brackets that are not in
2639    fact referenced, the optimization can be applied. It is unlikely that a
2640    relevant occurrence of .* (i.e. one which might indicate anchoring or forcing
2641    the match to follow \n) will appear inside brackets with a number greater than
2642    31, but if it does, any back reference > 31 suppresses the optimization.
2643    
2644    40. Added a new compile-time option PCRE_NO_AUTO_CAPTURE. This has the effect
2645    of disabling numbered capturing parentheses. Any opening parenthesis that is
2646    not followed by ? behaves as if it were followed by ?: but named parentheses
2647    can still be used for capturing (and they will acquire numbers in the usual
2648    way).
2649    
2650    41. Redesigned the return codes from the match() function into yes/no/error so
2651    that errors can be passed back from deep inside the nested calls. A malloc
2652    failure while inside a recursive subpattern call now causes the
2653    PCRE_ERROR_NOMEMORY return instead of quietly going wrong.
2654    
2655    42. It is now possible to set a limit on the number of times the match()
2656    function is called in a call to pcre_exec(). This facility makes it possible to
2657    limit the amount of recursion and backtracking, though not in a directly
2658    obvious way, because the match() function is used in a number of different
2659    circumstances. The count starts from zero for each position in the subject
2660    string (for non-anchored patterns). The default limit is, for compatibility, a
2661    large number, namely 10 000 000. You can change this in two ways:
2662    
2663    (a) When configuring PCRE before making, you can use --with-match-limit=n
2664        to set a default value for the compiled library.
2665    
2666    (b) For each call to pcre_exec(), you can pass a pcre_extra block in which
2667        a different value is set. See 45 below.
2668    
2669    If the limit is exceeded, pcre_exec() returns PCRE_ERROR_MATCHLIMIT.
2670    
2671    43. Added a new function pcre_config(int, void *) to enable run-time extraction
2672    of things that can be changed at compile time. The first argument specifies
2673    what is wanted and the second points to where the information is to be placed.
2674    The current list of available information is:
2675    
2676      PCRE_CONFIG_UTF8
2677    
2678    The output is an integer that is set to one if UTF-8 support is available;
2679    otherwise it is set to zero.
2680    
2681      PCRE_CONFIG_NEWLINE
2682    
2683    The output is an integer that it set to the value of the code that is used for
2684    newline. It is either LF (10) or CR (13).
2685    
2686      PCRE_CONFIG_LINK_SIZE
2687    
2688    The output is an integer that contains the number of bytes used for internal
2689    linkage in compiled expressions. The value is 2, 3, or 4. See item 32 above.
2690    
2691      PCRE_CONFIG_POSIX_MALLOC_THRESHOLD
2692    
2693    The output is an integer that contains the threshold above which the POSIX
2694    interface uses malloc() for output vectors. See item 31 above.
2695    
2696      PCRE_CONFIG_MATCH_LIMIT
2697    
2698    The output is an unsigned integer that contains the default limit of the number
2699    of match() calls in a pcre_exec() execution. See 42 above.
2700    
2701    44. pcretest has been upgraded by the addition of the -C option. This causes it
2702    to extract all the available output from the new pcre_config() function, and to
2703    output it. The program then exits immediately.
2704    
2705    45. A need has arisen to pass over additional data with calls to pcre_exec() in
2706    order to support additional features. One way would have been to define
2707    pcre_exec2() (for example) with extra arguments, but this would not have been
2708    extensible, and would also have required all calls to the original function to
2709    be mapped to the new one. Instead, I have chosen to extend the mechanism that
2710    is used for passing in "extra" data from pcre_study().
2711    
2712    The pcre_extra structure is now exposed and defined in pcre.h. It currently
2713    contains the following fields:
2714    
2715      flags         a bitmap indicating which of the following fields are set
2716      study_data    opaque data from pcre_study()
2717      match_limit   a way of specifying a limit on match() calls for a specific
2718                      call to pcre_exec()
2719      callout_data  data for callouts (see 49 below)
2720    
2721    The flag bits are also defined in pcre.h, and are
2722    
2723      PCRE_EXTRA_STUDY_DATA
2724      PCRE_EXTRA_MATCH_LIMIT
2725      PCRE_EXTRA_CALLOUT_DATA
2726    
2727    The pcre_study() function now returns one of these new pcre_extra blocks, with
2728    the actual study data pointed to by the study_data field, and the
2729    PCRE_EXTRA_STUDY_DATA flag set. This can be passed directly to pcre_exec() as
2730    before. That is, this change is entirely upwards-compatible and requires no
2731    change to existing code.
2732    
2733    If you want to pass in additional data to pcre_exec(), you can either place it
2734    in a pcre_extra block provided by pcre_study(), or create your own pcre_extra
2735    block.
2736    
2737    46. pcretest has been extended to test the PCRE_EXTRA_MATCH_LIMIT feature. If a
2738    data string contains the escape sequence \M, pcretest calls pcre_exec() several
2739    times with different match limits, until it finds the minimum value needed for
2740    pcre_exec() to complete. The value is then output. This can be instructive; for
2741    most simple matches the number is quite small, but for pathological cases it
2742    gets very large very quickly.
2743    
2744    47. There's a new option for pcre_fullinfo() called PCRE_INFO_STUDYSIZE. It
2745    returns the size of the data block pointed to by the study_data field in a
2746    pcre_extra block, that is, the value that was passed as the argument to
2747    pcre_malloc() when PCRE was getting memory in which to place the information
2748    created by pcre_study(). The fourth argument should point to a size_t variable.
2749    pcretest has been extended so that this information is shown after a successful
2750    pcre_study() call when information about the compiled regex is being displayed.
2751    
2752    48. Cosmetic change to Makefile: there's no need to have / after $(DESTDIR)
2753    because what follows is always an absolute path. (Later: it turns out that this
2754    is more than cosmetic for MinGW, because it doesn't like empty path
2755    components.)
2756    
2757    49. Some changes have been made to the callout feature (see 28 above):
2758    
2759    (i)  A callout function now has three choices for what it returns:
2760    
2761           0  =>  success, carry on matching
2762         > 0  =>  failure at this point, but backtrack if possible
2763         < 0  =>  serious error, return this value from pcre_exec()
2764    
2765         Negative values should normally be chosen from the set of PCRE_ERROR_xxx
2766         values. In particular, returning PCRE_ERROR_NOMATCH forces a standard
2767         "match failed" error. The error number PCRE_ERROR_CALLOUT is reserved for
2768         use by callout functions. It will never be used by PCRE itself.
2769    
2770    (ii) The pcre_extra structure (see 45 above) has a void * field called
2771         callout_data, with corresponding flag bit PCRE_EXTRA_CALLOUT_DATA. The
2772         pcre_callout_block structure has a field of the same name. The contents of
2773         the field passed in the pcre_extra structure are passed to the callout
2774         function in the corresponding field in the callout block. This makes it
2775         easier to use the same callout-containing regex from multiple threads. For
2776         testing, the pcretest program has a new data escape
2777    
2778           \C*n        pass the number n (may be negative) as callout_data
2779    
2780         If the callout function in pcretest receives a non-zero value as
2781         callout_data, it returns that value.
2782    
2783    50. Makefile wasn't handling CFLAGS properly when compiling dftables. Also,
2784    there were some redundant $(CFLAGS) in commands that are now specified as
2785    $(LINK), which already includes $(CFLAGS).
2786    
2787    51. Extensions to UTF-8 support are listed below. These all apply when (a) PCRE
2788    has been compiled with UTF-8 support *and* pcre_compile() has been compiled
2789    with the PCRE_UTF8 flag. Patterns that are compiled without that flag assume
2790    one-byte characters throughout. Note that case-insensitive matching applies
2791    only to characters whose values are less than 256. PCRE doesn't support the
2792    notion of cases for higher-valued characters.
2793    
2794    (i)   A character class whose characters are all within 0-255 is handled as
2795          a bit map, and the map is inverted for negative classes. Previously, a
2796          character > 255 always failed to match such a class; however it should
2797          match if the class was a negative one (e.g. [^ab]). This has been fixed.
2798    
2799    (ii)  A negated character class with a single character < 255 is coded as
2800          "not this character" (OP_NOT). This wasn't working properly when the test
2801          character was multibyte, either singly or repeated.
2802    
2803    (iii) Repeats of multibyte characters are now handled correctly in UTF-8
2804          mode, for example: \x{100}{2,3}.
2805    
2806    (iv)  The character escapes \b, \B, \d, \D, \s, \S, \w, and \W (either
2807          singly or repeated) now correctly test multibyte characters. However,
2808          PCRE doesn't recognize any characters with values greater than 255 as
2809          digits, spaces, or word characters. Such characters always match \D, \S,
2810          and \W, and never match \d, \s, or \w.
2811    
2812    (v)   Classes may now contain characters and character ranges with values
2813          greater than 255. For example: [ab\x{100}-\x{400}].
2814    
2815    (vi)  pcregrep now has a --utf-8 option (synonym -u) which makes it call
2816          PCRE in UTF-8 mode.
2817    
2818    52. The info request value PCRE_INFO_FIRSTCHAR has been renamed
2819    PCRE_INFO_FIRSTBYTE because it is a byte value. However, the old name is
2820    retained for backwards compatibility. (Note that LASTLITERAL is also a byte
2821    value.)
2822    
2823    53. The single man page has become too large. I have therefore split it up into
2824    a number of separate man pages. These also give rise to individual HTML pages;
2825    these are now put in a separate directory, and there is an index.html page that
2826    lists them all. Some hyperlinking between the pages has been installed.
2827    
2828    54. Added convenience functions for handling named capturing parentheses.
2829    
2830    55. Unknown escapes inside character classes (e.g. [\M]) and escapes that
2831    aren't interpreted therein (e.g. [\C]) are literals in Perl. This is now also
2832    true in PCRE, except when the PCRE_EXTENDED option is set, in which case they
2833    are faulted.
2834    
2835    56. Introduced HOST_CC and HOST_CFLAGS which can be set in the environment when
2836    calling configure. These values are used when compiling the dftables.c program
2837    which is run to generate the source of the default character tables. They
2838    default to the values of CC and CFLAGS. If you are cross-compiling PCRE,
2839    you will need to set these values.
2840    
2841    57. Updated the building process for Windows DLL, as provided by Fred Cox.
2842    
2843    
2844    Version 3.9 02-Jan-02
2845    ---------------------
2846    
2847    1. A bit of extraneous text had somehow crept into the pcregrep documentation.
2848    
2849    2. If --disable-static was given, the building process failed when trying to
2850    build pcretest and pcregrep. (For some reason it was using libtool to compile
2851    them, which is not right, as they aren't part of the library.)
2852    
2853    
2854    Version 3.8 18-Dec-01
2855    ---------------------
2856    
2857    1. The experimental UTF-8 code was completely screwed up. It was packing the
2858    bytes in the wrong order. How dumb can you get?
2859    
2860    
2861    Version 3.7 29-Oct-01
2862    ---------------------
2863    
2864    1. In updating pcretest to check change 1 of version 3.6, I screwed up.
2865    This caused pcretest, when used on the test data, to segfault. Unfortunately,
2866    this didn't happen under Solaris 8, where I normally test things.
2867    
2868    2. The Makefile had to be changed to make it work on BSD systems, where 'make'
2869    doesn't seem to recognize that ./xxx and xxx are the same file. (This entry
2870    isn't in ChangeLog distributed with 3.7 because I forgot when I hastily made
2871    this fix an hour or so after the initial 3.7 release.)
2872    
2873    
2874    Version 3.6 23-Oct-01
2875    ---------------------
2876    
2877    1. Crashed with /(sens|respons)e and \1ibility/ and "sense and sensibility" if
2878    offsets passed as NULL with zero offset count.
2879    
2880    2. The config.guess and config.sub files had not been updated when I moved to
2881    the latest autoconf.
2882    
2883    
2884    Version 3.5 15-Aug-01
2885    ---------------------
2886    
2887    1. Added some missing #if !defined NOPOSIX conditionals in pcretest.c that
2888    had been forgotten.
2889    
2890    2. By using declared but undefined structures, we can avoid using "void"
2891    definitions in pcre.h while keeping the internal definitions of the structures
2892    private.
2893    
2894    3. The distribution is now built using autoconf 2.50 and libtool 1.4. From a
2895    user point of view, this means that both static and shared libraries are built
2896    by default, but this can be individually controlled. More of the work of
2897    handling this static/shared cases is now inside libtool instead of PCRE's make
2898    file.
2899    
2900    4. The pcretest utility is now installed along with pcregrep because it is
2901    useful for users (to test regexs) and by doing this, it automatically gets
2902    relinked by libtool. The documentation has been turned into a man page, so
2903    there are now .1, .txt, and .html versions in /doc.
2904    
2905    5. Upgrades to pcregrep:
2906       (i)   Added long-form option names like gnu grep.
2907       (ii)  Added --help to list all options with an explanatory phrase.
2908       (iii) Added -r, --recursive to recurse into sub-directories.
2909       (iv)  Added -f, --file to read patterns from a file.
2910    
2911    6. pcre_exec() was referring to its "code" argument before testing that
2912    argument for NULL (and giving an error if it was NULL).
2913    
2914    7. Upgraded Makefile.in to allow for compiling in a different directory from
2915    the source directory.
2916    
2917    8. Tiny buglet in pcretest: when pcre_fullinfo() was called to retrieve the
2918    options bits, the pointer it was passed was to an int instead of to an unsigned
2919    long int. This mattered only on 64-bit systems.
2920    
2921    9. Fixed typo (3.4/1) in pcre.h again. Sigh. I had changed pcre.h (which is
2922    generated) instead of pcre.in, which it its source. Also made the same change
2923    in several of the .c files.
2924    
2925    10. A new release of gcc defines printf() as a macro, which broke pcretest
2926    because it had an ifdef in the middle of a string argument for printf(). Fixed
2927    by using separate calls to printf().
2928    
2929    11. Added --enable-newline-is-cr and --enable-newline-is-lf to the configure
2930    script, to force use of CR or LF instead of \n in the source. On non-Unix
2931    systems, the value can be set in config.h.
2932    
2933    12. The limit of 200 on non-capturing parentheses is a _nesting_ limit, not an
2934    absolute limit. Changed the text of the error message to make this clear, and
2935    likewise updated the man page.
2936    
2937    13. The limit of 99 on the number of capturing subpatterns has been removed.
2938    The new limit is 65535, which I hope will not be a "real" limit.
2939    
2940    
2941    Version 3.4 22-Aug-00
2942    ---------------------
2943    
2944    1. Fixed typo in pcre.h: unsigned const char * changed to const unsigned char *.
2945    
2946    2. Diagnose condition (?(0) as an error instead of crashing on matching.
2947    
2948    
2949    Version 3.3 01-Aug-00
2950    ---------------------
2951    
2952    1. If an octal character was given, but the value was greater than \377, it
2953    was not getting masked to the least significant bits, as documented. This could
2954    lead to crashes in some systems.
2955    
2956    2. Perl 5.6 (if not earlier versions) accepts classes like [a-\d] and treats
2957    the hyphen as a literal. PCRE used to give an error; it now behaves like Perl.
2958    
2959    3. Added the functions pcre_free_substring() and pcre_free_substring_list().
2960    These just pass their arguments on to (pcre_free)(), but they are provided
2961    because some uses of PCRE bind it to non-C systems that can call its functions,
2962    but cannot call free() or pcre_free() directly.
2963    
2964    4. Add "make test" as a synonym for "make check". Corrected some comments in
2965    the Makefile.
2966    
2967    5. Add $(DESTDIR)/ in front of all the paths in the "install" target in the
2968    Makefile.
2969    
2970    6. Changed the name of pgrep to pcregrep, because Solaris has introduced a
2971    command called pgrep for grepping around the active processes.
2972    
2973    7. Added the beginnings of support for UTF-8 character strings.
2974    
2975    8. Arranged for the Makefile to pass over the settings of CC, CFLAGS, and
2976    RANLIB to ./ltconfig so that they are used by libtool. I think these are all
2977    the relevant ones. (AR is not passed because ./ltconfig does its own figuring
2978    out for the ar command.)
2979    
2980    
2981    Version 3.2 12-May-00
2982    ---------------------
2983    
2984    This is purely a bug fixing release.
2985    
2986    1. If the pattern /((Z)+|A)*/ was matched agained ZABCDEFG it matched Z instead
2987    of ZA. This was just one example of several cases that could provoke this bug,
2988    which was introduced by change 9 of version 2.00. The code for breaking
2989    infinite loops after an iteration that matches an empty string was't working
2990    correctly.
2991    
2992    2. The pcretest program was not imitating Perl correctly for the pattern /a*/g
2993    when matched against abbab (for example). After matching an empty string, it
2994    wasn't forcing anchoring when setting PCRE_NOTEMPTY for the next attempt; this
2995    caused it to match further down the string than it should.
2996    
2997    3. The code contained an inclusion of sys/types.h. It isn't clear why this
2998    was there because it doesn't seem to be needed, and it causes trouble on some
2999    systems, as it is not a Standard C header. It has been removed.
3000    
3001    4. Made 4 silly changes to the source to avoid stupid compiler warnings that
3002    were reported on the Macintosh. The changes were from
3003    
3004      while ((c = *(++ptr)) != 0 && c != '\n');
3005    to
3006      while ((c = *(++ptr)) != 0 && c != '\n') ;
3007    
3008    Totally extraordinary, but if that's what it takes...
3009    
3010    5. PCRE is being used in one environment where neither memmove() nor bcopy() is
3011    available. Added HAVE_BCOPY and an autoconf test for it; if neither
3012    HAVE_MEMMOVE nor HAVE_BCOPY is set, use a built-in emulation function which
3013    assumes the way PCRE uses memmove() (always moving upwards).
3014    
3015    6. PCRE is being used in one environment where strchr() is not available. There
3016    was only one use in pcre.c, and writing it out to avoid strchr() probably gives
3017    faster code anyway.
3018    
3019    
3020    Version 3.1 09-Feb-00
3021    ---------------------
3022    
3023    The only change in this release is the fixing of some bugs in Makefile.in for
3024    the "install" target:
3025    
3026    (1) It was failing to install pcreposix.h.
3027    
3028    (2) It was overwriting the pcre.3 man page with the pcreposix.3 man page.
3029    
3030    
3031    Version 3.0 01-Feb-00
3032    ---------------------
3033    
3034    1. Add support for the /+ modifier to perltest (to output $` like it does in
3035    pcretest).
3036    
3037    2. Add support for the /g modifier to perltest.
3038    
3039    3. Fix pcretest so that it behaves even more like Perl for /g when the pattern
3040    matches null strings.
3041    
3042    4. Fix perltest so that it doesn't do unwanted things when fed an empty
3043    pattern. Perl treats empty patterns specially - it reuses the most recent
3044    pattern, which is not what we want. Replace // by /(?#)/ in order to avoid this
3045    effect.
3046    
3047    5. The POSIX interface was broken in that it was just handing over the POSIX
3048    captured string vector to pcre_exec(), but (since release 2.00) PCRE has
3049    required a bigger vector, with some working space on the end. This means that
3050    the POSIX wrapper now has to get and free some memory, and copy the results.
3051    
3052    6. Added some simple autoconf support, placing the test data and the
3053    documentation in separate directories, re-organizing some of the
3054    information files, and making it build pcre-config (a GNU standard). Also added
3055    libtool support for building PCRE as a shared library, which is now the
3056    default.
3057    
3058    7. Got rid of the leading zero in the definition of PCRE_MINOR because 08 and
3059    09 are not valid octal constants. Single digits will be used for minor values
3060    less than 10.
3061    
3062    8. Defined REG_EXTENDED and REG_NOSUB as zero in the POSIX header, so that
3063    existing programs that set these in the POSIX interface can use PCRE without
3064    modification.
3065    
3066    9. Added a new function, pcre_fullinfo() with an extensible interface. It can
3067    return all that pcre_info() returns, plus additional data. The pcre_info()
3068    function is retained for compatibility, but is considered to be obsolete.
3069    
3070    10. Added experimental recursion feature (?R) to handle one common case that
3071    Perl 5.6 will be able to do with (?p{...}).
3072    
3073    11. Added support for POSIX character classes like [:alpha:], which Perl is
3074    adopting.
3075    
3076    
3077    Version 2.08 31-Aug-99
3078    ----------------------
3079    
3080    1. When startoffset was not zero and the pattern began with ".*", PCRE was not
3081    trying to match at the startoffset position, but instead was moving forward to
3082    the next newline as if a previous match had failed.
3083    
3084    2. pcretest was not making use of PCRE_NOTEMPTY when repeating for /g and /G,
3085    and could get into a loop if a null string was matched other than at the start
3086    of the subject.
3087    
3088    3. Added definitions of PCRE_MAJOR and PCRE_MINOR to pcre.h so the version can
3089    be distinguished at compile time, and for completeness also added PCRE_DATE.
3090    
3091    5. Added Paul Sokolovsky's minor changes to make it easy to compile a Win32 DLL
3092    in GnuWin32 environments.
3093    
3094    
3095    Version 2.07 29-Jul-99
3096    ----------------------
3097    
3098    1. The documentation is now supplied in plain text form and HTML as well as in
3099    the form of man page sources.
3100    
3101    2. C++ compilers don't like assigning (void *) values to other pointer types.
3102    In particular this affects malloc(). Although there is no problem in Standard
3103    C, I've put in casts to keep C++ compilers happy.
3104    
3105    3. Typo on pcretest.c; a cast of (unsigned char *) in the POSIX regexec() call
3106    should be (const char *).
3107    
3108    4. If NOPOSIX is defined, pcretest.c compiles without POSIX support. This may
3109    be useful for non-Unix systems who don't want to bother with the POSIX stuff.
3110    However, I haven't made this a standard facility. The documentation doesn't
3111    mention it, and the Makefile doesn't support it.
3112    
3113    5. The Makefile now contains an "install" target, with editable destinations at
3114    the top of the file. The pcretest program is not installed.
3115    
3116    6. pgrep -V now gives the PCRE version number and date.
3117    
3118    7. Fixed bug: a zero repetition after a literal string (e.g. /abcde{0}/) was
3119    causing the entire string to be ignored, instead of just the last character.
3120    
3121    8. If a pattern like /"([^\\"]+|\\.)*"/ is applied in the normal way to a
3122    non-matching string, it can take a very, very long time, even for strings of
3123    quite modest length, because of the nested recursion. PCRE now does better in
3124    some of these cases. It does this by remembering the last required literal
3125    character in the pattern, and pre-searching the subject to ensure it is present
3126    before running the real match. In other words, it applies a heuristic to detect
3127    some types of certain failure quickly, and in the above example, if presented
3128    with a string that has no trailing " it gives "no match" very quickly.
3129    
3130    9. A new runtime option PCRE_NOTEMPTY causes null string matches to be ignored;
3131    other alternatives are tried instead.
3132    
3133    
3134    Version 2.06 09-Jun-99
3135    ----------------------
3136    
3137    1. Change pcretest's output for amount of store used to show just the code
3138    space, because the remainder (the data block) varies in size between 32-bit and
3139    64-bit systems.
3140    
3141    2. Added an extra argument to pcre_exec() to supply an offset in the subject to
3142    start matching at. This allows lookbehinds to work when searching for multiple
3143    occurrences in a string.
3144    
3145    3. Added additional options to pcretest for testing multiple occurrences:
3146    
3147       /+   outputs the rest of the string that follows a match
3148       /g   loops for multiple occurrences, using the new startoffset argument
3149       /G   loops for multiple occurrences by passing an incremented pointer
3150    
3151    4. PCRE wasn't doing the "first character" optimization for patterns starting
3152    with \b or \B, though it was doing it for other lookbehind assertions. That is,
3153    it wasn't noticing that a match for a pattern such as /\bxyz/ has to start with
3154    the letter 'x'. On long subject strings, this gives a significant speed-up.
3155    
3156    
3157    Version 2.05 21-Apr-99
3158    ----------------------
3159    
3160    1. Changed the type of magic_number from int to long int so that it works
3161    properly on 16-bit systems.
3162    
3163    2. Fixed a bug which caused patterns starting with .* not to work correctly
3164    when the subject string contained newline characters. PCRE was assuming
3165    anchoring for such patterns in all cases, which is not correct because .* will
3166    not pass a newline unless PCRE_DOTALL is set. It now assumes anchoring only if
3167    DOTALL is set at top level; otherwise it knows that patterns starting with .*
3168    must be retried after every newline in the subject.
3169    
3170    
3171    Version 2.04 18-Feb-99
3172    ----------------------
3173    
3174    1. For parenthesized subpatterns with repeats whose minimum was zero, the
3175    computation of the store needed to hold the pattern was incorrect (too large).
3176    If such patterns were nested a few deep, this could multiply and become a real
3177    problem.
3178    
3179    2. Added /M option to pcretest to show the memory requirement of a specific
3180    pattern. Made -m a synonym of -s (which does this globally) for compatibility.
3181    
3182    3. Subpatterns of the form (regex){n,m} (i.e. limited maximum) were being
3183    compiled in such a way that the backtracking after subsequent failure was
3184    pessimal. Something like (a){0,3} was compiled as (a)?(a)?(a)? instead of
3185    ((a)((a)(a)?)?)? with disastrous performance if the maximum was of any size.
3186    
3187    
3188    Version 2.03 02-Feb-99
3189    ----------------------
3190    
3191    1. Fixed typo and small mistake in man page.
3192    
3193    2. Added 4th condition (GPL supersedes if conflict) and created separate
3194    LICENCE file containing the conditions.
3195    
3196    3. Updated pcretest so that patterns such as /abc\/def/ work like they do in
3197    Perl, that is the internal \ allows the delimiter to be included in the
3198    pattern. Locked out the use of \ as a delimiter. If \ immediately follows
3199    the final delimiter, add \ to the end of the pattern (to test the error).
3200    
3201    4. Added the convenience functions for extracting substrings after a successful
3202    match. Updated pcretest to make it able to test these functions.
3203    
3204    
3205    Version 2.02 14-Jan-99
3206    ----------------------
3207    
3208    1. Initialized the working variables associated with each extraction so that
3209    their saving and restoring doesn't refer to uninitialized store.
3210    
3211    2. Put dummy code into study.c in order to trick the optimizer of the IBM C
3212    compiler for OS/2 into generating correct code. Apparently IBM isn't going to
3213    fix the problem.
3214    
3215    3. Pcretest: the timing code wasn't using LOOPREPEAT for timing execution
3216    calls, and wasn't printing the correct value for compiling calls. Increased the
3217    default value of LOOPREPEAT, and the number of significant figures in the
3218    times.
3219    
3220    4. Changed "/bin/rm" in the Makefile to "-rm" so it works on Windows NT.
3221    
3222    5. Renamed "deftables" as "dftables" to get it down to 8 characters, to avoid
3223    a building problem on Windows NT with a FAT file system.
3224    
3225    
3226    Version 2.01 21-Oct-98
3227    ----------------------
3228    
3229    1. Changed the API for pcre_compile() to allow for the provision of a pointer
3230    to character tables built by pcre_maketables() in the current locale. If NULL
3231    is passed, the default tables are used.
3232    
3233    
3234    Version 2.00 24-Sep-98
3235    ----------------------
3236    
3237    1. Since the (>?) facility is in Perl 5.005, don't require PCRE_EXTRA to enable
3238    it any more.
3239    
3240    2. Allow quantification of (?>) groups, and make it work correctly.
3241    
3242    3. The first character computation wasn't working for (?>) groups.
3243    
3244    4. Correct the implementation of \Z (it is permitted to match on the \n at the
3245    end of the subject) and add 5.005's \z, which really does match only at the
3246    very end of the subject.
3247    
3248    5. Remove the \X "cut" facility; Perl doesn't have it, and (?> is neater.
3249    
3250    6. Remove the ability to specify CASELESS, MULTILINE, DOTALL, and
3251    DOLLAR_END_ONLY at runtime, to make it possible to implement the Perl 5.005
3252    localized options. All options to pcre_study() were also removed.
3253    
3254    7. Add other new features from 5.005:
3255    
3256       $(?<=           positive lookbehind
3257       $(?<!           negative lookbehind
3258       (?imsx-imsx)    added the unsetting capability
3259                       such a setting is global if at outer level; local otherwise
3260       (?imsx-imsx:)   non-capturing groups with option setting
3261       (?(cond)re|re)  conditional pattern matching
3262    
3263       A backreference to itself in a repeated group matches the previous
3264       captured string.
3265    
3266    8. General tidying up of studying (both automatic and via "study")
3267    consequential on the addition of new assertions.
3268    
3269    9. As in 5.005, unlimited repeated groups that could match an empty substring
3270    are no longer faulted at compile time. Instead, the loop is forcibly broken at
3271    runtime if any iteration does actually match an empty substring.
3272    
3273    10. Include the RunTest script in the distribution.
3274    
3275    11. Added tests from the Perl 5.005_02 distribution. This showed up a few
3276    discrepancies, some of which were old and were also with respect to 5.004. They
3277    have now been fixed.
3278    
3279    
3280    Version 1.09 28-Apr-98
3281    ----------------------
3282    
3283    1. A negated single character class followed by a quantifier with a minimum
3284    value of one (e.g.  [^x]{1,6}  ) was not compiled correctly. This could lead to
3285    program crashes, or just wrong answers. This did not apply to negated classes
3286    containing more than one character, or to minima other than one.
3287    
3288    
3289    Version 1.08 27-Mar-98
3290    ----------------------
3291    
3292    1. Add PCRE_UNGREEDY to invert the greediness of quantifiers.
3293    
3294    2. Add (?U) and (?X) to set PCRE_UNGREEDY and PCRE_EXTRA respectively. The
3295    latter must appear before anything that relies on it in the pattern.
3296    
3297    
3298    Version 1.07 16-Feb-98
3299    ----------------------
3300    
3301    1. A pattern such as /((a)*)*/ was not being diagnosed as in error (unlimited
3302    repeat of a potentially empty string).
3303    
3304    
3305    Version 1.06 23-Jan-98
3306    ----------------------
3307    
3308    1. Added Markus Oberhumer's little patches for C++.
3309    
3310    2. Literal strings longer than 255 characters were broken.
3311    
3312    
3313    Version 1.05 23-Dec-97
3314    ----------------------
3315    
3316    1. Negated character classes containing more than one character were failing if
3317    PCRE_CASELESS was set at run time.
3318    
3319    
3320    Version 1.04 19-Dec-97
3321    ----------------------
3322    
3323    1. Corrected the man page, where some "const" qualifiers had been omitted.
3324    
3325    2. Made debugging output print "{0,xxx}" instead of just "{,xxx}" to agree with
3326    input syntax.
3327    
3328    3. Fixed memory leak which occurred when a regex with back references was
3329    matched with an offsets vector that wasn't big enough. The temporary memory
3330    that is used in this case wasn't being freed if the match failed.
3331    
3332    4. Tidied pcretest to ensure it frees memory that it gets.
3333    
3334    5. Temporary memory was being obtained in the case where the passed offsets
3335    vector was exactly big enough.
3336    
3337    6. Corrected definition of offsetof() from change 5 below.
3338    
3339    7. I had screwed up change 6 below and broken the rules for the use of
3340    setjmp(). Now fixed.
3341    
3342    
3343    Version 1.03 18-Dec-97
3344    ----------------------
3345    
3346    1. A erroneous regex with a missing opening parenthesis was correctly
3347    diagnosed, but PCRE attempted to access brastack[-1], which could cause crashes
3348    on some systems.
3349    
3350    2. Replaced offsetof(real_pcre, code) by offsetof(real_pcre, code[0]) because
3351    it was reported that one broken compiler failed on the former because "code" is
3352    also an independent variable.
3353    
3354    3. The erroneous regex a[]b caused an array overrun reference.
3355    
3356    4. A regex ending with a one-character negative class (e.g. /[^k]$/) did not
3357    fail on data ending with that character. (It was going on too far, and checking
3358    the next character, typically a binary zero.) This was specific to the
3359    optimized code for single-character negative classes.
3360    
3361    5. Added a contributed patch from the TIN world which does the following:
3362    
3363      + Add an undef for memmove, in case the the system defines a macro for it.
3364    
3365      + Add a definition of offsetof(), in case there isn't one. (I don't know
3366        the reason behind this - offsetof() is part of the ANSI standard - but
3367        it does no harm).
3368    
3369      + Reduce the ifdef's in pcre.c using macro DPRINTF, thereby eliminating
3370        most of the places where whitespace preceded '#'. I have given up and
3371        allowed the remaining 2 cases to be at the margin.
3372    
3373      + Rename some variables in pcre to eliminate shadowing. This seems very
3374        pedantic, but does no harm, of course.
3375    
3376    6. Moved the call to setjmp() into its own function, to get rid of warnings
3377    from gcc -Wall, and avoided calling it at all unless PCRE_EXTRA is used.
3378    
3379    7. Constructs such as \d{8,} were compiling into the equivalent of
3380    \d{8}\d{0,65527} instead of \d{8}\d* which didn't make much difference to the
3381    outcome, but in this particular case used more store than had been allocated,
3382    which caused the bug to be discovered because it threw up an internal error.
3383    
3384    8. The debugging code in both pcre and pcretest for outputting the compiled
3385    form of a regex was going wrong in the case of back references followed by
3386    curly-bracketed repeats.
3387    
3388    
3389    Version 1.02 12-Dec-97
3390    ----------------------
3391    
3392    1. Typos in pcre.3 and comments in the source fixed.
3393    
3394    2. Applied a contributed patch to get rid of places where it used to remove
3395    'const' from variables, and fixed some signed/unsigned and uninitialized
3396    variable warnings.
3397    
3398    3. Added the "runtest" target to Makefile.
3399    
3400    4. Set default compiler flag to -O2 rather than just -O.
3401    
3402    
3403  Version 1.01 19-Nov-97  Version 1.01 19-Nov-97
3404  ----------------------  ----------------------

Legend:
Removed from v.5  
changed lines
  Added in v.431

webmaster@exim.org
ViewVC Help
Powered by ViewVC 1.1.12