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

Diff of /code/trunk/ChangeLog

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

revision 19 by nigel, Sat Feb 24 21:38:33 2007 UTC revision 480 by ph10, Sun Jan 3 17:29:59 2010 UTC
# Line 1  Line 1 
1  ChangeLog for PCRE  ChangeLog for PCRE
2  ------------------  ------------------
3    
4    Version 8.01 11-Dec-09
5    ----------------------
6    
7    1.  If a pattern contained a conditional subpattern with only one branch (in
8        particular, this includes all (DEFINE) patterns), a call to pcre_study()
9        computed the wrong minimum data length (which is of course zero for such
10        subpatterns).
11    
12    2.  For patterns such as (?i)a(?-i)b|c where an option setting at the start of
13        the pattern is reset in the first branch, pcre_compile() failed with
14        "internal error: code overflow at offset...". This happened only when
15        the reset was to the original external option setting. (An optimization
16        abstracts leading options settings into an external setting, which was the
17        cause of this.)
18    
19    3.  A pattern such as ^(?!a(*SKIP)b) where a negative assertion contained one
20        of the verbs SKIP, PRUNE, or COMMIT, did not work correctly. When the
21        assertion pattern did not match (meaning that the assertion was true), it
22        was incorrectly treated as false if the SKIP had been reached during the
23        matching. This also applied to assertions used as conditions.
24    
25    4.  If an item that is not supported by pcre_dfa_exec() was encountered in an
26        assertion subpattern, including such a pattern used as a condition,
27        unpredictable results occurred, instead of the error return
28        PCRE_ERROR_DFA_UITEM.
29    
30    5.  The C++ GlobalReplace function was not working like Perl for the special
31        situation when an empty string is matched. It now does the fancy magic
32        stuff that is necessary.
33    
34    6.  In pcre_internal.h, obsolete includes to setjmp.h and stdarg.h have been
35        removed. (These were left over from very, very early versions of PCRE.)
36    
37    7.  Some cosmetic changes to the code to make life easier when compiling it
38        as part of something else:
39    
40        (a) Change DEBUG to PCRE_DEBUG.
41    
42        (b) In pcre_compile(), rename the member of the "branch_chain" structure
43            called "current" as "current_branch", to prevent a collision with the
44            Linux macro when compiled as a kernel module.
45    
46        (c) In pcre_study(), rename the function set_bit() as set_table_bit(), to
47            prevent a collision with the Linux macro when compiled as a kernel
48            module.
49    
50    8.  In pcre_compile() there are some checks for integer overflows that used to
51        cast potentially large values to (double). This has been changed to that
52        when building, a check for int64_t is made, and if it is found, it is used
53        instead, thus avoiding the use of floating point arithmetic. (There is no
54        other use of FP in PCRE.) If int64_t is not found, the fallback is to
55        double.
56    
57    9.  Added two casts to avoid signed/unsigned warnings from VS Studio Express
58        2005 (difference between two addresses compared to an unsigned value).
59    
60    10. Change the standard AC_CHECK_LIB test for libbz2 in configure.ac to a
61        custom one, because of the following reported problem in Windows:
62    
63          - libbz2 uses the Pascal calling convention (WINAPI) for the functions
64            under Win32.
65          - The standard autoconf AC_CHECK_LIB fails to include "bzlib.h",
66            therefore missing the function definition.
67          - The compiler thus generates a "C" signature for the test function.
68          - The linker fails to find the "C" function.
69          - PCRE fails to configure if asked to do so against libbz2.
70    
71    11. When running libtoolize from libtool-2.2.6b as part of autogen.sh, these
72        messages were output:
73    
74          Consider adding `AC_CONFIG_MACRO_DIR([m4])' to configure.ac and
75          rerunning libtoolize, to keep the correct libtool macros in-tree.
76          Consider adding `-I m4' to ACLOCAL_AMFLAGS in Makefile.am.
77    
78        I have done both of these things.
79    
80    12. Although pcre_dfa_exec() does not use nearly as much stack as pcre_exec()
81        most of the time, it *can* run out if it is given a pattern that contains a
82        runaway infinite recursion. I updated the discussion in the pcrestack man
83        page.
84    
85    
86    
87    Version 8.00 19-Oct-09
88    ----------------------
89    
90    1.  The table for translating pcre_compile() error codes into POSIX error codes
91        was out-of-date, and there was no check on the pcre_compile() error code
92        being within the table. This could lead to an OK return being given in
93        error.
94    
95    2.  Changed the call to open a subject file in pcregrep from fopen(pathname,
96        "r") to fopen(pathname, "rb"), which fixed a problem with some of the tests
97        in a Windows environment.
98    
99    3.  The pcregrep --count option prints the count for each file even when it is
100        zero, as does GNU grep. However, pcregrep was also printing all files when
101        --files-with-matches was added. Now, when both options are given, it prints
102        counts only for those files that have at least one match. (GNU grep just
103        prints the file name in this circumstance, but including the count seems
104        more useful - otherwise, why use --count?) Also ensured that the
105        combination -clh just lists non-zero counts, with no names.
106    
107    4.  The long form of the pcregrep -F option was incorrectly implemented as
108        --fixed_strings instead of --fixed-strings. This is an incompatible change,
109        but it seems right to fix it, and I didn't think it was worth preserving
110        the old behaviour.
111    
112    5.  The command line items --regex=pattern and --regexp=pattern were not
113        recognized by pcregrep, which required --regex pattern or --regexp pattern
114        (with a space rather than an '='). The man page documented the '=' forms,
115        which are compatible with GNU grep; these now work.
116    
117    6.  No libpcreposix.pc file was created for pkg-config; there was just
118        libpcre.pc and libpcrecpp.pc. The omission has been rectified.
119    
120    7.  Added #ifndef SUPPORT_UCP into the pcre_ucd.c module, to reduce its size
121        when UCP support is not needed, by modifying the Python script that
122        generates it from Unicode data files. This should not matter if the module
123        is correctly used as a library, but I received one complaint about 50K of
124        unwanted data. My guess is that the person linked everything into his
125        program rather than using a library. Anyway, it does no harm.
126    
127    8.  A pattern such as /\x{123}{2,2}+/8 was incorrectly compiled; the trigger
128        was a minimum greater than 1 for a wide character in a possessive
129        repetition. The same bug could also affect patterns like /(\x{ff}{0,2})*/8
130        which had an unlimited repeat of a nested, fixed maximum repeat of a wide
131        character. Chaos in the form of incorrect output or a compiling loop could
132        result.
133    
134    9.  The restrictions on what a pattern can contain when partial matching is
135        requested for pcre_exec() have been removed. All patterns can now be
136        partially matched by this function. In addition, if there are at least two
137        slots in the offset vector, the offset of the earliest inspected character
138        for the match and the offset of the end of the subject are set in them when
139        PCRE_ERROR_PARTIAL is returned.
140    
141    10. Partial matching has been split into two forms: PCRE_PARTIAL_SOFT, which is
142        synonymous with PCRE_PARTIAL, for backwards compatibility, and
143        PCRE_PARTIAL_HARD, which causes a partial match to supersede a full match,
144        and may be more useful for multi-segment matching.
145    
146    11. Partial matching with pcre_exec() is now more intuitive. A partial match
147        used to be given if ever the end of the subject was reached; now it is
148        given only if matching could not proceed because another character was
149        needed. This makes a difference in some odd cases such as Z(*FAIL) with the
150        string "Z", which now yields "no match" instead of "partial match". In the
151        case of pcre_dfa_exec(), "no match" is given if every matching path for the
152        final character ended with (*FAIL).
153    
154    12. Restarting a match using pcre_dfa_exec() after a partial match did not work
155        if the pattern had a "must contain" character that was already found in the
156        earlier partial match, unless partial matching was again requested. For
157        example, with the pattern /dog.(body)?/, the "must contain" character is
158        "g". If the first part-match was for the string "dog", restarting with
159        "sbody" failed. This bug has been fixed.
160    
161    13. The string returned by pcre_dfa_exec() after a partial match has been
162        changed so that it starts at the first inspected character rather than the
163        first character of the match. This makes a difference only if the pattern
164        starts with a lookbehind assertion or \b or \B (\K is not supported by
165        pcre_dfa_exec()). It's an incompatible change, but it makes the two
166        matching functions compatible, and I think it's the right thing to do.
167    
168    14. Added a pcredemo man page, created automatically from the pcredemo.c file,
169        so that the demonstration program is easily available in environments where
170        PCRE has not been installed from source.
171    
172    15. Arranged to add -DPCRE_STATIC to cflags in libpcre.pc, libpcreposix.cp,
173        libpcrecpp.pc and pcre-config when PCRE is not compiled as a shared
174        library.
175    
176    16. Added REG_UNGREEDY to the pcreposix interface, at the request of a user.
177        It maps to PCRE_UNGREEDY. It is not, of course, POSIX-compatible, but it
178        is not the first non-POSIX option to be added. Clearly some people find
179        these options useful.
180    
181    17. If a caller to the POSIX matching function regexec() passes a non-zero
182        value for nmatch with a NULL value for pmatch, the value of
183        nmatch is forced to zero.
184    
185    18. RunGrepTest did not have a test for the availability of the -u option of
186        the diff command, as RunTest does. It now checks in the same way as
187        RunTest, and also checks for the -b option.
188    
189    19. If an odd number of negated classes containing just a single character
190        interposed, within parentheses, between a forward reference to a named
191        subpattern and the definition of the subpattern, compilation crashed with
192        an internal error, complaining that it could not find the referenced
193        subpattern. An example of a crashing pattern is /(?&A)(([^m])(?<A>))/.
194        [The bug was that it was starting one character too far in when skipping
195        over the character class, thus treating the ] as data rather than
196        terminating the class. This meant it could skip too much.]
197    
198    20. Added PCRE_NOTEMPTY_ATSTART in order to be able to correctly implement the
199        /g option in pcretest when the pattern contains \K, which makes it possible
200        to have an empty string match not at the start, even when the pattern is
201        anchored. Updated pcretest and pcredemo to use this option.
202    
203    21. If the maximum number of capturing subpatterns in a recursion was greater
204        than the maximum at the outer level, the higher number was returned, but
205        with unset values at the outer level. The correct (outer level) value is
206        now given.
207    
208    22. If (*ACCEPT) appeared inside capturing parentheses, previous releases of
209        PCRE did not set those parentheses (unlike Perl). I have now found a way to
210        make it do so. The string so far is captured, making this feature
211        compatible with Perl.
212    
213    23. The tests have been re-organized, adding tests 11 and 12, to make it
214        possible to check the Perl 5.10 features against Perl 5.10.
215    
216    24. Perl 5.10 allows subroutine calls in lookbehinds, as long as the subroutine
217        pattern matches a fixed length string. PCRE did not allow this; now it
218        does. Neither allows recursion.
219    
220    25. I finally figured out how to implement a request to provide the minimum
221        length of subject string that was needed in order to match a given pattern.
222        (It was back references and recursion that I had previously got hung up
223        on.) This code has now been added to pcre_study(); it finds a lower bound
224        to the length of subject needed. It is not necessarily the greatest lower
225        bound, but using it to avoid searching strings that are too short does give
226        some useful speed-ups. The value is available to calling programs via
227        pcre_fullinfo().
228    
229    26. While implementing 25, I discovered to my embarrassment that pcretest had
230        not been passing the result of pcre_study() to pcre_dfa_exec(), so the
231        study optimizations had never been tested with that matching function.
232        Oops. What is worse, even when it was passed study data, there was a bug in
233        pcre_dfa_exec() that meant it never actually used it. Double oops. There
234        were also very few tests of studied patterns with pcre_dfa_exec().
235    
236    27. If (?| is used to create subpatterns with duplicate numbers, they are now
237        allowed to have the same name, even if PCRE_DUPNAMES is not set. However,
238        on the other side of the coin, they are no longer allowed to have different
239        names, because these cannot be distinguished in PCRE, and this has caused
240        confusion. (This is a difference from Perl.)
241    
242    28. When duplicate subpattern names are present (necessarily with different
243        numbers, as required by 27 above), and a test is made by name in a
244        conditional pattern, either for a subpattern having been matched, or for
245        recursion in such a pattern, all the associated numbered subpatterns are
246        tested, and the overall condition is true if the condition is true for any
247        one of them. This is the way Perl works, and is also more like the way
248        testing by number works.
249    
250    
251    Version 7.9 11-Apr-09
252    ---------------------
253    
254    1.  When building with support for bzlib/zlib (pcregrep) and/or readline
255        (pcretest), all targets were linked against these libraries. This included
256        libpcre, libpcreposix, and libpcrecpp, even though they do not use these
257        libraries. This caused unwanted dependencies to be created. This problem
258        has been fixed, and now only pcregrep is linked with bzlib/zlib and only
259        pcretest is linked with readline.
260    
261    2.  The "typedef int BOOL" in pcre_internal.h that was included inside the
262        "#ifndef FALSE" condition by an earlier change (probably 7.8/18) has been
263        moved outside it again, because FALSE and TRUE are already defined in AIX,
264        but BOOL is not.
265    
266    3.  The pcre_config() function was treating the PCRE_MATCH_LIMIT and
267        PCRE_MATCH_LIMIT_RECURSION values as ints, when they should be long ints.
268    
269    4.  The pcregrep documentation said spaces were inserted as well as colons (or
270        hyphens) following file names and line numbers when outputting matching
271        lines. This is not true; no spaces are inserted. I have also clarified the
272        wording for the --colour (or --color) option.
273    
274    5.  In pcregrep, when --colour was used with -o, the list of matching strings
275        was not coloured; this is different to GNU grep, so I have changed it to be
276        the same.
277    
278    6.  When --colo(u)r was used in pcregrep, only the first matching substring in
279        each matching line was coloured. Now it goes on to look for further matches
280        of any of the test patterns, which is the same behaviour as GNU grep.
281    
282    7.  A pattern that could match an empty string could cause pcregrep to loop; it
283        doesn't make sense to accept an empty string match in pcregrep, so I have
284        locked it out (using PCRE's PCRE_NOTEMPTY option). By experiment, this
285        seems to be how GNU grep behaves.
286    
287    8.  The pattern (?(?=.*b)b|^) was incorrectly compiled as "match must be at
288        start or after a newline", because the conditional assertion was not being
289        correctly handled. The rule now is that both the assertion and what follows
290        in the first alternative must satisfy the test.
291    
292    9.  If auto-callout was enabled in a pattern with a conditional group whose
293        condition was an assertion, PCRE could crash during matching, both with
294        pcre_exec() and pcre_dfa_exec().
295    
296    10. The PCRE_DOLLAR_ENDONLY option was not working when pcre_dfa_exec() was
297        used for matching.
298    
299    11. Unicode property support in character classes was not working for
300        characters (bytes) greater than 127 when not in UTF-8 mode.
301    
302    12. Added the -M command line option to pcretest.
303    
304    14. Added the non-standard REG_NOTEMPTY option to the POSIX interface.
305    
306    15. Added the PCRE_NO_START_OPTIMIZE match-time option.
307    
308    16. Added comments and documentation about mis-use of no_arg in the C++
309        wrapper.
310    
311    17. Implemented support for UTF-8 encoding in EBCDIC environments, a patch
312        from Martin Jerabek that uses macro names for all relevant character and
313        string constants.
314    
315    18. Added to pcre_internal.h two configuration checks: (a) If both EBCDIC and
316        SUPPORT_UTF8 are set, give an error; (b) If SUPPORT_UCP is set without
317        SUPPORT_UTF8, define SUPPORT_UTF8. The "configure" script handles both of
318        these, but not everybody uses configure.
319    
320    19. A conditional group that had only one branch was not being correctly
321        recognized as an item that could match an empty string. This meant that an
322        enclosing group might also not be so recognized, causing infinite looping
323        (and probably a segfault) for patterns such as ^"((?(?=[a])[^"])|b)*"$
324        with the subject "ab", where knowledge that the repeated group can match
325        nothing is needed in order to break the loop.
326    
327    20. If a pattern that was compiled with callouts was matched using pcre_dfa_
328        exec(), but without supplying a callout function, matching went wrong.
329    
330    21. If PCRE_ERROR_MATCHLIMIT occurred during a recursion, there was a memory
331        leak if the size of the offset vector was greater than 30. When the vector
332        is smaller, the saved offsets during recursion go onto a local stack
333        vector, but for larger vectors malloc() is used. It was failing to free
334        when the recursion yielded PCRE_ERROR_MATCH_LIMIT (or any other "abnormal"
335        error, in fact).
336    
337    22. There was a missing #ifdef SUPPORT_UTF8 round one of the variables in the
338        heapframe that is used only when UTF-8 support is enabled. This caused no
339        problem, but was untidy.
340    
341    23. Steven Van Ingelgem's patch to CMakeLists.txt to change the name
342        CMAKE_BINARY_DIR to PROJECT_BINARY_DIR so that it works when PCRE is
343        included within another project.
344    
345    24. Steven Van Ingelgem's patches to add more options to the CMake support,
346        slightly modified by me:
347    
348          (a) PCRE_BUILD_TESTS can be set OFF not to build the tests, including
349              not building pcregrep.
350    
351          (b) PCRE_BUILD_PCREGREP can be see OFF not to build pcregrep, but only
352              if PCRE_BUILD_TESTS is also set OFF, because the tests use pcregrep.
353    
354    25. Forward references, both numeric and by name, in patterns that made use of
355        duplicate group numbers, could behave incorrectly or give incorrect errors,
356        because when scanning forward to find the reference group, PCRE was not
357        taking into account the duplicate group numbers. A pattern such as
358        ^X(?3)(a)(?|(b)|(q))(Y) is an example.
359    
360    26. Changed a few more instances of "const unsigned char *" to USPTR, making
361        the feature of a custom pointer more persuasive (as requested by a user).
362    
363    27. Wrapped the definitions of fileno and isatty for Windows, which appear in
364        pcretest.c, inside #ifndefs, because it seems they are sometimes already
365        pre-defined.
366    
367    28. Added support for (*UTF8) at the start of a pattern.
368    
369    29. Arrange for flags added by the "release type" setting in CMake to be shown
370        in the configuration summary.
371    
372    
373    Version 7.8 05-Sep-08
374    ---------------------
375    
376    1.  Replaced UCP searching code with optimized version as implemented for Ad
377        Muncher (http://www.admuncher.com/) by Peter Kankowski. This uses a two-
378        stage table and inline lookup instead of a function, giving speed ups of 2
379        to 5 times on some simple patterns that I tested. Permission was given to
380        distribute the MultiStage2.py script that generates the tables (it's not in
381        the tarball, but is in the Subversion repository).
382    
383    2.  Updated the Unicode datatables to Unicode 5.1.0. This adds yet more
384        scripts.
385    
386    3.  Change 12 for 7.7 introduced a bug in pcre_study() when a pattern contained
387        a group with a zero qualifier. The result of the study could be incorrect,
388        or the function might crash, depending on the pattern.
389    
390    4.  Caseless matching was not working for non-ASCII characters in back
391        references. For example, /(\x{de})\1/8i was not matching \x{de}\x{fe}.
392        It now works when Unicode Property Support is available.
393    
394    5.  In pcretest, an escape such as \x{de} in the data was always generating
395        a UTF-8 string, even in non-UTF-8 mode. Now it generates a single byte in
396        non-UTF-8 mode. If the value is greater than 255, it gives a warning about
397        truncation.
398    
399    6.  Minor bugfix in pcrecpp.cc (change "" == ... to NULL == ...).
400    
401    7.  Added two (int) casts to pcregrep when printing the difference of two
402        pointers, in case they are 64-bit values.
403    
404    8.  Added comments about Mac OS X stack usage to the pcrestack man page and to
405        test 2 if it fails.
406    
407    9.  Added PCRE_CALL_CONVENTION just before the names of all exported functions,
408        and a #define of that name to empty if it is not externally set. This is to
409        allow users of MSVC to set it if necessary.
410    
411    10. The PCRE_EXP_DEFN macro which precedes exported functions was missing from
412        the convenience functions in the pcre_get.c source file.
413    
414    11. An option change at the start of a pattern that had top-level alternatives
415        could cause overwriting and/or a crash. This command provoked a crash in
416        some environments:
417    
418          printf "/(?i)[\xc3\xa9\xc3\xbd]|[\xc3\xa9\xc3\xbdA]/8\n" | pcretest
419    
420        This potential security problem was recorded as CVE-2008-2371.
421    
422    12. For a pattern where the match had to start at the beginning or immediately
423        after a newline (e.g /.*anything/ without the DOTALL flag), pcre_exec() and
424        pcre_dfa_exec() could read past the end of the passed subject if there was
425        no match. To help with detecting such bugs (e.g. with valgrind), I modified
426        pcretest so that it places the subject at the end of its malloc-ed buffer.
427    
428    13. The change to pcretest in 12 above threw up a couple more cases when pcre_
429        exec() might read past the end of the data buffer in UTF-8 mode.
430    
431    14. A similar bug to 7.3/2 existed when the PCRE_FIRSTLINE option was set and
432        the data contained the byte 0x85 as part of a UTF-8 character within its
433        first line. This applied both to normal and DFA matching.
434    
435    15. Lazy qualifiers were not working in some cases in UTF-8 mode. For example,
436        /^[^d]*?$/8 failed to match "abc".
437    
438    16. Added a missing copyright notice to pcrecpp_internal.h.
439    
440    17. Make it more clear in the documentation that values returned from
441        pcre_exec() in ovector are byte offsets, not character counts.
442    
443    18. Tidied a few places to stop certain compilers from issuing warnings.
444    
445    19. Updated the Virtual Pascal + BCC files to compile the latest v7.7, as
446        supplied by Stefan Weber. I made a further small update for 7.8 because
447        there is a change of source arrangements: the pcre_searchfuncs.c module is
448        replaced by pcre_ucd.c.
449    
450    
451    Version 7.7 07-May-08
452    ---------------------
453    
454    1.  Applied Craig's patch to sort out a long long problem: "If we can't convert
455        a string to a long long, pretend we don't even have a long long." This is
456        done by checking for the strtoq, strtoll, and _strtoi64 functions.
457    
458    2.  Applied Craig's patch to pcrecpp.cc to restore ABI compatibility with
459        pre-7.6 versions, which defined a global no_arg variable instead of putting
460        it in the RE class. (See also #8 below.)
461    
462    3.  Remove a line of dead code, identified by coverity and reported by Nuno
463        Lopes.
464    
465    4.  Fixed two related pcregrep bugs involving -r with --include or --exclude:
466    
467        (1) The include/exclude patterns were being applied to the whole pathnames
468            of files, instead of just to the final components.
469    
470        (2) If there was more than one level of directory, the subdirectories were
471            skipped unless they satisfied the include/exclude conditions. This is
472            inconsistent with GNU grep (and could even be seen as contrary to the
473            pcregrep specification - which I improved to make it absolutely clear).
474            The action now is always to scan all levels of directory, and just
475            apply the include/exclude patterns to regular files.
476    
477    5.  Added the --include_dir and --exclude_dir patterns to pcregrep, and used
478        --exclude_dir in the tests to avoid scanning .svn directories.
479    
480    6.  Applied Craig's patch to the QuoteMeta function so that it escapes the
481        NUL character as backslash + 0 rather than backslash + NUL, because PCRE
482        doesn't support NULs in patterns.
483    
484    7.  Added some missing "const"s to declarations of static tables in
485        pcre_compile.c and pcre_dfa_exec.c.
486    
487    8.  Applied Craig's patch to pcrecpp.cc to fix a problem in OS X that was
488        caused by fix #2  above. (Subsequently also a second patch to fix the
489        first patch. And a third patch - this was a messy problem.)
490    
491    9.  Applied Craig's patch to remove the use of push_back().
492    
493    10. Applied Alan Lehotsky's patch to add REG_STARTEND support to the POSIX
494        matching function regexec().
495    
496    11. Added support for the Oniguruma syntax \g<name>, \g<n>, \g'name', \g'n',
497        which, however, unlike Perl's \g{...}, are subroutine calls, not back
498        references. PCRE supports relative numbers with this syntax (I don't think
499        Oniguruma does).
500    
501    12. Previously, a group with a zero repeat such as (...){0} was completely
502        omitted from the compiled regex. However, this means that if the group
503        was called as a subroutine from elsewhere in the pattern, things went wrong
504        (an internal error was given). Such groups are now left in the compiled
505        pattern, with a new opcode that causes them to be skipped at execution
506        time.
507    
508    13. Added the PCRE_JAVASCRIPT_COMPAT option. This makes the following changes
509        to the way PCRE behaves:
510    
511        (a) A lone ] character is dis-allowed (Perl treats it as data).
512    
513        (b) A back reference to an unmatched subpattern matches an empty string
514            (Perl fails the current match path).
515    
516        (c) A data ] in a character class must be notated as \] because if the
517            first data character in a class is ], it defines an empty class. (In
518            Perl it is not possible to have an empty class.) The empty class []
519            never matches; it forces failure and is equivalent to (*FAIL) or (?!).
520            The negative empty class [^] matches any one character, independently
521            of the DOTALL setting.
522    
523    14. A pattern such as /(?2)[]a()b](abc)/ which had a forward reference to a
524        non-existent subpattern following a character class starting with ']' and
525        containing () gave an internal compiling error instead of "reference to
526        non-existent subpattern". Fortunately, when the pattern did exist, the
527        compiled code was correct. (When scanning forwards to check for the
528        existencd of the subpattern, it was treating the data ']' as terminating
529        the class, so got the count wrong. When actually compiling, the reference
530        was subsequently set up correctly.)
531    
532    15. The "always fail" assertion (?!) is optimzed to (*FAIL) by pcre_compile;
533        it was being rejected as not supported by pcre_dfa_exec(), even though
534        other assertions are supported. I have made pcre_dfa_exec() support
535        (*FAIL).
536    
537    16. The implementation of 13c above involved the invention of a new opcode,
538        OP_ALLANY, which is like OP_ANY but doesn't check the /s flag. Since /s
539        cannot be changed at match time, I realized I could make a small
540        improvement to matching performance by compiling OP_ALLANY instead of
541        OP_ANY for "." when DOTALL was set, and then removing the runtime tests
542        on the OP_ANY path.
543    
544    17. Compiling pcretest on Windows with readline support failed without the
545        following two fixes: (1) Make the unistd.h include conditional on
546        HAVE_UNISTD_H; (2) #define isatty and fileno as _isatty and _fileno.
547    
548    18. Changed CMakeLists.txt and cmake/FindReadline.cmake to arrange for the
549        ncurses library to be included for pcretest when ReadLine support is
550        requested, but also to allow for it to be overridden. This patch came from
551        Daniel Bergström.
552    
553    19. There was a typo in the file ucpinternal.h where f0_rangeflag was defined
554        as 0x00f00000 instead of 0x00800000. Luckily, this would not have caused
555        any errors with the current Unicode tables. Thanks to Peter Kankowski for
556        spotting this.
557    
558    
559    Version 7.6 28-Jan-08
560    ---------------------
561    
562    1.  A character class containing a very large number of characters with
563        codepoints greater than 255 (in UTF-8 mode, of course) caused a buffer
564        overflow.
565    
566    2.  Patch to cut out the "long long" test in pcrecpp_unittest when
567        HAVE_LONG_LONG is not defined.
568    
569    3.  Applied Christian Ehrlicher's patch to update the CMake build files to
570        bring them up to date and include new features. This patch includes:
571    
572        - Fixed PH's badly added libz and libbz2 support.
573        - Fixed a problem with static linking.
574        - Added pcredemo. [But later removed - see 7 below.]
575        - Fixed dftables problem and added an option.
576        - Added a number of HAVE_XXX tests, including HAVE_WINDOWS_H and
577            HAVE_LONG_LONG.
578        - Added readline support for pcretest.
579        - Added an listing of the option settings after cmake has run.
580    
581    4.  A user submitted a patch to Makefile that makes it easy to create
582        "pcre.dll" under mingw when using Configure/Make. I added stuff to
583        Makefile.am that cause it to include this special target, without
584        affecting anything else. Note that the same mingw target plus all
585        the other distribution libraries and programs are now supported
586        when configuring with CMake (see 6 below) instead of with
587        Configure/Make.
588    
589    5.  Applied Craig's patch that moves no_arg into the RE class in the C++ code.
590        This is an attempt to solve the reported problem "pcrecpp::no_arg is not
591        exported in the Windows port". It has not yet been confirmed that the patch
592        solves the problem, but it does no harm.
593    
594    6.  Applied Sheri's patch to CMakeLists.txt to add NON_STANDARD_LIB_PREFIX and
595        NON_STANDARD_LIB_SUFFIX for dll names built with mingw when configured
596        with CMake, and also correct the comment about stack recursion.
597    
598    7.  Remove the automatic building of pcredemo from the ./configure system and
599        from CMakeLists.txt. The whole idea of pcredemo.c is that it is an example
600        of a program that users should build themselves after PCRE is installed, so
601        building it automatically is not really right. What is more, it gave
602        trouble in some build environments.
603    
604    8.  Further tidies to CMakeLists.txt from Sheri and Christian.
605    
606    
607    Version 7.5 10-Jan-08
608    ---------------------
609    
610    1.  Applied a patch from Craig: "This patch makes it possible to 'ignore'
611        values in parens when parsing an RE using the C++ wrapper."
612    
613    2.  Negative specials like \S did not work in character classes in UTF-8 mode.
614        Characters greater than 255 were excluded from the class instead of being
615        included.
616    
617    3.  The same bug as (2) above applied to negated POSIX classes such as
618        [:^space:].
619    
620    4.  PCRECPP_STATIC was referenced in pcrecpp_internal.h, but nowhere was it
621        defined or documented. It seems to have been a typo for PCRE_STATIC, so
622        I have changed it.
623    
624    5.  The construct (?&) was not diagnosed as a syntax error (it referenced the
625        first named subpattern) and a construct such as (?&a) would reference the
626        first named subpattern whose name started with "a" (in other words, the
627        length check was missing). Both these problems are fixed. "Subpattern name
628        expected" is now given for (?&) (a zero-length name), and this patch also
629        makes it give the same error for \k'' (previously it complained that that
630        was a reference to a non-existent subpattern).
631    
632    6.  The erroneous patterns (?+-a) and (?-+a) give different error messages;
633        this is right because (?- can be followed by option settings as well as by
634        digits. I have, however, made the messages clearer.
635    
636    7.  Patterns such as (?(1)a|b) (a pattern that contains fewer subpatterns
637        than the number used in the conditional) now cause a compile-time error.
638        This is actually not compatible with Perl, which accepts such patterns, but
639        treats the conditional as always being FALSE (as PCRE used to), but it
640        seems to me that giving a diagnostic is better.
641    
642    8.  Change "alphameric" to the more common word "alphanumeric" in comments
643        and messages.
644    
645    9.  Fix two occurrences of "backslash" in comments that should have been
646        "backspace".
647    
648    10. Remove two redundant lines of code that can never be obeyed (their function
649        was moved elsewhere).
650    
651    11. The program that makes PCRE's Unicode character property table had a bug
652        which caused it to generate incorrect table entries for sequences of
653        characters that have the same character type, but are in different scripts.
654        It amalgamated them into a single range, with the script of the first of
655        them. In other words, some characters were in the wrong script. There were
656        thirteen such cases, affecting characters in the following ranges:
657    
658          U+002b0 - U+002c1
659          U+0060c - U+0060d
660          U+0061e - U+00612
661          U+0064b - U+0065e
662          U+0074d - U+0076d
663          U+01800 - U+01805
664          U+01d00 - U+01d77
665          U+01d9b - U+01dbf
666          U+0200b - U+0200f
667          U+030fc - U+030fe
668          U+03260 - U+0327f
669          U+0fb46 - U+0fbb1
670          U+10450 - U+1049d
671    
672    12. The -o option (show only the matching part of a line) for pcregrep was not
673        compatible with GNU grep in that, if there was more than one match in a
674        line, it showed only the first of them. It now behaves in the same way as
675        GNU grep.
676    
677    13. If the -o and -v options were combined for pcregrep, it printed a blank
678        line for every non-matching line. GNU grep prints nothing, and pcregrep now
679        does the same. The return code can be used to tell if there were any
680        non-matching lines.
681    
682    14. Added --file-offsets and --line-offsets to pcregrep.
683    
684    15. The pattern (?=something)(?R) was not being diagnosed as a potentially
685        infinitely looping recursion. The bug was that positive lookaheads were not
686        being skipped when checking for a possible empty match (negative lookaheads
687        and both kinds of lookbehind were skipped).
688    
689    16. Fixed two typos in the Windows-only code in pcregrep.c, and moved the
690        inclusion of <windows.h> to before rather than after the definition of
691        INVALID_FILE_ATTRIBUTES (patch from David Byron).
692    
693    17. Specifying a possessive quantifier with a specific limit for a Unicode
694        character property caused pcre_compile() to compile bad code, which led at
695        runtime to PCRE_ERROR_INTERNAL (-14). Examples of patterns that caused this
696        are: /\p{Zl}{2,3}+/8 and /\p{Cc}{2}+/8. It was the possessive "+" that
697        caused the error; without that there was no problem.
698    
699    18. Added --enable-pcregrep-libz and --enable-pcregrep-libbz2.
700    
701    19. Added --enable-pcretest-libreadline.
702    
703    20. In pcrecpp.cc, the variable 'count' was incremented twice in
704        RE::GlobalReplace(). As a result, the number of replacements returned was
705        double what it should be. I removed one of the increments, but Craig sent a
706        later patch that removed the other one (the right fix) and added unit tests
707        that check the return values (which was not done before).
708    
709    21. Several CMake things:
710    
711        (1) Arranged that, when cmake is used on Unix, the libraries end up with
712            the names libpcre and libpcreposix, not just pcre and pcreposix.
713    
714        (2) The above change means that pcretest and pcregrep are now correctly
715            linked with the newly-built libraries, not previously installed ones.
716    
717        (3) Added PCRE_SUPPORT_LIBREADLINE, PCRE_SUPPORT_LIBZ, PCRE_SUPPORT_LIBBZ2.
718    
719    22. In UTF-8 mode, with newline set to "any", a pattern such as .*a.*=.b.*
720        crashed when matching a string such as a\x{2029}b (note that \x{2029} is a
721        UTF-8 newline character). The key issue is that the pattern starts .*;
722        this means that the match must be either at the beginning, or after a
723        newline. The bug was in the code for advancing after a failed match and
724        checking that the new position followed a newline. It was not taking
725        account of UTF-8 characters correctly.
726    
727    23. PCRE was behaving differently from Perl in the way it recognized POSIX
728        character classes. PCRE was not treating the sequence [:...:] as a
729        character class unless the ... were all letters. Perl, however, seems to
730        allow any characters between [: and :], though of course it rejects as
731        unknown any "names" that contain non-letters, because all the known class
732        names consist only of letters. Thus, Perl gives an error for [[:1234:]],
733        for example, whereas PCRE did not - it did not recognize a POSIX character
734        class. This seemed a bit dangerous, so the code has been changed to be
735        closer to Perl. The behaviour is not identical to Perl, because PCRE will
736        diagnose an unknown class for, for example, [[:l\ower:]] where Perl will
737        treat it as [[:lower:]]. However, PCRE does now give "unknown" errors where
738        Perl does, and where it didn't before.
739    
740    24. Rewrite so as to remove the single use of %n from pcregrep because in some
741        Windows environments %n is disabled by default.
742    
743    
744    Version 7.4 21-Sep-07
745    ---------------------
746    
747    1.  Change 7.3/28 was implemented for classes by looking at the bitmap. This
748        means that a class such as [\s] counted as "explicit reference to CR or
749        LF". That isn't really right - the whole point of the change was to try to
750        help when there was an actual mention of one of the two characters. So now
751        the change happens only if \r or \n (or a literal CR or LF) character is
752        encountered.
753    
754    2.  The 32-bit options word was also used for 6 internal flags, but the numbers
755        of both had grown to the point where there were only 3 bits left.
756        Fortunately, there was spare space in the data structure, and so I have
757        moved the internal flags into a new 16-bit field to free up more option
758        bits.
759    
760    3.  The appearance of (?J) at the start of a pattern set the DUPNAMES option,
761        but did not set the internal JCHANGED flag - either of these is enough to
762        control the way the "get" function works - but the PCRE_INFO_JCHANGED
763        facility is supposed to tell if (?J) was ever used, so now (?J) at the
764        start sets both bits.
765    
766    4.  Added options (at build time, compile time, exec time) to change \R from
767        matching any Unicode line ending sequence to just matching CR, LF, or CRLF.
768    
769    5.  doc/pcresyntax.html was missing from the distribution.
770    
771    6.  Put back the definition of PCRE_ERROR_NULLWSLIMIT, for backward
772        compatibility, even though it is no longer used.
773    
774    7.  Added macro for snprintf to pcrecpp_unittest.cc and also for strtoll and
775        strtoull to pcrecpp.cc to select the available functions in WIN32 when the
776        windows.h file is present (where different names are used). [This was
777        reversed later after testing - see 16 below.]
778    
779    8.  Changed all #include <config.h> to #include "config.h". There were also
780        some further <pcre.h> cases that I changed to "pcre.h".
781    
782    9.  When pcregrep was used with the --colour option, it missed the line ending
783        sequence off the lines that it output.
784    
785    10. It was pointed out to me that arrays of string pointers cause lots of
786        relocations when a shared library is dynamically loaded. A technique of
787        using a single long string with a table of offsets can drastically reduce
788        these. I have refactored PCRE in four places to do this. The result is
789        dramatic:
790    
791          Originally:                          290
792          After changing UCP table:            187
793          After changing error message table:   43
794          After changing table of "verbs"       36
795          After changing table of Posix names   22
796    
797        Thanks to the folks working on Gregex for glib for this insight.
798    
799    11. --disable-stack-for-recursion caused compiling to fail unless -enable-
800        unicode-properties was also set.
801    
802    12. Updated the tests so that they work when \R is defaulted to ANYCRLF.
803    
804    13. Added checks for ANY and ANYCRLF to pcrecpp.cc where it previously
805        checked only for CRLF.
806    
807    14. Added casts to pcretest.c to avoid compiler warnings.
808    
809    15. Added Craig's patch to various pcrecpp modules to avoid compiler warnings.
810    
811    16. Added Craig's patch to remove the WINDOWS_H tests, that were not working,
812        and instead check for _strtoi64 explicitly, and avoid the use of snprintf()
813        entirely. This removes changes made in 7 above.
814    
815    17. The CMake files have been updated, and there is now more information about
816        building with CMake in the NON-UNIX-USE document.
817    
818    
819    Version 7.3 28-Aug-07
820    ---------------------
821    
822     1. In the rejigging of the build system that eventually resulted in 7.1, the
823        line "#include <pcre.h>" was included in pcre_internal.h. The use of angle
824        brackets there is not right, since it causes compilers to look for an
825        installed pcre.h, not the version that is in the source that is being
826        compiled (which of course may be different). I have changed it back to:
827    
828          #include "pcre.h"
829    
830        I have a vague recollection that the change was concerned with compiling in
831        different directories, but in the new build system, that is taken care of
832        by the VPATH setting the Makefile.
833    
834     2. The pattern .*$ when run in not-DOTALL UTF-8 mode with newline=any failed
835        when the subject happened to end in the byte 0x85 (e.g. if the last
836        character was \x{1ec5}). *Character* 0x85 is one of the "any" newline
837        characters but of course it shouldn't be taken as a newline when it is part
838        of another character. The bug was that, for an unlimited repeat of . in
839        not-DOTALL UTF-8 mode, PCRE was advancing by bytes rather than by
840        characters when looking for a newline.
841    
842     3. A small performance improvement in the DOTALL UTF-8 mode .* case.
843    
844     4. Debugging: adjusted the names of opcodes for different kinds of parentheses
845        in debug output.
846    
847     5. Arrange to use "%I64d" instead of "%lld" and "%I64u" instead of "%llu" for
848        long printing in the pcrecpp unittest when running under MinGW.
849    
850     6. ESC_K was left out of the EBCDIC table.
851    
852     7. Change 7.0/38 introduced a new limit on the number of nested non-capturing
853        parentheses; I made it 1000, which seemed large enough. Unfortunately, the
854        limit also applies to "virtual nesting" when a pattern is recursive, and in
855        this case 1000 isn't so big. I have been able to remove this limit at the
856        expense of backing off one optimization in certain circumstances. Normally,
857        when pcre_exec() would call its internal match() function recursively and
858        immediately return the result unconditionally, it uses a "tail recursion"
859        feature to save stack. However, when a subpattern that can match an empty
860        string has an unlimited repetition quantifier, it no longer makes this
861        optimization. That gives it a stack frame in which to save the data for
862        checking that an empty string has been matched. Previously this was taken
863        from the 1000-entry workspace that had been reserved. So now there is no
864        explicit limit, but more stack is used.
865    
866     8. Applied Daniel's patches to solve problems with the import/export magic
867        syntax that is required for Windows, and which was going wrong for the
868        pcreposix and pcrecpp parts of the library. These were overlooked when this
869        problem was solved for the main library.
870    
871     9. There were some crude static tests to avoid integer overflow when computing
872        the size of patterns that contain repeated groups with explicit upper
873        limits. As the maximum quantifier is 65535, the maximum group length was
874        set at 30,000 so that the product of these two numbers did not overflow a
875        32-bit integer. However, it turns out that people want to use groups that
876        are longer than 30,000 bytes (though not repeat them that many times).
877        Change 7.0/17 (the refactoring of the way the pattern size is computed) has
878        made it possible to implement the integer overflow checks in a much more
879        dynamic way, which I have now done. The artificial limitation on group
880        length has been removed - we now have only the limit on the total length of
881        the compiled pattern, which depends on the LINK_SIZE setting.
882    
883    10. Fixed a bug in the documentation for get/copy named substring when
884        duplicate names are permitted. If none of the named substrings are set, the
885        functions return PCRE_ERROR_NOSUBSTRING (7); the doc said they returned an
886        empty string.
887    
888    11. Because Perl interprets \Q...\E at a high level, and ignores orphan \E
889        instances, patterns such as [\Q\E] or [\E] or even [^\E] cause an error,
890        because the ] is interpreted as the first data character and the
891        terminating ] is not found. PCRE has been made compatible with Perl in this
892        regard. Previously, it interpreted [\Q\E] as an empty class, and [\E] could
893        cause memory overwriting.
894    
895    10. Like Perl, PCRE automatically breaks an unlimited repeat after an empty
896        string has been matched (to stop an infinite loop). It was not recognizing
897        a conditional subpattern that could match an empty string if that
898        subpattern was within another subpattern. For example, it looped when
899        trying to match  (((?(1)X|))*)  but it was OK with  ((?(1)X|)*)  where the
900        condition was not nested. This bug has been fixed.
901    
902    12. A pattern like \X?\d or \P{L}?\d in non-UTF-8 mode could cause a backtrack
903        past the start of the subject in the presence of bytes with the top bit
904        set, for example "\x8aBCD".
905    
906    13. Added Perl 5.10 experimental backtracking controls (*FAIL), (*F), (*PRUNE),
907        (*SKIP), (*THEN), (*COMMIT), and (*ACCEPT).
908    
909    14. Optimized (?!) to (*FAIL).
910    
911    15. Updated the test for a valid UTF-8 string to conform to the later RFC 3629.
912        This restricts code points to be within the range 0 to 0x10FFFF, excluding
913        the "low surrogate" sequence 0xD800 to 0xDFFF. Previously, PCRE allowed the
914        full range 0 to 0x7FFFFFFF, as defined by RFC 2279. Internally, it still
915        does: it's just the validity check that is more restrictive.
916    
917    16. Inserted checks for integer overflows during escape sequence (backslash)
918        processing, and also fixed erroneous offset values for syntax errors during
919        backslash processing.
920    
921    17. Fixed another case of looking too far back in non-UTF-8 mode (cf 12 above)
922        for patterns like [\PPP\x8a]{1,}\x80 with the subject "A\x80".
923    
924    18. An unterminated class in a pattern like (?1)\c[ with a "forward reference"
925        caused an overrun.
926    
927    19. A pattern like (?:[\PPa*]*){8,} which had an "extended class" (one with
928        something other than just ASCII characters) inside a group that had an
929        unlimited repeat caused a loop at compile time (while checking to see
930        whether the group could match an empty string).
931    
932    20. Debugging a pattern containing \p or \P could cause a crash. For example,
933        [\P{Any}] did so. (Error in the code for printing property names.)
934    
935    21. An orphan \E inside a character class could cause a crash.
936    
937    22. A repeated capturing bracket such as (A)? could cause a wild memory
938        reference during compilation.
939    
940    23. There are several functions in pcre_compile() that scan along a compiled
941        expression for various reasons (e.g. to see if it's fixed length for look
942        behind). There were bugs in these functions when a repeated \p or \P was
943        present in the pattern. These operators have additional parameters compared
944        with \d, etc, and these were not being taken into account when moving along
945        the compiled data. Specifically:
946    
947        (a) A item such as \p{Yi}{3} in a lookbehind was not treated as fixed
948            length.
949    
950        (b) An item such as \pL+ within a repeated group could cause crashes or
951            loops.
952    
953        (c) A pattern such as \p{Yi}+(\P{Yi}+)(?1) could give an incorrect
954            "reference to non-existent subpattern" error.
955    
956        (d) A pattern like (\P{Yi}{2}\277)? could loop at compile time.
957    
958    24. A repeated \S or \W in UTF-8 mode could give wrong answers when multibyte
959        characters were involved (for example /\S{2}/8g with "A\x{a3}BC").
960    
961    25. Using pcregrep in multiline, inverted mode (-Mv) caused it to loop.
962    
963    26. Patterns such as [\P{Yi}A] which include \p or \P and just one other
964        character were causing crashes (broken optimization).
965    
966    27. Patterns such as (\P{Yi}*\277)* (group with possible zero repeat containing
967        \p or \P) caused a compile-time loop.
968    
969    28. More problems have arisen in unanchored patterns when CRLF is a valid line
970        break. For example, the unstudied pattern [\r\n]A does not match the string
971        "\r\nA" because change 7.0/46 below moves the current point on by two
972        characters after failing to match at the start. However, the pattern \nA
973        *does* match, because it doesn't start till \n, and if [\r\n]A is studied,
974        the same is true. There doesn't seem any very clean way out of this, but
975        what I have chosen to do makes the common cases work: PCRE now takes note
976        of whether there can be an explicit match for \r or \n anywhere in the
977        pattern, and if so, 7.0/46 no longer applies. As part of this change,
978        there's a new PCRE_INFO_HASCRORLF option for finding out whether a compiled
979        pattern has explicit CR or LF references.
980    
981    29. Added (*CR) etc for changing newline setting at start of pattern.
982    
983    
984    Version 7.2 19-Jun-07
985    ---------------------
986    
987     1. If the fr_FR locale cannot be found for test 3, try the "french" locale,
988        which is apparently normally available under Windows.
989    
990     2. Re-jig the pcregrep tests with different newline settings in an attempt
991        to make them independent of the local environment's newline setting.
992    
993     3. Add code to configure.ac to remove -g from the CFLAGS default settings.
994    
995     4. Some of the "internals" tests were previously cut out when the link size
996        was not 2, because the output contained actual offsets. The recent new
997        "Z" feature of pcretest means that these can be cut out, making the tests
998        usable with all link sizes.
999    
1000     5. Implemented Stan Switzer's goto replacement for longjmp() when not using
1001        stack recursion. This gives a massive performance boost under BSD, but just
1002        a small improvement under Linux. However, it saves one field in the frame
1003        in all cases.
1004    
1005     6. Added more features from the forthcoming Perl 5.10:
1006    
1007        (a) (?-n) (where n is a string of digits) is a relative subroutine or
1008            recursion call. It refers to the nth most recently opened parentheses.
1009    
1010        (b) (?+n) is also a relative subroutine call; it refers to the nth next
1011            to be opened parentheses.
1012    
1013        (c) Conditions that refer to capturing parentheses can be specified
1014            relatively, for example, (?(-2)... or (?(+3)...
1015    
1016        (d) \K resets the start of the current match so that everything before
1017            is not part of it.
1018    
1019        (e) \k{name} is synonymous with \k<name> and \k'name' (.NET compatible).
1020    
1021        (f) \g{name} is another synonym - part of Perl 5.10's unification of
1022            reference syntax.
1023    
1024        (g) (?| introduces a group in which the numbering of parentheses in each
1025            alternative starts with the same number.
1026    
1027        (h) \h, \H, \v, and \V match horizontal and vertical whitespace.
1028    
1029     7. Added two new calls to pcre_fullinfo(): PCRE_INFO_OKPARTIAL and
1030        PCRE_INFO_JCHANGED.
1031    
1032     8. A pattern such as  (.*(.)?)*  caused pcre_exec() to fail by either not
1033        terminating or by crashing. Diagnosed by Viktor Griph; it was in the code
1034        for detecting groups that can match an empty string.
1035    
1036     9. A pattern with a very large number of alternatives (more than several
1037        hundred) was running out of internal workspace during the pre-compile
1038        phase, where pcre_compile() figures out how much memory will be needed. A
1039        bit of new cunning has reduced the workspace needed for groups with
1040        alternatives. The 1000-alternative test pattern now uses 12 bytes of
1041        workspace instead of running out of the 4096 that are available.
1042    
1043    10. Inserted some missing (unsigned int) casts to get rid of compiler warnings.
1044    
1045    11. Applied patch from Google to remove an optimization that didn't quite work.
1046        The report of the bug said:
1047    
1048          pcrecpp::RE("a*").FullMatch("aaa") matches, while
1049          pcrecpp::RE("a*?").FullMatch("aaa") does not, and
1050          pcrecpp::RE("a*?\\z").FullMatch("aaa") does again.
1051    
1052    12. If \p or \P was used in non-UTF-8 mode on a character greater than 127
1053        it matched the wrong number of bytes.
1054    
1055    
1056    Version 7.1 24-Apr-07
1057    ---------------------
1058    
1059     1. Applied Bob Rossi and Daniel G's patches to convert the build system to one
1060        that is more "standard", making use of automake and other Autotools. There
1061        is some re-arrangement of the files and adjustment of comments consequent
1062        on this.
1063    
1064     2. Part of the patch fixed a problem with the pcregrep tests. The test of -r
1065        for recursive directory scanning broke on some systems because the files
1066        are not scanned in any specific order and on different systems the order
1067        was different. A call to "sort" has been inserted into RunGrepTest for the
1068        approprate test as a short-term fix. In the longer term there may be an
1069        alternative.
1070    
1071     3. I had an email from Eric Raymond about problems translating some of PCRE's
1072        man pages to HTML (despite the fact that I distribute HTML pages, some
1073        people do their own conversions for various reasons). The problems
1074        concerned the use of low-level troff macros .br and .in. I have therefore
1075        removed all such uses from the man pages (some were redundant, some could
1076        be replaced by .nf/.fi pairs). The 132html script that I use to generate
1077        HTML has been updated to handle .nf/.fi and to complain if it encounters
1078        .br or .in.
1079    
1080     4. Updated comments in configure.ac that get placed in config.h.in and also
1081        arranged for config.h to be included in the distribution, with the name
1082        config.h.generic, for the benefit of those who have to compile without
1083        Autotools (compare pcre.h, which is now distributed as pcre.h.generic).
1084    
1085     5. Updated the support (such as it is) for Virtual Pascal, thanks to Stefan
1086        Weber: (1) pcre_internal.h was missing some function renames; (2) updated
1087        makevp.bat for the current PCRE, using the additional files
1088        makevp_c.txt, makevp_l.txt, and pcregexp.pas.
1089    
1090     6. A Windows user reported a minor discrepancy with test 2, which turned out
1091        to be caused by a trailing space on an input line that had got lost in his
1092        copy. The trailing space was an accident, so I've just removed it.
1093    
1094     7. Add -Wl,-R... flags in pcre-config.in for *BSD* systems, as I'm told
1095        that is needed.
1096    
1097     8. Mark ucp_table (in ucptable.h) and ucp_gentype (in pcre_ucp_searchfuncs.c)
1098        as "const" (a) because they are and (b) because it helps the PHP
1099        maintainers who have recently made a script to detect big data structures
1100        in the php code that should be moved to the .rodata section. I remembered
1101        to update Builducptable as well, so it won't revert if ucptable.h is ever
1102        re-created.
1103    
1104     9. Added some extra #ifdef SUPPORT_UTF8 conditionals into pcretest.c,
1105        pcre_printint.src, pcre_compile.c, pcre_study.c, and pcre_tables.c, in
1106        order to be able to cut out the UTF-8 tables in the latter when UTF-8
1107        support is not required. This saves 1.5-2K of code, which is important in
1108        some applications.
1109    
1110        Later: more #ifdefs are needed in pcre_ord2utf8.c and pcre_valid_utf8.c
1111        so as not to refer to the tables, even though these functions will never be
1112        called when UTF-8 support is disabled. Otherwise there are problems with a
1113        shared library.
1114    
1115    10. Fixed two bugs in the emulated memmove() function in pcre_internal.h:
1116    
1117        (a) It was defining its arguments as char * instead of void *.
1118    
1119        (b) It was assuming that all moves were upwards in memory; this was true
1120            a long time ago when I wrote it, but is no longer the case.
1121    
1122        The emulated memove() is provided for those environments that have neither
1123        memmove() nor bcopy(). I didn't think anyone used it these days, but that
1124        is clearly not the case, as these two bugs were recently reported.
1125    
1126    11. The script PrepareRelease is now distributed: it calls 132html, CleanTxt,
1127        and Detrail to create the HTML documentation, the .txt form of the man
1128        pages, and it removes trailing spaces from listed files. It also creates
1129        pcre.h.generic and config.h.generic from pcre.h and config.h. In the latter
1130        case, it wraps all the #defines with #ifndefs. This script should be run
1131        before "make dist".
1132    
1133    12. Fixed two fairly obscure bugs concerned with quantified caseless matching
1134        with Unicode property support.
1135    
1136        (a) For a maximizing quantifier, if the two different cases of the
1137            character were of different lengths in their UTF-8 codings (there are
1138            some cases like this - I found 11), and the matching function had to
1139            back up over a mixture of the two cases, it incorrectly assumed they
1140            were both the same length.
1141    
1142        (b) When PCRE was configured to use the heap rather than the stack for
1143            recursion during matching, it was not correctly preserving the data for
1144            the other case of a UTF-8 character when checking ahead for a match
1145            while processing a minimizing repeat. If the check also involved
1146            matching a wide character, but failed, corruption could cause an
1147            erroneous result when trying to check for a repeat of the original
1148            character.
1149    
1150    13. Some tidying changes to the testing mechanism:
1151    
1152        (a) The RunTest script now detects the internal link size and whether there
1153            is UTF-8 and UCP support by running ./pcretest -C instead of relying on
1154            values substituted by "configure". (The RunGrepTest script already did
1155            this for UTF-8.) The configure.ac script no longer substitutes the
1156            relevant variables.
1157    
1158        (b) The debugging options /B and /D in pcretest show the compiled bytecode
1159            with length and offset values. This means that the output is different
1160            for different internal link sizes. Test 2 is skipped for link sizes
1161            other than 2 because of this, bypassing the problem. Unfortunately,
1162            there was also a test in test 3 (the locale tests) that used /B and
1163            failed for link sizes other than 2. Rather than cut the whole test out,
1164            I have added a new /Z option to pcretest that replaces the length and
1165            offset values with spaces. This is now used to make test 3 independent
1166            of link size. (Test 2 will be tidied up later.)
1167    
1168    14. If erroroffset was passed as NULL to pcre_compile, it provoked a
1169        segmentation fault instead of returning the appropriate error message.
1170    
1171    15. In multiline mode when the newline sequence was set to "any", the pattern
1172        ^$ would give a match between the \r and \n of a subject such as "A\r\nB".
1173        This doesn't seem right; it now treats the CRLF combination as the line
1174        ending, and so does not match in that case. It's only a pattern such as ^$
1175        that would hit this one: something like ^ABC$ would have failed after \r
1176        and then tried again after \r\n.
1177    
1178    16. Changed the comparison command for RunGrepTest from "diff -u" to "diff -ub"
1179        in an attempt to make files that differ only in their line terminators
1180        compare equal. This works on Linux.
1181    
1182    17. Under certain error circumstances pcregrep might try to free random memory
1183        as it exited. This is now fixed, thanks to valgrind.
1184    
1185    19. In pcretest, if the pattern /(?m)^$/g<any> was matched against the string
1186        "abc\r\n\r\n", it found an unwanted second match after the second \r. This
1187        was because its rules for how to advance for /g after matching an empty
1188        string at the end of a line did not allow for this case. They now check for
1189        it specially.
1190    
1191    20. pcretest is supposed to handle patterns and data of any length, by
1192        extending its buffers when necessary. It was getting this wrong when the
1193        buffer for a data line had to be extended.
1194    
1195    21. Added PCRE_NEWLINE_ANYCRLF which is like ANY, but matches only CR, LF, or
1196        CRLF as a newline sequence.
1197    
1198    22. Code for handling Unicode properties in pcre_dfa_exec() wasn't being cut
1199        out by #ifdef SUPPORT_UCP. This did no harm, as it could never be used, but
1200        I have nevertheless tidied it up.
1201    
1202    23. Added some casts to kill warnings from HP-UX ia64 compiler.
1203    
1204    24. Added a man page for pcre-config.
1205    
1206    
1207    Version 7.0 19-Dec-06
1208    ---------------------
1209    
1210     1. Fixed a signed/unsigned compiler warning in pcre_compile.c, shown up by
1211        moving to gcc 4.1.1.
1212    
1213     2. The -S option for pcretest uses setrlimit(); I had omitted to #include
1214        sys/time.h, which is documented as needed for this function. It doesn't
1215        seem to matter on Linux, but it showed up on some releases of OS X.
1216    
1217     3. It seems that there are systems where bytes whose values are greater than
1218        127 match isprint() in the "C" locale. The "C" locale should be the
1219        default when a C program starts up. In most systems, only ASCII printing
1220        characters match isprint(). This difference caused the output from pcretest
1221        to vary, making some of the tests fail. I have changed pcretest so that:
1222    
1223        (a) When it is outputting text in the compiled version of a pattern, bytes
1224            other than 32-126 are always shown as hex escapes.
1225    
1226        (b) When it is outputting text that is a matched part of a subject string,
1227            it does the same, unless a different locale has been set for the match
1228            (using the /L modifier). In this case, it uses isprint() to decide.
1229    
1230     4. Fixed a major bug that caused incorrect computation of the amount of memory
1231        required for a compiled pattern when options that changed within the
1232        pattern affected the logic of the preliminary scan that determines the
1233        length. The relevant options are -x, and -i in UTF-8 mode. The result was
1234        that the computed length was too small. The symptoms of this bug were
1235        either the PCRE error "internal error: code overflow" from pcre_compile(),
1236        or a glibc crash with a message such as "pcretest: free(): invalid next
1237        size (fast)". Examples of patterns that provoked this bug (shown in
1238        pcretest format) are:
1239    
1240          /(?-x: )/x
1241          /(?x)(?-x: \s*#\s*)/
1242          /((?i)[\x{c0}])/8
1243          /(?i:[\x{c0}])/8
1244    
1245        HOWEVER: Change 17 below makes this fix obsolete as the memory computation
1246        is now done differently.
1247    
1248     5. Applied patches from Google to: (a) add a QuoteMeta function to the C++
1249        wrapper classes; (b) implement a new function in the C++ scanner that is
1250        more efficient than the old way of doing things because it avoids levels of
1251        recursion in the regex matching; (c) add a paragraph to the documentation
1252        for the FullMatch() function.
1253    
1254     6. The escape sequence \n was being treated as whatever was defined as
1255        "newline". Not only was this contrary to the documentation, which states
1256        that \n is character 10 (hex 0A), but it also went horribly wrong when
1257        "newline" was defined as CRLF. This has been fixed.
1258    
1259     7. In pcre_dfa_exec.c the value of an unsigned integer (the variable called c)
1260        was being set to -1 for the "end of line" case (supposedly a value that no
1261        character can have). Though this value is never used (the check for end of
1262        line is "zero bytes in current character"), it caused compiler complaints.
1263        I've changed it to 0xffffffff.
1264    
1265     8. In pcre_version.c, the version string was being built by a sequence of
1266        C macros that, in the event of PCRE_PRERELEASE being defined as an empty
1267        string (as it is for production releases) called a macro with an empty
1268        argument. The C standard says the result of this is undefined. The gcc
1269        compiler treats it as an empty string (which was what was wanted) but it is
1270        reported that Visual C gives an error. The source has been hacked around to
1271        avoid this problem.
1272    
1273     9. On the advice of a Windows user, included <io.h> and <fcntl.h> in Windows
1274        builds of pcretest, and changed the call to _setmode() to use _O_BINARY
1275        instead of 0x8000. Made all the #ifdefs test both _WIN32 and WIN32 (not all
1276        of them did).
1277    
1278    10. Originally, pcretest opened its input and output without "b"; then I was
1279        told that "b" was needed in some environments, so it was added for release
1280        5.0 to both the input and output. (It makes no difference on Unix-like
1281        systems.) Later I was told that it is wrong for the input on Windows. I've
1282        now abstracted the modes into two macros, to make it easier to fiddle with
1283        them, and removed "b" from the input mode under Windows.
1284    
1285    11. Added pkgconfig support for the C++ wrapper library, libpcrecpp.
1286    
1287    12. Added -help and --help to pcretest as an official way of being reminded
1288        of the options.
1289    
1290    13. Removed some redundant semicolons after macro calls in pcrecpparg.h.in
1291        and pcrecpp.cc because they annoy compilers at high warning levels.
1292    
1293    14. A bit of tidying/refactoring in pcre_exec.c in the main bumpalong loop.
1294    
1295    15. Fixed an occurrence of == in configure.ac that should have been = (shell
1296        scripts are not C programs :-) and which was not noticed because it works
1297        on Linux.
1298    
1299    16. pcretest is supposed to handle any length of pattern and data line (as one
1300        line or as a continued sequence of lines) by extending its input buffer if
1301        necessary. This feature was broken for very long pattern lines, leading to
1302        a string of junk being passed to pcre_compile() if the pattern was longer
1303        than about 50K.
1304    
1305    17. I have done a major re-factoring of the way pcre_compile() computes the
1306        amount of memory needed for a compiled pattern. Previously, there was code
1307        that made a preliminary scan of the pattern in order to do this. That was
1308        OK when PCRE was new, but as the facilities have expanded, it has become
1309        harder and harder to keep it in step with the real compile phase, and there
1310        have been a number of bugs (see for example, 4 above). I have now found a
1311        cunning way of running the real compile function in a "fake" mode that
1312        enables it to compute how much memory it would need, while actually only
1313        ever using a few hundred bytes of working memory and without too many
1314        tests of the mode. This should make future maintenance and development
1315        easier. A side effect of this work is that the limit of 200 on the nesting
1316        depth of parentheses has been removed (though this was never a serious
1317        limitation, I suspect). However, there is a downside: pcre_compile() now
1318        runs more slowly than before (30% or more, depending on the pattern). I
1319        hope this isn't a big issue. There is no effect on runtime performance.
1320    
1321    18. Fixed a minor bug in pcretest: if a pattern line was not terminated by a
1322        newline (only possible for the last line of a file) and it was a
1323        pattern that set a locale (followed by /Lsomething), pcretest crashed.
1324    
1325    19. Added additional timing features to pcretest. (1) The -tm option now times
1326        matching only, not compiling. (2) Both -t and -tm can be followed, as a
1327        separate command line item, by a number that specifies the number of
1328        repeats to use when timing. The default is 50000; this gives better
1329        precision, but takes uncomfortably long for very large patterns.
1330    
1331    20. Extended pcre_study() to be more clever in cases where a branch of a
1332        subpattern has no definite first character. For example, (a*|b*)[cd] would
1333        previously give no result from pcre_study(). Now it recognizes that the
1334        first character must be a, b, c, or d.
1335    
1336    21. There was an incorrect error "recursive call could loop indefinitely" if
1337        a subpattern (or the entire pattern) that was being tested for matching an
1338        empty string contained only one non-empty item after a nested subpattern.
1339        For example, the pattern (?>\x{100}*)\d(?R) provoked this error
1340        incorrectly, because the \d was being skipped in the check.
1341    
1342    22. The pcretest program now has a new pattern option /B and a command line
1343        option -b, which is equivalent to adding /B to every pattern. This causes
1344        it to show the compiled bytecode, without the additional information that
1345        -d shows. The effect of -d is now the same as -b with -i (and similarly, /D
1346        is the same as /B/I).
1347    
1348    23. A new optimization is now able automatically to treat some sequences such
1349        as a*b as a*+b. More specifically, if something simple (such as a character
1350        or a simple class like \d) has an unlimited quantifier, and is followed by
1351        something that cannot possibly match the quantified thing, the quantifier
1352        is automatically "possessified".
1353    
1354    24. A recursive reference to a subpattern whose number was greater than 39
1355        went wrong under certain circumstances in UTF-8 mode. This bug could also
1356        have affected the operation of pcre_study().
1357    
1358    25. Realized that a little bit of performance could be had by replacing
1359        (c & 0xc0) == 0xc0 with c >= 0xc0 when processing UTF-8 characters.
1360    
1361    26. Timing data from pcretest is now shown to 4 decimal places instead of 3.
1362    
1363    27. Possessive quantifiers such as a++ were previously implemented by turning
1364        them into atomic groups such as ($>a+). Now they have their own opcodes,
1365        which improves performance. This includes the automatically created ones
1366        from 23 above.
1367    
1368    28. A pattern such as (?=(\w+))\1: which simulates an atomic group using a
1369        lookahead was broken if it was not anchored. PCRE was mistakenly expecting
1370        the first matched character to be a colon. This applied both to named and
1371        numbered groups.
1372    
1373    29. The ucpinternal.h header file was missing its idempotency #ifdef.
1374    
1375    30. I was sent a "project" file called libpcre.a.dev which I understand makes
1376        building PCRE on Windows easier, so I have included it in the distribution.
1377    
1378    31. There is now a check in pcretest against a ridiculously large number being
1379        returned by pcre_exec() or pcre_dfa_exec(). If this happens in a /g or /G
1380        loop, the loop is abandoned.
1381    
1382    32. Forward references to subpatterns in conditions such as (?(2)...) where
1383        subpattern 2 is defined later cause pcre_compile() to search forwards in
1384        the pattern for the relevant set of parentheses. This search went wrong
1385        when there were unescaped parentheses in a character class, parentheses
1386        escaped with \Q...\E, or parentheses in a #-comment in /x mode.
1387    
1388    33. "Subroutine" calls and backreferences were previously restricted to
1389        referencing subpatterns earlier in the regex. This restriction has now
1390        been removed.
1391    
1392    34. Added a number of extra features that are going to be in Perl 5.10. On the
1393        whole, these are just syntactic alternatives for features that PCRE had
1394        previously implemented using the Python syntax or my own invention. The
1395        other formats are all retained for compatibility.
1396    
1397        (a) Named groups can now be defined as (?<name>...) or (?'name'...) as well
1398            as (?P<name>...). The new forms, as well as being in Perl 5.10, are
1399            also .NET compatible.
1400    
1401        (b) A recursion or subroutine call to a named group can now be defined as
1402            (?&name) as well as (?P>name).
1403    
1404        (c) A backreference to a named group can now be defined as \k<name> or
1405            \k'name' as well as (?P=name). The new forms, as well as being in Perl
1406            5.10, are also .NET compatible.
1407    
1408        (d) A conditional reference to a named group can now use the syntax
1409            (?(<name>) or (?('name') as well as (?(name).
1410    
1411        (e) A "conditional group" of the form (?(DEFINE)...) can be used to define
1412            groups (named and numbered) that are never evaluated inline, but can be
1413            called as "subroutines" from elsewhere. In effect, the DEFINE condition
1414            is always false. There may be only one alternative in such a group.
1415    
1416        (f) A test for recursion can be given as (?(R1).. or (?(R&name)... as well
1417            as the simple (?(R). The condition is true only if the most recent
1418            recursion is that of the given number or name. It does not search out
1419            through the entire recursion stack.
1420    
1421        (g) The escape \gN or \g{N} has been added, where N is a positive or
1422            negative number, specifying an absolute or relative reference.
1423    
1424    35. Tidied to get rid of some further signed/unsigned compiler warnings and
1425        some "unreachable code" warnings.
1426    
1427    36. Updated the Unicode property tables to Unicode version 5.0.0. Amongst other
1428        things, this adds five new scripts.
1429    
1430    37. Perl ignores orphaned \E escapes completely. PCRE now does the same.
1431        There were also incompatibilities regarding the handling of \Q..\E inside
1432        character classes, for example with patterns like [\Qa\E-\Qz\E] where the
1433        hyphen was adjacent to \Q or \E. I hope I've cleared all this up now.
1434    
1435    38. Like Perl, PCRE detects when an indefinitely repeated parenthesized group
1436        matches an empty string, and forcibly breaks the loop. There were bugs in
1437        this code in non-simple cases. For a pattern such as  ^(a()*)*  matched
1438        against  aaaa  the result was just "a" rather than "aaaa", for example. Two
1439        separate and independent bugs (that affected different cases) have been
1440        fixed.
1441    
1442    39. Refactored the code to abolish the use of different opcodes for small
1443        capturing bracket numbers. This is a tidy that I avoided doing when I
1444        removed the limit on the number of capturing brackets for 3.5 back in 2001.
1445        The new approach is not only tidier, it makes it possible to reduce the
1446        memory needed to fix the previous bug (38).
1447    
1448    40. Implemented PCRE_NEWLINE_ANY to recognize any of the Unicode newline
1449        sequences (http://unicode.org/unicode/reports/tr18/) as "newline" when
1450        processing dot, circumflex, or dollar metacharacters, or #-comments in /x
1451        mode.
1452    
1453    41. Add \R to match any Unicode newline sequence, as suggested in the Unicode
1454        report.
1455    
1456    42. Applied patch, originally from Ari Pollak, modified by Google, to allow
1457        copy construction and assignment in the C++ wrapper.
1458    
1459    43. Updated pcregrep to support "--newline=any". In the process, I fixed a
1460        couple of bugs that could have given wrong results in the "--newline=crlf"
1461        case.
1462    
1463    44. Added a number of casts and did some reorganization of signed/unsigned int
1464        variables following suggestions from Dair Grant. Also renamed the variable
1465        "this" as "item" because it is a C++ keyword.
1466    
1467    45. Arranged for dftables to add
1468    
1469          #include "pcre_internal.h"
1470    
1471        to pcre_chartables.c because without it, gcc 4.x may remove the array
1472        definition from the final binary if PCRE is built into a static library and
1473        dead code stripping is activated.
1474    
1475    46. For an unanchored pattern, if a match attempt fails at the start of a
1476        newline sequence, and the newline setting is CRLF or ANY, and the next two
1477        characters are CRLF, advance by two characters instead of one.
1478    
1479    
1480    Version 6.7 04-Jul-06
1481    ---------------------
1482    
1483     1. In order to handle tests when input lines are enormously long, pcretest has
1484        been re-factored so that it automatically extends its buffers when
1485        necessary. The code is crude, but this _is_ just a test program. The
1486        default size has been increased from 32K to 50K.
1487    
1488     2. The code in pcre_study() was using the value of the re argument before
1489        testing it for NULL. (Of course, in any sensible call of the function, it
1490        won't be NULL.)
1491    
1492     3. The memmove() emulation function in pcre_internal.h, which is used on
1493        systems that lack both memmove() and bcopy() - that is, hardly ever -
1494        was missing a "static" storage class specifier.
1495    
1496     4. When UTF-8 mode was not set, PCRE looped when compiling certain patterns
1497        containing an extended class (one that cannot be represented by a bitmap
1498        because it contains high-valued characters or Unicode property items, e.g.
1499        [\pZ]). Almost always one would set UTF-8 mode when processing such a
1500        pattern, but PCRE should not loop if you do not (it no longer does).
1501        [Detail: two cases were found: (a) a repeated subpattern containing an
1502        extended class; (b) a recursive reference to a subpattern that followed a
1503        previous extended class. It wasn't skipping over the extended class
1504        correctly when UTF-8 mode was not set.]
1505    
1506     5. A negated single-character class was not being recognized as fixed-length
1507        in lookbehind assertions such as (?<=[^f]), leading to an incorrect
1508        compile error "lookbehind assertion is not fixed length".
1509    
1510     6. The RunPerlTest auxiliary script was showing an unexpected difference
1511        between PCRE and Perl for UTF-8 tests. It turns out that it is hard to
1512        write a Perl script that can interpret lines of an input file either as
1513        byte characters or as UTF-8, which is what "perltest" was being required to
1514        do for the non-UTF-8 and UTF-8 tests, respectively. Essentially what you
1515        can't do is switch easily at run time between having the "use utf8;" pragma
1516        or not. In the end, I fudged it by using the RunPerlTest script to insert
1517        "use utf8;" explicitly for the UTF-8 tests.
1518    
1519     7. In multiline (/m) mode, PCRE was matching ^ after a terminating newline at
1520        the end of the subject string, contrary to the documentation and to what
1521        Perl does. This was true of both matching functions. Now it matches only at
1522        the start of the subject and immediately after *internal* newlines.
1523    
1524     8. A call of pcre_fullinfo() from pcretest to get the option bits was passing
1525        a pointer to an int instead of a pointer to an unsigned long int. This
1526        caused problems on 64-bit systems.
1527    
1528     9. Applied a patch from the folks at Google to pcrecpp.cc, to fix "another
1529        instance of the 'standard' template library not being so standard".
1530    
1531    10. There was no check on the number of named subpatterns nor the maximum
1532        length of a subpattern name. The product of these values is used to compute
1533        the size of the memory block for a compiled pattern. By supplying a very
1534        long subpattern name and a large number of named subpatterns, the size
1535        computation could be caused to overflow. This is now prevented by limiting
1536        the length of names to 32 characters, and the number of named subpatterns
1537        to 10,000.
1538    
1539    11. Subpatterns that are repeated with specific counts have to be replicated in
1540        the compiled pattern. The size of memory for this was computed from the
1541        length of the subpattern and the repeat count. The latter is limited to
1542        65535, but there was no limit on the former, meaning that integer overflow
1543        could in principle occur. The compiled length of a repeated subpattern is
1544        now limited to 30,000 bytes in order to prevent this.
1545    
1546    12. Added the optional facility to have named substrings with the same name.
1547    
1548    13. Added the ability to use a named substring as a condition, using the
1549        Python syntax: (?(name)yes|no). This overloads (?(R)... and names that
1550        are numbers (not recommended). Forward references are permitted.
1551    
1552    14. Added forward references in named backreferences (if you see what I mean).
1553    
1554    15. In UTF-8 mode, with the PCRE_DOTALL option set, a quantified dot in the
1555        pattern could run off the end of the subject. For example, the pattern
1556        "(?s)(.{1,5})"8 did this with the subject "ab".
1557    
1558    16. If PCRE_DOTALL or PCRE_MULTILINE were set, pcre_dfa_exec() behaved as if
1559        PCRE_CASELESS was set when matching characters that were quantified with ?
1560        or *.
1561    
1562    17. A character class other than a single negated character that had a minimum
1563        but no maximum quantifier - for example [ab]{6,} - was not handled
1564        correctly by pce_dfa_exec(). It would match only one character.
1565    
1566    18. A valid (though odd) pattern that looked like a POSIX character
1567        class but used an invalid character after [ (for example [[,abc,]]) caused
1568        pcre_compile() to give the error "Failed: internal error: code overflow" or
1569        in some cases to crash with a glibc free() error. This could even happen if
1570        the pattern terminated after [[ but there just happened to be a sequence of
1571        letters, a binary zero, and a closing ] in the memory that followed.
1572    
1573    19. Perl's treatment of octal escapes in the range \400 to \777 has changed
1574        over the years. Originally (before any Unicode support), just the bottom 8
1575        bits were taken. Thus, for example, \500 really meant \100. Nowadays the
1576        output from "man perlunicode" includes this:
1577    
1578          The regular expression compiler produces polymorphic opcodes.  That
1579          is, the pattern adapts to the data and automatically switches to
1580          the Unicode character scheme when presented with Unicode data--or
1581          instead uses a traditional byte scheme when presented with byte
1582          data.
1583    
1584        Sadly, a wide octal escape does not cause a switch, and in a string with
1585        no other multibyte characters, these octal escapes are treated as before.
1586        Thus, in Perl, the pattern  /\500/ actually matches \100 but the pattern
1587        /\500|\x{1ff}/ matches \500 or \777 because the whole thing is treated as a
1588        Unicode string.
1589    
1590        I have not perpetrated such confusion in PCRE. Up till now, it took just
1591        the bottom 8 bits, as in old Perl. I have now made octal escapes with
1592        values greater than \377 illegal in non-UTF-8 mode. In UTF-8 mode they
1593        translate to the appropriate multibyte character.
1594    
1595    29. Applied some refactoring to reduce the number of warnings from Microsoft
1596        and Borland compilers. This has included removing the fudge introduced
1597        seven years ago for the OS/2 compiler (see 2.02/2 below) because it caused
1598        a warning about an unused variable.
1599    
1600    21. PCRE has not included VT (character 0x0b) in the set of whitespace
1601        characters since release 4.0, because Perl (from release 5.004) does not.
1602        [Or at least, is documented not to: some releases seem to be in conflict
1603        with the documentation.] However, when a pattern was studied with
1604        pcre_study() and all its branches started with \s, PCRE still included VT
1605        as a possible starting character. Of course, this did no harm; it just
1606        caused an unnecessary match attempt.
1607    
1608    22. Removed a now-redundant internal flag bit that recorded the fact that case
1609        dependency changed within the pattern. This was once needed for "required
1610        byte" processing, but is no longer used. This recovers a now-scarce options
1611        bit. Also moved the least significant internal flag bit to the most-
1612        significant bit of the word, which was not previously used (hangover from
1613        the days when it was an int rather than a uint) to free up another bit for
1614        the future.
1615    
1616    23. Added support for CRLF line endings as well as CR and LF. As well as the
1617        default being selectable at build time, it can now be changed at runtime
1618        via the PCRE_NEWLINE_xxx flags. There are now options for pcregrep to
1619        specify that it is scanning data with non-default line endings.
1620    
1621    24. Changed the definition of CXXLINK to make it agree with the definition of
1622        LINK in the Makefile, by replacing LDFLAGS to CXXFLAGS.
1623    
1624    25. Applied Ian Taylor's patches to avoid using another stack frame for tail
1625        recursions. This makes a big different to stack usage for some patterns.
1626    
1627    26. If a subpattern containing a named recursion or subroutine reference such
1628        as (?P>B) was quantified, for example (xxx(?P>B)){3}, the calculation of
1629        the space required for the compiled pattern went wrong and gave too small a
1630        value. Depending on the environment, this could lead to "Failed: internal
1631        error: code overflow at offset 49" or "glibc detected double free or
1632        corruption" errors.
1633    
1634    27. Applied patches from Google (a) to support the new newline modes and (b) to
1635        advance over multibyte UTF-8 characters in GlobalReplace.
1636    
1637    28. Change free() to pcre_free() in pcredemo.c. Apparently this makes a
1638        difference for some implementation of PCRE in some Windows version.
1639    
1640    29. Added some extra testing facilities to pcretest:
1641    
1642        \q<number>   in a data line sets the "match limit" value
1643        \Q<number>   in a data line sets the "match recursion limt" value
1644        -S <number>  sets the stack size, where <number> is in megabytes
1645    
1646        The -S option isn't available for Windows.
1647    
1648    
1649    Version 6.6 06-Feb-06
1650    ---------------------
1651    
1652     1. Change 16(a) for 6.5 broke things, because PCRE_DATA_SCOPE was not defined
1653        in pcreposix.h. I have copied the definition from pcre.h.
1654    
1655     2. Change 25 for 6.5 broke compilation in a build directory out-of-tree
1656        because pcre.h is no longer a built file.
1657    
1658     3. Added Jeff Friedl's additional debugging patches to pcregrep. These are
1659        not normally included in the compiled code.
1660    
1661    
1662    Version 6.5 01-Feb-06
1663    ---------------------
1664    
1665     1. When using the partial match feature with pcre_dfa_exec(), it was not
1666        anchoring the second and subsequent partial matches at the new starting
1667        point. This could lead to incorrect results. For example, with the pattern
1668        /1234/, partially matching against "123" and then "a4" gave a match.
1669    
1670     2. Changes to pcregrep:
1671    
1672        (a) All non-match returns from pcre_exec() were being treated as failures
1673            to match the line. Now, unless the error is PCRE_ERROR_NOMATCH, an
1674            error message is output. Some extra information is given for the
1675            PCRE_ERROR_MATCHLIMIT and PCRE_ERROR_RECURSIONLIMIT errors, which are
1676            probably the only errors that are likely to be caused by users (by
1677            specifying a regex that has nested indefinite repeats, for instance).
1678            If there are more than 20 of these errors, pcregrep is abandoned.
1679    
1680        (b) A binary zero was treated as data while matching, but terminated the
1681            output line if it was written out. This has been fixed: binary zeroes
1682            are now no different to any other data bytes.
1683    
1684        (c) Whichever of the LC_ALL or LC_CTYPE environment variables is set is
1685            used to set a locale for matching. The --locale=xxxx long option has
1686            been added (no short equivalent) to specify a locale explicitly on the
1687            pcregrep command, overriding the environment variables.
1688    
1689        (d) When -B was used with -n, some line numbers in the output were one less
1690            than they should have been.
1691    
1692        (e) Added the -o (--only-matching) option.
1693    
1694        (f) If -A or -C was used with -c (count only), some lines of context were
1695            accidentally printed for the final match.
1696    
1697        (g) Added the -H (--with-filename) option.
1698    
1699        (h) The combination of options -rh failed to suppress file names for files
1700            that were found from directory arguments.
1701    
1702        (i) Added the -D (--devices) and -d (--directories) options.
1703    
1704        (j) Added the -F (--fixed-strings) option.
1705    
1706        (k) Allow "-" to be used as a file name for -f as well as for a data file.
1707    
1708        (l) Added the --colo(u)r option.
1709    
1710        (m) Added Jeffrey Friedl's -S testing option, but within #ifdefs so that it
1711            is not present by default.
1712    
1713     3. A nasty bug was discovered in the handling of recursive patterns, that is,
1714        items such as (?R) or (?1), when the recursion could match a number of
1715        alternatives. If it matched one of the alternatives, but subsequently,
1716        outside the recursion, there was a failure, the code tried to back up into
1717        the recursion. However, because of the way PCRE is implemented, this is not
1718        possible, and the result was an incorrect result from the match.
1719    
1720        In order to prevent this happening, the specification of recursion has
1721        been changed so that all such subpatterns are automatically treated as
1722        atomic groups. Thus, for example, (?R) is treated as if it were (?>(?R)).
1723    
1724     4. I had overlooked the fact that, in some locales, there are characters for
1725        which isalpha() is true but neither isupper() nor islower() are true. In
1726        the fr_FR locale, for instance, the \xAA and \xBA characters (ordmasculine
1727        and ordfeminine) are like this. This affected the treatment of \w and \W
1728        when they appeared in character classes, but not when they appeared outside
1729        a character class. The bit map for "word" characters is now created
1730        separately from the results of isalnum() instead of just taking it from the
1731        upper, lower, and digit maps. (Plus the underscore character, of course.)
1732    
1733     5. The above bug also affected the handling of POSIX character classes such as
1734        [[:alpha:]] and [[:alnum:]]. These do not have their own bit maps in PCRE's
1735        permanent tables. Instead, the bit maps for such a class were previously
1736        created as the appropriate unions of the upper, lower, and digit bitmaps.
1737        Now they are created by subtraction from the [[:word:]] class, which has
1738        its own bitmap.
1739    
1740     6. The [[:blank:]] character class matches horizontal, but not vertical space.
1741        It is created by subtracting the vertical space characters (\x09, \x0a,
1742        \x0b, \x0c) from the [[:space:]] bitmap. Previously, however, the
1743        subtraction was done in the overall bitmap for a character class, meaning
1744        that a class such as [\x0c[:blank:]] was incorrect because \x0c would not
1745        be recognized. This bug has been fixed.
1746    
1747     7. Patches from the folks at Google:
1748    
1749          (a) pcrecpp.cc: "to handle a corner case that may or may not happen in
1750          real life, but is still worth protecting against".
1751    
1752          (b) pcrecpp.cc: "corrects a bug when negative radixes are used with
1753          regular expressions".
1754    
1755          (c) pcre_scanner.cc: avoid use of std::count() because not all systems
1756          have it.
1757    
1758          (d) Split off pcrecpparg.h from pcrecpp.h and had the former built by
1759          "configure" and the latter not, in order to fix a problem somebody had
1760          with compiling the Arg class on HP-UX.
1761    
1762          (e) Improve the error-handling of the C++ wrapper a little bit.
1763    
1764          (f) New tests for checking recursion limiting.
1765    
1766     8. The pcre_memmove() function, which is used only if the environment does not
1767        have a standard memmove() function (and is therefore rarely compiled),
1768        contained two bugs: (a) use of int instead of size_t, and (b) it was not
1769        returning a result (though PCRE never actually uses the result).
1770    
1771     9. In the POSIX regexec() interface, if nmatch is specified as a ridiculously
1772        large number - greater than INT_MAX/(3*sizeof(int)) - REG_ESPACE is
1773        returned instead of calling malloc() with an overflowing number that would
1774        most likely cause subsequent chaos.
1775    
1776    10. The debugging option of pcretest was not showing the NO_AUTO_CAPTURE flag.
1777    
1778    11. The POSIX flag REG_NOSUB is now supported. When a pattern that was compiled
1779        with this option is matched, the nmatch and pmatch options of regexec() are
1780        ignored.
1781    
1782    12. Added REG_UTF8 to the POSIX interface. This is not defined by POSIX, but is
1783        provided in case anyone wants to the the POSIX interface with UTF-8
1784        strings.
1785    
1786    13. Added CXXLDFLAGS to the Makefile parameters to provide settings only on the
1787        C++ linking (needed for some HP-UX environments).
1788    
1789    14. Avoid compiler warnings in get_ucpname() when compiled without UCP support
1790        (unused parameter) and in the pcre_printint() function (omitted "default"
1791        switch label when the default is to do nothing).
1792    
1793    15. Added some code to make it possible, when PCRE is compiled as a C++
1794        library, to replace subject pointers for pcre_exec() with a smart pointer
1795        class, thus making it possible to process discontinuous strings.
1796    
1797    16. The two macros PCRE_EXPORT and PCRE_DATA_SCOPE are confusing, and perform
1798        much the same function. They were added by different people who were trying
1799        to make PCRE easy to compile on non-Unix systems. It has been suggested
1800        that PCRE_EXPORT be abolished now that there is more automatic apparatus
1801        for compiling on Windows systems. I have therefore replaced it with
1802        PCRE_DATA_SCOPE. This is set automatically for Windows; if not set it
1803        defaults to "extern" for C or "extern C" for C++, which works fine on
1804        Unix-like systems. It is now possible to override the value of PCRE_DATA_
1805        SCOPE with something explicit in config.h. In addition:
1806    
1807        (a) pcreposix.h still had just "extern" instead of either of these macros;
1808            I have replaced it with PCRE_DATA_SCOPE.
1809    
1810        (b) Functions such as _pcre_xclass(), which are internal to the library,
1811            but external in the C sense, all had PCRE_EXPORT in their definitions.
1812            This is apparently wrong for the Windows case, so I have removed it.
1813            (It makes no difference on Unix-like systems.)
1814    
1815    17. Added a new limit, MATCH_LIMIT_RECURSION, which limits the depth of nesting
1816        of recursive calls to match(). This is different to MATCH_LIMIT because
1817        that limits the total number of calls to match(), not all of which increase
1818        the depth of recursion. Limiting the recursion depth limits the amount of
1819        stack (or heap if NO_RECURSE is set) that is used. The default can be set
1820        when PCRE is compiled, and changed at run time. A patch from Google adds
1821        this functionality to the C++ interface.
1822    
1823    18. Changes to the handling of Unicode character properties:
1824    
1825        (a) Updated the table to Unicode 4.1.0.
1826    
1827        (b) Recognize characters that are not in the table as "Cn" (undefined).
1828    
1829        (c) I revised the way the table is implemented to a much improved format
1830            which includes recognition of ranges. It now supports the ranges that
1831            are defined in UnicodeData.txt, and it also amalgamates other
1832            characters into ranges. This has reduced the number of entries in the
1833            table from around 16,000 to around 3,000, thus reducing its size
1834            considerably. I realized I did not need to use a tree structure after
1835            all - a binary chop search is just as efficient. Having reduced the
1836            number of entries, I extended their size from 6 bytes to 8 bytes to
1837            allow for more data.
1838    
1839        (d) Added support for Unicode script names via properties such as \p{Han}.
1840    
1841    19. In UTF-8 mode, a backslash followed by a non-Ascii character was not
1842        matching that character.
1843    
1844    20. When matching a repeated Unicode property with a minimum greater than zero,
1845        (for example \pL{2,}), PCRE could look past the end of the subject if it
1846        reached it while seeking the minimum number of characters. This could
1847        happen only if some of the characters were more than one byte long, because
1848        there is a check for at least the minimum number of bytes.
1849    
1850    21. Refactored the implementation of \p and \P so as to be more general, to
1851        allow for more different types of property in future. This has changed the
1852        compiled form incompatibly. Anybody with saved compiled patterns that use
1853        \p or \P will have to recompile them.
1854    
1855    22. Added "Any" and "L&" to the supported property types.
1856    
1857    23. Recognize \x{...} as a code point specifier, even when not in UTF-8 mode,
1858        but give a compile time error if the value is greater than 0xff.
1859    
1860    24. The man pages for pcrepartial, pcreprecompile, and pcre_compile2 were
1861        accidentally not being installed or uninstalled.
1862    
1863    25. The pcre.h file was built from pcre.h.in, but the only changes that were
1864        made were to insert the current release number. This seemed silly, because
1865        it made things harder for people building PCRE on systems that don't run
1866        "configure". I have turned pcre.h into a distributed file, no longer built
1867        by "configure", with the version identification directly included. There is
1868        no longer a pcre.h.in file.
1869    
1870        However, this change necessitated a change to the pcre-config script as
1871        well. It is built from pcre-config.in, and one of the substitutions was the
1872        release number. I have updated configure.ac so that ./configure now finds
1873        the release number by grepping pcre.h.
1874    
1875    26. Added the ability to run the tests under valgrind.
1876    
1877    
1878    Version 6.4 05-Sep-05
1879    ---------------------
1880    
1881     1. Change 6.0/10/(l) to pcregrep introduced a bug that caused separator lines
1882        "--" to be printed when multiple files were scanned, even when none of the
1883        -A, -B, or -C options were used. This is not compatible with Gnu grep, so I
1884        consider it to be a bug, and have restored the previous behaviour.
1885    
1886     2. A couple of code tidies to get rid of compiler warnings.
1887    
1888     3. The pcretest program used to cheat by referring to symbols in the library
1889        whose names begin with _pcre_. These are internal symbols that are not
1890        really supposed to be visible externally, and in some environments it is
1891        possible to suppress them. The cheating is now confined to including
1892        certain files from the library's source, which is a bit cleaner.
1893    
1894     4. Renamed pcre.in as pcre.h.in to go with pcrecpp.h.in; it also makes the
1895        file's purpose clearer.
1896    
1897     5. Reorganized pcre_ucp_findchar().
1898    
1899    
1900    Version 6.3 15-Aug-05
1901    ---------------------
1902    
1903     1. The file libpcre.pc.in did not have general read permission in the tarball.
1904    
1905     2. There were some problems when building without C++ support:
1906    
1907        (a) If C++ support was not built, "make install" and "make test" still
1908            tried to test it.
1909    
1910        (b) There were problems when the value of CXX was explicitly set. Some
1911            changes have been made to try to fix these, and ...
1912    
1913        (c) --disable-cpp can now be used to explicitly disable C++ support.
1914    
1915        (d) The use of @CPP_OBJ@ directly caused a blank line preceded by a
1916            backslash in a target when C++ was disabled. This confuses some
1917            versions of "make", apparently. Using an intermediate variable solves
1918            this. (Same for CPP_LOBJ.)
1919    
1920     3. $(LINK_FOR_BUILD) now includes $(CFLAGS_FOR_BUILD) and $(LINK)
1921        (non-Windows) now includes $(CFLAGS) because these flags are sometimes
1922        necessary on certain architectures.
1923    
1924     4. Added a setting of -export-symbols-regex to the link command to remove
1925        those symbols that are exported in the C sense, but actually are local
1926        within the library, and not documented. Their names all begin with
1927        "_pcre_". This is not a perfect job, because (a) we have to except some
1928        symbols that pcretest ("illegally") uses, and (b) the facility isn't always
1929        available (and never for static libraries). I have made a note to try to
1930        find a way round (a) in the future.
1931    
1932    
1933    Version 6.2 01-Aug-05
1934    ---------------------
1935    
1936     1. There was no test for integer overflow of quantifier values. A construction
1937        such as {1111111111111111} would give undefined results. What is worse, if
1938        a minimum quantifier for a parenthesized subpattern overflowed and became
1939        negative, the calculation of the memory size went wrong. This could have
1940        led to memory overwriting.
1941    
1942     2. Building PCRE using VPATH was broken. Hopefully it is now fixed.
1943    
1944     3. Added "b" to the 2nd argument of fopen() in dftables.c, for non-Unix-like
1945        operating environments where this matters.
1946    
1947     4. Applied Giuseppe Maxia's patch to add additional features for controlling
1948        PCRE options from within the C++ wrapper.
1949    
1950     5. Named capturing subpatterns were not being correctly counted when a pattern
1951        was compiled. This caused two problems: (a) If there were more than 100
1952        such subpatterns, the calculation of the memory needed for the whole
1953        compiled pattern went wrong, leading to an overflow error. (b) Numerical
1954        back references of the form \12, where the number was greater than 9, were
1955        not recognized as back references, even though there were sufficient
1956        previous subpatterns.
1957    
1958     6. Two minor patches to pcrecpp.cc in order to allow it to compile on older
1959        versions of gcc, e.g. 2.95.4.
1960    
1961    
1962    Version 6.1 21-Jun-05
1963    ---------------------
1964    
1965     1. There was one reference to the variable "posix" in pcretest.c that was not
1966        surrounded by "#if !defined NOPOSIX".
1967    
1968     2. Make it possible to compile pcretest without DFA support, UTF8 support, or
1969        the cross-check on the old pcre_info() function, for the benefit of the
1970        cut-down version of PCRE that is currently imported into Exim.
1971    
1972     3. A (silly) pattern starting with (?i)(?-i) caused an internal space
1973        allocation error. I've done the easy fix, which wastes 2 bytes for sensible
1974        patterns that start (?i) but I don't think that matters. The use of (?i) is
1975        just an example; this all applies to the other options as well.
1976    
1977     4. Since libtool seems to echo the compile commands it is issuing, the output
1978        from "make" can be reduced a bit by putting "@" in front of each libtool
1979        compile command.
1980    
1981     5. Patch from the folks at Google for configure.in to be a bit more thorough
1982        in checking for a suitable C++ installation before trying to compile the
1983        C++ stuff. This should fix a reported problem when a compiler was present,
1984        but no suitable headers.
1985    
1986     6. The man pages all had just "PCRE" as their title. I have changed them to
1987        be the relevant file name. I have also arranged that these names are
1988        retained in the file doc/pcre.txt, which is a concatenation in text format
1989        of all the man pages except the little individual ones for each function.
1990    
1991     7. The NON-UNIX-USE file had not been updated for the different set of source
1992        files that come with release 6. I also added a few comments about the C++
1993        wrapper.
1994    
1995    
1996    Version 6.0 07-Jun-05
1997    ---------------------
1998    
1999     1. Some minor internal re-organization to help with my DFA experiments.
2000    
2001     2. Some missing #ifdef SUPPORT_UCP conditionals in pcretest and printint that
2002        didn't matter for the library itself when fully configured, but did matter
2003        when compiling without UCP support, or within Exim, where the ucp files are
2004        not imported.
2005    
2006     3. Refactoring of the library code to split up the various functions into
2007        different source modules. The addition of the new DFA matching code (see
2008        below) to a single monolithic source would have made it really too
2009        unwieldy, quite apart from causing all the code to be include in a
2010        statically linked application, when only some functions are used. This is
2011        relevant even without the DFA addition now that patterns can be compiled in
2012        one application and matched in another.
2013    
2014        The downside of splitting up is that there have to be some external
2015        functions and data tables that are used internally in different modules of
2016        the library but which are not part of the API. These have all had their
2017        names changed to start with "_pcre_" so that they are unlikely to clash
2018        with other external names.
2019    
2020     4. Added an alternate matching function, pcre_dfa_exec(), which matches using
2021        a different (DFA) algorithm. Although it is slower than the original
2022        function, it does have some advantages for certain types of matching
2023        problem.
2024    
2025     5. Upgrades to pcretest in order to test the features of pcre_dfa_exec(),
2026        including restarting after a partial match.
2027    
2028     6. A patch for pcregrep that defines INVALID_FILE_ATTRIBUTES if it is not
2029        defined when compiling for Windows was sent to me. I have put it into the
2030        code, though I have no means of testing or verifying it.
2031    
2032     7. Added the pcre_refcount() auxiliary function.
2033    
2034     8. Added the PCRE_FIRSTLINE option. This constrains an unanchored pattern to
2035        match before or at the first newline in the subject string. In pcretest,
2036        the /f option on a pattern can be used to set this.
2037    
2038     9. A repeated \w when used in UTF-8 mode with characters greater than 256
2039        would behave wrongly. This has been present in PCRE since release 4.0.
2040    
2041    10. A number of changes to the pcregrep command:
2042    
2043        (a) Refactored how -x works; insert ^(...)$ instead of setting
2044            PCRE_ANCHORED and checking the length, in preparation for adding
2045            something similar for -w.
2046    
2047        (b) Added the -w (match as a word) option.
2048    
2049        (c) Refactored the way lines are read and buffered so as to have more
2050            than one at a time available.
2051    
2052        (d) Implemented a pcregrep test script.
2053    
2054        (e) Added the -M (multiline match) option. This allows patterns to match
2055            over several lines of the subject. The buffering ensures that at least
2056            8K, or the rest of the document (whichever is the shorter) is available
2057            for matching (and similarly the previous 8K for lookbehind assertions).
2058    
2059        (f) Changed the --help output so that it now says
2060    
2061              -w, --word-regex(p)
2062    
2063            instead of two lines, one with "regex" and the other with "regexp"
2064            because that confused at least one person since the short forms are the
2065            same. (This required a bit of code, as the output is generated
2066            automatically from a table. It wasn't just a text change.)
2067    
2068        (g) -- can be used to terminate pcregrep options if the next thing isn't an
2069            option but starts with a hyphen. Could be a pattern or a path name
2070            starting with a hyphen, for instance.
2071    
2072        (h) "-" can be given as a file name to represent stdin.
2073    
2074        (i) When file names are being printed, "(standard input)" is used for
2075            the standard input, for compatibility with GNU grep. Previously
2076            "<stdin>" was used.
2077    
2078        (j) The option --label=xxx can be used to supply a name to be used for
2079            stdin when file names are being printed. There is no short form.
2080    
2081        (k) Re-factored the options decoding logic because we are going to add
2082            two more options that take data. Such options can now be given in four
2083            different ways, e.g. "-fname", "-f name", "--file=name", "--file name".
2084    
2085        (l) Added the -A, -B, and -C options for requesting that lines of context
2086            around matches be printed.
2087    
2088        (m) Added the -L option to print the names of files that do not contain
2089            any matching lines, that is, the complement of -l.
2090    
2091        (n) The return code is 2 if any file cannot be opened, but pcregrep does
2092            continue to scan other files.
2093    
2094        (o) The -s option was incorrectly implemented. For compatibility with other
2095            greps, it now suppresses the error message for a non-existent or non-
2096            accessible file (but not the return code). There is a new option called
2097            -q that suppresses the output of matching lines, which was what -s was
2098            previously doing.
2099    
2100        (p) Added --include and --exclude options to specify files for inclusion
2101            and exclusion when recursing.
2102    
2103    11. The Makefile was not using the Autoconf-supported LDFLAGS macro properly.
2104        Hopefully, it now does.
2105    
2106    12. Missing cast in pcre_study().
2107    
2108    13. Added an "uninstall" target to the makefile.
2109    
2110    14. Replaced "extern" in the function prototypes in Makefile.in with
2111        "PCRE_DATA_SCOPE", which defaults to 'extern' or 'extern "C"' in the Unix
2112        world, but is set differently for Windows.
2113    
2114    15. Added a second compiling function called pcre_compile2(). The only
2115        difference is that it has an extra argument, which is a pointer to an
2116        integer error code. When there is a compile-time failure, this is set
2117        non-zero, in addition to the error test pointer being set to point to an
2118        error message. The new argument may be NULL if no error number is required
2119        (but then you may as well call pcre_compile(), which is now just a
2120        wrapper). This facility is provided because some applications need a
2121        numeric error indication, but it has also enabled me to tidy up the way
2122        compile-time errors are handled in the POSIX wrapper.
2123    
2124    16. Added VPATH=.libs to the makefile; this should help when building with one
2125        prefix path and installing with another. (Or so I'm told by someone who
2126        knows more about this stuff than I do.)
2127    
2128    17. Added a new option, REG_DOTALL, to the POSIX function regcomp(). This
2129        passes PCRE_DOTALL to the pcre_compile() function, making the "." character
2130        match everything, including newlines. This is not POSIX-compatible, but
2131        somebody wanted the feature. From pcretest it can be activated by using
2132        both the P and the s flags.
2133    
2134    18. AC_PROG_LIBTOOL appeared twice in Makefile.in. Removed one.
2135    
2136    19. libpcre.pc was being incorrectly installed as executable.
2137    
2138    20. A couple of places in pcretest check for end-of-line by looking for '\n';
2139        it now also looks for '\r' so that it will work unmodified on Windows.
2140    
2141    21. Added Google's contributed C++ wrapper to the distribution.
2142    
2143    22. Added some untidy missing memory free() calls in pcretest, to keep
2144        Electric Fence happy when testing.
2145    
2146    
2147    
2148    Version 5.0 13-Sep-04
2149    ---------------------
2150    
2151     1. Internal change: literal characters are no longer packed up into items
2152        containing multiple characters in a single byte-string. Each character
2153        is now matched using a separate opcode. However, there may be more than one
2154        byte in the character in UTF-8 mode.
2155    
2156     2. The pcre_callout_block structure has two new fields: pattern_position and
2157        next_item_length. These contain the offset in the pattern to the next match
2158        item, and its length, respectively.
2159    
2160     3. The PCRE_AUTO_CALLOUT option for pcre_compile() requests the automatic
2161        insertion of callouts before each pattern item. Added the /C option to
2162        pcretest to make use of this.
2163    
2164     4. On the advice of a Windows user, the lines
2165    
2166          #if defined(_WIN32) || defined(WIN32)
2167          _setmode( _fileno( stdout ), 0x8000 );
2168          #endif  /* defined(_WIN32) || defined(WIN32) */
2169    
2170        have been added to the source of pcretest. This apparently does useful
2171        magic in relation to line terminators.
2172    
2173     5. Changed "r" and "w" in the calls to fopen() in pcretest to "rb" and "wb"
2174        for the benefit of those environments where the "b" makes a difference.
2175    
2176     6. The icc compiler has the same options as gcc, but "configure" doesn't seem
2177        to know about it. I have put a hack into configure.in that adds in code
2178        to set GCC=yes if CC=icc. This seems to end up at a point in the
2179        generated configure script that is early enough to affect the setting of
2180        compiler options, which is what is needed, but I have no means of testing
2181        whether it really works. (The user who reported this had patched the
2182        generated configure script, which of course I cannot do.)
2183    
2184        LATER: After change 22 below (new libtool files), the configure script
2185        seems to know about icc (and also ecc). Therefore, I have commented out
2186        this hack in configure.in.
2187    
2188     7. Added support for pkg-config (2 patches were sent in).
2189    
2190     8. Negated POSIX character classes that used a combination of internal tables
2191        were completely broken. These were [[:^alpha:]], [[:^alnum:]], and
2192        [[:^ascii]]. Typically, they would match almost any characters. The other
2193        POSIX classes were not broken in this way.
2194    
2195     9. Matching the pattern "\b.*?" against "ab cd", starting at offset 1, failed
2196        to find the match, as PCRE was deluded into thinking that the match had to
2197        start at the start point or following a newline. The same bug applied to
2198        patterns with negative forward assertions or any backward assertions
2199        preceding ".*" at the start, unless the pattern required a fixed first
2200        character. This was a failing pattern: "(?!.bcd).*". The bug is now fixed.
2201    
2202    10. In UTF-8 mode, when moving forwards in the subject after a failed match
2203        starting at the last subject character, bytes beyond the end of the subject
2204        string were read.
2205    
2206    11. Renamed the variable "class" as "classbits" to make life easier for C++
2207        users. (Previously there was a macro definition, but it apparently wasn't
2208        enough.)
2209    
2210    12. Added the new field "tables" to the extra data so that tables can be passed
2211        in at exec time, or the internal tables can be re-selected. This allows
2212        a compiled regex to be saved and re-used at a later time by a different
2213        program that might have everything at different addresses.
2214    
2215    13. Modified the pcre-config script so that, when run on Solaris, it shows a
2216        -R library as well as a -L library.
2217    
2218    14. The debugging options of pcretest (-d on the command line or D on a
2219        pattern) showed incorrect output for anything following an extended class
2220        that contained multibyte characters and which was followed by a quantifier.
2221    
2222    15. Added optional support for general category Unicode character properties
2223        via the \p, \P, and \X escapes. Unicode property support implies UTF-8
2224        support. It adds about 90K to the size of the library. The meanings of the
2225        inbuilt class escapes such as \d and \s have NOT been changed.
2226    
2227    16. Updated pcredemo.c to include calls to free() to release the memory for the
2228        compiled pattern.
2229    
2230    17. The generated file chartables.c was being created in the source directory
2231        instead of in the building directory. This caused the build to fail if the
2232        source directory was different from the building directory, and was
2233        read-only.
2234    
2235    18. Added some sample Win commands from Mark Tetrode into the NON-UNIX-USE
2236        file. No doubt somebody will tell me if they don't make sense... Also added
2237        Dan Mooney's comments about building on OpenVMS.
2238    
2239    19. Added support for partial matching via the PCRE_PARTIAL option for
2240        pcre_exec() and the \P data escape in pcretest.
2241    
2242    20. Extended pcretest with 3 new pattern features:
2243    
2244        (i)   A pattern option of the form ">rest-of-line" causes pcretest to
2245              write the compiled pattern to the file whose name is "rest-of-line".
2246              This is a straight binary dump of the data, with the saved pointer to
2247              the character tables forced to be NULL. The study data, if any, is
2248              written too. After writing, pcretest reads a new pattern.
2249    
2250        (ii)  If, instead of a pattern, "<rest-of-line" is given, pcretest reads a
2251              compiled pattern from the given file. There must not be any
2252              occurrences of "<" in the file name (pretty unlikely); if there are,
2253              pcretest will instead treat the initial "<" as a pattern delimiter.
2254              After reading in the pattern, pcretest goes on to read data lines as
2255              usual.
2256    
2257        (iii) The F pattern option causes pcretest to flip the bytes in the 32-bit
2258              and 16-bit fields in a compiled pattern, to simulate a pattern that
2259              was compiled on a host of opposite endianness.
2260    
2261    21. The pcre-exec() function can now cope with patterns that were compiled on
2262        hosts of opposite endianness, with this restriction:
2263    
2264          As for any compiled expression that is saved and used later, the tables
2265          pointer field cannot be preserved; the extra_data field in the arguments
2266          to pcre_exec() should be used to pass in a tables address if a value
2267          other than the default internal tables were used at compile time.
2268    
2269    22. Calling pcre_exec() with a negative value of the "ovecsize" parameter is
2270        now diagnosed as an error. Previously, most of the time, a negative number
2271        would have been treated as zero, but if in addition "ovector" was passed as
2272        NULL, a crash could occur.
2273    
2274    23. Updated the files ltmain.sh, config.sub, config.guess, and aclocal.m4 with
2275        new versions from the libtool 1.5 distribution (the last one is a copy of
2276        a file called libtool.m4). This seems to have fixed the need to patch
2277        "configure" to support Darwin 1.3 (which I used to do). However, I still
2278        had to patch ltmain.sh to ensure that ${SED} is set (it isn't on my
2279        workstation).
2280    
2281    24. Changed the PCRE licence to be the more standard "BSD" licence.
2282    
2283    
2284    Version 4.5 01-Dec-03
2285    ---------------------
2286    
2287     1. There has been some re-arrangement of the code for the match() function so
2288        that it can be compiled in a version that does not call itself recursively.
2289        Instead, it keeps those local variables that need separate instances for
2290        each "recursion" in a frame on the heap, and gets/frees frames whenever it
2291        needs to "recurse". Keeping track of where control must go is done by means
2292        of setjmp/longjmp. The whole thing is implemented by a set of macros that
2293        hide most of the details from the main code, and operates only if
2294        NO_RECURSE is defined while compiling pcre.c. If PCRE is built using the
2295        "configure" mechanism, "--disable-stack-for-recursion" turns on this way of
2296        operating.
2297    
2298        To make it easier for callers to provide specially tailored get/free
2299        functions for this usage, two new functions, pcre_stack_malloc, and
2300        pcre_stack_free, are used. They are always called in strict stacking order,
2301        and the size of block requested is always the same.
2302    
2303        The PCRE_CONFIG_STACKRECURSE info parameter can be used to find out whether
2304        PCRE has been compiled to use the stack or the heap for recursion. The
2305        -C option of pcretest uses this to show which version is compiled.
2306    
2307        A new data escape \S, is added to pcretest; it causes the amounts of store
2308        obtained and freed by both kinds of malloc/free at match time to be added
2309        to the output.
2310    
2311     2. Changed the locale test to use "fr_FR" instead of "fr" because that's
2312        what's available on my current Linux desktop machine.
2313    
2314     3. When matching a UTF-8 string, the test for a valid string at the start has
2315        been extended. If start_offset is not zero, PCRE now checks that it points
2316        to a byte that is the start of a UTF-8 character. If not, it returns
2317        PCRE_ERROR_BADUTF8_OFFSET (-11). Note: the whole string is still checked;
2318        this is necessary because there may be backward assertions in the pattern.
2319        When matching the same subject several times, it may save resources to use
2320        PCRE_NO_UTF8_CHECK on all but the first call if the string is long.
2321    
2322     4. The code for checking the validity of UTF-8 strings has been tightened so
2323        that it rejects (a) strings containing 0xfe or 0xff bytes and (b) strings
2324        containing "overlong sequences".
2325    
2326     5. Fixed a bug (appearing twice) that I could not find any way of exploiting!
2327        I had written "if ((digitab[*p++] && chtab_digit) == 0)" where the "&&"
2328        should have been "&", but it just so happened that all the cases this let
2329        through by mistake were picked up later in the function.
2330    
2331     6. I had used a variable called "isblank" - this is a C99 function, causing
2332        some compilers to warn. To avoid this, I renamed it (as "blankclass").
2333    
2334     7. Cosmetic: (a) only output another newline at the end of pcretest if it is
2335        prompting; (b) run "./pcretest /dev/null" at the start of the test script
2336        so the version is shown; (c) stop "make test" echoing "./RunTest".
2337    
2338     8. Added patches from David Burgess to enable PCRE to run on EBCDIC systems.
2339    
2340     9. The prototype for memmove() for systems that don't have it was using
2341        size_t, but the inclusion of the header that defines size_t was later. I've
2342        moved the #includes for the C headers earlier to avoid this.
2343    
2344    10. Added some adjustments to the code to make it easier to compiler on certain
2345        special systems:
2346    
2347          (a) Some "const" qualifiers were missing.
2348          (b) Added the macro EXPORT before all exported functions; by default this
2349              is defined to be empty.
2350          (c) Changed the dftables auxiliary program (that builds chartables.c) so
2351              that it reads its output file name as an argument instead of writing
2352              to the standard output and assuming this can be redirected.
2353    
2354    11. In UTF-8 mode, if a recursive reference (e.g. (?1)) followed a character
2355        class containing characters with values greater than 255, PCRE compilation
2356        went into a loop.
2357    
2358    12. A recursive reference to a subpattern that was within another subpattern
2359        that had a minimum quantifier of zero caused PCRE to crash. For example,
2360        (x(y(?2))z)? provoked this bug with a subject that got as far as the
2361        recursion. If the recursively-called subpattern itself had a zero repeat,
2362        that was OK.
2363    
2364    13. In pcretest, the buffer for reading a data line was set at 30K, but the
2365        buffer into which it was copied (for escape processing) was still set at
2366        1024, so long lines caused crashes.
2367    
2368    14. A pattern such as /[ab]{1,3}+/ failed to compile, giving the error
2369        "internal error: code overflow...". This applied to any character class
2370        that was followed by a possessive quantifier.
2371    
2372    15. Modified the Makefile to add libpcre.la as a prerequisite for
2373        libpcreposix.la because I was told this is needed for a parallel build to
2374        work.
2375    
2376    16. If a pattern that contained .* following optional items at the start was
2377        studied, the wrong optimizing data was generated, leading to matching
2378        errors. For example, studying /[ab]*.*c/ concluded, erroneously, that any
2379        matching string must start with a or b or c. The correct conclusion for
2380        this pattern is that a match can start with any character.
2381    
2382    
2383    Version 4.4 13-Aug-03
2384    ---------------------
2385    
2386     1. In UTF-8 mode, a character class containing characters with values between
2387        127 and 255 was not handled correctly if the compiled pattern was studied.
2388        In fixing this, I have also improved the studying algorithm for such
2389        classes (slightly).
2390    
2391     2. Three internal functions had redundant arguments passed to them. Removal
2392        might give a very teeny performance improvement.
2393    
2394     3. Documentation bug: the value of the capture_top field in a callout is *one
2395        more than* the number of the hightest numbered captured substring.
2396    
2397     4. The Makefile linked pcretest and pcregrep with -lpcre, which could result
2398        in incorrectly linking with a previously installed version. They now link
2399        explicitly with libpcre.la.
2400    
2401     5. configure.in no longer needs to recognize Cygwin specially.
2402    
2403     6. A problem in pcre.in for Windows platforms is fixed.
2404    
2405     7. If a pattern was successfully studied, and the -d (or /D) flag was given to
2406        pcretest, it used to include the size of the study block as part of its
2407        output. Unfortunately, the structure contains a field that has a different
2408        size on different hardware architectures. This meant that the tests that
2409        showed this size failed. As the block is currently always of a fixed size,
2410        this information isn't actually particularly useful in pcretest output, so
2411        I have just removed it.
2412    
2413     8. Three pre-processor statements accidentally did not start in column 1.
2414        Sadly, there are *still* compilers around that complain, even though
2415        standard C has not required this for well over a decade. Sigh.
2416    
2417     9. In pcretest, the code for checking callouts passed small integers in the
2418        callout_data field, which is a void * field. However, some picky compilers
2419        complained about the casts involved for this on 64-bit systems. Now
2420        pcretest passes the address of the small integer instead, which should get
2421        rid of the warnings.
2422    
2423    10. By default, when in UTF-8 mode, PCRE now checks for valid UTF-8 strings at
2424        both compile and run time, and gives an error if an invalid UTF-8 sequence
2425        is found. There is a option for disabling this check in cases where the
2426        string is known to be correct and/or the maximum performance is wanted.
2427    
2428    11. In response to a bug report, I changed one line in Makefile.in from
2429    
2430            -Wl,--out-implib,.libs/lib@WIN_PREFIX@pcreposix.dll.a \
2431        to
2432            -Wl,--out-implib,.libs/@WIN_PREFIX@libpcreposix.dll.a \
2433    
2434        to look similar to other lines, but I have no way of telling whether this
2435        is the right thing to do, as I do not use Windows. No doubt I'll get told
2436        if it's wrong...
2437    
2438    
2439    Version 4.3 21-May-03
2440    ---------------------
2441    
2442    1. Two instances of @WIN_PREFIX@ omitted from the Windows targets in the
2443       Makefile.
2444    
2445    2. Some refactoring to improve the quality of the code:
2446    
2447       (i)   The utf8_table... variables are now declared "const".
2448    
2449       (ii)  The code for \cx, which used the "case flipping" table to upper case
2450             lower case letters, now just substracts 32. This is ASCII-specific,
2451             but the whole concept of \cx is ASCII-specific, so it seems
2452             reasonable.
2453    
2454       (iii) PCRE was using its character types table to recognize decimal and
2455             hexadecimal digits in the pattern. This is silly, because it handles
2456             only 0-9, a-f, and A-F, but the character types table is locale-
2457             specific, which means strange things might happen. A private
2458             table is now used for this - though it costs 256 bytes, a table is
2459             much faster than multiple explicit tests. Of course, the standard
2460             character types table is still used for matching digits in subject
2461             strings against \d.
2462    
2463       (iv)  Strictly, the identifier ESC_t is reserved by POSIX (all identifiers
2464             ending in _t are). So I've renamed it as ESC_tee.
2465    
2466    3. The first argument for regexec() in the POSIX wrapper should have been
2467       defined as "const".
2468    
2469    4. Changed pcretest to use malloc() for its buffers so that they can be
2470       Electric Fenced for debugging.
2471    
2472    5. There were several places in the code where, in UTF-8 mode, PCRE would try
2473       to read one or more bytes before the start of the subject string. Often this
2474       had no effect on PCRE's behaviour, but in some circumstances it could
2475       provoke a segmentation fault.
2476    
2477    6. A lookbehind at the start of a pattern in UTF-8 mode could also cause PCRE
2478       to try to read one or more bytes before the start of the subject string.
2479    
2480    7. A lookbehind in a pattern matched in non-UTF-8 mode on a PCRE compiled with
2481       UTF-8 support could misbehave in various ways if the subject string
2482       contained bytes with the 0x80 bit set and the 0x40 bit unset in a lookbehind
2483       area. (PCRE was not checking for the UTF-8 mode flag, and trying to move
2484       back over UTF-8 characters.)
2485    
2486    
2487    Version 4.2 14-Apr-03
2488    ---------------------
2489    
2490    1. Typo "#if SUPPORT_UTF8" instead of "#ifdef SUPPORT_UTF8" fixed.
2491    
2492    2. Changes to the building process, supplied by Ronald Landheer-Cieslak
2493         [ON_WINDOWS]: new variable, "#" on non-Windows platforms
2494         [NOT_ON_WINDOWS]: new variable, "#" on Windows platforms
2495         [WIN_PREFIX]: new variable, "cyg" for Cygwin
2496         * Makefile.in: use autoconf substitution for OBJEXT, EXEEXT, BUILD_OBJEXT
2497           and BUILD_EXEEXT
2498         Note: automatic setting of the BUILD variables is not yet working
2499         set CPPFLAGS and BUILD_CPPFLAGS (but don't use yet) - should be used at
2500           compile-time but not at link-time
2501         [LINK]: use for linking executables only
2502         make different versions for Windows and non-Windows
2503         [LINKLIB]: new variable, copy of UNIX-style LINK, used for linking
2504           libraries
2505         [LINK_FOR_BUILD]: new variable
2506         [OBJEXT]: use throughout
2507         [EXEEXT]: use throughout
2508         <winshared>: new target
2509         <wininstall>: new target
2510         <dftables.o>: use native compiler
2511         <dftables>: use native linker
2512         <install>: handle Windows platform correctly
2513         <clean>: ditto
2514         <check>: ditto
2515         copy DLL to top builddir before testing
2516    
2517       As part of these changes, -no-undefined was removed again. This was reported
2518       to give trouble on HP-UX 11.0, so getting rid of it seems like a good idea
2519       in any case.
2520    
2521    3. Some tidies to get rid of compiler warnings:
2522    
2523       . In the match_data structure, match_limit was an unsigned long int, whereas
2524         match_call_count was an int. I've made them both unsigned long ints.
2525    
2526       . In pcretest the fact that a const uschar * doesn't automatically cast to
2527         a void * provoked a warning.
2528    
2529       . Turning on some more compiler warnings threw up some "shadow" variables
2530         and a few more missing casts.
2531    
2532    4. If PCRE was complied with UTF-8 support, but called without the PCRE_UTF8
2533       option, a class that contained a single character with a value between 128
2534       and 255 (e.g. /[\xFF]/) caused PCRE to crash.
2535    
2536    5. If PCRE was compiled with UTF-8 support, but called without the PCRE_UTF8
2537       option, a class that contained several characters, but with at least one
2538       whose value was between 128 and 255 caused PCRE to crash.
2539    
2540    
2541    Version 4.1 12-Mar-03
2542    ---------------------
2543    
2544    1. Compiling with gcc -pedantic found a couple of places where casts were
2545    needed, and a string in dftables.c that was longer than standard compilers are
2546    required to support.
2547    
2548    2. Compiling with Sun's compiler found a few more places where the code could
2549    be tidied up in order to avoid warnings.
2550    
2551    3. The variables for cross-compiling were called HOST_CC and HOST_CFLAGS; the
2552    first of these names is deprecated in the latest Autoconf in favour of the name
2553    CC_FOR_BUILD, because "host" is typically used to mean the system on which the
2554    compiled code will be run. I can't find a reference for HOST_CFLAGS, but by
2555    analogy I have changed it to CFLAGS_FOR_BUILD.
2556    
2557    4. Added -no-undefined to the linking command in the Makefile, because this is
2558    apparently helpful for Windows. To make it work, also added "-L. -lpcre" to the
2559    linking step for the pcreposix library.
2560    
2561    5. PCRE was failing to diagnose the case of two named groups with the same
2562    name.
2563    
2564    6. A problem with one of PCRE's optimizations was discovered. PCRE remembers a
2565    literal character that is needed in the subject for a match, and scans along to
2566    ensure that it is present before embarking on the full matching process. This
2567    saves time in cases of nested unlimited repeats that are never going to match.
2568    Problem: the scan can take a lot of time if the subject is very long (e.g.
2569    megabytes), thus penalizing straightforward matches. It is now done only if the
2570    amount of subject to be scanned is less than 1000 bytes.
2571    
2572    7. A lesser problem with the same optimization is that it was recording the
2573    first character of an anchored pattern as "needed", thus provoking a search
2574    right along the subject, even when the first match of the pattern was going to
2575    fail. The "needed" character is now not set for anchored patterns, unless it
2576    follows something in the pattern that is of non-fixed length. Thus, it still
2577    fulfils its original purpose of finding quick non-matches in cases of nested
2578    unlimited repeats, but isn't used for simple anchored patterns such as /^abc/.
2579    
2580    
2581    Version 4.0 17-Feb-03
2582    ---------------------
2583    
2584    1. If a comment in an extended regex that started immediately after a meta-item
2585    extended to the end of string, PCRE compiled incorrect data. This could lead to
2586    all kinds of weird effects. Example: /#/ was bad; /()#/ was bad; /a#/ was not.
2587    
2588    2. Moved to autoconf 2.53 and libtool 1.4.2.
2589    
2590    3. Perl 5.8 no longer needs "use utf8" for doing UTF-8 things. Consequently,
2591    the special perltest8 script is no longer needed - all the tests can be run
2592    from a single perltest script.
2593    
2594    4. From 5.004, Perl has not included the VT character (0x0b) in the set defined
2595    by \s. It has now been removed in PCRE. This means it isn't recognized as
2596    whitespace in /x regexes too, which is the same as Perl. Note that the POSIX
2597    class [:space:] *does* include VT, thereby creating a mess.
2598    
2599    5. Added the class [:blank:] (a GNU extension from Perl 5.8) to match only
2600    space and tab.
2601    
2602    6. Perl 5.005 was a long time ago. It's time to amalgamate the tests that use
2603    its new features into the main test script, reducing the number of scripts.
2604    
2605    7. Perl 5.8 has changed the meaning of patterns like /a(?i)b/. Earlier versions
2606    were backward compatible, and made the (?i) apply to the whole pattern, as if
2607    /i were given. Now it behaves more logically, and applies the option setting
2608    only to what follows. PCRE has been changed to follow suit. However, if it
2609    finds options settings right at the start of the pattern, it extracts them into
2610    the global options, as before. Thus, they show up in the info data.
2611    
2612    8. Added support for the \Q...\E escape sequence. Characters in between are
2613    treated as literals. This is slightly different from Perl in that $ and @ are
2614    also handled as literals inside the quotes. In Perl, they will cause variable
2615    interpolation. Note the following examples:
2616    
2617        Pattern            PCRE matches      Perl matches
2618    
2619        \Qabc$xyz\E        abc$xyz           abc followed by the contents of $xyz
2620        \Qabc\$xyz\E       abc\$xyz          abc\$xyz
2621        \Qabc\E\$\Qxyz\E   abc$xyz           abc$xyz
2622    
2623    For compatibility with Perl, \Q...\E sequences are recognized inside character
2624    classes as well as outside them.
2625    
2626    9. Re-organized 3 code statements in pcretest to avoid "overflow in
2627    floating-point constant arithmetic" warnings from a Microsoft compiler. Added a
2628    (size_t) cast to one statement in pcretest and one in pcreposix to avoid
2629    signed/unsigned warnings.
2630    
2631    10. SunOS4 doesn't have strtoul(). This was used only for unpicking the -o
2632    option for pcretest, so I've replaced it by a simple function that does just
2633    that job.
2634    
2635    11. pcregrep was ending with code 0 instead of 2 for the commands "pcregrep" or
2636    "pcregrep -".
2637    
2638    12. Added "possessive quantifiers" ?+, *+, ++, and {,}+ which come from Sun's
2639    Java package. This provides some syntactic sugar for simple cases of what my
2640    documentation calls "once-only subpatterns". A pattern such as x*+ is the same
2641    as (?>x*). In other words, if what is inside (?>...) is just a single repeated
2642    item, you can use this simplified notation. Note that only makes sense with
2643    greedy quantifiers. Consequently, the use of the possessive quantifier forces
2644    greediness, whatever the setting of the PCRE_UNGREEDY option.
2645    
2646    13. A change of greediness default within a pattern was not taking effect at
2647    the current level for patterns like /(b+(?U)a+)/. It did apply to parenthesized
2648    subpatterns that followed. Patterns like /b+(?U)a+/ worked because the option
2649    was abstracted outside.
2650    
2651    14. PCRE now supports the \G assertion. It is true when the current matching
2652    position is at the start point of the match. This differs from \A when the
2653    starting offset is non-zero. Used with the /g option of pcretest (or similar
2654    code), it works in the same way as it does for Perl's /g option. If all
2655    alternatives of a regex begin with \G, the expression is anchored to the start
2656    match position, and the "anchored" flag is set in the compiled expression.
2657    
2658    15. Some bugs concerning the handling of certain option changes within patterns
2659    have been fixed. These applied to options other than (?ims). For example,
2660    "a(?x: b c )d" did not match "XabcdY" but did match "Xa b c dY". It should have
2661    been the other way round. Some of this was related to change 7 above.
2662    
2663    16. PCRE now gives errors for /[.x.]/ and /[=x=]/ as unsupported POSIX
2664    features, as Perl does. Previously, PCRE gave the warnings only for /[[.x.]]/
2665    and /[[=x=]]/. PCRE now also gives an error for /[:name:]/ because it supports
2666    POSIX classes only within a class (e.g. /[[:alpha:]]/).
2667    
2668    17. Added support for Perl's \C escape. This matches one byte, even in UTF8
2669    mode. Unlike ".", it always matches newline, whatever the setting of
2670    PCRE_DOTALL. However, PCRE does not permit \C to appear in lookbehind
2671    assertions. Perl allows it, but it doesn't (in general) work because it can't
2672    calculate the length of the lookbehind. At least, that's the case for Perl
2673    5.8.0 - I've been told they are going to document that it doesn't work in
2674    future.
2675    
2676    18. Added an error diagnosis for escapes that PCRE does not support: these are
2677    \L, \l, \N, \P, \p, \U, \u, and \X.
2678    
2679    19. Although correctly diagnosing a missing ']' in a character class, PCRE was
2680    reading past the end of the pattern in cases such as /[abcd/.
2681    
2682    20. PCRE was getting more memory than necessary for patterns with classes that
2683    contained both POSIX named classes and other characters, e.g. /[[:space:]abc/.
2684    
2685    21. Added some code, conditional on #ifdef VPCOMPAT, to make life easier for
2686    compiling PCRE for use with Virtual Pascal.
2687    
2688    22. Small fix to the Makefile to make it work properly if the build is done
2689    outside the source tree.
2690    
2691    23. Added a new extension: a condition to go with recursion. If a conditional
2692    subpattern starts with (?(R) the "true" branch is used if recursion has
2693    happened, whereas the "false" branch is used only at the top level.
2694    
2695    24. When there was a very long string of literal characters (over 255 bytes
2696    without UTF support, over 250 bytes with UTF support), the computation of how
2697    much memory was required could be incorrect, leading to segfaults or other
2698    strange effects.
2699    
2700    25. PCRE was incorrectly assuming anchoring (either to start of subject or to
2701    start of line for a non-DOTALL pattern) when a pattern started with (.*) and
2702    there was a subsequent back reference to those brackets. This meant that, for
2703    example, /(.*)\d+\1/ failed to match "abc123bc". Unfortunately, it isn't
2704    possible to check for precisely this case. All we can do is abandon the
2705    optimization if .* occurs inside capturing brackets when there are any back
2706    references whatsoever. (See below for a better fix that came later.)
2707    
2708    26. The handling of the optimization for finding the first character of a
2709    non-anchored pattern, and for finding a character that is required later in the
2710    match were failing in some cases. This didn't break the matching; it just
2711    failed to optimize when it could. The way this is done has been re-implemented.
2712    
2713    27. Fixed typo in error message for invalid (?R item (it said "(?p").
2714    
2715    28. Added a new feature that provides some of the functionality that Perl
2716    provides with (?{...}). The facility is termed a "callout". The way it is done
2717    in PCRE is for the caller to provide an optional function, by setting
2718    pcre_callout to its entry point. Like pcre_malloc and pcre_free, this is a
2719    global variable. By default it is unset, which disables all calling out. To get
2720    the function called, the regex must include (?C) at appropriate points. This
2721    is, in fact, equivalent to (?C0), and any number <= 255 may be given with (?C).
2722    This provides a means of identifying different callout points. When PCRE
2723    reaches such a point in the regex, if pcre_callout has been set, the external
2724    function is called. It is provided with data in a structure called
2725    pcre_callout_block, which is defined in pcre.h. If the function returns 0,
2726    matching continues; if it returns a non-zero value, the match at the current
2727    point fails. However, backtracking will occur if possible. [This was changed
2728    later and other features added - see item 49 below.]
2729    
2730    29. pcretest is upgraded to test the callout functionality. It provides a
2731    callout function that displays information. By default, it shows the start of
2732    the match and the current position in the text. There are some new data escapes
2733    to vary what happens:
2734    
2735        \C+         in addition, show current contents of captured substrings
2736        \C-         do not supply a callout function
2737        \C!n        return 1 when callout number n is reached
2738        \C!n!m      return 1 when callout number n is reached for the mth time
2739    
2740    30. If pcregrep was called with the -l option and just a single file name, it
2741    output "<stdin>" if a match was found, instead of the file name.
2742    
2743    31. Improve the efficiency of the POSIX API to PCRE. If the number of capturing
2744    slots is less than POSIX_MALLOC_THRESHOLD, use a block on the stack to pass to
2745    pcre_exec(). This saves a malloc/free per call. The default value of
2746    POSIX_MALLOC_THRESHOLD is 10; it can be changed by --with-posix-malloc-threshold
2747    when configuring.
2748    
2749    32. The default maximum size of a compiled pattern is 64K. There have been a
2750    few cases of people hitting this limit. The code now uses macros to handle the
2751    storing of links as offsets within the compiled pattern. It defaults to 2-byte
2752    links, but this can be changed to 3 or 4 bytes by --with-link-size when
2753    configuring. Tests 2 and 5 work only with 2-byte links because they output
2754    debugging information about compiled patterns.
2755    
2756    33. Internal code re-arrangements:
2757    
2758    (a) Moved the debugging function for printing out a compiled regex into
2759        its own source file (printint.c) and used #include to pull it into
2760        pcretest.c and, when DEBUG is defined, into pcre.c, instead of having two
2761        separate copies.
2762    
2763    (b) Defined the list of op-code names for debugging as a macro in
2764        internal.h so that it is next to the definition of the opcodes.
2765    
2766    (c) Defined a table of op-code lengths for simpler skipping along compiled
2767        code. This is again a macro in internal.h so that it is next to the
2768        definition of the opcodes.
2769    
2770    34. Added support for recursive calls to individual subpatterns, along the
2771    lines of Robin Houston's patch (but implemented somewhat differently).
2772    
2773    35. Further mods to the Makefile to help Win32. Also, added code to pcregrep to
2774    allow it to read and process whole directories in Win32. This code was
2775    contributed by Lionel Fourquaux; it has not been tested by me.
2776    
2777    36. Added support for named subpatterns. The Python syntax (?P<name>...) is
2778    used to name a group. Names consist of alphanumerics and underscores, and must
2779    be unique. Back references use the syntax (?P=name) and recursive calls use
2780    (?P>name) which is a PCRE extension to the Python extension. Groups still have
2781    numbers. The function pcre_fullinfo() can be used after compilation to extract
2782    a name/number map. There are three relevant calls:
2783    
2784      PCRE_INFO_NAMEENTRYSIZE        yields the size of each entry in the map
2785      PCRE_INFO_NAMECOUNT            yields the number of entries
2786      PCRE_INFO_NAMETABLE            yields a pointer to the map.
2787    
2788    The map is a vector of fixed-size entries. The size of each entry depends on
2789    the length of the longest name used. The first two bytes of each entry are the
2790    group number, most significant byte first. There follows the corresponding
2791    name, zero terminated. The names are in alphabetical order.
2792    
2793    37. Make the maximum literal string in the compiled code 250 for the non-UTF-8
2794    case instead of 255. Making it the same both with and without UTF-8 support
2795    means that the same test output works with both.
2796    
2797    38. There was a case of malloc(0) in the POSIX testing code in pcretest. Avoid
2798    calling malloc() with a zero argument.
2799    
2800    39. Change 25 above had to resort to a heavy-handed test for the .* anchoring
2801    optimization. I've improved things by keeping a bitmap of backreferences with
2802    numbers 1-31 so that if .* occurs inside capturing brackets that are not in
2803    fact referenced, the optimization can be applied. It is unlikely that a
2804    relevant occurrence of .* (i.e. one which might indicate anchoring or forcing
2805    the match to follow \n) will appear inside brackets with a number greater than
2806    31, but if it does, any back reference > 31 suppresses the optimization.
2807    
2808    40. Added a new compile-time option PCRE_NO_AUTO_CAPTURE. This has the effect
2809    of disabling numbered capturing parentheses. Any opening parenthesis that is
2810    not followed by ? behaves as if it were followed by ?: but named parentheses
2811    can still be used for capturing (and they will acquire numbers in the usual
2812    way).
2813    
2814    41. Redesigned the return codes from the match() function into yes/no/error so
2815    that errors can be passed back from deep inside the nested calls. A malloc
2816    failure while inside a recursive subpattern call now causes the
2817    PCRE_ERROR_NOMEMORY return instead of quietly going wrong.
2818    
2819    42. It is now possible to set a limit on the number of times the match()
2820    function is called in a call to pcre_exec(). This facility makes it possible to
2821    limit the amount of recursion and backtracking, though not in a directly
2822    obvious way, because the match() function is used in a number of different
2823    circumstances. The count starts from zero for each position in the subject
2824    string (for non-anchored patterns). The default limit is, for compatibility, a
2825    large number, namely 10 000 000. You can change this in two ways:
2826    
2827    (a) When configuring PCRE before making, you can use --with-match-limit=n
2828        to set a default value for the compiled library.
2829    
2830    (b) For each call to pcre_exec(), you can pass a pcre_extra block in which
2831        a different value is set. See 45 below.
2832    
2833    If the limit is exceeded, pcre_exec() returns PCRE_ERROR_MATCHLIMIT.
2834    
2835    43. Added a new function pcre_config(int, void *) to enable run-time extraction
2836    of things that can be changed at compile time. The first argument specifies
2837    what is wanted and the second points to where the information is to be placed.
2838    The current list of available information is:
2839    
2840      PCRE_CONFIG_UTF8
2841    
2842    The output is an integer that is set to one if UTF-8 support is available;
2843    otherwise it is set to zero.
2844    
2845      PCRE_CONFIG_NEWLINE
2846    
2847    The output is an integer that it set to the value of the code that is used for
2848    newline. It is either LF (10) or CR (13).
2849    
2850      PCRE_CONFIG_LINK_SIZE
2851    
2852    The output is an integer that contains the number of bytes used for internal
2853    linkage in compiled expressions. The value is 2, 3, or 4. See item 32 above.
2854    
2855      PCRE_CONFIG_POSIX_MALLOC_THRESHOLD
2856    
2857    The output is an integer that contains the threshold above which the POSIX
2858    interface uses malloc() for output vectors. See item 31 above.
2859    
2860      PCRE_CONFIG_MATCH_LIMIT
2861    
2862    The output is an unsigned integer that contains the default limit of the number
2863    of match() calls in a pcre_exec() execution. See 42 above.
2864    
2865    44. pcretest has been upgraded by the addition of the -C option. This causes it
2866    to extract all the available output from the new pcre_config() function, and to
2867    output it. The program then exits immediately.
2868    
2869    45. A need has arisen to pass over additional data with calls to pcre_exec() in
2870    order to support additional features. One way would have been to define
2871    pcre_exec2() (for example) with extra arguments, but this would not have been
2872    extensible, and would also have required all calls to the original function to
2873    be mapped to the new one. Instead, I have chosen to extend the mechanism that
2874    is used for passing in "extra" data from pcre_study().
2875    
2876    The pcre_extra structure is now exposed and defined in pcre.h. It currently
2877    contains the following fields:
2878    
2879      flags         a bitmap indicating which of the following fields are set
2880      study_data    opaque data from pcre_study()
2881      match_limit   a way of specifying a limit on match() calls for a specific
2882                      call to pcre_exec()
2883      callout_data  data for callouts (see 49 below)
2884    
2885    The flag bits are also defined in pcre.h, and are
2886    
2887      PCRE_EXTRA_STUDY_DATA
2888      PCRE_EXTRA_MATCH_LIMIT
2889      PCRE_EXTRA_CALLOUT_DATA
2890    
2891    The pcre_study() function now returns one of these new pcre_extra blocks, with
2892    the actual study data pointed to by the study_data field, and the
2893    PCRE_EXTRA_STUDY_DATA flag set. This can be passed directly to pcre_exec() as
2894    before. That is, this change is entirely upwards-compatible and requires no
2895    change to existing code.
2896    
2897    If you want to pass in additional data to pcre_exec(), you can either place it
2898    in a pcre_extra block provided by pcre_study(), or create your own pcre_extra
2899    block.
2900    
2901    46. pcretest has been extended to test the PCRE_EXTRA_MATCH_LIMIT feature. If a
2902    data string contains the escape sequence \M, pcretest calls pcre_exec() several
2903    times with different match limits, until it finds the minimum value needed for
2904    pcre_exec() to complete. The value is then output. This can be instructive; for
2905    most simple matches the number is quite small, but for pathological cases it
2906    gets very large very quickly.
2907    
2908    47. There's a new option for pcre_fullinfo() called PCRE_INFO_STUDYSIZE. It
2909    returns the size of the data block pointed to by the study_data field in a
2910    pcre_extra block, that is, the value that was passed as the argument to
2911    pcre_malloc() when PCRE was getting memory in which to place the information
2912    created by pcre_study(). The fourth argument should point to a size_t variable.
2913    pcretest has been extended so that this information is shown after a successful
2914    pcre_study() call when information about the compiled regex is being displayed.
2915    
2916    48. Cosmetic change to Makefile: there's no need to have / after $(DESTDIR)
2917    because what follows is always an absolute path. (Later: it turns out that this
2918    is more than cosmetic for MinGW, because it doesn't like empty path
2919    components.)
2920    
2921    49. Some changes have been made to the callout feature (see 28 above):
2922    
2923    (i)  A callout function now has three choices for what it returns:
2924    
2925           0  =>  success, carry on matching
2926         > 0  =>  failure at this point, but backtrack if possible
2927         < 0  =>  serious error, return this value from pcre_exec()
2928    
2929         Negative values should normally be chosen from the set of PCRE_ERROR_xxx
2930         values. In particular, returning PCRE_ERROR_NOMATCH forces a standard
2931         "match failed" error. The error number PCRE_ERROR_CALLOUT is reserved for
2932         use by callout functions. It will never be used by PCRE itself.
2933    
2934    (ii) The pcre_extra structure (see 45 above) has a void * field called
2935         callout_data, with corresponding flag bit PCRE_EXTRA_CALLOUT_DATA. The
2936         pcre_callout_block structure has a field of the same name. The contents of
2937         the field passed in the pcre_extra structure are passed to the callout
2938         function in the corresponding field in the callout block. This makes it
2939         easier to use the same callout-containing regex from multiple threads. For
2940         testing, the pcretest program has a new data escape
2941    
2942           \C*n        pass the number n (may be negative) as callout_data
2943    
2944         If the callout function in pcretest receives a non-zero value as
2945         callout_data, it returns that value.
2946    
2947    50. Makefile wasn't handling CFLAGS properly when compiling dftables. Also,
2948    there were some redundant $(CFLAGS) in commands that are now specified as
2949    $(LINK), which already includes $(CFLAGS).
2950    
2951    51. Extensions to UTF-8 support are listed below. These all apply when (a) PCRE
2952    has been compiled with UTF-8 support *and* pcre_compile() has been compiled
2953    with the PCRE_UTF8 flag. Patterns that are compiled without that flag assume
2954    one-byte characters throughout. Note that case-insensitive matching applies
2955    only to characters whose values are less than 256. PCRE doesn't support the
2956    notion of cases for higher-valued characters.
2957    
2958    (i)   A character class whose characters are all within 0-255 is handled as
2959          a bit map, and the map is inverted for negative classes. Previously, a
2960          character > 255 always failed to match such a class; however it should
2961          match if the class was a negative one (e.g. [^ab]). This has been fixed.
2962    
2963    (ii)  A negated character class with a single character < 255 is coded as
2964          "not this character" (OP_NOT). This wasn't working properly when the test
2965          character was multibyte, either singly or repeated.
2966    
2967    (iii) Repeats of multibyte characters are now handled correctly in UTF-8
2968          mode, for example: \x{100}{2,3}.
2969    
2970    (iv)  The character escapes \b, \B, \d, \D, \s, \S, \w, and \W (either
2971          singly or repeated) now correctly test multibyte characters. However,
2972          PCRE doesn't recognize any characters with values greater than 255 as
2973          digits, spaces, or word characters. Such characters always match \D, \S,
2974          and \W, and never match \d, \s, or \w.
2975    
2976    (v)   Classes may now contain characters and character ranges with values
2977          greater than 255. For example: [ab\x{100}-\x{400}].
2978    
2979    (vi)  pcregrep now has a --utf-8 option (synonym -u) which makes it call
2980          PCRE in UTF-8 mode.
2981    
2982    52. The info request value PCRE_INFO_FIRSTCHAR has been renamed
2983    PCRE_INFO_FIRSTBYTE because it is a byte value. However, the old name is
2984    retained for backwards compatibility. (Note that LASTLITERAL is also a byte
2985    value.)
2986    
2987    53. The single man page has become too large. I have therefore split it up into
2988    a number of separate man pages. These also give rise to individual HTML pages;
2989    these are now put in a separate directory, and there is an index.html page that
2990    lists them all. Some hyperlinking between the pages has been installed.
2991    
2992    54. Added convenience functions for handling named capturing parentheses.
2993    
2994    55. Unknown escapes inside character classes (e.g. [\M]) and escapes that
2995    aren't interpreted therein (e.g. [\C]) are literals in Perl. This is now also
2996    true in PCRE, except when the PCRE_EXTENDED option is set, in which case they
2997    are faulted.
2998    
2999    56. Introduced HOST_CC and HOST_CFLAGS which can be set in the environment when
3000    calling configure. These values are used when compiling the dftables.c program
3001    which is run to generate the source of the default character tables. They
3002    default to the values of CC and CFLAGS. If you are cross-compiling PCRE,
3003    you will need to set these values.
3004    
3005    57. Updated the building process for Windows DLL, as provided by Fred Cox.
3006    
3007    
3008    Version 3.9 02-Jan-02
3009    ---------------------
3010    
3011    1. A bit of extraneous text had somehow crept into the pcregrep documentation.
3012    
3013    2. If --disable-static was given, the building process failed when trying to
3014    build pcretest and pcregrep. (For some reason it was using libtool to compile
3015    them, which is not right, as they aren't part of the library.)
3016    
3017    
3018    Version 3.8 18-Dec-01
3019    ---------------------
3020    
3021    1. The experimental UTF-8 code was completely screwed up. It was packing the
3022    bytes in the wrong order. How dumb can you get?
3023    
3024    
3025    Version 3.7 29-Oct-01
3026    ---------------------
3027    
3028    1. In updating pcretest to check change 1 of version 3.6, I screwed up.
3029    This caused pcretest, when used on the test data, to segfault. Unfortunately,
3030    this didn't happen under Solaris 8, where I normally test things.
3031    
3032    2. The Makefile had to be changed to make it work on BSD systems, where 'make'
3033    doesn't seem to recognize that ./xxx and xxx are the same file. (This entry
3034    isn't in ChangeLog distributed with 3.7 because I forgot when I hastily made
3035    this fix an hour or so after the initial 3.7 release.)
3036    
3037    
3038    Version 3.6 23-Oct-01
3039    ---------------------
3040    
3041    1. Crashed with /(sens|respons)e and \1ibility/ and "sense and sensibility" if
3042    offsets passed as NULL with zero offset count.
3043    
3044    2. The config.guess and config.sub files had not been updated when I moved to
3045    the latest autoconf.
3046    
3047    
3048    Version 3.5 15-Aug-01
3049    ---------------------
3050    
3051    1. Added some missing #if !defined NOPOSIX conditionals in pcretest.c that
3052    had been forgotten.
3053    
3054    2. By using declared but undefined structures, we can avoid using "void"
3055    definitions in pcre.h while keeping the internal definitions of the structures
3056    private.
3057    
3058    3. The distribution is now built using autoconf 2.50 and libtool 1.4. From a
3059    user point of view, this means that both static and shared libraries are built
3060    by default, but this can be individually controlled. More of the work of
3061    handling this static/shared cases is now inside libtool instead of PCRE's make
3062    file.
3063    
3064    4. The pcretest utility is now installed along with pcregrep because it is
3065    useful for users (to test regexs) and by doing this, it automatically gets
3066    relinked by libtool. The documentation has been turned into a man page, so
3067    there are now .1, .txt, and .html versions in /doc.
3068    
3069    5. Upgrades to pcregrep:
3070       (i)   Added long-form option names like gnu grep.
3071       (ii)  Added --help to list all options with an explanatory phrase.
3072       (iii) Added -r, --recursive to recurse into sub-directories.
3073       (iv)  Added -f, --file to read patterns from a file.
3074    
3075    6. pcre_exec() was referring to its "code" argument before testing that
3076    argument for NULL (and giving an error if it was NULL).
3077    
3078    7. Upgraded Makefile.in to allow for compiling in a different directory from
3079    the source directory.
3080    
3081    8. Tiny buglet in pcretest: when pcre_fullinfo() was called to retrieve the
3082    options bits, the pointer it was passed was to an int instead of to an unsigned
3083    long int. This mattered only on 64-bit systems.
3084    
3085    9. Fixed typo (3.4/1) in pcre.h again. Sigh. I had changed pcre.h (which is
3086    generated) instead of pcre.in, which it its source. Also made the same change
3087    in several of the .c files.
3088    
3089    10. A new release of gcc defines printf() as a macro, which broke pcretest
3090    because it had an ifdef in the middle of a string argument for printf(). Fixed
3091    by using separate calls to printf().
3092    
3093    11. Added --enable-newline-is-cr and --enable-newline-is-lf to the configure
3094    script, to force use of CR or LF instead of \n in the source. On non-Unix
3095    systems, the value can be set in config.h.
3096    
3097    12. The limit of 200 on non-capturing parentheses is a _nesting_ limit, not an
3098    absolute limit. Changed the text of the error message to make this clear, and
3099    likewise updated the man page.
3100    
3101    13. The limit of 99 on the number of capturing subpatterns has been removed.
3102    The new limit is 65535, which I hope will not be a "real" limit.
3103    
3104    
3105    Version 3.4 22-Aug-00
3106    ---------------------
3107    
3108    1. Fixed typo in pcre.h: unsigned const char * changed to const unsigned char *.
3109    
3110    2. Diagnose condition (?(0) as an error instead of crashing on matching.
3111    
3112    
3113    Version 3.3 01-Aug-00
3114    ---------------------
3115    
3116    1. If an octal character was given, but the value was greater than \377, it
3117    was not getting masked to the least significant bits, as documented. This could
3118    lead to crashes in some systems.
3119    
3120    2. Perl 5.6 (if not earlier versions) accepts classes like [a-\d] and treats
3121    the hyphen as a literal. PCRE used to give an error; it now behaves like Perl.
3122    
3123    3. Added the functions pcre_free_substring() and pcre_free_substring_list().
3124    These just pass their arguments on to (pcre_free)(), but they are provided
3125    because some uses of PCRE bind it to non-C systems that can call its functions,
3126    but cannot call free() or pcre_free() directly.
3127    
3128    4. Add "make test" as a synonym for "make check". Corrected some comments in
3129    the Makefile.
3130    
3131    5. Add $(DESTDIR)/ in front of all the paths in the "install" target in the
3132    Makefile.
3133    
3134    6. Changed the name of pgrep to pcregrep, because Solaris has introduced a
3135    command called pgrep for grepping around the active processes.
3136    
3137    7. Added the beginnings of support for UTF-8 character strings.
3138    
3139    8. Arranged for the Makefile to pass over the settings of CC, CFLAGS, and
3140    RANLIB to ./ltconfig so that they are used by libtool. I think these are all
3141    the relevant ones. (AR is not passed because ./ltconfig does its own figuring
3142    out for the ar command.)
3143    
3144    
3145    Version 3.2 12-May-00
3146    ---------------------
3147    
3148    This is purely a bug fixing release.
3149    
3150    1. If the pattern /((Z)+|A)*/ was matched agained ZABCDEFG it matched Z instead
3151    of ZA. This was just one example of several cases that could provoke this bug,
3152    which was introduced by change 9 of version 2.00. The code for breaking
3153    infinite loops after an iteration that matches an empty string was't working
3154    correctly.
3155    
3156    2. The pcretest program was not imitating Perl correctly for the pattern /a*/g
3157    when matched against abbab (for example). After matching an empty string, it
3158    wasn't forcing anchoring when setting PCRE_NOTEMPTY for the next attempt; this
3159    caused it to match further down the string than it should.
3160    
3161    3. The code contained an inclusion of sys/types.h. It isn't clear why this
3162    was there because it doesn't seem to be needed, and it causes trouble on some
3163    systems, as it is not a Standard C header. It has been removed.
3164    
3165    4. Made 4 silly changes to the source to avoid stupid compiler warnings that
3166    were reported on the Macintosh. The changes were from
3167    
3168      while ((c = *(++ptr)) != 0 && c != '\n');
3169    to
3170      while ((c = *(++ptr)) != 0 && c != '\n') ;
3171    
3172    Totally extraordinary, but if that's what it takes...
3173    
3174    5. PCRE is being used in one environment where neither memmove() nor bcopy() is
3175    available. Added HAVE_BCOPY and an autoconf test for it; if neither
3176    HAVE_MEMMOVE nor HAVE_BCOPY is set, use a built-in emulation function which
3177    assumes the way PCRE uses memmove() (always moving upwards).
3178    
3179    6. PCRE is being used in one environment where strchr() is not available. There
3180    was only one use in pcre.c, and writing it out to avoid strchr() probably gives
3181    faster code anyway.
3182    
3183    
3184    Version 3.1 09-Feb-00
3185    ---------------------
3186    
3187    The only change in this release is the fixing of some bugs in Makefile.in for
3188    the "install" target:
3189    
3190    (1) It was failing to install pcreposix.h.
3191    
3192    (2) It was overwriting the pcre.3 man page with the pcreposix.3 man page.
3193    
3194    
3195    Version 3.0 01-Feb-00
3196    ---------------------
3197    
3198    1. Add support for the /+ modifier to perltest (to output $` like it does in
3199    pcretest).
3200    
3201    2. Add support for the /g modifier to perltest.
3202    
3203    3. Fix pcretest so that it behaves even more like Perl for /g when the pattern
3204    matches null strings.
3205    
3206    4. Fix perltest so that it doesn't do unwanted things when fed an empty
3207    pattern. Perl treats empty patterns specially - it reuses the most recent
3208    pattern, which is not what we want. Replace // by /(?#)/ in order to avoid this
3209    effect.
3210    
3211    5. The POSIX interface was broken in that it was just handing over the POSIX
3212    captured string vector to pcre_exec(), but (since release 2.00) PCRE has
3213    required a bigger vector, with some working space on the end. This means that
3214    the POSIX wrapper now has to get and free some memory, and copy the results.
3215    
3216    6. Added some simple autoconf support, placing the test data and the
3217    documentation in separate directories, re-organizing some of the
3218    information files, and making it build pcre-config (a GNU standard). Also added
3219    libtool support for building PCRE as a shared library, which is now the
3220    default.
3221    
3222    7. Got rid of the leading zero in the definition of PCRE_MINOR because 08 and
3223    09 are not valid octal constants. Single digits will be used for minor values
3224    less than 10.
3225    
3226    8. Defined REG_EXTENDED and REG_NOSUB as zero in the POSIX header, so that
3227    existing programs that set these in the POSIX interface can use PCRE without
3228    modification.
3229    
3230    9. Added a new function, pcre_fullinfo() with an extensible interface. It can
3231    return all that pcre_info() returns, plus additional data. The pcre_info()
3232    function is retained for compatibility, but is considered to be obsolete.
3233    
3234    10. Added experimental recursion feature (?R) to handle one common case that
3235    Perl 5.6 will be able to do with (?p{...}).
3236    
3237    11. Added support for POSIX character classes like [:alpha:], which Perl is
3238    adopting.
3239    
3240    
3241    Version 2.08 31-Aug-99
3242    ----------------------
3243    
3244    1. When startoffset was not zero and the pattern began with ".*", PCRE was not
3245    trying to match at the startoffset position, but instead was moving forward to
3246    the next newline as if a previous match had failed.
3247    
3248    2. pcretest was not making use of PCRE_NOTEMPTY when repeating for /g and /G,
3249    and could get into a loop if a null string was matched other than at the start
3250    of the subject.
3251    
3252    3. Added definitions of PCRE_MAJOR and PCRE_MINOR to pcre.h so the version can
3253    be distinguished at compile time, and for completeness also added PCRE_DATE.
3254    
3255    5. Added Paul Sokolovsky's minor changes to make it easy to compile a Win32 DLL
3256    in GnuWin32 environments.
3257    
3258    
3259    Version 2.07 29-Jul-99
3260    ----------------------
3261    
3262    1. The documentation is now supplied in plain text form and HTML as well as in
3263    the form of man page sources.
3264    
3265    2. C++ compilers don't like assigning (void *) values to other pointer types.
3266    In particular this affects malloc(). Although there is no problem in Standard
3267    C, I've put in casts to keep C++ compilers happy.
3268    
3269    3. Typo on pcretest.c; a cast of (unsigned char *) in the POSIX regexec() call
3270    should be (const char *).
3271    
3272    4. If NOPOSIX is defined, pcretest.c compiles without POSIX support. This may
3273    be useful for non-Unix systems who don't want to bother with the POSIX stuff.
3274    However, I haven't made this a standard facility. The documentation doesn't
3275    mention it, and the Makefile doesn't support it.
3276    
3277    5. The Makefile now contains an "install" target, with editable destinations at
3278    the top of the file. The pcretest program is not installed.
3279    
3280    6. pgrep -V now gives the PCRE version number and date.
3281    
3282    7. Fixed bug: a zero repetition after a literal string (e.g. /abcde{0}/) was
3283    causing the entire string to be ignored, instead of just the last character.
3284    
3285    8. If a pattern like /"([^\\"]+|\\.)*"/ is applied in the normal way to a
3286    non-matching string, it can take a very, very long time, even for strings of
3287    quite modest length, because of the nested recursion. PCRE now does better in
3288    some of these cases. It does this by remembering the last required literal
3289    character in the pattern, and pre-searching the subject to ensure it is present
3290    before running the real match. In other words, it applies a heuristic to detect
3291    some types of certain failure quickly, and in the above example, if presented
3292    with a string that has no trailing " it gives "no match" very quickly.
3293    
3294    9. A new runtime option PCRE_NOTEMPTY causes null string matches to be ignored;
3295    other alternatives are tried instead.
3296    
3297    
3298    Version 2.06 09-Jun-99
3299    ----------------------
3300    
3301    1. Change pcretest's output for amount of store used to show just the code
3302    space, because the remainder (the data block) varies in size between 32-bit and
3303    64-bit systems.
3304    
3305    2. Added an extra argument to pcre_exec() to supply an offset in the subject to
3306    start matching at. This allows lookbehinds to work when searching for multiple
3307    occurrences in a string.
3308    
3309    3. Added additional options to pcretest for testing multiple occurrences:
3310    
3311       /+   outputs the rest of the string that follows a match
3312       /g   loops for multiple occurrences, using the new startoffset argument
3313       /G   loops for multiple occurrences by passing an incremented pointer
3314    
3315    4. PCRE wasn't doing the "first character" optimization for patterns starting
3316    with \b or \B, though it was doing it for other lookbehind assertions. That is,
3317    it wasn't noticing that a match for a pattern such as /\bxyz/ has to start with
3318    the letter 'x'. On long subject strings, this gives a significant speed-up.
3319    
3320    
3321    Version 2.05 21-Apr-99
3322    ----------------------
3323    
3324    1. Changed the type of magic_number from int to long int so that it works
3325    properly on 16-bit systems.
3326    
3327    2. Fixed a bug which caused patterns starting with .* not to work correctly
3328    when the subject string contained newline characters. PCRE was assuming
3329    anchoring for such patterns in all cases, which is not correct because .* will
3330    not pass a newline unless PCRE_DOTALL is set. It now assumes anchoring only if
3331    DOTALL is set at top level; otherwise it knows that patterns starting with .*
3332    must be retried after every newline in the subject.
3333    
3334    
3335    Version 2.04 18-Feb-99
3336    ----------------------
3337    
3338    1. For parenthesized subpatterns with repeats whose minimum was zero, the
3339    computation of the store needed to hold the pattern was incorrect (too large).
3340    If such patterns were nested a few deep, this could multiply and become a real
3341    problem.
3342    
3343    2. Added /M option to pcretest to show the memory requirement of a specific
3344    pattern. Made -m a synonym of -s (which does this globally) for compatibility.
3345    
3346    3. Subpatterns of the form (regex){n,m} (i.e. limited maximum) were being
3347    compiled in such a way that the backtracking after subsequent failure was
3348    pessimal. Something like (a){0,3} was compiled as (a)?(a)?(a)? instead of
3349    ((a)((a)(a)?)?)? with disastrous performance if the maximum was of any size.
3350    
3351    
3352    Version 2.03 02-Feb-99
3353    ----------------------
3354    
3355    1. Fixed typo and small mistake in man page.
3356    
3357    2. Added 4th condition (GPL supersedes if conflict) and created separate
3358    LICENCE file containing the conditions.
3359    
3360    3. Updated pcretest so that patterns such as /abc\/def/ work like they do in
3361    Perl, that is the internal \ allows the delimiter to be included in the
3362    pattern. Locked out the use of \ as a delimiter. If \ immediately follows
3363    the final delimiter, add \ to the end of the pattern (to test the error).
3364    
3365    4. Added the convenience functions for extracting substrings after a successful
3366    match. Updated pcretest to make it able to test these functions.
3367    
3368    
3369    Version 2.02 14-Jan-99
3370    ----------------------
3371    
3372    1. Initialized the working variables associated with each extraction so that
3373    their saving and restoring doesn't refer to uninitialized store.
3374    
3375    2. Put dummy code into study.c in order to trick the optimizer of the IBM C
3376    compiler for OS/2 into generating correct code. Apparently IBM isn't going to
3377    fix the problem.
3378    
3379    3. Pcretest: the timing code wasn't using LOOPREPEAT for timing execution
3380    calls, and wasn't printing the correct value for compiling calls. Increased the
3381    default value of LOOPREPEAT, and the number of significant figures in the
3382    times.
3383    
3384    4. Changed "/bin/rm" in the Makefile to "-rm" so it works on Windows NT.
3385    
3386    5. Renamed "deftables" as "dftables" to get it down to 8 characters, to avoid
3387    a building problem on Windows NT with a FAT file system.
3388    
3389    
3390    Version 2.01 21-Oct-98
3391    ----------------------
3392    
3393    1. Changed the API for pcre_compile() to allow for the provision of a pointer
3394    to character tables built by pcre_maketables() in the current locale. If NULL
3395    is passed, the default tables are used.
3396    
3397    
3398    Version 2.00 24-Sep-98
3399    ----------------------
3400    
3401    1. Since the (>?) facility is in Perl 5.005, don't require PCRE_EXTRA to enable
3402    it any more.
3403    
3404    2. Allow quantification of (?>) groups, and make it work correctly.
3405    
3406    3. The first character computation wasn't working for (?>) groups.
3407    
3408    4. Correct the implementation of \Z (it is permitted to match on the \n at the
3409    end of the subject) and add 5.005's \z, which really does match only at the
3410    very end of the subject.
3411    
3412    5. Remove the \X "cut" facility; Perl doesn't have it, and (?> is neater.
3413    
3414    6. Remove the ability to specify CASELESS, MULTILINE, DOTALL, and
3415    DOLLAR_END_ONLY at runtime, to make it possible to implement the Perl 5.005
3416    localized options. All options to pcre_study() were also removed.
3417    
3418    7. Add other new features from 5.005:
3419    
3420       $(?<=           positive lookbehind
3421       $(?<!           negative lookbehind
3422       (?imsx-imsx)    added the unsetting capability
3423                       such a setting is global if at outer level; local otherwise
3424       (?imsx-imsx:)   non-capturing groups with option setting
3425       (?(cond)re|re)  conditional pattern matching
3426    
3427       A backreference to itself in a repeated group matches the previous
3428       captured string.
3429    
3430    8. General tidying up of studying (both automatic and via "study")
3431    consequential on the addition of new assertions.
3432    
3433    9. As in 5.005, unlimited repeated groups that could match an empty substring
3434    are no longer faulted at compile time. Instead, the loop is forcibly broken at
3435    runtime if any iteration does actually match an empty substring.
3436    
3437    10. Include the RunTest script in the distribution.
3438    
3439    11. Added tests from the Perl 5.005_02 distribution. This showed up a few
3440    discrepancies, some of which were old and were also with respect to 5.004. They
3441    have now been fixed.
3442    
3443    
3444    Version 1.09 28-Apr-98
3445    ----------------------
3446    
3447    1. A negated single character class followed by a quantifier with a minimum
3448    value of one (e.g.  [^x]{1,6}  ) was not compiled correctly. This could lead to
3449    program crashes, or just wrong answers. This did not apply to negated classes
3450    containing more than one character, or to minima other than one.
3451    
3452    
3453  Version 1.08 27-Mar-98  Version 1.08 27-Mar-98
3454  ----------------------  ----------------------

Legend:
Removed from v.19  
changed lines
  Added in v.480

  ViewVC Help
Powered by ViewVC 1.1.5