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

Diff of /code/trunk/ChangeLog

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

revision 59 by nigel, Sat Feb 24 21:39:54 2007 UTC revision 528 by ph10, Sat May 29 16:40:22 2010 UTC
# Line 1  Line 1 
1  ChangeLog for PCRE  ChangeLog for PCRE
2  ------------------  ------------------
3    
4    Version 8.10 03 May-2010
5    ------------------------
6    
7    1.  Added support for (*MARK:ARG) and for ARG additions to PRUNE, SKIP, and
8        THEN.
9    
10    2.  (*ACCEPT) was not working when inside an atomic group.
11    
12    3.  Inside a character class, \B is treated as a literal by default, but
13        faulted if PCRE_EXTRA is set. This mimics Perl's behaviour (the -w option
14        causes the error). The code is unchanged, but I tidied the documentation.
15    
16    4.  Inside a character class, PCRE always treated \R and \X as literals,
17        whereas Perl faults them if its -w option is set. I have changed PCRE so
18        that it faults them when PCRE_EXTRA is set.
19    
20    5.  Added support for \N, which always matches any character other than
21        newline. (It is the same as "." when PCRE_DOTALL is not set.)
22    
23    6.  When compiling pcregrep with newer versions of gcc which may have
24        FORTIFY_SOURCE set, several warnings "ignoring return value of 'fwrite',
25        declared with attribute warn_unused_result" were given. Just casting the
26        result to (void) does not stop the warnings; a more elaborate fudge is
27        needed. I've used a macro to implement this.
28    
29    7.  Minor change to pcretest.c to avoid a compiler warning.
30    
31    8.  Added four artifical Unicode properties to help with an option to make
32        \s etc use properties (see next item). The new properties are: Xan
33        (alphanumeric), Xsp (Perl space), Xps (POSIX space), and Xwd (word).
34    
35    9.  Added PCRE_UCP to make \b, \d, \s, \w, and certain POSIX character classes
36        use Unicode properties. (*UCP) at the start of a pattern can be used to set
37        this option. Modified pcretest to add /W to test this facility. Added
38        REG_UCP to make it available via the POSIX interface.
39    
40    10. Added --line-buffered to pcregrep.
41    
42    11. In UTF-8 mode, if a pattern that was compiled with PCRE_CASELESS was
43        studied, and the match started with a letter with a code point greater than
44        127 whose first byte was different to the first byte of the other case of
45        the letter, the other case of this starting letter was not recognized.
46    
47    12. pcreposix.c included pcre.h before including pcre_internal.h. This caused a
48        conflict in the definition of PCRE_EXP_DECL. I have removed the include of
49        pcre.h as pcre_internal.h includes pcre.h itself. (This may be a bit of
50        historical tidying that never got done.)
51    
52    13. If a pattern that was studied started with a repeated Unicode property
53        test, for example, \p{Nd}+, there was the theoretical possibility of
54        setting up an incorrect bitmap of starting bytes, but fortunately it could
55        not have actually happened in practice until change 8 above was made (it
56        added property types that matched character-matching opcodes).
57    
58    14. pcre_study() now recognizes \h, \v, and \R when constructing a bit map of
59        possible starting bytes for non-anchored patterns.
60    
61    15. The "auto-possessify" feature of pcre_compile() now recognizes \R.
62    
63    
64    
65    Version 8.02 19-Mar-2010
66    ------------------------
67    
68    1.  The Unicode data tables have been updated to Unicode 5.2.0.
69    
70    2.  Added the option --libs-cpp to pcre-config, but only when C++ support is
71        configured.
72    
73    3.  Updated the licensing terms in the pcregexp.pas file, as agreed with the
74        original author of that file, following a query about its status.
75    
76    4.  On systems that do not have stdint.h (e.g. Solaris), check for and include
77        inttypes.h instead. This fixes a bug that was introduced by change 8.01/8.
78    
79    5.  A pattern such as (?&t)*+(?(DEFINE)(?<t>.)) which has a possessive
80        quantifier applied to a forward-referencing subroutine call, could compile
81        incorrect code or give the error "internal error: previously-checked
82        referenced subpattern not found".
83    
84    6.  Both MS Visual Studio and Symbian OS have problems with initializing
85        variables to point to external functions. For these systems, therefore,
86        pcre_malloc etc. are now initialized to local functions that call the
87        relevant global functions.
88    
89    7.  There were two entries missing in the vectors called coptable and poptable
90        in pcre_dfa_exec.c. This could lead to memory accesses outsize the vectors.
91        I've fixed the data, and added a kludgy way of testing at compile time that
92        the lengths are correct (equal to the number of opcodes).
93    
94    8.  Following on from 7, I added a similar kludge to check the length of the
95        eint vector in pcreposix.c.
96    
97    9.  Error texts for pcre_compile() are held as one long string to avoid too
98        much relocation at load time. To find a text, the string is searched,
99        counting zeros. There was no check for running off the end of the string,
100        which could happen if a new error number was added without updating the
101        string.
102    
103    10. \K gave a compile-time error if it appeared in a lookbehind assersion.
104    
105    11. \K was not working if it appeared in an atomic group or in a group that
106        was called as a "subroutine", or in an assertion. Perl 5.11 documents that
107        \K is "not well defined" if used in an assertion. PCRE now accepts it if
108        the assertion is positive, but not if it is negative.
109    
110    12. Change 11 fortuitously reduced the size of the stack frame used in the
111        "match()" function of pcre_exec.c by one pointer. Forthcoming
112        implementation of support for (*MARK) will need an extra pointer on the
113        stack; I have reserved it now, so that the stack frame size does not
114        decrease.
115    
116    13. A pattern such as (?P<L1>(?P<L2>0)|(?P>L2)(?P>L1)) in which the only other
117        item in branch that calls a recursion is a subroutine call - as in the
118        second branch in the above example - was incorrectly given the compile-
119        time error "recursive call could loop indefinitely" because pcre_compile()
120        was not correctly checking the subroutine for matching a non-empty string.
121    
122    14. The checks for overrunning compiling workspace could trigger after an
123        overrun had occurred. This is a "should never occur" error, but it can be
124        triggered by pathological patterns such as hundreds of nested parentheses.
125        The checks now trigger 100 bytes before the end of the workspace.
126    
127    15. Fix typo in configure.ac: "srtoq" should be "strtoq".
128    
129    
130    Version 8.01 19-Jan-2010
131    ------------------------
132    
133    1.  If a pattern contained a conditional subpattern with only one branch (in
134        particular, this includes all (*DEFINE) patterns), a call to pcre_study()
135        computed the wrong minimum data length (which is of course zero for such
136        subpatterns). This could cause incorrect "no match" results.
137    
138    2.  For patterns such as (?i)a(?-i)b|c where an option setting at the start of
139        the pattern is reset in the first branch, pcre_compile() failed with
140        "internal error: code overflow at offset...". This happened only when
141        the reset was to the original external option setting. (An optimization
142        abstracts leading options settings into an external setting, which was the
143        cause of this.)
144    
145    3.  A pattern such as ^(?!a(*SKIP)b) where a negative assertion contained one
146        of the verbs SKIP, PRUNE, or COMMIT, did not work correctly. When the
147        assertion pattern did not match (meaning that the assertion was true), it
148        was incorrectly treated as false if the SKIP had been reached during the
149        matching. This also applied to assertions used as conditions.
150    
151    4.  If an item that is not supported by pcre_dfa_exec() was encountered in an
152        assertion subpattern, including such a pattern used as a condition,
153        unpredictable results occurred, instead of the error return
154        PCRE_ERROR_DFA_UITEM.
155    
156    5.  The C++ GlobalReplace function was not working like Perl for the special
157        situation when an empty string is matched. It now does the fancy magic
158        stuff that is necessary.
159    
160    6.  In pcre_internal.h, obsolete includes to setjmp.h and stdarg.h have been
161        removed. (These were left over from very, very early versions of PCRE.)
162    
163    7.  Some cosmetic changes to the code to make life easier when compiling it
164        as part of something else:
165    
166        (a) Change DEBUG to PCRE_DEBUG.
167    
168        (b) In pcre_compile(), rename the member of the "branch_chain" structure
169            called "current" as "current_branch", to prevent a collision with the
170            Linux macro when compiled as a kernel module.
171    
172        (c) In pcre_study(), rename the function set_bit() as set_table_bit(), to
173            prevent a collision with the Linux macro when compiled as a kernel
174            module.
175    
176    8.  In pcre_compile() there are some checks for integer overflows that used to
177        cast potentially large values to (double). This has been changed to that
178        when building, a check for int64_t is made, and if it is found, it is used
179        instead, thus avoiding the use of floating point arithmetic. (There is no
180        other use of FP in PCRE.) If int64_t is not found, the fallback is to
181        double.
182    
183    9.  Added two casts to avoid signed/unsigned warnings from VS Studio Express
184        2005 (difference between two addresses compared to an unsigned value).
185    
186    10. Change the standard AC_CHECK_LIB test for libbz2 in configure.ac to a
187        custom one, because of the following reported problem in Windows:
188    
189          - libbz2 uses the Pascal calling convention (WINAPI) for the functions
190              under Win32.
191          - The standard autoconf AC_CHECK_LIB fails to include "bzlib.h",
192              therefore missing the function definition.
193          - The compiler thus generates a "C" signature for the test function.
194          - The linker fails to find the "C" function.
195          - PCRE fails to configure if asked to do so against libbz2.
196    
197    11. When running libtoolize from libtool-2.2.6b as part of autogen.sh, these
198        messages were output:
199    
200          Consider adding `AC_CONFIG_MACRO_DIR([m4])' to configure.ac and
201          rerunning libtoolize, to keep the correct libtool macros in-tree.
202          Consider adding `-I m4' to ACLOCAL_AMFLAGS in Makefile.am.
203    
204        I have done both of these things.
205    
206    12. Although pcre_dfa_exec() does not use nearly as much stack as pcre_exec()
207        most of the time, it *can* run out if it is given a pattern that contains a
208        runaway infinite recursion. I updated the discussion in the pcrestack man
209        page.
210    
211    13. Now that we have gone to the x.xx style of version numbers, the minor
212        version may start with zero. Using 08 or 09 is a bad idea because users
213        might check the value of PCRE_MINOR in their code, and 08 or 09 may be
214        interpreted as invalid octal numbers. I've updated the previous comment in
215        configure.ac, and also added a check that gives an error if 08 or 09 are
216        used.
217    
218    14. Change 8.00/11 was not quite complete: code had been accidentally omitted,
219        causing partial matching to fail when the end of the subject matched \W
220        in a UTF-8 pattern where \W was quantified with a minimum of 3.
221    
222    15. There were some discrepancies between the declarations in pcre_internal.h
223        of _pcre_is_newline(), _pcre_was_newline(), and _pcre_valid_utf8() and
224        their definitions. The declarations used "const uschar *" and the
225        definitions used USPTR. Even though USPTR is normally defined as "const
226        unsigned char *" (and uschar is typedeffed as "unsigned char"), it was
227        reported that: "This difference in casting confuses some C++ compilers, for
228        example, SunCC recognizes above declarations as different functions and
229        generates broken code for hbpcre." I have changed the declarations to use
230        USPTR.
231    
232    16. GNU libtool is named differently on some systems. The autogen.sh script now
233        tries several variants such as glibtoolize (MacOSX) and libtoolize1x
234        (FreeBSD).
235    
236    17. Applied Craig's patch that fixes an HP aCC compile error in pcre 8.00
237        (strtoXX undefined when compiling pcrecpp.cc). The patch contains this
238        comment: "Figure out how to create a longlong from a string: strtoll and
239        equivalent. It's not enough to call AC_CHECK_FUNCS: hpux has a strtoll, for
240        instance, but it only takes 2 args instead of 3!"
241    
242    18. A subtle bug concerned with back references has been fixed by a change of
243        specification, with a corresponding code fix. A pattern such as
244        ^(xa|=?\1a)+$ which contains a back reference inside the group to which it
245        refers, was giving matches when it shouldn't. For example, xa=xaaa would
246        match that pattern. Interestingly, Perl (at least up to 5.11.3) has the
247        same bug. Such groups have to be quantified to be useful, or contained
248        inside another quantified group. (If there's no repetition, the reference
249        can never match.) The problem arises because, having left the group and
250        moved on to the rest of the pattern, a later failure that backtracks into
251        the group uses the captured value from the final iteration of the group
252        rather than the correct earlier one. I have fixed this in PCRE by forcing
253        any group that contains a reference to itself to be an atomic group; that
254        is, there cannot be any backtracking into it once it has completed. This is
255        similar to recursive and subroutine calls.
256    
257    
258    Version 8.00 19-Oct-09
259    ----------------------
260    
261    1.  The table for translating pcre_compile() error codes into POSIX error codes
262        was out-of-date, and there was no check on the pcre_compile() error code
263        being within the table. This could lead to an OK return being given in
264        error.
265    
266    2.  Changed the call to open a subject file in pcregrep from fopen(pathname,
267        "r") to fopen(pathname, "rb"), which fixed a problem with some of the tests
268        in a Windows environment.
269    
270    3.  The pcregrep --count option prints the count for each file even when it is
271        zero, as does GNU grep. However, pcregrep was also printing all files when
272        --files-with-matches was added. Now, when both options are given, it prints
273        counts only for those files that have at least one match. (GNU grep just
274        prints the file name in this circumstance, but including the count seems
275        more useful - otherwise, why use --count?) Also ensured that the
276        combination -clh just lists non-zero counts, with no names.
277    
278    4.  The long form of the pcregrep -F option was incorrectly implemented as
279        --fixed_strings instead of --fixed-strings. This is an incompatible change,
280        but it seems right to fix it, and I didn't think it was worth preserving
281        the old behaviour.
282    
283    5.  The command line items --regex=pattern and --regexp=pattern were not
284        recognized by pcregrep, which required --regex pattern or --regexp pattern
285        (with a space rather than an '='). The man page documented the '=' forms,
286        which are compatible with GNU grep; these now work.
287    
288    6.  No libpcreposix.pc file was created for pkg-config; there was just
289        libpcre.pc and libpcrecpp.pc. The omission has been rectified.
290    
291    7.  Added #ifndef SUPPORT_UCP into the pcre_ucd.c module, to reduce its size
292        when UCP support is not needed, by modifying the Python script that
293        generates it from Unicode data files. This should not matter if the module
294        is correctly used as a library, but I received one complaint about 50K of
295        unwanted data. My guess is that the person linked everything into his
296        program rather than using a library. Anyway, it does no harm.
297    
298    8.  A pattern such as /\x{123}{2,2}+/8 was incorrectly compiled; the trigger
299        was a minimum greater than 1 for a wide character in a possessive
300        repetition. The same bug could also affect patterns like /(\x{ff}{0,2})*/8
301        which had an unlimited repeat of a nested, fixed maximum repeat of a wide
302        character. Chaos in the form of incorrect output or a compiling loop could
303        result.
304    
305    9.  The restrictions on what a pattern can contain when partial matching is
306        requested for pcre_exec() have been removed. All patterns can now be
307        partially matched by this function. In addition, if there are at least two
308        slots in the offset vector, the offset of the earliest inspected character
309        for the match and the offset of the end of the subject are set in them when
310        PCRE_ERROR_PARTIAL is returned.
311    
312    10. Partial matching has been split into two forms: PCRE_PARTIAL_SOFT, which is
313        synonymous with PCRE_PARTIAL, for backwards compatibility, and
314        PCRE_PARTIAL_HARD, which causes a partial match to supersede a full match,
315        and may be more useful for multi-segment matching.
316    
317    11. Partial matching with pcre_exec() is now more intuitive. A partial match
318        used to be given if ever the end of the subject was reached; now it is
319        given only if matching could not proceed because another character was
320        needed. This makes a difference in some odd cases such as Z(*FAIL) with the
321        string "Z", which now yields "no match" instead of "partial match". In the
322        case of pcre_dfa_exec(), "no match" is given if every matching path for the
323        final character ended with (*FAIL).
324    
325    12. Restarting a match using pcre_dfa_exec() after a partial match did not work
326        if the pattern had a "must contain" character that was already found in the
327        earlier partial match, unless partial matching was again requested. For
328        example, with the pattern /dog.(body)?/, the "must contain" character is
329        "g". If the first part-match was for the string "dog", restarting with
330        "sbody" failed. This bug has been fixed.
331    
332    13. The string returned by pcre_dfa_exec() after a partial match has been
333        changed so that it starts at the first inspected character rather than the
334        first character of the match. This makes a difference only if the pattern
335        starts with a lookbehind assertion or \b or \B (\K is not supported by
336        pcre_dfa_exec()). It's an incompatible change, but it makes the two
337        matching functions compatible, and I think it's the right thing to do.
338    
339    14. Added a pcredemo man page, created automatically from the pcredemo.c file,
340        so that the demonstration program is easily available in environments where
341        PCRE has not been installed from source.
342    
343    15. Arranged to add -DPCRE_STATIC to cflags in libpcre.pc, libpcreposix.cp,
344        libpcrecpp.pc and pcre-config when PCRE is not compiled as a shared
345        library.
346    
347    16. Added REG_UNGREEDY to the pcreposix interface, at the request of a user.
348        It maps to PCRE_UNGREEDY. It is not, of course, POSIX-compatible, but it
349        is not the first non-POSIX option to be added. Clearly some people find
350        these options useful.
351    
352    17. If a caller to the POSIX matching function regexec() passes a non-zero
353        value for nmatch with a NULL value for pmatch, the value of
354        nmatch is forced to zero.
355    
356    18. RunGrepTest did not have a test for the availability of the -u option of
357        the diff command, as RunTest does. It now checks in the same way as
358        RunTest, and also checks for the -b option.
359    
360    19. If an odd number of negated classes containing just a single character
361        interposed, within parentheses, between a forward reference to a named
362        subpattern and the definition of the subpattern, compilation crashed with
363        an internal error, complaining that it could not find the referenced
364        subpattern. An example of a crashing pattern is /(?&A)(([^m])(?<A>))/.
365        [The bug was that it was starting one character too far in when skipping
366        over the character class, thus treating the ] as data rather than
367        terminating the class. This meant it could skip too much.]
368    
369    20. Added PCRE_NOTEMPTY_ATSTART in order to be able to correctly implement the
370        /g option in pcretest when the pattern contains \K, which makes it possible
371        to have an empty string match not at the start, even when the pattern is
372        anchored. Updated pcretest and pcredemo to use this option.
373    
374    21. If the maximum number of capturing subpatterns in a recursion was greater
375        than the maximum at the outer level, the higher number was returned, but
376        with unset values at the outer level. The correct (outer level) value is
377        now given.
378    
379    22. If (*ACCEPT) appeared inside capturing parentheses, previous releases of
380        PCRE did not set those parentheses (unlike Perl). I have now found a way to
381        make it do so. The string so far is captured, making this feature
382        compatible with Perl.
383    
384    23. The tests have been re-organized, adding tests 11 and 12, to make it
385        possible to check the Perl 5.10 features against Perl 5.10.
386    
387    24. Perl 5.10 allows subroutine calls in lookbehinds, as long as the subroutine
388        pattern matches a fixed length string. PCRE did not allow this; now it
389        does. Neither allows recursion.
390    
391    25. I finally figured out how to implement a request to provide the minimum
392        length of subject string that was needed in order to match a given pattern.
393        (It was back references and recursion that I had previously got hung up
394        on.) This code has now been added to pcre_study(); it finds a lower bound
395        to the length of subject needed. It is not necessarily the greatest lower
396        bound, but using it to avoid searching strings that are too short does give
397        some useful speed-ups. The value is available to calling programs via
398        pcre_fullinfo().
399    
400    26. While implementing 25, I discovered to my embarrassment that pcretest had
401        not been passing the result of pcre_study() to pcre_dfa_exec(), so the
402        study optimizations had never been tested with that matching function.
403        Oops. What is worse, even when it was passed study data, there was a bug in
404        pcre_dfa_exec() that meant it never actually used it. Double oops. There
405        were also very few tests of studied patterns with pcre_dfa_exec().
406    
407    27. If (?| is used to create subpatterns with duplicate numbers, they are now
408        allowed to have the same name, even if PCRE_DUPNAMES is not set. However,
409        on the other side of the coin, they are no longer allowed to have different
410        names, because these cannot be distinguished in PCRE, and this has caused
411        confusion. (This is a difference from Perl.)
412    
413    28. When duplicate subpattern names are present (necessarily with different
414        numbers, as required by 27 above), and a test is made by name in a
415        conditional pattern, either for a subpattern having been matched, or for
416        recursion in such a pattern, all the associated numbered subpatterns are
417        tested, and the overall condition is true if the condition is true for any
418        one of them. This is the way Perl works, and is also more like the way
419        testing by number works.
420    
421    
422    Version 7.9 11-Apr-09
423    ---------------------
424    
425    1.  When building with support for bzlib/zlib (pcregrep) and/or readline
426        (pcretest), all targets were linked against these libraries. This included
427        libpcre, libpcreposix, and libpcrecpp, even though they do not use these
428        libraries. This caused unwanted dependencies to be created. This problem
429        has been fixed, and now only pcregrep is linked with bzlib/zlib and only
430        pcretest is linked with readline.
431    
432    2.  The "typedef int BOOL" in pcre_internal.h that was included inside the
433        "#ifndef FALSE" condition by an earlier change (probably 7.8/18) has been
434        moved outside it again, because FALSE and TRUE are already defined in AIX,
435        but BOOL is not.
436    
437    3.  The pcre_config() function was treating the PCRE_MATCH_LIMIT and
438        PCRE_MATCH_LIMIT_RECURSION values as ints, when they should be long ints.
439    
440    4.  The pcregrep documentation said spaces were inserted as well as colons (or
441        hyphens) following file names and line numbers when outputting matching
442        lines. This is not true; no spaces are inserted. I have also clarified the
443        wording for the --colour (or --color) option.
444    
445    5.  In pcregrep, when --colour was used with -o, the list of matching strings
446        was not coloured; this is different to GNU grep, so I have changed it to be
447        the same.
448    
449    6.  When --colo(u)r was used in pcregrep, only the first matching substring in
450        each matching line was coloured. Now it goes on to look for further matches
451        of any of the test patterns, which is the same behaviour as GNU grep.
452    
453    7.  A pattern that could match an empty string could cause pcregrep to loop; it
454        doesn't make sense to accept an empty string match in pcregrep, so I have
455        locked it out (using PCRE's PCRE_NOTEMPTY option). By experiment, this
456        seems to be how GNU grep behaves.
457    
458    8.  The pattern (?(?=.*b)b|^) was incorrectly compiled as "match must be at
459        start or after a newline", because the conditional assertion was not being
460        correctly handled. The rule now is that both the assertion and what follows
461        in the first alternative must satisfy the test.
462    
463    9.  If auto-callout was enabled in a pattern with a conditional group whose
464        condition was an assertion, PCRE could crash during matching, both with
465        pcre_exec() and pcre_dfa_exec().
466    
467    10. The PCRE_DOLLAR_ENDONLY option was not working when pcre_dfa_exec() was
468        used for matching.
469    
470    11. Unicode property support in character classes was not working for
471        characters (bytes) greater than 127 when not in UTF-8 mode.
472    
473    12. Added the -M command line option to pcretest.
474    
475    14. Added the non-standard REG_NOTEMPTY option to the POSIX interface.
476    
477    15. Added the PCRE_NO_START_OPTIMIZE match-time option.
478    
479    16. Added comments and documentation about mis-use of no_arg in the C++
480        wrapper.
481    
482    17. Implemented support for UTF-8 encoding in EBCDIC environments, a patch
483        from Martin Jerabek that uses macro names for all relevant character and
484        string constants.
485    
486    18. Added to pcre_internal.h two configuration checks: (a) If both EBCDIC and
487        SUPPORT_UTF8 are set, give an error; (b) If SUPPORT_UCP is set without
488        SUPPORT_UTF8, define SUPPORT_UTF8. The "configure" script handles both of
489        these, but not everybody uses configure.
490    
491    19. A conditional group that had only one branch was not being correctly
492        recognized as an item that could match an empty string. This meant that an
493        enclosing group might also not be so recognized, causing infinite looping
494        (and probably a segfault) for patterns such as ^"((?(?=[a])[^"])|b)*"$
495        with the subject "ab", where knowledge that the repeated group can match
496        nothing is needed in order to break the loop.
497    
498    20. If a pattern that was compiled with callouts was matched using pcre_dfa_
499        exec(), but without supplying a callout function, matching went wrong.
500    
501    21. If PCRE_ERROR_MATCHLIMIT occurred during a recursion, there was a memory
502        leak if the size of the offset vector was greater than 30. When the vector
503        is smaller, the saved offsets during recursion go onto a local stack
504        vector, but for larger vectors malloc() is used. It was failing to free
505        when the recursion yielded PCRE_ERROR_MATCH_LIMIT (or any other "abnormal"
506        error, in fact).
507    
508    22. There was a missing #ifdef SUPPORT_UTF8 round one of the variables in the
509        heapframe that is used only when UTF-8 support is enabled. This caused no
510        problem, but was untidy.
511    
512    23. Steven Van Ingelgem's patch to CMakeLists.txt to change the name
513        CMAKE_BINARY_DIR to PROJECT_BINARY_DIR so that it works when PCRE is
514        included within another project.
515    
516    24. Steven Van Ingelgem's patches to add more options to the CMake support,
517        slightly modified by me:
518    
519          (a) PCRE_BUILD_TESTS can be set OFF not to build the tests, including
520              not building pcregrep.
521    
522          (b) PCRE_BUILD_PCREGREP can be see OFF not to build pcregrep, but only
523              if PCRE_BUILD_TESTS is also set OFF, because the tests use pcregrep.
524    
525    25. Forward references, both numeric and by name, in patterns that made use of
526        duplicate group numbers, could behave incorrectly or give incorrect errors,
527        because when scanning forward to find the reference group, PCRE was not
528        taking into account the duplicate group numbers. A pattern such as
529        ^X(?3)(a)(?|(b)|(q))(Y) is an example.
530    
531    26. Changed a few more instances of "const unsigned char *" to USPTR, making
532        the feature of a custom pointer more persuasive (as requested by a user).
533    
534    27. Wrapped the definitions of fileno and isatty for Windows, which appear in
535        pcretest.c, inside #ifndefs, because it seems they are sometimes already
536        pre-defined.
537    
538    28. Added support for (*UTF8) at the start of a pattern.
539    
540    29. Arrange for flags added by the "release type" setting in CMake to be shown
541        in the configuration summary.
542    
543    
544    Version 7.8 05-Sep-08
545    ---------------------
546    
547    1.  Replaced UCP searching code with optimized version as implemented for Ad
548        Muncher (http://www.admuncher.com/) by Peter Kankowski. This uses a two-
549        stage table and inline lookup instead of a function, giving speed ups of 2
550        to 5 times on some simple patterns that I tested. Permission was given to
551        distribute the MultiStage2.py script that generates the tables (it's not in
552        the tarball, but is in the Subversion repository).
553    
554    2.  Updated the Unicode datatables to Unicode 5.1.0. This adds yet more
555        scripts.
556    
557    3.  Change 12 for 7.7 introduced a bug in pcre_study() when a pattern contained
558        a group with a zero qualifier. The result of the study could be incorrect,
559        or the function might crash, depending on the pattern.
560    
561    4.  Caseless matching was not working for non-ASCII characters in back
562        references. For example, /(\x{de})\1/8i was not matching \x{de}\x{fe}.
563        It now works when Unicode Property Support is available.
564    
565    5.  In pcretest, an escape such as \x{de} in the data was always generating
566        a UTF-8 string, even in non-UTF-8 mode. Now it generates a single byte in
567        non-UTF-8 mode. If the value is greater than 255, it gives a warning about
568        truncation.
569    
570    6.  Minor bugfix in pcrecpp.cc (change "" == ... to NULL == ...).
571    
572    7.  Added two (int) casts to pcregrep when printing the difference of two
573        pointers, in case they are 64-bit values.
574    
575    8.  Added comments about Mac OS X stack usage to the pcrestack man page and to
576        test 2 if it fails.
577    
578    9.  Added PCRE_CALL_CONVENTION just before the names of all exported functions,
579        and a #define of that name to empty if it is not externally set. This is to
580        allow users of MSVC to set it if necessary.
581    
582    10. The PCRE_EXP_DEFN macro which precedes exported functions was missing from
583        the convenience functions in the pcre_get.c source file.
584    
585    11. An option change at the start of a pattern that had top-level alternatives
586        could cause overwriting and/or a crash. This command provoked a crash in
587        some environments:
588    
589          printf "/(?i)[\xc3\xa9\xc3\xbd]|[\xc3\xa9\xc3\xbdA]/8\n" | pcretest
590    
591        This potential security problem was recorded as CVE-2008-2371.
592    
593    12. For a pattern where the match had to start at the beginning or immediately
594        after a newline (e.g /.*anything/ without the DOTALL flag), pcre_exec() and
595        pcre_dfa_exec() could read past the end of the passed subject if there was
596        no match. To help with detecting such bugs (e.g. with valgrind), I modified
597        pcretest so that it places the subject at the end of its malloc-ed buffer.
598    
599    13. The change to pcretest in 12 above threw up a couple more cases when pcre_
600        exec() might read past the end of the data buffer in UTF-8 mode.
601    
602    14. A similar bug to 7.3/2 existed when the PCRE_FIRSTLINE option was set and
603        the data contained the byte 0x85 as part of a UTF-8 character within its
604        first line. This applied both to normal and DFA matching.
605    
606    15. Lazy qualifiers were not working in some cases in UTF-8 mode. For example,
607        /^[^d]*?$/8 failed to match "abc".
608    
609    16. Added a missing copyright notice to pcrecpp_internal.h.
610    
611    17. Make it more clear in the documentation that values returned from
612        pcre_exec() in ovector are byte offsets, not character counts.
613    
614    18. Tidied a few places to stop certain compilers from issuing warnings.
615    
616    19. Updated the Virtual Pascal + BCC files to compile the latest v7.7, as
617        supplied by Stefan Weber. I made a further small update for 7.8 because
618        there is a change of source arrangements: the pcre_searchfuncs.c module is
619        replaced by pcre_ucd.c.
620    
621    
622    Version 7.7 07-May-08
623    ---------------------
624    
625    1.  Applied Craig's patch to sort out a long long problem: "If we can't convert
626        a string to a long long, pretend we don't even have a long long." This is
627        done by checking for the strtoq, strtoll, and _strtoi64 functions.
628    
629    2.  Applied Craig's patch to pcrecpp.cc to restore ABI compatibility with
630        pre-7.6 versions, which defined a global no_arg variable instead of putting
631        it in the RE class. (See also #8 below.)
632    
633    3.  Remove a line of dead code, identified by coverity and reported by Nuno
634        Lopes.
635    
636    4.  Fixed two related pcregrep bugs involving -r with --include or --exclude:
637    
638        (1) The include/exclude patterns were being applied to the whole pathnames
639            of files, instead of just to the final components.
640    
641        (2) If there was more than one level of directory, the subdirectories were
642            skipped unless they satisfied the include/exclude conditions. This is
643            inconsistent with GNU grep (and could even be seen as contrary to the
644            pcregrep specification - which I improved to make it absolutely clear).
645            The action now is always to scan all levels of directory, and just
646            apply the include/exclude patterns to regular files.
647    
648    5.  Added the --include_dir and --exclude_dir patterns to pcregrep, and used
649        --exclude_dir in the tests to avoid scanning .svn directories.
650    
651    6.  Applied Craig's patch to the QuoteMeta function so that it escapes the
652        NUL character as backslash + 0 rather than backslash + NUL, because PCRE
653        doesn't support NULs in patterns.
654    
655    7.  Added some missing "const"s to declarations of static tables in
656        pcre_compile.c and pcre_dfa_exec.c.
657    
658    8.  Applied Craig's patch to pcrecpp.cc to fix a problem in OS X that was
659        caused by fix #2  above. (Subsequently also a second patch to fix the
660        first patch. And a third patch - this was a messy problem.)
661    
662    9.  Applied Craig's patch to remove the use of push_back().
663    
664    10. Applied Alan Lehotsky's patch to add REG_STARTEND support to the POSIX
665        matching function regexec().
666    
667    11. Added support for the Oniguruma syntax \g<name>, \g<n>, \g'name', \g'n',
668        which, however, unlike Perl's \g{...}, are subroutine calls, not back
669        references. PCRE supports relative numbers with this syntax (I don't think
670        Oniguruma does).
671    
672    12. Previously, a group with a zero repeat such as (...){0} was completely
673        omitted from the compiled regex. However, this means that if the group
674        was called as a subroutine from elsewhere in the pattern, things went wrong
675        (an internal error was given). Such groups are now left in the compiled
676        pattern, with a new opcode that causes them to be skipped at execution
677        time.
678    
679    13. Added the PCRE_JAVASCRIPT_COMPAT option. This makes the following changes
680        to the way PCRE behaves:
681    
682        (a) A lone ] character is dis-allowed (Perl treats it as data).
683    
684        (b) A back reference to an unmatched subpattern matches an empty string
685            (Perl fails the current match path).
686    
687        (c) A data ] in a character class must be notated as \] because if the
688            first data character in a class is ], it defines an empty class. (In
689            Perl it is not possible to have an empty class.) The empty class []
690            never matches; it forces failure and is equivalent to (*FAIL) or (?!).
691            The negative empty class [^] matches any one character, independently
692            of the DOTALL setting.
693    
694    14. A pattern such as /(?2)[]a()b](abc)/ which had a forward reference to a
695        non-existent subpattern following a character class starting with ']' and
696        containing () gave an internal compiling error instead of "reference to
697        non-existent subpattern". Fortunately, when the pattern did exist, the
698        compiled code was correct. (When scanning forwards to check for the
699        existencd of the subpattern, it was treating the data ']' as terminating
700        the class, so got the count wrong. When actually compiling, the reference
701        was subsequently set up correctly.)
702    
703    15. The "always fail" assertion (?!) is optimzed to (*FAIL) by pcre_compile;
704        it was being rejected as not supported by pcre_dfa_exec(), even though
705        other assertions are supported. I have made pcre_dfa_exec() support
706        (*FAIL).
707    
708    16. The implementation of 13c above involved the invention of a new opcode,
709        OP_ALLANY, which is like OP_ANY but doesn't check the /s flag. Since /s
710        cannot be changed at match time, I realized I could make a small
711        improvement to matching performance by compiling OP_ALLANY instead of
712        OP_ANY for "." when DOTALL was set, and then removing the runtime tests
713        on the OP_ANY path.
714    
715    17. Compiling pcretest on Windows with readline support failed without the
716        following two fixes: (1) Make the unistd.h include conditional on
717        HAVE_UNISTD_H; (2) #define isatty and fileno as _isatty and _fileno.
718    
719    18. Changed CMakeLists.txt and cmake/FindReadline.cmake to arrange for the
720        ncurses library to be included for pcretest when ReadLine support is
721        requested, but also to allow for it to be overridden. This patch came from
722        Daniel Bergström.
723    
724    19. There was a typo in the file ucpinternal.h where f0_rangeflag was defined
725        as 0x00f00000 instead of 0x00800000. Luckily, this would not have caused
726        any errors with the current Unicode tables. Thanks to Peter Kankowski for
727        spotting this.
728    
729    
730    Version 7.6 28-Jan-08
731    ---------------------
732    
733    1.  A character class containing a very large number of characters with
734        codepoints greater than 255 (in UTF-8 mode, of course) caused a buffer
735        overflow.
736    
737    2.  Patch to cut out the "long long" test in pcrecpp_unittest when
738        HAVE_LONG_LONG is not defined.
739    
740    3.  Applied Christian Ehrlicher's patch to update the CMake build files to
741        bring them up to date and include new features. This patch includes:
742    
743        - Fixed PH's badly added libz and libbz2 support.
744        - Fixed a problem with static linking.
745        - Added pcredemo. [But later removed - see 7 below.]
746        - Fixed dftables problem and added an option.
747        - Added a number of HAVE_XXX tests, including HAVE_WINDOWS_H and
748            HAVE_LONG_LONG.
749        - Added readline support for pcretest.
750        - Added an listing of the option settings after cmake has run.
751    
752    4.  A user submitted a patch to Makefile that makes it easy to create
753        "pcre.dll" under mingw when using Configure/Make. I added stuff to
754        Makefile.am that cause it to include this special target, without
755        affecting anything else. Note that the same mingw target plus all
756        the other distribution libraries and programs are now supported
757        when configuring with CMake (see 6 below) instead of with
758        Configure/Make.
759    
760    5.  Applied Craig's patch that moves no_arg into the RE class in the C++ code.
761        This is an attempt to solve the reported problem "pcrecpp::no_arg is not
762        exported in the Windows port". It has not yet been confirmed that the patch
763        solves the problem, but it does no harm.
764    
765    6.  Applied Sheri's patch to CMakeLists.txt to add NON_STANDARD_LIB_PREFIX and
766        NON_STANDARD_LIB_SUFFIX for dll names built with mingw when configured
767        with CMake, and also correct the comment about stack recursion.
768    
769    7.  Remove the automatic building of pcredemo from the ./configure system and
770        from CMakeLists.txt. The whole idea of pcredemo.c is that it is an example
771        of a program that users should build themselves after PCRE is installed, so
772        building it automatically is not really right. What is more, it gave
773        trouble in some build environments.
774    
775    8.  Further tidies to CMakeLists.txt from Sheri and Christian.
776    
777    
778    Version 7.5 10-Jan-08
779    ---------------------
780    
781    1.  Applied a patch from Craig: "This patch makes it possible to 'ignore'
782        values in parens when parsing an RE using the C++ wrapper."
783    
784    2.  Negative specials like \S did not work in character classes in UTF-8 mode.
785        Characters greater than 255 were excluded from the class instead of being
786        included.
787    
788    3.  The same bug as (2) above applied to negated POSIX classes such as
789        [:^space:].
790    
791    4.  PCRECPP_STATIC was referenced in pcrecpp_internal.h, but nowhere was it
792        defined or documented. It seems to have been a typo for PCRE_STATIC, so
793        I have changed it.
794    
795    5.  The construct (?&) was not diagnosed as a syntax error (it referenced the
796        first named subpattern) and a construct such as (?&a) would reference the
797        first named subpattern whose name started with "a" (in other words, the
798        length check was missing). Both these problems are fixed. "Subpattern name
799        expected" is now given for (?&) (a zero-length name), and this patch also
800        makes it give the same error for \k'' (previously it complained that that
801        was a reference to a non-existent subpattern).
802    
803    6.  The erroneous patterns (?+-a) and (?-+a) give different error messages;
804        this is right because (?- can be followed by option settings as well as by
805        digits. I have, however, made the messages clearer.
806    
807    7.  Patterns such as (?(1)a|b) (a pattern that contains fewer subpatterns
808        than the number used in the conditional) now cause a compile-time error.
809        This is actually not compatible with Perl, which accepts such patterns, but
810        treats the conditional as always being FALSE (as PCRE used to), but it
811        seems to me that giving a diagnostic is better.
812    
813    8.  Change "alphameric" to the more common word "alphanumeric" in comments
814        and messages.
815    
816    9.  Fix two occurrences of "backslash" in comments that should have been
817        "backspace".
818    
819    10. Remove two redundant lines of code that can never be obeyed (their function
820        was moved elsewhere).
821    
822    11. The program that makes PCRE's Unicode character property table had a bug
823        which caused it to generate incorrect table entries for sequences of
824        characters that have the same character type, but are in different scripts.
825        It amalgamated them into a single range, with the script of the first of
826        them. In other words, some characters were in the wrong script. There were
827        thirteen such cases, affecting characters in the following ranges:
828    
829          U+002b0 - U+002c1
830          U+0060c - U+0060d
831          U+0061e - U+00612
832          U+0064b - U+0065e
833          U+0074d - U+0076d
834          U+01800 - U+01805
835          U+01d00 - U+01d77
836          U+01d9b - U+01dbf
837          U+0200b - U+0200f
838          U+030fc - U+030fe
839          U+03260 - U+0327f
840          U+0fb46 - U+0fbb1
841          U+10450 - U+1049d
842    
843    12. The -o option (show only the matching part of a line) for pcregrep was not
844        compatible with GNU grep in that, if there was more than one match in a
845        line, it showed only the first of them. It now behaves in the same way as
846        GNU grep.
847    
848    13. If the -o and -v options were combined for pcregrep, it printed a blank
849        line for every non-matching line. GNU grep prints nothing, and pcregrep now
850        does the same. The return code can be used to tell if there were any
851        non-matching lines.
852    
853    14. Added --file-offsets and --line-offsets to pcregrep.
854    
855    15. The pattern (?=something)(?R) was not being diagnosed as a potentially
856        infinitely looping recursion. The bug was that positive lookaheads were not
857        being skipped when checking for a possible empty match (negative lookaheads
858        and both kinds of lookbehind were skipped).
859    
860    16. Fixed two typos in the Windows-only code in pcregrep.c, and moved the
861        inclusion of <windows.h> to before rather than after the definition of
862        INVALID_FILE_ATTRIBUTES (patch from David Byron).
863    
864    17. Specifying a possessive quantifier with a specific limit for a Unicode
865        character property caused pcre_compile() to compile bad code, which led at
866        runtime to PCRE_ERROR_INTERNAL (-14). Examples of patterns that caused this
867        are: /\p{Zl}{2,3}+/8 and /\p{Cc}{2}+/8. It was the possessive "+" that
868        caused the error; without that there was no problem.
869    
870    18. Added --enable-pcregrep-libz and --enable-pcregrep-libbz2.
871    
872    19. Added --enable-pcretest-libreadline.
873    
874    20. In pcrecpp.cc, the variable 'count' was incremented twice in
875        RE::GlobalReplace(). As a result, the number of replacements returned was
876        double what it should be. I removed one of the increments, but Craig sent a
877        later patch that removed the other one (the right fix) and added unit tests
878        that check the return values (which was not done before).
879    
880    21. Several CMake things:
881    
882        (1) Arranged that, when cmake is used on Unix, the libraries end up with
883            the names libpcre and libpcreposix, not just pcre and pcreposix.
884    
885        (2) The above change means that pcretest and pcregrep are now correctly
886            linked with the newly-built libraries, not previously installed ones.
887    
888        (3) Added PCRE_SUPPORT_LIBREADLINE, PCRE_SUPPORT_LIBZ, PCRE_SUPPORT_LIBBZ2.
889    
890    22. In UTF-8 mode, with newline set to "any", a pattern such as .*a.*=.b.*
891        crashed when matching a string such as a\x{2029}b (note that \x{2029} is a
892        UTF-8 newline character). The key issue is that the pattern starts .*;
893        this means that the match must be either at the beginning, or after a
894        newline. The bug was in the code for advancing after a failed match and
895        checking that the new position followed a newline. It was not taking
896        account of UTF-8 characters correctly.
897    
898    23. PCRE was behaving differently from Perl in the way it recognized POSIX
899        character classes. PCRE was not treating the sequence [:...:] as a
900        character class unless the ... were all letters. Perl, however, seems to
901        allow any characters between [: and :], though of course it rejects as
902        unknown any "names" that contain non-letters, because all the known class
903        names consist only of letters. Thus, Perl gives an error for [[:1234:]],
904        for example, whereas PCRE did not - it did not recognize a POSIX character
905        class. This seemed a bit dangerous, so the code has been changed to be
906        closer to Perl. The behaviour is not identical to Perl, because PCRE will
907        diagnose an unknown class for, for example, [[:l\ower:]] where Perl will
908        treat it as [[:lower:]]. However, PCRE does now give "unknown" errors where
909        Perl does, and where it didn't before.
910    
911    24. Rewrite so as to remove the single use of %n from pcregrep because in some
912        Windows environments %n is disabled by default.
913    
914    
915    Version 7.4 21-Sep-07
916    ---------------------
917    
918    1.  Change 7.3/28 was implemented for classes by looking at the bitmap. This
919        means that a class such as [\s] counted as "explicit reference to CR or
920        LF". That isn't really right - the whole point of the change was to try to
921        help when there was an actual mention of one of the two characters. So now
922        the change happens only if \r or \n (or a literal CR or LF) character is
923        encountered.
924    
925    2.  The 32-bit options word was also used for 6 internal flags, but the numbers
926        of both had grown to the point where there were only 3 bits left.
927        Fortunately, there was spare space in the data structure, and so I have
928        moved the internal flags into a new 16-bit field to free up more option
929        bits.
930    
931    3.  The appearance of (?J) at the start of a pattern set the DUPNAMES option,
932        but did not set the internal JCHANGED flag - either of these is enough to
933        control the way the "get" function works - but the PCRE_INFO_JCHANGED
934        facility is supposed to tell if (?J) was ever used, so now (?J) at the
935        start sets both bits.
936    
937    4.  Added options (at build time, compile time, exec time) to change \R from
938        matching any Unicode line ending sequence to just matching CR, LF, or CRLF.
939    
940    5.  doc/pcresyntax.html was missing from the distribution.
941    
942    6.  Put back the definition of PCRE_ERROR_NULLWSLIMIT, for backward
943        compatibility, even though it is no longer used.
944    
945    7.  Added macro for snprintf to pcrecpp_unittest.cc and also for strtoll and
946        strtoull to pcrecpp.cc to select the available functions in WIN32 when the
947        windows.h file is present (where different names are used). [This was
948        reversed later after testing - see 16 below.]
949    
950    8.  Changed all #include <config.h> to #include "config.h". There were also
951        some further <pcre.h> cases that I changed to "pcre.h".
952    
953    9.  When pcregrep was used with the --colour option, it missed the line ending
954        sequence off the lines that it output.
955    
956    10. It was pointed out to me that arrays of string pointers cause lots of
957        relocations when a shared library is dynamically loaded. A technique of
958        using a single long string with a table of offsets can drastically reduce
959        these. I have refactored PCRE in four places to do this. The result is
960        dramatic:
961    
962          Originally:                          290
963          After changing UCP table:            187
964          After changing error message table:   43
965          After changing table of "verbs"       36
966          After changing table of Posix names   22
967    
968        Thanks to the folks working on Gregex for glib for this insight.
969    
970    11. --disable-stack-for-recursion caused compiling to fail unless -enable-
971        unicode-properties was also set.
972    
973    12. Updated the tests so that they work when \R is defaulted to ANYCRLF.
974    
975    13. Added checks for ANY and ANYCRLF to pcrecpp.cc where it previously
976        checked only for CRLF.
977    
978    14. Added casts to pcretest.c to avoid compiler warnings.
979    
980    15. Added Craig's patch to various pcrecpp modules to avoid compiler warnings.
981    
982    16. Added Craig's patch to remove the WINDOWS_H tests, that were not working,
983        and instead check for _strtoi64 explicitly, and avoid the use of snprintf()
984        entirely. This removes changes made in 7 above.
985    
986    17. The CMake files have been updated, and there is now more information about
987        building with CMake in the NON-UNIX-USE document.
988    
989    
990    Version 7.3 28-Aug-07
991    ---------------------
992    
993     1. In the rejigging of the build system that eventually resulted in 7.1, the
994        line "#include <pcre.h>" was included in pcre_internal.h. The use of angle
995        brackets there is not right, since it causes compilers to look for an
996        installed pcre.h, not the version that is in the source that is being
997        compiled (which of course may be different). I have changed it back to:
998    
999          #include "pcre.h"
1000    
1001        I have a vague recollection that the change was concerned with compiling in
1002        different directories, but in the new build system, that is taken care of
1003        by the VPATH setting the Makefile.
1004    
1005     2. The pattern .*$ when run in not-DOTALL UTF-8 mode with newline=any failed
1006        when the subject happened to end in the byte 0x85 (e.g. if the last
1007        character was \x{1ec5}). *Character* 0x85 is one of the "any" newline
1008        characters but of course it shouldn't be taken as a newline when it is part
1009        of another character. The bug was that, for an unlimited repeat of . in
1010        not-DOTALL UTF-8 mode, PCRE was advancing by bytes rather than by
1011        characters when looking for a newline.
1012    
1013     3. A small performance improvement in the DOTALL UTF-8 mode .* case.
1014    
1015     4. Debugging: adjusted the names of opcodes for different kinds of parentheses
1016        in debug output.
1017    
1018     5. Arrange to use "%I64d" instead of "%lld" and "%I64u" instead of "%llu" for
1019        long printing in the pcrecpp unittest when running under MinGW.
1020    
1021     6. ESC_K was left out of the EBCDIC table.
1022    
1023     7. Change 7.0/38 introduced a new limit on the number of nested non-capturing
1024        parentheses; I made it 1000, which seemed large enough. Unfortunately, the
1025        limit also applies to "virtual nesting" when a pattern is recursive, and in
1026        this case 1000 isn't so big. I have been able to remove this limit at the
1027        expense of backing off one optimization in certain circumstances. Normally,
1028        when pcre_exec() would call its internal match() function recursively and
1029        immediately return the result unconditionally, it uses a "tail recursion"
1030        feature to save stack. However, when a subpattern that can match an empty
1031        string has an unlimited repetition quantifier, it no longer makes this
1032        optimization. That gives it a stack frame in which to save the data for
1033        checking that an empty string has been matched. Previously this was taken
1034        from the 1000-entry workspace that had been reserved. So now there is no
1035        explicit limit, but more stack is used.
1036    
1037     8. Applied Daniel's patches to solve problems with the import/export magic
1038        syntax that is required for Windows, and which was going wrong for the
1039        pcreposix and pcrecpp parts of the library. These were overlooked when this
1040        problem was solved for the main library.
1041    
1042     9. There were some crude static tests to avoid integer overflow when computing
1043        the size of patterns that contain repeated groups with explicit upper
1044        limits. As the maximum quantifier is 65535, the maximum group length was
1045        set at 30,000 so that the product of these two numbers did not overflow a
1046        32-bit integer. However, it turns out that people want to use groups that
1047        are longer than 30,000 bytes (though not repeat them that many times).
1048        Change 7.0/17 (the refactoring of the way the pattern size is computed) has
1049        made it possible to implement the integer overflow checks in a much more
1050        dynamic way, which I have now done. The artificial limitation on group
1051        length has been removed - we now have only the limit on the total length of
1052        the compiled pattern, which depends on the LINK_SIZE setting.
1053    
1054    10. Fixed a bug in the documentation for get/copy named substring when
1055        duplicate names are permitted. If none of the named substrings are set, the
1056        functions return PCRE_ERROR_NOSUBSTRING (7); the doc said they returned an
1057        empty string.
1058    
1059    11. Because Perl interprets \Q...\E at a high level, and ignores orphan \E
1060        instances, patterns such as [\Q\E] or [\E] or even [^\E] cause an error,
1061        because the ] is interpreted as the first data character and the
1062        terminating ] is not found. PCRE has been made compatible with Perl in this
1063        regard. Previously, it interpreted [\Q\E] as an empty class, and [\E] could
1064        cause memory overwriting.
1065    
1066    10. Like Perl, PCRE automatically breaks an unlimited repeat after an empty
1067        string has been matched (to stop an infinite loop). It was not recognizing
1068        a conditional subpattern that could match an empty string if that
1069        subpattern was within another subpattern. For example, it looped when
1070        trying to match  (((?(1)X|))*)  but it was OK with  ((?(1)X|)*)  where the
1071        condition was not nested. This bug has been fixed.
1072    
1073    12. A pattern like \X?\d or \P{L}?\d in non-UTF-8 mode could cause a backtrack
1074        past the start of the subject in the presence of bytes with the top bit
1075        set, for example "\x8aBCD".
1076    
1077    13. Added Perl 5.10 experimental backtracking controls (*FAIL), (*F), (*PRUNE),
1078        (*SKIP), (*THEN), (*COMMIT), and (*ACCEPT).
1079    
1080    14. Optimized (?!) to (*FAIL).
1081    
1082    15. Updated the test for a valid UTF-8 string to conform to the later RFC 3629.
1083        This restricts code points to be within the range 0 to 0x10FFFF, excluding
1084        the "low surrogate" sequence 0xD800 to 0xDFFF. Previously, PCRE allowed the
1085        full range 0 to 0x7FFFFFFF, as defined by RFC 2279. Internally, it still
1086        does: it's just the validity check that is more restrictive.
1087    
1088    16. Inserted checks for integer overflows during escape sequence (backslash)
1089        processing, and also fixed erroneous offset values for syntax errors during
1090        backslash processing.
1091    
1092    17. Fixed another case of looking too far back in non-UTF-8 mode (cf 12 above)
1093        for patterns like [\PPP\x8a]{1,}\x80 with the subject "A\x80".
1094    
1095    18. An unterminated class in a pattern like (?1)\c[ with a "forward reference"
1096        caused an overrun.
1097    
1098    19. A pattern like (?:[\PPa*]*){8,} which had an "extended class" (one with
1099        something other than just ASCII characters) inside a group that had an
1100        unlimited repeat caused a loop at compile time (while checking to see
1101        whether the group could match an empty string).
1102    
1103    20. Debugging a pattern containing \p or \P could cause a crash. For example,
1104        [\P{Any}] did so. (Error in the code for printing property names.)
1105    
1106    21. An orphan \E inside a character class could cause a crash.
1107    
1108    22. A repeated capturing bracket such as (A)? could cause a wild memory
1109        reference during compilation.
1110    
1111    23. There are several functions in pcre_compile() that scan along a compiled
1112        expression for various reasons (e.g. to see if it's fixed length for look
1113        behind). There were bugs in these functions when a repeated \p or \P was
1114        present in the pattern. These operators have additional parameters compared
1115        with \d, etc, and these were not being taken into account when moving along
1116        the compiled data. Specifically:
1117    
1118        (a) A item such as \p{Yi}{3} in a lookbehind was not treated as fixed
1119            length.
1120    
1121        (b) An item such as \pL+ within a repeated group could cause crashes or
1122            loops.
1123    
1124        (c) A pattern such as \p{Yi}+(\P{Yi}+)(?1) could give an incorrect
1125            "reference to non-existent subpattern" error.
1126    
1127        (d) A pattern like (\P{Yi}{2}\277)? could loop at compile time.
1128    
1129    24. A repeated \S or \W in UTF-8 mode could give wrong answers when multibyte
1130        characters were involved (for example /\S{2}/8g with "A\x{a3}BC").
1131    
1132    25. Using pcregrep in multiline, inverted mode (-Mv) caused it to loop.
1133    
1134    26. Patterns such as [\P{Yi}A] which include \p or \P and just one other
1135        character were causing crashes (broken optimization).
1136    
1137    27. Patterns such as (\P{Yi}*\277)* (group with possible zero repeat containing
1138        \p or \P) caused a compile-time loop.
1139    
1140    28. More problems have arisen in unanchored patterns when CRLF is a valid line
1141        break. For example, the unstudied pattern [\r\n]A does not match the string
1142        "\r\nA" because change 7.0/46 below moves the current point on by two
1143        characters after failing to match at the start. However, the pattern \nA
1144        *does* match, because it doesn't start till \n, and if [\r\n]A is studied,
1145        the same is true. There doesn't seem any very clean way out of this, but
1146        what I have chosen to do makes the common cases work: PCRE now takes note
1147        of whether there can be an explicit match for \r or \n anywhere in the
1148        pattern, and if so, 7.0/46 no longer applies. As part of this change,
1149        there's a new PCRE_INFO_HASCRORLF option for finding out whether a compiled
1150        pattern has explicit CR or LF references.
1151    
1152    29. Added (*CR) etc for changing newline setting at start of pattern.
1153    
1154    
1155    Version 7.2 19-Jun-07
1156    ---------------------
1157    
1158     1. If the fr_FR locale cannot be found for test 3, try the "french" locale,
1159        which is apparently normally available under Windows.
1160    
1161     2. Re-jig the pcregrep tests with different newline settings in an attempt
1162        to make them independent of the local environment's newline setting.
1163    
1164     3. Add code to configure.ac to remove -g from the CFLAGS default settings.
1165    
1166     4. Some of the "internals" tests were previously cut out when the link size
1167        was not 2, because the output contained actual offsets. The recent new
1168        "Z" feature of pcretest means that these can be cut out, making the tests
1169        usable with all link sizes.
1170    
1171     5. Implemented Stan Switzer's goto replacement for longjmp() when not using
1172        stack recursion. This gives a massive performance boost under BSD, but just
1173        a small improvement under Linux. However, it saves one field in the frame
1174        in all cases.
1175    
1176     6. Added more features from the forthcoming Perl 5.10:
1177    
1178        (a) (?-n) (where n is a string of digits) is a relative subroutine or
1179            recursion call. It refers to the nth most recently opened parentheses.
1180    
1181        (b) (?+n) is also a relative subroutine call; it refers to the nth next
1182            to be opened parentheses.
1183    
1184        (c) Conditions that refer to capturing parentheses can be specified
1185            relatively, for example, (?(-2)... or (?(+3)...
1186    
1187        (d) \K resets the start of the current match so that everything before
1188            is not part of it.
1189    
1190        (e) \k{name} is synonymous with \k<name> and \k'name' (.NET compatible).
1191    
1192        (f) \g{name} is another synonym - part of Perl 5.10's unification of
1193            reference syntax.
1194    
1195        (g) (?| introduces a group in which the numbering of parentheses in each
1196            alternative starts with the same number.
1197    
1198        (h) \h, \H, \v, and \V match horizontal and vertical whitespace.
1199    
1200     7. Added two new calls to pcre_fullinfo(): PCRE_INFO_OKPARTIAL and
1201        PCRE_INFO_JCHANGED.
1202    
1203     8. A pattern such as  (.*(.)?)*  caused pcre_exec() to fail by either not
1204        terminating or by crashing. Diagnosed by Viktor Griph; it was in the code
1205        for detecting groups that can match an empty string.
1206    
1207     9. A pattern with a very large number of alternatives (more than several
1208        hundred) was running out of internal workspace during the pre-compile
1209        phase, where pcre_compile() figures out how much memory will be needed. A
1210        bit of new cunning has reduced the workspace needed for groups with
1211        alternatives. The 1000-alternative test pattern now uses 12 bytes of
1212        workspace instead of running out of the 4096 that are available.
1213    
1214    10. Inserted some missing (unsigned int) casts to get rid of compiler warnings.
1215    
1216    11. Applied patch from Google to remove an optimization that didn't quite work.
1217        The report of the bug said:
1218    
1219          pcrecpp::RE("a*").FullMatch("aaa") matches, while
1220          pcrecpp::RE("a*?").FullMatch("aaa") does not, and
1221          pcrecpp::RE("a*?\\z").FullMatch("aaa") does again.
1222    
1223    12. If \p or \P was used in non-UTF-8 mode on a character greater than 127
1224        it matched the wrong number of bytes.
1225    
1226    
1227    Version 7.1 24-Apr-07
1228    ---------------------
1229    
1230     1. Applied Bob Rossi and Daniel G's patches to convert the build system to one
1231        that is more "standard", making use of automake and other Autotools. There
1232        is some re-arrangement of the files and adjustment of comments consequent
1233        on this.
1234    
1235     2. Part of the patch fixed a problem with the pcregrep tests. The test of -r
1236        for recursive directory scanning broke on some systems because the files
1237        are not scanned in any specific order and on different systems the order
1238        was different. A call to "sort" has been inserted into RunGrepTest for the
1239        approprate test as a short-term fix. In the longer term there may be an
1240        alternative.
1241    
1242     3. I had an email from Eric Raymond about problems translating some of PCRE's
1243        man pages to HTML (despite the fact that I distribute HTML pages, some
1244        people do their own conversions for various reasons). The problems
1245        concerned the use of low-level troff macros .br and .in. I have therefore
1246        removed all such uses from the man pages (some were redundant, some could
1247        be replaced by .nf/.fi pairs). The 132html script that I use to generate
1248        HTML has been updated to handle .nf/.fi and to complain if it encounters
1249        .br or .in.
1250    
1251     4. Updated comments in configure.ac that get placed in config.h.in and also
1252        arranged for config.h to be included in the distribution, with the name
1253        config.h.generic, for the benefit of those who have to compile without
1254        Autotools (compare pcre.h, which is now distributed as pcre.h.generic).
1255    
1256     5. Updated the support (such as it is) for Virtual Pascal, thanks to Stefan
1257        Weber: (1) pcre_internal.h was missing some function renames; (2) updated
1258        makevp.bat for the current PCRE, using the additional files
1259        makevp_c.txt, makevp_l.txt, and pcregexp.pas.
1260    
1261     6. A Windows user reported a minor discrepancy with test 2, which turned out
1262        to be caused by a trailing space on an input line that had got lost in his
1263        copy. The trailing space was an accident, so I've just removed it.
1264    
1265     7. Add -Wl,-R... flags in pcre-config.in for *BSD* systems, as I'm told
1266        that is needed.
1267    
1268     8. Mark ucp_table (in ucptable.h) and ucp_gentype (in pcre_ucp_searchfuncs.c)
1269        as "const" (a) because they are and (b) because it helps the PHP
1270        maintainers who have recently made a script to detect big data structures
1271        in the php code that should be moved to the .rodata section. I remembered
1272        to update Builducptable as well, so it won't revert if ucptable.h is ever
1273        re-created.
1274    
1275     9. Added some extra #ifdef SUPPORT_UTF8 conditionals into pcretest.c,
1276        pcre_printint.src, pcre_compile.c, pcre_study.c, and pcre_tables.c, in
1277        order to be able to cut out the UTF-8 tables in the latter when UTF-8
1278        support is not required. This saves 1.5-2K of code, which is important in
1279        some applications.
1280    
1281        Later: more #ifdefs are needed in pcre_ord2utf8.c and pcre_valid_utf8.c
1282        so as not to refer to the tables, even though these functions will never be
1283        called when UTF-8 support is disabled. Otherwise there are problems with a
1284        shared library.
1285    
1286    10. Fixed two bugs in the emulated memmove() function in pcre_internal.h:
1287    
1288        (a) It was defining its arguments as char * instead of void *.
1289    
1290        (b) It was assuming that all moves were upwards in memory; this was true
1291            a long time ago when I wrote it, but is no longer the case.
1292    
1293        The emulated memove() is provided for those environments that have neither
1294        memmove() nor bcopy(). I didn't think anyone used it these days, but that
1295        is clearly not the case, as these two bugs were recently reported.
1296    
1297    11. The script PrepareRelease is now distributed: it calls 132html, CleanTxt,
1298        and Detrail to create the HTML documentation, the .txt form of the man
1299        pages, and it removes trailing spaces from listed files. It also creates
1300        pcre.h.generic and config.h.generic from pcre.h and config.h. In the latter
1301        case, it wraps all the #defines with #ifndefs. This script should be run
1302        before "make dist".
1303    
1304    12. Fixed two fairly obscure bugs concerned with quantified caseless matching
1305        with Unicode property support.
1306    
1307        (a) For a maximizing quantifier, if the two different cases of the
1308            character were of different lengths in their UTF-8 codings (there are
1309            some cases like this - I found 11), and the matching function had to
1310            back up over a mixture of the two cases, it incorrectly assumed they
1311            were both the same length.
1312    
1313        (b) When PCRE was configured to use the heap rather than the stack for
1314            recursion during matching, it was not correctly preserving the data for
1315            the other case of a UTF-8 character when checking ahead for a match
1316            while processing a minimizing repeat. If the check also involved
1317            matching a wide character, but failed, corruption could cause an
1318            erroneous result when trying to check for a repeat of the original
1319            character.
1320    
1321    13. Some tidying changes to the testing mechanism:
1322    
1323        (a) The RunTest script now detects the internal link size and whether there
1324            is UTF-8 and UCP support by running ./pcretest -C instead of relying on
1325            values substituted by "configure". (The RunGrepTest script already did
1326            this for UTF-8.) The configure.ac script no longer substitutes the
1327            relevant variables.
1328    
1329        (b) The debugging options /B and /D in pcretest show the compiled bytecode
1330            with length and offset values. This means that the output is different
1331            for different internal link sizes. Test 2 is skipped for link sizes
1332            other than 2 because of this, bypassing the problem. Unfortunately,
1333            there was also a test in test 3 (the locale tests) that used /B and
1334            failed for link sizes other than 2. Rather than cut the whole test out,
1335            I have added a new /Z option to pcretest that replaces the length and
1336            offset values with spaces. This is now used to make test 3 independent
1337            of link size. (Test 2 will be tidied up later.)
1338    
1339    14. If erroroffset was passed as NULL to pcre_compile, it provoked a
1340        segmentation fault instead of returning the appropriate error message.
1341    
1342    15. In multiline mode when the newline sequence was set to "any", the pattern
1343        ^$ would give a match between the \r and \n of a subject such as "A\r\nB".
1344        This doesn't seem right; it now treats the CRLF combination as the line
1345        ending, and so does not match in that case. It's only a pattern such as ^$
1346        that would hit this one: something like ^ABC$ would have failed after \r
1347        and then tried again after \r\n.
1348    
1349    16. Changed the comparison command for RunGrepTest from "diff -u" to "diff -ub"
1350        in an attempt to make files that differ only in their line terminators
1351        compare equal. This works on Linux.
1352    
1353    17. Under certain error circumstances pcregrep might try to free random memory
1354        as it exited. This is now fixed, thanks to valgrind.
1355    
1356    19. In pcretest, if the pattern /(?m)^$/g<any> was matched against the string
1357        "abc\r\n\r\n", it found an unwanted second match after the second \r. This
1358        was because its rules for how to advance for /g after matching an empty
1359        string at the end of a line did not allow for this case. They now check for
1360        it specially.
1361    
1362    20. pcretest is supposed to handle patterns and data of any length, by
1363        extending its buffers when necessary. It was getting this wrong when the
1364        buffer for a data line had to be extended.
1365    
1366    21. Added PCRE_NEWLINE_ANYCRLF which is like ANY, but matches only CR, LF, or
1367        CRLF as a newline sequence.
1368    
1369    22. Code for handling Unicode properties in pcre_dfa_exec() wasn't being cut
1370        out by #ifdef SUPPORT_UCP. This did no harm, as it could never be used, but
1371        I have nevertheless tidied it up.
1372    
1373    23. Added some casts to kill warnings from HP-UX ia64 compiler.
1374    
1375    24. Added a man page for pcre-config.
1376    
1377    
1378    Version 7.0 19-Dec-06
1379    ---------------------
1380    
1381     1. Fixed a signed/unsigned compiler warning in pcre_compile.c, shown up by
1382        moving to gcc 4.1.1.
1383    
1384     2. The -S option for pcretest uses setrlimit(); I had omitted to #include
1385        sys/time.h, which is documented as needed for this function. It doesn't
1386        seem to matter on Linux, but it showed up on some releases of OS X.
1387    
1388     3. It seems that there are systems where bytes whose values are greater than
1389        127 match isprint() in the "C" locale. The "C" locale should be the
1390        default when a C program starts up. In most systems, only ASCII printing
1391        characters match isprint(). This difference caused the output from pcretest
1392        to vary, making some of the tests fail. I have changed pcretest so that:
1393    
1394        (a) When it is outputting text in the compiled version of a pattern, bytes
1395            other than 32-126 are always shown as hex escapes.
1396    
1397        (b) When it is outputting text that is a matched part of a subject string,
1398            it does the same, unless a different locale has been set for the match
1399            (using the /L modifier). In this case, it uses isprint() to decide.
1400    
1401     4. Fixed a major bug that caused incorrect computation of the amount of memory
1402        required for a compiled pattern when options that changed within the
1403        pattern affected the logic of the preliminary scan that determines the
1404        length. The relevant options are -x, and -i in UTF-8 mode. The result was
1405        that the computed length was too small. The symptoms of this bug were
1406        either the PCRE error "internal error: code overflow" from pcre_compile(),
1407        or a glibc crash with a message such as "pcretest: free(): invalid next
1408        size (fast)". Examples of patterns that provoked this bug (shown in
1409        pcretest format) are:
1410    
1411          /(?-x: )/x
1412          /(?x)(?-x: \s*#\s*)/
1413          /((?i)[\x{c0}])/8
1414          /(?i:[\x{c0}])/8
1415    
1416        HOWEVER: Change 17 below makes this fix obsolete as the memory computation
1417        is now done differently.
1418    
1419     5. Applied patches from Google to: (a) add a QuoteMeta function to the C++
1420        wrapper classes; (b) implement a new function in the C++ scanner that is
1421        more efficient than the old way of doing things because it avoids levels of
1422        recursion in the regex matching; (c) add a paragraph to the documentation
1423        for the FullMatch() function.
1424    
1425     6. The escape sequence \n was being treated as whatever was defined as
1426        "newline". Not only was this contrary to the documentation, which states
1427        that \n is character 10 (hex 0A), but it also went horribly wrong when
1428        "newline" was defined as CRLF. This has been fixed.
1429    
1430     7. In pcre_dfa_exec.c the value of an unsigned integer (the variable called c)
1431        was being set to -1 for the "end of line" case (supposedly a value that no
1432        character can have). Though this value is never used (the check for end of
1433        line is "zero bytes in current character"), it caused compiler complaints.
1434        I've changed it to 0xffffffff.
1435    
1436     8. In pcre_version.c, the version string was being built by a sequence of
1437        C macros that, in the event of PCRE_PRERELEASE being defined as an empty
1438        string (as it is for production releases) called a macro with an empty
1439        argument. The C standard says the result of this is undefined. The gcc
1440        compiler treats it as an empty string (which was what was wanted) but it is
1441        reported that Visual C gives an error. The source has been hacked around to
1442        avoid this problem.
1443    
1444     9. On the advice of a Windows user, included <io.h> and <fcntl.h> in Windows
1445        builds of pcretest, and changed the call to _setmode() to use _O_BINARY
1446        instead of 0x8000. Made all the #ifdefs test both _WIN32 and WIN32 (not all
1447        of them did).
1448    
1449    10. Originally, pcretest opened its input and output without "b"; then I was
1450        told that "b" was needed in some environments, so it was added for release
1451        5.0 to both the input and output. (It makes no difference on Unix-like
1452        systems.) Later I was told that it is wrong for the input on Windows. I've
1453        now abstracted the modes into two macros, to make it easier to fiddle with
1454        them, and removed "b" from the input mode under Windows.
1455    
1456    11. Added pkgconfig support for the C++ wrapper library, libpcrecpp.
1457    
1458    12. Added -help and --help to pcretest as an official way of being reminded
1459        of the options.
1460    
1461    13. Removed some redundant semicolons after macro calls in pcrecpparg.h.in
1462        and pcrecpp.cc because they annoy compilers at high warning levels.
1463    
1464    14. A bit of tidying/refactoring in pcre_exec.c in the main bumpalong loop.
1465    
1466    15. Fixed an occurrence of == in configure.ac that should have been = (shell
1467        scripts are not C programs :-) and which was not noticed because it works
1468        on Linux.
1469    
1470    16. pcretest is supposed to handle any length of pattern and data line (as one
1471        line or as a continued sequence of lines) by extending its input buffer if
1472        necessary. This feature was broken for very long pattern lines, leading to
1473        a string of junk being passed to pcre_compile() if the pattern was longer
1474        than about 50K.
1475    
1476    17. I have done a major re-factoring of the way pcre_compile() computes the
1477        amount of memory needed for a compiled pattern. Previously, there was code
1478        that made a preliminary scan of the pattern in order to do this. That was
1479        OK when PCRE was new, but as the facilities have expanded, it has become
1480        harder and harder to keep it in step with the real compile phase, and there
1481        have been a number of bugs (see for example, 4 above). I have now found a
1482        cunning way of running the real compile function in a "fake" mode that
1483        enables it to compute how much memory it would need, while actually only
1484        ever using a few hundred bytes of working memory and without too many
1485        tests of the mode. This should make future maintenance and development
1486        easier. A side effect of this work is that the limit of 200 on the nesting
1487        depth of parentheses has been removed (though this was never a serious
1488        limitation, I suspect). However, there is a downside: pcre_compile() now
1489        runs more slowly than before (30% or more, depending on the pattern). I
1490        hope this isn't a big issue. There is no effect on runtime performance.
1491    
1492    18. Fixed a minor bug in pcretest: if a pattern line was not terminated by a
1493        newline (only possible for the last line of a file) and it was a
1494        pattern that set a locale (followed by /Lsomething), pcretest crashed.
1495    
1496    19. Added additional timing features to pcretest. (1) The -tm option now times
1497        matching only, not compiling. (2) Both -t and -tm can be followed, as a
1498        separate command line item, by a number that specifies the number of
1499        repeats to use when timing. The default is 50000; this gives better
1500        precision, but takes uncomfortably long for very large patterns.
1501    
1502    20. Extended pcre_study() to be more clever in cases where a branch of a
1503        subpattern has no definite first character. For example, (a*|b*)[cd] would
1504        previously give no result from pcre_study(). Now it recognizes that the
1505        first character must be a, b, c, or d.
1506    
1507    21. There was an incorrect error "recursive call could loop indefinitely" if
1508        a subpattern (or the entire pattern) that was being tested for matching an
1509        empty string contained only one non-empty item after a nested subpattern.
1510        For example, the pattern (?>\x{100}*)\d(?R) provoked this error
1511        incorrectly, because the \d was being skipped in the check.
1512    
1513    22. The pcretest program now has a new pattern option /B and a command line
1514        option -b, which is equivalent to adding /B to every pattern. This causes
1515        it to show the compiled bytecode, without the additional information that
1516        -d shows. The effect of -d is now the same as -b with -i (and similarly, /D
1517        is the same as /B/I).
1518    
1519    23. A new optimization is now able automatically to treat some sequences such
1520        as a*b as a*+b. More specifically, if something simple (such as a character
1521        or a simple class like \d) has an unlimited quantifier, and is followed by
1522        something that cannot possibly match the quantified thing, the quantifier
1523        is automatically "possessified".
1524    
1525    24. A recursive reference to a subpattern whose number was greater than 39
1526        went wrong under certain circumstances in UTF-8 mode. This bug could also
1527        have affected the operation of pcre_study().
1528    
1529    25. Realized that a little bit of performance could be had by replacing
1530        (c & 0xc0) == 0xc0 with c >= 0xc0 when processing UTF-8 characters.
1531    
1532    26. Timing data from pcretest is now shown to 4 decimal places instead of 3.
1533    
1534    27. Possessive quantifiers such as a++ were previously implemented by turning
1535        them into atomic groups such as ($>a+). Now they have their own opcodes,
1536        which improves performance. This includes the automatically created ones
1537        from 23 above.
1538    
1539    28. A pattern such as (?=(\w+))\1: which simulates an atomic group using a
1540        lookahead was broken if it was not anchored. PCRE was mistakenly expecting
1541        the first matched character to be a colon. This applied both to named and
1542        numbered groups.
1543    
1544    29. The ucpinternal.h header file was missing its idempotency #ifdef.
1545    
1546    30. I was sent a "project" file called libpcre.a.dev which I understand makes
1547        building PCRE on Windows easier, so I have included it in the distribution.
1548    
1549    31. There is now a check in pcretest against a ridiculously large number being
1550        returned by pcre_exec() or pcre_dfa_exec(). If this happens in a /g or /G
1551        loop, the loop is abandoned.
1552    
1553    32. Forward references to subpatterns in conditions such as (?(2)...) where
1554        subpattern 2 is defined later cause pcre_compile() to search forwards in
1555        the pattern for the relevant set of parentheses. This search went wrong
1556        when there were unescaped parentheses in a character class, parentheses
1557        escaped with \Q...\E, or parentheses in a #-comment in /x mode.
1558    
1559    33. "Subroutine" calls and backreferences were previously restricted to
1560        referencing subpatterns earlier in the regex. This restriction has now
1561        been removed.
1562    
1563    34. Added a number of extra features that are going to be in Perl 5.10. On the
1564        whole, these are just syntactic alternatives for features that PCRE had
1565        previously implemented using the Python syntax or my own invention. The
1566        other formats are all retained for compatibility.
1567    
1568        (a) Named groups can now be defined as (?<name>...) or (?'name'...) as well
1569            as (?P<name>...). The new forms, as well as being in Perl 5.10, are
1570            also .NET compatible.
1571    
1572        (b) A recursion or subroutine call to a named group can now be defined as
1573            (?&name) as well as (?P>name).
1574    
1575        (c) A backreference to a named group can now be defined as \k<name> or
1576            \k'name' as well as (?P=name). The new forms, as well as being in Perl
1577            5.10, are also .NET compatible.
1578    
1579        (d) A conditional reference to a named group can now use the syntax
1580            (?(<name>) or (?('name') as well as (?(name).
1581    
1582        (e) A "conditional group" of the form (?(DEFINE)...) can be used to define
1583            groups (named and numbered) that are never evaluated inline, but can be
1584            called as "subroutines" from elsewhere. In effect, the DEFINE condition
1585            is always false. There may be only one alternative in such a group.
1586    
1587        (f) A test for recursion can be given as (?(R1).. or (?(R&name)... as well
1588            as the simple (?(R). The condition is true only if the most recent
1589            recursion is that of the given number or name. It does not search out
1590            through the entire recursion stack.
1591    
1592        (g) The escape \gN or \g{N} has been added, where N is a positive or
1593            negative number, specifying an absolute or relative reference.
1594    
1595    35. Tidied to get rid of some further signed/unsigned compiler warnings and
1596        some "unreachable code" warnings.
1597    
1598    36. Updated the Unicode property tables to Unicode version 5.0.0. Amongst other
1599        things, this adds five new scripts.
1600    
1601    37. Perl ignores orphaned \E escapes completely. PCRE now does the same.
1602        There were also incompatibilities regarding the handling of \Q..\E inside
1603        character classes, for example with patterns like [\Qa\E-\Qz\E] where the
1604        hyphen was adjacent to \Q or \E. I hope I've cleared all this up now.
1605    
1606    38. Like Perl, PCRE detects when an indefinitely repeated parenthesized group
1607        matches an empty string, and forcibly breaks the loop. There were bugs in
1608        this code in non-simple cases. For a pattern such as  ^(a()*)*  matched
1609        against  aaaa  the result was just "a" rather than "aaaa", for example. Two
1610        separate and independent bugs (that affected different cases) have been
1611        fixed.
1612    
1613    39. Refactored the code to abolish the use of different opcodes for small
1614        capturing bracket numbers. This is a tidy that I avoided doing when I
1615        removed the limit on the number of capturing brackets for 3.5 back in 2001.
1616        The new approach is not only tidier, it makes it possible to reduce the
1617        memory needed to fix the previous bug (38).
1618    
1619    40. Implemented PCRE_NEWLINE_ANY to recognize any of the Unicode newline
1620        sequences (http://unicode.org/unicode/reports/tr18/) as "newline" when
1621        processing dot, circumflex, or dollar metacharacters, or #-comments in /x
1622        mode.
1623    
1624    41. Add \R to match any Unicode newline sequence, as suggested in the Unicode
1625        report.
1626    
1627    42. Applied patch, originally from Ari Pollak, modified by Google, to allow
1628        copy construction and assignment in the C++ wrapper.
1629    
1630    43. Updated pcregrep to support "--newline=any". In the process, I fixed a
1631        couple of bugs that could have given wrong results in the "--newline=crlf"
1632        case.
1633    
1634    44. Added a number of casts and did some reorganization of signed/unsigned int
1635        variables following suggestions from Dair Grant. Also renamed the variable
1636        "this" as "item" because it is a C++ keyword.
1637    
1638    45. Arranged for dftables to add
1639    
1640          #include "pcre_internal.h"
1641    
1642        to pcre_chartables.c because without it, gcc 4.x may remove the array
1643        definition from the final binary if PCRE is built into a static library and
1644        dead code stripping is activated.
1645    
1646    46. For an unanchored pattern, if a match attempt fails at the start of a
1647        newline sequence, and the newline setting is CRLF or ANY, and the next two
1648        characters are CRLF, advance by two characters instead of one.
1649    
1650    
1651    Version 6.7 04-Jul-06
1652    ---------------------
1653    
1654     1. In order to handle tests when input lines are enormously long, pcretest has
1655        been re-factored so that it automatically extends its buffers when
1656        necessary. The code is crude, but this _is_ just a test program. The
1657        default size has been increased from 32K to 50K.
1658    
1659     2. The code in pcre_study() was using the value of the re argument before
1660        testing it for NULL. (Of course, in any sensible call of the function, it
1661        won't be NULL.)
1662    
1663     3. The memmove() emulation function in pcre_internal.h, which is used on
1664        systems that lack both memmove() and bcopy() - that is, hardly ever -
1665        was missing a "static" storage class specifier.
1666    
1667     4. When UTF-8 mode was not set, PCRE looped when compiling certain patterns
1668        containing an extended class (one that cannot be represented by a bitmap
1669        because it contains high-valued characters or Unicode property items, e.g.
1670        [\pZ]). Almost always one would set UTF-8 mode when processing such a
1671        pattern, but PCRE should not loop if you do not (it no longer does).
1672        [Detail: two cases were found: (a) a repeated subpattern containing an
1673        extended class; (b) a recursive reference to a subpattern that followed a
1674        previous extended class. It wasn't skipping over the extended class
1675        correctly when UTF-8 mode was not set.]
1676    
1677     5. A negated single-character class was not being recognized as fixed-length
1678        in lookbehind assertions such as (?<=[^f]), leading to an incorrect
1679        compile error "lookbehind assertion is not fixed length".
1680    
1681     6. The RunPerlTest auxiliary script was showing an unexpected difference
1682        between PCRE and Perl for UTF-8 tests. It turns out that it is hard to
1683        write a Perl script that can interpret lines of an input file either as
1684        byte characters or as UTF-8, which is what "perltest" was being required to
1685        do for the non-UTF-8 and UTF-8 tests, respectively. Essentially what you
1686        can't do is switch easily at run time between having the "use utf8;" pragma
1687        or not. In the end, I fudged it by using the RunPerlTest script to insert
1688        "use utf8;" explicitly for the UTF-8 tests.
1689    
1690     7. In multiline (/m) mode, PCRE was matching ^ after a terminating newline at
1691        the end of the subject string, contrary to the documentation and to what
1692        Perl does. This was true of both matching functions. Now it matches only at
1693        the start of the subject and immediately after *internal* newlines.
1694    
1695     8. A call of pcre_fullinfo() from pcretest to get the option bits was passing
1696        a pointer to an int instead of a pointer to an unsigned long int. This
1697        caused problems on 64-bit systems.
1698    
1699     9. Applied a patch from the folks at Google to pcrecpp.cc, to fix "another
1700        instance of the 'standard' template library not being so standard".
1701    
1702    10. There was no check on the number of named subpatterns nor the maximum
1703        length of a subpattern name. The product of these values is used to compute
1704        the size of the memory block for a compiled pattern. By supplying a very
1705        long subpattern name and a large number of named subpatterns, the size
1706        computation could be caused to overflow. This is now prevented by limiting
1707        the length of names to 32 characters, and the number of named subpatterns
1708        to 10,000.
1709    
1710    11. Subpatterns that are repeated with specific counts have to be replicated in
1711        the compiled pattern. The size of memory for this was computed from the
1712        length of the subpattern and the repeat count. The latter is limited to
1713        65535, but there was no limit on the former, meaning that integer overflow
1714        could in principle occur. The compiled length of a repeated subpattern is
1715        now limited to 30,000 bytes in order to prevent this.
1716    
1717    12. Added the optional facility to have named substrings with the same name.
1718    
1719    13. Added the ability to use a named substring as a condition, using the
1720        Python syntax: (?(name)yes|no). This overloads (?(R)... and names that
1721        are numbers (not recommended). Forward references are permitted.
1722    
1723    14. Added forward references in named backreferences (if you see what I mean).
1724    
1725    15. In UTF-8 mode, with the PCRE_DOTALL option set, a quantified dot in the
1726        pattern could run off the end of the subject. For example, the pattern
1727        "(?s)(.{1,5})"8 did this with the subject "ab".
1728    
1729    16. If PCRE_DOTALL or PCRE_MULTILINE were set, pcre_dfa_exec() behaved as if
1730        PCRE_CASELESS was set when matching characters that were quantified with ?
1731        or *.
1732    
1733    17. A character class other than a single negated character that had a minimum
1734        but no maximum quantifier - for example [ab]{6,} - was not handled
1735        correctly by pce_dfa_exec(). It would match only one character.
1736    
1737    18. A valid (though odd) pattern that looked like a POSIX character
1738        class but used an invalid character after [ (for example [[,abc,]]) caused
1739        pcre_compile() to give the error "Failed: internal error: code overflow" or
1740        in some cases to crash with a glibc free() error. This could even happen if
1741        the pattern terminated after [[ but there just happened to be a sequence of
1742        letters, a binary zero, and a closing ] in the memory that followed.
1743    
1744    19. Perl's treatment of octal escapes in the range \400 to \777 has changed
1745        over the years. Originally (before any Unicode support), just the bottom 8
1746        bits were taken. Thus, for example, \500 really meant \100. Nowadays the
1747        output from "man perlunicode" includes this:
1748    
1749          The regular expression compiler produces polymorphic opcodes.  That
1750          is, the pattern adapts to the data and automatically switches to
1751          the Unicode character scheme when presented with Unicode data--or
1752          instead uses a traditional byte scheme when presented with byte
1753          data.
1754    
1755        Sadly, a wide octal escape does not cause a switch, and in a string with
1756        no other multibyte characters, these octal escapes are treated as before.
1757        Thus, in Perl, the pattern  /\500/ actually matches \100 but the pattern
1758        /\500|\x{1ff}/ matches \500 or \777 because the whole thing is treated as a
1759        Unicode string.
1760    
1761        I have not perpetrated such confusion in PCRE. Up till now, it took just
1762        the bottom 8 bits, as in old Perl. I have now made octal escapes with
1763        values greater than \377 illegal in non-UTF-8 mode. In UTF-8 mode they
1764        translate to the appropriate multibyte character.
1765    
1766    29. Applied some refactoring to reduce the number of warnings from Microsoft
1767        and Borland compilers. This has included removing the fudge introduced
1768        seven years ago for the OS/2 compiler (see 2.02/2 below) because it caused
1769        a warning about an unused variable.
1770    
1771    21. PCRE has not included VT (character 0x0b) in the set of whitespace
1772        characters since release 4.0, because Perl (from release 5.004) does not.
1773        [Or at least, is documented not to: some releases seem to be in conflict
1774        with the documentation.] However, when a pattern was studied with
1775        pcre_study() and all its branches started with \s, PCRE still included VT
1776        as a possible starting character. Of course, this did no harm; it just
1777        caused an unnecessary match attempt.
1778    
1779    22. Removed a now-redundant internal flag bit that recorded the fact that case
1780        dependency changed within the pattern. This was once needed for "required
1781        byte" processing, but is no longer used. This recovers a now-scarce options
1782        bit. Also moved the least significant internal flag bit to the most-
1783        significant bit of the word, which was not previously used (hangover from
1784        the days when it was an int rather than a uint) to free up another bit for
1785        the future.
1786    
1787    23. Added support for CRLF line endings as well as CR and LF. As well as the
1788        default being selectable at build time, it can now be changed at runtime
1789        via the PCRE_NEWLINE_xxx flags. There are now options for pcregrep to
1790        specify that it is scanning data with non-default line endings.
1791    
1792    24. Changed the definition of CXXLINK to make it agree with the definition of
1793        LINK in the Makefile, by replacing LDFLAGS to CXXFLAGS.
1794    
1795    25. Applied Ian Taylor's patches to avoid using another stack frame for tail
1796        recursions. This makes a big different to stack usage for some patterns.
1797    
1798    26. If a subpattern containing a named recursion or subroutine reference such
1799        as (?P>B) was quantified, for example (xxx(?P>B)){3}, the calculation of
1800        the space required for the compiled pattern went wrong and gave too small a
1801        value. Depending on the environment, this could lead to "Failed: internal
1802        error: code overflow at offset 49" or "glibc detected double free or
1803        corruption" errors.
1804    
1805    27. Applied patches from Google (a) to support the new newline modes and (b) to
1806        advance over multibyte UTF-8 characters in GlobalReplace.
1807    
1808    28. Change free() to pcre_free() in pcredemo.c. Apparently this makes a
1809        difference for some implementation of PCRE in some Windows version.
1810    
1811    29. Added some extra testing facilities to pcretest:
1812    
1813        \q<number>   in a data line sets the "match limit" value
1814        \Q<number>   in a data line sets the "match recursion limt" value
1815        -S <number>  sets the stack size, where <number> is in megabytes
1816    
1817        The -S option isn't available for Windows.
1818    
1819    
1820    Version 6.6 06-Feb-06
1821    ---------------------
1822    
1823     1. Change 16(a) for 6.5 broke things, because PCRE_DATA_SCOPE was not defined
1824        in pcreposix.h. I have copied the definition from pcre.h.
1825    
1826     2. Change 25 for 6.5 broke compilation in a build directory out-of-tree
1827        because pcre.h is no longer a built file.
1828    
1829     3. Added Jeff Friedl's additional debugging patches to pcregrep. These are
1830        not normally included in the compiled code.
1831    
1832    
1833    Version 6.5 01-Feb-06
1834    ---------------------
1835    
1836     1. When using the partial match feature with pcre_dfa_exec(), it was not
1837        anchoring the second and subsequent partial matches at the new starting
1838        point. This could lead to incorrect results. For example, with the pattern
1839        /1234/, partially matching against "123" and then "a4" gave a match.
1840    
1841     2. Changes to pcregrep:
1842    
1843        (a) All non-match returns from pcre_exec() were being treated as failures
1844            to match the line. Now, unless the error is PCRE_ERROR_NOMATCH, an
1845            error message is output. Some extra information is given for the
1846            PCRE_ERROR_MATCHLIMIT and PCRE_ERROR_RECURSIONLIMIT errors, which are
1847            probably the only errors that are likely to be caused by users (by
1848            specifying a regex that has nested indefinite repeats, for instance).
1849            If there are more than 20 of these errors, pcregrep is abandoned.
1850    
1851        (b) A binary zero was treated as data while matching, but terminated the
1852            output line if it was written out. This has been fixed: binary zeroes
1853            are now no different to any other data bytes.
1854    
1855        (c) Whichever of the LC_ALL or LC_CTYPE environment variables is set is
1856            used to set a locale for matching. The --locale=xxxx long option has
1857            been added (no short equivalent) to specify a locale explicitly on the
1858            pcregrep command, overriding the environment variables.
1859    
1860        (d) When -B was used with -n, some line numbers in the output were one less
1861            than they should have been.
1862    
1863        (e) Added the -o (--only-matching) option.
1864    
1865        (f) If -A or -C was used with -c (count only), some lines of context were
1866            accidentally printed for the final match.
1867    
1868        (g) Added the -H (--with-filename) option.
1869    
1870        (h) The combination of options -rh failed to suppress file names for files
1871            that were found from directory arguments.
1872    
1873        (i) Added the -D (--devices) and -d (--directories) options.
1874    
1875        (j) Added the -F (--fixed-strings) option.
1876    
1877        (k) Allow "-" to be used as a file name for -f as well as for a data file.
1878    
1879        (l) Added the --colo(u)r option.
1880    
1881        (m) Added Jeffrey Friedl's -S testing option, but within #ifdefs so that it
1882            is not present by default.
1883    
1884     3. A nasty bug was discovered in the handling of recursive patterns, that is,
1885        items such as (?R) or (?1), when the recursion could match a number of
1886        alternatives. If it matched one of the alternatives, but subsequently,
1887        outside the recursion, there was a failure, the code tried to back up into
1888        the recursion. However, because of the way PCRE is implemented, this is not
1889        possible, and the result was an incorrect result from the match.
1890    
1891        In order to prevent this happening, the specification of recursion has
1892        been changed so that all such subpatterns are automatically treated as
1893        atomic groups. Thus, for example, (?R) is treated as if it were (?>(?R)).
1894    
1895     4. I had overlooked the fact that, in some locales, there are characters for
1896        which isalpha() is true but neither isupper() nor islower() are true. In
1897        the fr_FR locale, for instance, the \xAA and \xBA characters (ordmasculine
1898        and ordfeminine) are like this. This affected the treatment of \w and \W
1899        when they appeared in character classes, but not when they appeared outside
1900        a character class. The bit map for "word" characters is now created
1901        separately from the results of isalnum() instead of just taking it from the
1902        upper, lower, and digit maps. (Plus the underscore character, of course.)
1903    
1904     5. The above bug also affected the handling of POSIX character classes such as
1905        [[:alpha:]] and [[:alnum:]]. These do not have their own bit maps in PCRE's
1906        permanent tables. Instead, the bit maps for such a class were previously
1907        created as the appropriate unions of the upper, lower, and digit bitmaps.
1908        Now they are created by subtraction from the [[:word:]] class, which has
1909        its own bitmap.
1910    
1911     6. The [[:blank:]] character class matches horizontal, but not vertical space.
1912        It is created by subtracting the vertical space characters (\x09, \x0a,
1913        \x0b, \x0c) from the [[:space:]] bitmap. Previously, however, the
1914        subtraction was done in the overall bitmap for a character class, meaning
1915        that a class such as [\x0c[:blank:]] was incorrect because \x0c would not
1916        be recognized. This bug has been fixed.
1917    
1918     7. Patches from the folks at Google:
1919    
1920          (a) pcrecpp.cc: "to handle a corner case that may or may not happen in
1921          real life, but is still worth protecting against".
1922    
1923          (b) pcrecpp.cc: "corrects a bug when negative radixes are used with
1924          regular expressions".
1925    
1926          (c) pcre_scanner.cc: avoid use of std::count() because not all systems
1927          have it.
1928    
1929          (d) Split off pcrecpparg.h from pcrecpp.h and had the former built by
1930          "configure" and the latter not, in order to fix a problem somebody had
1931          with compiling the Arg class on HP-UX.
1932    
1933          (e) Improve the error-handling of the C++ wrapper a little bit.
1934    
1935          (f) New tests for checking recursion limiting.
1936    
1937     8. The pcre_memmove() function, which is used only if the environment does not
1938        have a standard memmove() function (and is therefore rarely compiled),
1939        contained two bugs: (a) use of int instead of size_t, and (b) it was not
1940        returning a result (though PCRE never actually uses the result).
1941    
1942     9. In the POSIX regexec() interface, if nmatch is specified as a ridiculously
1943        large number - greater than INT_MAX/(3*sizeof(int)) - REG_ESPACE is
1944        returned instead of calling malloc() with an overflowing number that would
1945        most likely cause subsequent chaos.
1946    
1947    10. The debugging option of pcretest was not showing the NO_AUTO_CAPTURE flag.
1948    
1949    11. The POSIX flag REG_NOSUB is now supported. When a pattern that was compiled
1950        with this option is matched, the nmatch and pmatch options of regexec() are
1951        ignored.
1952    
1953    12. Added REG_UTF8 to the POSIX interface. This is not defined by POSIX, but is
1954        provided in case anyone wants to the the POSIX interface with UTF-8
1955        strings.
1956    
1957    13. Added CXXLDFLAGS to the Makefile parameters to provide settings only on the
1958        C++ linking (needed for some HP-UX environments).
1959    
1960    14. Avoid compiler warnings in get_ucpname() when compiled without UCP support
1961        (unused parameter) and in the pcre_printint() function (omitted "default"
1962        switch label when the default is to do nothing).
1963    
1964    15. Added some code to make it possible, when PCRE is compiled as a C++
1965        library, to replace subject pointers for pcre_exec() with a smart pointer
1966        class, thus making it possible to process discontinuous strings.
1967    
1968    16. The two macros PCRE_EXPORT and PCRE_DATA_SCOPE are confusing, and perform
1969        much the same function. They were added by different people who were trying
1970        to make PCRE easy to compile on non-Unix systems. It has been suggested
1971        that PCRE_EXPORT be abolished now that there is more automatic apparatus
1972        for compiling on Windows systems. I have therefore replaced it with
1973        PCRE_DATA_SCOPE. This is set automatically for Windows; if not set it
1974        defaults to "extern" for C or "extern C" for C++, which works fine on
1975        Unix-like systems. It is now possible to override the value of PCRE_DATA_
1976        SCOPE with something explicit in config.h. In addition:
1977    
1978        (a) pcreposix.h still had just "extern" instead of either of these macros;
1979            I have replaced it with PCRE_DATA_SCOPE.
1980    
1981        (b) Functions such as _pcre_xclass(), which are internal to the library,
1982            but external in the C sense, all had PCRE_EXPORT in their definitions.
1983            This is apparently wrong for the Windows case, so I have removed it.
1984            (It makes no difference on Unix-like systems.)
1985    
1986    17. Added a new limit, MATCH_LIMIT_RECURSION, which limits the depth of nesting
1987        of recursive calls to match(). This is different to MATCH_LIMIT because
1988        that limits the total number of calls to match(), not all of which increase
1989        the depth of recursion. Limiting the recursion depth limits the amount of
1990        stack (or heap if NO_RECURSE is set) that is used. The default can be set
1991        when PCRE is compiled, and changed at run time. A patch from Google adds
1992        this functionality to the C++ interface.
1993    
1994    18. Changes to the handling of Unicode character properties:
1995    
1996        (a) Updated the table to Unicode 4.1.0.
1997    
1998        (b) Recognize characters that are not in the table as "Cn" (undefined).
1999    
2000        (c) I revised the way the table is implemented to a much improved format
2001            which includes recognition of ranges. It now supports the ranges that
2002            are defined in UnicodeData.txt, and it also amalgamates other
2003            characters into ranges. This has reduced the number of entries in the
2004            table from around 16,000 to around 3,000, thus reducing its size
2005            considerably. I realized I did not need to use a tree structure after
2006            all - a binary chop search is just as efficient. Having reduced the
2007            number of entries, I extended their size from 6 bytes to 8 bytes to
2008            allow for more data.
2009    
2010        (d) Added support for Unicode script names via properties such as \p{Han}.
2011    
2012    19. In UTF-8 mode, a backslash followed by a non-Ascii character was not
2013        matching that character.
2014    
2015    20. When matching a repeated Unicode property with a minimum greater than zero,
2016        (for example \pL{2,}), PCRE could look past the end of the subject if it
2017        reached it while seeking the minimum number of characters. This could
2018        happen only if some of the characters were more than one byte long, because
2019        there is a check for at least the minimum number of bytes.
2020    
2021    21. Refactored the implementation of \p and \P so as to be more general, to
2022        allow for more different types of property in future. This has changed the
2023        compiled form incompatibly. Anybody with saved compiled patterns that use
2024        \p or \P will have to recompile them.
2025    
2026    22. Added "Any" and "L&" to the supported property types.
2027    
2028    23. Recognize \x{...} as a code point specifier, even when not in UTF-8 mode,
2029        but give a compile time error if the value is greater than 0xff.
2030    
2031    24. The man pages for pcrepartial, pcreprecompile, and pcre_compile2 were
2032        accidentally not being installed or uninstalled.
2033    
2034    25. The pcre.h file was built from pcre.h.in, but the only changes that were
2035        made were to insert the current release number. This seemed silly, because
2036        it made things harder for people building PCRE on systems that don't run
2037        "configure". I have turned pcre.h into a distributed file, no longer built
2038        by "configure", with the version identification directly included. There is
2039        no longer a pcre.h.in file.
2040    
2041        However, this change necessitated a change to the pcre-config script as
2042        well. It is built from pcre-config.in, and one of the substitutions was the
2043        release number. I have updated configure.ac so that ./configure now finds
2044        the release number by grepping pcre.h.
2045    
2046    26. Added the ability to run the tests under valgrind.
2047    
2048    
2049    Version 6.4 05-Sep-05
2050    ---------------------
2051    
2052     1. Change 6.0/10/(l) to pcregrep introduced a bug that caused separator lines
2053        "--" to be printed when multiple files were scanned, even when none of the
2054        -A, -B, or -C options were used. This is not compatible with Gnu grep, so I
2055        consider it to be a bug, and have restored the previous behaviour.
2056    
2057     2. A couple of code tidies to get rid of compiler warnings.
2058    
2059     3. The pcretest program used to cheat by referring to symbols in the library
2060        whose names begin with _pcre_. These are internal symbols that are not
2061        really supposed to be visible externally, and in some environments it is
2062        possible to suppress them. The cheating is now confined to including
2063        certain files from the library's source, which is a bit cleaner.
2064    
2065     4. Renamed pcre.in as pcre.h.in to go with pcrecpp.h.in; it also makes the
2066        file's purpose clearer.
2067    
2068     5. Reorganized pcre_ucp_findchar().
2069    
2070    
2071    Version 6.3 15-Aug-05
2072    ---------------------
2073    
2074     1. The file libpcre.pc.in did not have general read permission in the tarball.
2075    
2076     2. There were some problems when building without C++ support:
2077    
2078        (a) If C++ support was not built, "make install" and "make test" still
2079            tried to test it.
2080    
2081        (b) There were problems when the value of CXX was explicitly set. Some
2082            changes have been made to try to fix these, and ...
2083    
2084        (c) --disable-cpp can now be used to explicitly disable C++ support.
2085    
2086        (d) The use of @CPP_OBJ@ directly caused a blank line preceded by a
2087            backslash in a target when C++ was disabled. This confuses some
2088            versions of "make", apparently. Using an intermediate variable solves
2089            this. (Same for CPP_LOBJ.)
2090    
2091     3. $(LINK_FOR_BUILD) now includes $(CFLAGS_FOR_BUILD) and $(LINK)
2092        (non-Windows) now includes $(CFLAGS) because these flags are sometimes
2093        necessary on certain architectures.
2094    
2095     4. Added a setting of -export-symbols-regex to the link command to remove
2096        those symbols that are exported in the C sense, but actually are local
2097        within the library, and not documented. Their names all begin with
2098        "_pcre_". This is not a perfect job, because (a) we have to except some
2099        symbols that pcretest ("illegally") uses, and (b) the facility isn't always
2100        available (and never for static libraries). I have made a note to try to
2101        find a way round (a) in the future.
2102    
2103    
2104    Version 6.2 01-Aug-05
2105    ---------------------
2106    
2107     1. There was no test for integer overflow of quantifier values. A construction
2108        such as {1111111111111111} would give undefined results. What is worse, if
2109        a minimum quantifier for a parenthesized subpattern overflowed and became
2110        negative, the calculation of the memory size went wrong. This could have
2111        led to memory overwriting.
2112    
2113     2. Building PCRE using VPATH was broken. Hopefully it is now fixed.
2114    
2115     3. Added "b" to the 2nd argument of fopen() in dftables.c, for non-Unix-like
2116        operating environments where this matters.
2117    
2118     4. Applied Giuseppe Maxia's patch to add additional features for controlling
2119        PCRE options from within the C++ wrapper.
2120    
2121     5. Named capturing subpatterns were not being correctly counted when a pattern
2122        was compiled. This caused two problems: (a) If there were more than 100
2123        such subpatterns, the calculation of the memory needed for the whole
2124        compiled pattern went wrong, leading to an overflow error. (b) Numerical
2125        back references of the form \12, where the number was greater than 9, were
2126        not recognized as back references, even though there were sufficient
2127        previous subpatterns.
2128    
2129     6. Two minor patches to pcrecpp.cc in order to allow it to compile on older
2130        versions of gcc, e.g. 2.95.4.
2131    
2132    
2133    Version 6.1 21-Jun-05
2134    ---------------------
2135    
2136     1. There was one reference to the variable "posix" in pcretest.c that was not
2137        surrounded by "#if !defined NOPOSIX".
2138    
2139     2. Make it possible to compile pcretest without DFA support, UTF8 support, or
2140        the cross-check on the old pcre_info() function, for the benefit of the
2141        cut-down version of PCRE that is currently imported into Exim.
2142    
2143     3. A (silly) pattern starting with (?i)(?-i) caused an internal space
2144        allocation error. I've done the easy fix, which wastes 2 bytes for sensible
2145        patterns that start (?i) but I don't think that matters. The use of (?i) is
2146        just an example; this all applies to the other options as well.
2147    
2148     4. Since libtool seems to echo the compile commands it is issuing, the output
2149        from "make" can be reduced a bit by putting "@" in front of each libtool
2150        compile command.
2151    
2152     5. Patch from the folks at Google for configure.in to be a bit more thorough
2153        in checking for a suitable C++ installation before trying to compile the
2154        C++ stuff. This should fix a reported problem when a compiler was present,
2155        but no suitable headers.
2156    
2157     6. The man pages all had just "PCRE" as their title. I have changed them to
2158        be the relevant file name. I have also arranged that these names are
2159        retained in the file doc/pcre.txt, which is a concatenation in text format
2160        of all the man pages except the little individual ones for each function.
2161    
2162     7. The NON-UNIX-USE file had not been updated for the different set of source
2163        files that come with release 6. I also added a few comments about the C++
2164        wrapper.
2165    
2166    
2167    Version 6.0 07-Jun-05
2168    ---------------------
2169    
2170     1. Some minor internal re-organization to help with my DFA experiments.
2171    
2172     2. Some missing #ifdef SUPPORT_UCP conditionals in pcretest and printint that
2173        didn't matter for the library itself when fully configured, but did matter
2174        when compiling without UCP support, or within Exim, where the ucp files are
2175        not imported.
2176    
2177     3. Refactoring of the library code to split up the various functions into
2178        different source modules. The addition of the new DFA matching code (see
2179        below) to a single monolithic source would have made it really too
2180        unwieldy, quite apart from causing all the code to be include in a
2181        statically linked application, when only some functions are used. This is
2182        relevant even without the DFA addition now that patterns can be compiled in
2183        one application and matched in another.
2184    
2185        The downside of splitting up is that there have to be some external
2186        functions and data tables that are used internally in different modules of
2187        the library but which are not part of the API. These have all had their
2188        names changed to start with "_pcre_" so that they are unlikely to clash
2189        with other external names.
2190    
2191     4. Added an alternate matching function, pcre_dfa_exec(), which matches using
2192        a different (DFA) algorithm. Although it is slower than the original
2193        function, it does have some advantages for certain types of matching
2194        problem.
2195    
2196     5. Upgrades to pcretest in order to test the features of pcre_dfa_exec(),
2197        including restarting after a partial match.
2198    
2199     6. A patch for pcregrep that defines INVALID_FILE_ATTRIBUTES if it is not
2200        defined when compiling for Windows was sent to me. I have put it into the
2201        code, though I have no means of testing or verifying it.
2202    
2203     7. Added the pcre_refcount() auxiliary function.
2204    
2205     8. Added the PCRE_FIRSTLINE option. This constrains an unanchored pattern to
2206        match before or at the first newline in the subject string. In pcretest,
2207        the /f option on a pattern can be used to set this.
2208    
2209     9. A repeated \w when used in UTF-8 mode with characters greater than 256
2210        would behave wrongly. This has been present in PCRE since release 4.0.
2211    
2212    10. A number of changes to the pcregrep command:
2213    
2214        (a) Refactored how -x works; insert ^(...)$ instead of setting
2215            PCRE_ANCHORED and checking the length, in preparation for adding
2216            something similar for -w.
2217    
2218        (b) Added the -w (match as a word) option.
2219    
2220        (c) Refactored the way lines are read and buffered so as to have more
2221            than one at a time available.
2222    
2223        (d) Implemented a pcregrep test script.
2224    
2225        (e) Added the -M (multiline match) option. This allows patterns to match
2226            over several lines of the subject. The buffering ensures that at least
2227            8K, or the rest of the document (whichever is the shorter) is available
2228            for matching (and similarly the previous 8K for lookbehind assertions).
2229    
2230        (f) Changed the --help output so that it now says
2231    
2232              -w, --word-regex(p)
2233    
2234            instead of two lines, one with "regex" and the other with "regexp"
2235            because that confused at least one person since the short forms are the
2236            same. (This required a bit of code, as the output is generated
2237            automatically from a table. It wasn't just a text change.)
2238    
2239        (g) -- can be used to terminate pcregrep options if the next thing isn't an
2240            option but starts with a hyphen. Could be a pattern or a path name
2241            starting with a hyphen, for instance.
2242    
2243        (h) "-" can be given as a file name to represent stdin.
2244    
2245        (i) When file names are being printed, "(standard input)" is used for
2246            the standard input, for compatibility with GNU grep. Previously
2247            "<stdin>" was used.
2248    
2249        (j) The option --label=xxx can be used to supply a name to be used for
2250            stdin when file names are being printed. There is no short form.
2251    
2252        (k) Re-factored the options decoding logic because we are going to add
2253            two more options that take data. Such options can now be given in four
2254            different ways, e.g. "-fname", "-f name", "--file=name", "--file name".
2255    
2256        (l) Added the -A, -B, and -C options for requesting that lines of context
2257            around matches be printed.
2258    
2259        (m) Added the -L option to print the names of files that do not contain
2260            any matching lines, that is, the complement of -l.
2261    
2262        (n) The return code is 2 if any file cannot be opened, but pcregrep does
2263            continue to scan other files.
2264    
2265        (o) The -s option was incorrectly implemented. For compatibility with other
2266            greps, it now suppresses the error message for a non-existent or non-
2267            accessible file (but not the return code). There is a new option called
2268            -q that suppresses the output of matching lines, which was what -s was
2269            previously doing.
2270    
2271        (p) Added --include and --exclude options to specify files for inclusion
2272            and exclusion when recursing.
2273    
2274    11. The Makefile was not using the Autoconf-supported LDFLAGS macro properly.
2275        Hopefully, it now does.
2276    
2277    12. Missing cast in pcre_study().
2278    
2279    13. Added an "uninstall" target to the makefile.
2280    
2281    14. Replaced "extern" in the function prototypes in Makefile.in with
2282        "PCRE_DATA_SCOPE", which defaults to 'extern' or 'extern "C"' in the Unix
2283        world, but is set differently for Windows.
2284    
2285    15. Added a second compiling function called pcre_compile2(). The only
2286        difference is that it has an extra argument, which is a pointer to an
2287        integer error code. When there is a compile-time failure, this is set
2288        non-zero, in addition to the error test pointer being set to point to an
2289        error message. The new argument may be NULL if no error number is required
2290        (but then you may as well call pcre_compile(), which is now just a
2291        wrapper). This facility is provided because some applications need a
2292        numeric error indication, but it has also enabled me to tidy up the way
2293        compile-time errors are handled in the POSIX wrapper.
2294    
2295    16. Added VPATH=.libs to the makefile; this should help when building with one
2296        prefix path and installing with another. (Or so I'm told by someone who
2297        knows more about this stuff than I do.)
2298    
2299    17. Added a new option, REG_DOTALL, to the POSIX function regcomp(). This
2300        passes PCRE_DOTALL to the pcre_compile() function, making the "." character
2301        match everything, including newlines. This is not POSIX-compatible, but
2302        somebody wanted the feature. From pcretest it can be activated by using
2303        both the P and the s flags.
2304    
2305    18. AC_PROG_LIBTOOL appeared twice in Makefile.in. Removed one.
2306    
2307    19. libpcre.pc was being incorrectly installed as executable.
2308    
2309    20. A couple of places in pcretest check for end-of-line by looking for '\n';
2310        it now also looks for '\r' so that it will work unmodified on Windows.
2311    
2312    21. Added Google's contributed C++ wrapper to the distribution.
2313    
2314    22. Added some untidy missing memory free() calls in pcretest, to keep
2315        Electric Fence happy when testing.
2316    
2317    
2318    
2319    Version 5.0 13-Sep-04
2320    ---------------------
2321    
2322     1. Internal change: literal characters are no longer packed up into items
2323        containing multiple characters in a single byte-string. Each character
2324        is now matched using a separate opcode. However, there may be more than one
2325        byte in the character in UTF-8 mode.
2326    
2327     2. The pcre_callout_block structure has two new fields: pattern_position and
2328        next_item_length. These contain the offset in the pattern to the next match
2329        item, and its length, respectively.
2330    
2331     3. The PCRE_AUTO_CALLOUT option for pcre_compile() requests the automatic
2332        insertion of callouts before each pattern item. Added the /C option to
2333        pcretest to make use of this.
2334    
2335     4. On the advice of a Windows user, the lines
2336    
2337          #if defined(_WIN32) || defined(WIN32)
2338          _setmode( _fileno( stdout ), 0x8000 );
2339          #endif  /* defined(_WIN32) || defined(WIN32) */
2340    
2341        have been added to the source of pcretest. This apparently does useful
2342        magic in relation to line terminators.
2343    
2344     5. Changed "r" and "w" in the calls to fopen() in pcretest to "rb" and "wb"
2345        for the benefit of those environments where the "b" makes a difference.
2346    
2347     6. The icc compiler has the same options as gcc, but "configure" doesn't seem
2348        to know about it. I have put a hack into configure.in that adds in code
2349        to set GCC=yes if CC=icc. This seems to end up at a point in the
2350        generated configure script that is early enough to affect the setting of
2351        compiler options, which is what is needed, but I have no means of testing
2352        whether it really works. (The user who reported this had patched the
2353        generated configure script, which of course I cannot do.)
2354    
2355        LATER: After change 22 below (new libtool files), the configure script
2356        seems to know about icc (and also ecc). Therefore, I have commented out
2357        this hack in configure.in.
2358    
2359     7. Added support for pkg-config (2 patches were sent in).
2360    
2361     8. Negated POSIX character classes that used a combination of internal tables
2362        were completely broken. These were [[:^alpha:]], [[:^alnum:]], and
2363        [[:^ascii]]. Typically, they would match almost any characters. The other
2364        POSIX classes were not broken in this way.
2365    
2366     9. Matching the pattern "\b.*?" against "ab cd", starting at offset 1, failed
2367        to find the match, as PCRE was deluded into thinking that the match had to
2368        start at the start point or following a newline. The same bug applied to
2369        patterns with negative forward assertions or any backward assertions
2370        preceding ".*" at the start, unless the pattern required a fixed first
2371        character. This was a failing pattern: "(?!.bcd).*". The bug is now fixed.
2372    
2373    10. In UTF-8 mode, when moving forwards in the subject after a failed match
2374        starting at the last subject character, bytes beyond the end of the subject
2375        string were read.
2376    
2377    11. Renamed the variable "class" as "classbits" to make life easier for C++
2378        users. (Previously there was a macro definition, but it apparently wasn't
2379        enough.)
2380    
2381    12. Added the new field "tables" to the extra data so that tables can be passed
2382        in at exec time, or the internal tables can be re-selected. This allows
2383        a compiled regex to be saved and re-used at a later time by a different
2384        program that might have everything at different addresses.
2385    
2386    13. Modified the pcre-config script so that, when run on Solaris, it shows a
2387        -R library as well as a -L library.
2388    
2389    14. The debugging options of pcretest (-d on the command line or D on a
2390        pattern) showed incorrect output for anything following an extended class
2391        that contained multibyte characters and which was followed by a quantifier.
2392    
2393    15. Added optional support for general category Unicode character properties
2394        via the \p, \P, and \X escapes. Unicode property support implies UTF-8
2395        support. It adds about 90K to the size of the library. The meanings of the
2396        inbuilt class escapes such as \d and \s have NOT been changed.
2397    
2398    16. Updated pcredemo.c to include calls to free() to release the memory for the
2399        compiled pattern.
2400    
2401    17. The generated file chartables.c was being created in the source directory
2402        instead of in the building directory. This caused the build to fail if the
2403        source directory was different from the building directory, and was
2404        read-only.
2405    
2406    18. Added some sample Win commands from Mark Tetrode into the NON-UNIX-USE
2407        file. No doubt somebody will tell me if they don't make sense... Also added
2408        Dan Mooney's comments about building on OpenVMS.
2409    
2410    19. Added support for partial matching via the PCRE_PARTIAL option for
2411        pcre_exec() and the \P data escape in pcretest.
2412    
2413    20. Extended pcretest with 3 new pattern features:
2414    
2415        (i)   A pattern option of the form ">rest-of-line" causes pcretest to
2416              write the compiled pattern to the file whose name is "rest-of-line".
2417              This is a straight binary dump of the data, with the saved pointer to
2418              the character tables forced to be NULL. The study data, if any, is
2419              written too. After writing, pcretest reads a new pattern.
2420    
2421        (ii)  If, instead of a pattern, "<rest-of-line" is given, pcretest reads a
2422              compiled pattern from the given file. There must not be any
2423              occurrences of "<" in the file name (pretty unlikely); if there are,
2424              pcretest will instead treat the initial "<" as a pattern delimiter.
2425              After reading in the pattern, pcretest goes on to read data lines as
2426              usual.
2427    
2428        (iii) The F pattern option causes pcretest to flip the bytes in the 32-bit
2429              and 16-bit fields in a compiled pattern, to simulate a pattern that
2430              was compiled on a host of opposite endianness.
2431    
2432    21. The pcre-exec() function can now cope with patterns that were compiled on
2433        hosts of opposite endianness, with this restriction:
2434    
2435          As for any compiled expression that is saved and used later, the tables
2436          pointer field cannot be preserved; the extra_data field in the arguments
2437          to pcre_exec() should be used to pass in a tables address if a value
2438          other than the default internal tables were used at compile time.
2439    
2440    22. Calling pcre_exec() with a negative value of the "ovecsize" parameter is
2441        now diagnosed as an error. Previously, most of the time, a negative number
2442        would have been treated as zero, but if in addition "ovector" was passed as
2443        NULL, a crash could occur.
2444    
2445    23. Updated the files ltmain.sh, config.sub, config.guess, and aclocal.m4 with
2446        new versions from the libtool 1.5 distribution (the last one is a copy of
2447        a file called libtool.m4). This seems to have fixed the need to patch
2448        "configure" to support Darwin 1.3 (which I used to do). However, I still
2449        had to patch ltmain.sh to ensure that ${SED} is set (it isn't on my
2450        workstation).
2451    
2452    24. Changed the PCRE licence to be the more standard "BSD" licence.
2453    
2454    
2455    Version 4.5 01-Dec-03
2456    ---------------------
2457    
2458     1. There has been some re-arrangement of the code for the match() function so
2459        that it can be compiled in a version that does not call itself recursively.
2460        Instead, it keeps those local variables that need separate instances for
2461        each "recursion" in a frame on the heap, and gets/frees frames whenever it
2462        needs to "recurse". Keeping track of where control must go is done by means
2463        of setjmp/longjmp. The whole thing is implemented by a set of macros that
2464        hide most of the details from the main code, and operates only if
2465        NO_RECURSE is defined while compiling pcre.c. If PCRE is built using the
2466        "configure" mechanism, "--disable-stack-for-recursion" turns on this way of
2467        operating.
2468    
2469        To make it easier for callers to provide specially tailored get/free
2470        functions for this usage, two new functions, pcre_stack_malloc, and
2471        pcre_stack_free, are used. They are always called in strict stacking order,
2472        and the size of block requested is always the same.
2473    
2474        The PCRE_CONFIG_STACKRECURSE info parameter can be used to find out whether
2475        PCRE has been compiled to use the stack or the heap for recursion. The
2476        -C option of pcretest uses this to show which version is compiled.
2477    
2478        A new data escape \S, is added to pcretest; it causes the amounts of store
2479        obtained and freed by both kinds of malloc/free at match time to be added
2480        to the output.
2481    
2482     2. Changed the locale test to use "fr_FR" instead of "fr" because that's
2483        what's available on my current Linux desktop machine.
2484    
2485     3. When matching a UTF-8 string, the test for a valid string at the start has
2486        been extended. If start_offset is not zero, PCRE now checks that it points
2487        to a byte that is the start of a UTF-8 character. If not, it returns
2488        PCRE_ERROR_BADUTF8_OFFSET (-11). Note: the whole string is still checked;
2489        this is necessary because there may be backward assertions in the pattern.
2490        When matching the same subject several times, it may save resources to use
2491        PCRE_NO_UTF8_CHECK on all but the first call if the string is long.
2492    
2493     4. The code for checking the validity of UTF-8 strings has been tightened so
2494        that it rejects (a) strings containing 0xfe or 0xff bytes and (b) strings
2495        containing "overlong sequences".
2496    
2497     5. Fixed a bug (appearing twice) that I could not find any way of exploiting!
2498        I had written "if ((digitab[*p++] && chtab_digit) == 0)" where the "&&"
2499        should have been "&", but it just so happened that all the cases this let
2500        through by mistake were picked up later in the function.
2501    
2502     6. I had used a variable called "isblank" - this is a C99 function, causing
2503        some compilers to warn. To avoid this, I renamed it (as "blankclass").
2504    
2505     7. Cosmetic: (a) only output another newline at the end of pcretest if it is
2506        prompting; (b) run "./pcretest /dev/null" at the start of the test script
2507        so the version is shown; (c) stop "make test" echoing "./RunTest".
2508    
2509     8. Added patches from David Burgess to enable PCRE to run on EBCDIC systems.
2510    
2511     9. The prototype for memmove() for systems that don't have it was using
2512        size_t, but the inclusion of the header that defines size_t was later. I've
2513        moved the #includes for the C headers earlier to avoid this.
2514    
2515    10. Added some adjustments to the code to make it easier to compiler on certain
2516        special systems:
2517    
2518          (a) Some "const" qualifiers were missing.
2519          (b) Added the macro EXPORT before all exported functions; by default this
2520              is defined to be empty.
2521          (c) Changed the dftables auxiliary program (that builds chartables.c) so
2522              that it reads its output file name as an argument instead of writing
2523              to the standard output and assuming this can be redirected.
2524    
2525    11. In UTF-8 mode, if a recursive reference (e.g. (?1)) followed a character
2526        class containing characters with values greater than 255, PCRE compilation
2527        went into a loop.
2528    
2529    12. A recursive reference to a subpattern that was within another subpattern
2530        that had a minimum quantifier of zero caused PCRE to crash. For example,
2531        (x(y(?2))z)? provoked this bug with a subject that got as far as the
2532        recursion. If the recursively-called subpattern itself had a zero repeat,
2533        that was OK.
2534    
2535    13. In pcretest, the buffer for reading a data line was set at 30K, but the
2536        buffer into which it was copied (for escape processing) was still set at
2537        1024, so long lines caused crashes.
2538    
2539    14. A pattern such as /[ab]{1,3}+/ failed to compile, giving the error
2540        "internal error: code overflow...". This applied to any character class
2541        that was followed by a possessive quantifier.
2542    
2543    15. Modified the Makefile to add libpcre.la as a prerequisite for
2544        libpcreposix.la because I was told this is needed for a parallel build to
2545        work.
2546    
2547    16. If a pattern that contained .* following optional items at the start was
2548        studied, the wrong optimizing data was generated, leading to matching
2549        errors. For example, studying /[ab]*.*c/ concluded, erroneously, that any
2550        matching string must start with a or b or c. The correct conclusion for
2551        this pattern is that a match can start with any character.
2552    
2553    
2554    Version 4.4 13-Aug-03
2555    ---------------------
2556    
2557     1. In UTF-8 mode, a character class containing characters with values between
2558        127 and 255 was not handled correctly if the compiled pattern was studied.
2559        In fixing this, I have also improved the studying algorithm for such
2560        classes (slightly).
2561    
2562     2. Three internal functions had redundant arguments passed to them. Removal
2563        might give a very teeny performance improvement.
2564    
2565     3. Documentation bug: the value of the capture_top field in a callout is *one
2566        more than* the number of the hightest numbered captured substring.
2567    
2568     4. The Makefile linked pcretest and pcregrep with -lpcre, which could result
2569        in incorrectly linking with a previously installed version. They now link
2570        explicitly with libpcre.la.
2571    
2572     5. configure.in no longer needs to recognize Cygwin specially.
2573    
2574     6. A problem in pcre.in for Windows platforms is fixed.
2575    
2576     7. If a pattern was successfully studied, and the -d (or /D) flag was given to
2577        pcretest, it used to include the size of the study block as part of its
2578        output. Unfortunately, the structure contains a field that has a different
2579        size on different hardware architectures. This meant that the tests that
2580        showed this size failed. As the block is currently always of a fixed size,
2581        this information isn't actually particularly useful in pcretest output, so
2582        I have just removed it.
2583    
2584     8. Three pre-processor statements accidentally did not start in column 1.
2585        Sadly, there are *still* compilers around that complain, even though
2586        standard C has not required this for well over a decade. Sigh.
2587    
2588     9. In pcretest, the code for checking callouts passed small integers in the
2589        callout_data field, which is a void * field. However, some picky compilers
2590        complained about the casts involved for this on 64-bit systems. Now
2591        pcretest passes the address of the small integer instead, which should get
2592        rid of the warnings.
2593    
2594    10. By default, when in UTF-8 mode, PCRE now checks for valid UTF-8 strings at
2595        both compile and run time, and gives an error if an invalid UTF-8 sequence
2596        is found. There is a option for disabling this check in cases where the
2597        string is known to be correct and/or the maximum performance is wanted.
2598    
2599    11. In response to a bug report, I changed one line in Makefile.in from
2600    
2601            -Wl,--out-implib,.libs/lib@WIN_PREFIX@pcreposix.dll.a \
2602        to
2603            -Wl,--out-implib,.libs/@WIN_PREFIX@libpcreposix.dll.a \
2604    
2605        to look similar to other lines, but I have no way of telling whether this
2606        is the right thing to do, as I do not use Windows. No doubt I'll get told
2607        if it's wrong...
2608    
2609    
2610    Version 4.3 21-May-03
2611    ---------------------
2612    
2613    1. Two instances of @WIN_PREFIX@ omitted from the Windows targets in the
2614       Makefile.
2615    
2616    2. Some refactoring to improve the quality of the code:
2617    
2618       (i)   The utf8_table... variables are now declared "const".
2619    
2620       (ii)  The code for \cx, which used the "case flipping" table to upper case
2621             lower case letters, now just substracts 32. This is ASCII-specific,
2622             but the whole concept of \cx is ASCII-specific, so it seems
2623             reasonable.
2624    
2625       (iii) PCRE was using its character types table to recognize decimal and
2626             hexadecimal digits in the pattern. This is silly, because it handles
2627             only 0-9, a-f, and A-F, but the character types table is locale-
2628             specific, which means strange things might happen. A private
2629             table is now used for this - though it costs 256 bytes, a table is
2630             much faster than multiple explicit tests. Of course, the standard
2631             character types table is still used for matching digits in subject
2632             strings against \d.
2633    
2634       (iv)  Strictly, the identifier ESC_t is reserved by POSIX (all identifiers
2635             ending in _t are). So I've renamed it as ESC_tee.
2636    
2637    3. The first argument for regexec() in the POSIX wrapper should have been
2638       defined as "const".
2639    
2640    4. Changed pcretest to use malloc() for its buffers so that they can be
2641       Electric Fenced for debugging.
2642    
2643    5. There were several places in the code where, in UTF-8 mode, PCRE would try
2644       to read one or more bytes before the start of the subject string. Often this
2645       had no effect on PCRE's behaviour, but in some circumstances it could
2646       provoke a segmentation fault.
2647    
2648    6. A lookbehind at the start of a pattern in UTF-8 mode could also cause PCRE
2649       to try to read one or more bytes before the start of the subject string.
2650    
2651    7. A lookbehind in a pattern matched in non-UTF-8 mode on a PCRE compiled with
2652       UTF-8 support could misbehave in various ways if the subject string
2653       contained bytes with the 0x80 bit set and the 0x40 bit unset in a lookbehind
2654       area. (PCRE was not checking for the UTF-8 mode flag, and trying to move
2655       back over UTF-8 characters.)
2656    
2657    
2658    Version 4.2 14-Apr-03
2659    ---------------------
2660    
2661    1. Typo "#if SUPPORT_UTF8" instead of "#ifdef SUPPORT_UTF8" fixed.
2662    
2663    2. Changes to the building process, supplied by Ronald Landheer-Cieslak
2664         [ON_WINDOWS]: new variable, "#" on non-Windows platforms
2665         [NOT_ON_WINDOWS]: new variable, "#" on Windows platforms
2666         [WIN_PREFIX]: new variable, "cyg" for Cygwin
2667         * Makefile.in: use autoconf substitution for OBJEXT, EXEEXT, BUILD_OBJEXT
2668           and BUILD_EXEEXT
2669         Note: automatic setting of the BUILD variables is not yet working
2670         set CPPFLAGS and BUILD_CPPFLAGS (but don't use yet) - should be used at
2671           compile-time but not at link-time
2672         [LINK]: use for linking executables only
2673         make different versions for Windows and non-Windows
2674         [LINKLIB]: new variable, copy of UNIX-style LINK, used for linking
2675           libraries
2676         [LINK_FOR_BUILD]: new variable
2677         [OBJEXT]: use throughout
2678         [EXEEXT]: use throughout
2679         <winshared>: new target
2680         <wininstall>: new target
2681         <dftables.o>: use native compiler
2682         <dftables>: use native linker
2683         <install>: handle Windows platform correctly
2684         <clean>: ditto
2685         <check>: ditto
2686         copy DLL to top builddir before testing
2687    
2688       As part of these changes, -no-undefined was removed again. This was reported
2689       to give trouble on HP-UX 11.0, so getting rid of it seems like a good idea
2690       in any case.
2691    
2692    3. Some tidies to get rid of compiler warnings:
2693    
2694       . In the match_data structure, match_limit was an unsigned long int, whereas
2695         match_call_count was an int. I've made them both unsigned long ints.
2696    
2697       . In pcretest the fact that a const uschar * doesn't automatically cast to
2698         a void * provoked a warning.
2699    
2700       . Turning on some more compiler warnings threw up some "shadow" variables
2701         and a few more missing casts.
2702    
2703    4. If PCRE was complied with UTF-8 support, but called without the PCRE_UTF8
2704       option, a class that contained a single character with a value between 128
2705       and 255 (e.g. /[\xFF]/) caused PCRE to crash.
2706    
2707    5. If PCRE was compiled with UTF-8 support, but called without the PCRE_UTF8
2708       option, a class that contained several characters, but with at least one
2709       whose value was between 128 and 255 caused PCRE to crash.
2710    
2711    
2712    Version 4.1 12-Mar-03
2713    ---------------------
2714    
2715    1. Compiling with gcc -pedantic found a couple of places where casts were
2716    needed, and a string in dftables.c that was longer than standard compilers are
2717    required to support.
2718    
2719    2. Compiling with Sun's compiler found a few more places where the code could
2720    be tidied up in order to avoid warnings.
2721    
2722    3. The variables for cross-compiling were called HOST_CC and HOST_CFLAGS; the
2723    first of these names is deprecated in the latest Autoconf in favour of the name
2724    CC_FOR_BUILD, because "host" is typically used to mean the system on which the
2725    compiled code will be run. I can't find a reference for HOST_CFLAGS, but by
2726    analogy I have changed it to CFLAGS_FOR_BUILD.
2727    
2728    4. Added -no-undefined to the linking command in the Makefile, because this is
2729    apparently helpful for Windows. To make it work, also added "-L. -lpcre" to the
2730    linking step for the pcreposix library.
2731    
2732    5. PCRE was failing to diagnose the case of two named groups with the same
2733    name.
2734    
2735    6. A problem with one of PCRE's optimizations was discovered. PCRE remembers a
2736    literal character that is needed in the subject for a match, and scans along to
2737    ensure that it is present before embarking on the full matching process. This
2738    saves time in cases of nested unlimited repeats that are never going to match.
2739    Problem: the scan can take a lot of time if the subject is very long (e.g.
2740    megabytes), thus penalizing straightforward matches. It is now done only if the
2741    amount of subject to be scanned is less than 1000 bytes.
2742    
2743    7. A lesser problem with the same optimization is that it was recording the
2744    first character of an anchored pattern as "needed", thus provoking a search
2745    right along the subject, even when the first match of the pattern was going to
2746    fail. The "needed" character is now not set for anchored patterns, unless it
2747    follows something in the pattern that is of non-fixed length. Thus, it still
2748    fulfils its original purpose of finding quick non-matches in cases of nested
2749    unlimited repeats, but isn't used for simple anchored patterns such as /^abc/.
2750    
2751    
2752    Version 4.0 17-Feb-03
2753    ---------------------
2754    
2755    1. If a comment in an extended regex that started immediately after a meta-item
2756    extended to the end of string, PCRE compiled incorrect data. This could lead to
2757    all kinds of weird effects. Example: /#/ was bad; /()#/ was bad; /a#/ was not.
2758    
2759    2. Moved to autoconf 2.53 and libtool 1.4.2.
2760    
2761    3. Perl 5.8 no longer needs "use utf8" for doing UTF-8 things. Consequently,
2762    the special perltest8 script is no longer needed - all the tests can be run
2763    from a single perltest script.
2764    
2765    4. From 5.004, Perl has not included the VT character (0x0b) in the set defined
2766    by \s. It has now been removed in PCRE. This means it isn't recognized as
2767    whitespace in /x regexes too, which is the same as Perl. Note that the POSIX
2768    class [:space:] *does* include VT, thereby creating a mess.
2769    
2770    5. Added the class [:blank:] (a GNU extension from Perl 5.8) to match only
2771    space and tab.
2772    
2773    6. Perl 5.005 was a long time ago. It's time to amalgamate the tests that use
2774    its new features into the main test script, reducing the number of scripts.
2775    
2776    7. Perl 5.8 has changed the meaning of patterns like /a(?i)b/. Earlier versions
2777    were backward compatible, and made the (?i) apply to the whole pattern, as if
2778    /i were given. Now it behaves more logically, and applies the option setting
2779    only to what follows. PCRE has been changed to follow suit. However, if it
2780    finds options settings right at the start of the pattern, it extracts them into
2781    the global options, as before. Thus, they show up in the info data.
2782    
2783    8. Added support for the \Q...\E escape sequence. Characters in between are
2784    treated as literals. This is slightly different from Perl in that $ and @ are
2785    also handled as literals inside the quotes. In Perl, they will cause variable
2786    interpolation. Note the following examples:
2787    
2788        Pattern            PCRE matches      Perl matches
2789    
2790        \Qabc$xyz\E        abc$xyz           abc followed by the contents of $xyz
2791        \Qabc\$xyz\E       abc\$xyz          abc\$xyz
2792        \Qabc\E\$\Qxyz\E   abc$xyz           abc$xyz
2793    
2794    For compatibility with Perl, \Q...\E sequences are recognized inside character
2795    classes as well as outside them.
2796    
2797    9. Re-organized 3 code statements in pcretest to avoid "overflow in
2798    floating-point constant arithmetic" warnings from a Microsoft compiler. Added a
2799    (size_t) cast to one statement in pcretest and one in pcreposix to avoid
2800    signed/unsigned warnings.
2801    
2802    10. SunOS4 doesn't have strtoul(). This was used only for unpicking the -o
2803    option for pcretest, so I've replaced it by a simple function that does just
2804    that job.
2805    
2806    11. pcregrep was ending with code 0 instead of 2 for the commands "pcregrep" or
2807    "pcregrep -".
2808    
2809    12. Added "possessive quantifiers" ?+, *+, ++, and {,}+ which come from Sun's
2810    Java package. This provides some syntactic sugar for simple cases of what my
2811    documentation calls "once-only subpatterns". A pattern such as x*+ is the same
2812    as (?>x*). In other words, if what is inside (?>...) is just a single repeated
2813    item, you can use this simplified notation. Note that only makes sense with
2814    greedy quantifiers. Consequently, the use of the possessive quantifier forces
2815    greediness, whatever the setting of the PCRE_UNGREEDY option.
2816    
2817    13. A change of greediness default within a pattern was not taking effect at
2818    the current level for patterns like /(b+(?U)a+)/. It did apply to parenthesized
2819    subpatterns that followed. Patterns like /b+(?U)a+/ worked because the option
2820    was abstracted outside.
2821    
2822    14. PCRE now supports the \G assertion. It is true when the current matching
2823    position is at the start point of the match. This differs from \A when the
2824    starting offset is non-zero. Used with the /g option of pcretest (or similar
2825    code), it works in the same way as it does for Perl's /g option. If all
2826    alternatives of a regex begin with \G, the expression is anchored to the start
2827    match position, and the "anchored" flag is set in the compiled expression.
2828    
2829    15. Some bugs concerning the handling of certain option changes within patterns
2830    have been fixed. These applied to options other than (?ims). For example,
2831    "a(?x: b c )d" did not match "XabcdY" but did match "Xa b c dY". It should have
2832    been the other way round. Some of this was related to change 7 above.
2833    
2834    16. PCRE now gives errors for /[.x.]/ and /[=x=]/ as unsupported POSIX
2835    features, as Perl does. Previously, PCRE gave the warnings only for /[[.x.]]/
2836    and /[[=x=]]/. PCRE now also gives an error for /[:name:]/ because it supports
2837    POSIX classes only within a class (e.g. /[[:alpha:]]/).
2838    
2839    17. Added support for Perl's \C escape. This matches one byte, even in UTF8
2840    mode. Unlike ".", it always matches newline, whatever the setting of
2841    PCRE_DOTALL. However, PCRE does not permit \C to appear in lookbehind
2842    assertions. Perl allows it, but it doesn't (in general) work because it can't
2843    calculate the length of the lookbehind. At least, that's the case for Perl
2844    5.8.0 - I've been told they are going to document that it doesn't work in
2845    future.
2846    
2847    18. Added an error diagnosis for escapes that PCRE does not support: these are
2848    \L, \l, \N, \P, \p, \U, \u, and \X.
2849    
2850    19. Although correctly diagnosing a missing ']' in a character class, PCRE was
2851    reading past the end of the pattern in cases such as /[abcd/.
2852    
2853    20. PCRE was getting more memory than necessary for patterns with classes that
2854    contained both POSIX named classes and other characters, e.g. /[[:space:]abc/.
2855    
2856    21. Added some code, conditional on #ifdef VPCOMPAT, to make life easier for
2857    compiling PCRE for use with Virtual Pascal.
2858    
2859    22. Small fix to the Makefile to make it work properly if the build is done
2860    outside the source tree.
2861    
2862    23. Added a new extension: a condition to go with recursion. If a conditional
2863    subpattern starts with (?(R) the "true" branch is used if recursion has
2864    happened, whereas the "false" branch is used only at the top level.
2865    
2866    24. When there was a very long string of literal characters (over 255 bytes
2867    without UTF support, over 250 bytes with UTF support), the computation of how
2868    much memory was required could be incorrect, leading to segfaults or other
2869    strange effects.
2870    
2871    25. PCRE was incorrectly assuming anchoring (either to start of subject or to
2872    start of line for a non-DOTALL pattern) when a pattern started with (.*) and
2873    there was a subsequent back reference to those brackets. This meant that, for
2874    example, /(.*)\d+\1/ failed to match "abc123bc". Unfortunately, it isn't
2875    possible to check for precisely this case. All we can do is abandon the
2876    optimization if .* occurs inside capturing brackets when there are any back
2877    references whatsoever. (See below for a better fix that came later.)
2878    
2879    26. The handling of the optimization for finding the first character of a
2880    non-anchored pattern, and for finding a character that is required later in the
2881    match were failing in some cases. This didn't break the matching; it just
2882    failed to optimize when it could. The way this is done has been re-implemented.
2883    
2884    27. Fixed typo in error message for invalid (?R item (it said "(?p").
2885    
2886    28. Added a new feature that provides some of the functionality that Perl
2887    provides with (?{...}). The facility is termed a "callout". The way it is done
2888    in PCRE is for the caller to provide an optional function, by setting
2889    pcre_callout to its entry point. Like pcre_malloc and pcre_free, this is a
2890    global variable. By default it is unset, which disables all calling out. To get
2891    the function called, the regex must include (?C) at appropriate points. This
2892    is, in fact, equivalent to (?C0), and any number <= 255 may be given with (?C).
2893    This provides a means of identifying different callout points. When PCRE
2894    reaches such a point in the regex, if pcre_callout has been set, the external
2895    function is called. It is provided with data in a structure called
2896    pcre_callout_block, which is defined in pcre.h. If the function returns 0,
2897    matching continues; if it returns a non-zero value, the match at the current
2898    point fails. However, backtracking will occur if possible. [This was changed
2899    later and other features added - see item 49 below.]
2900    
2901    29. pcretest is upgraded to test the callout functionality. It provides a
2902    callout function that displays information. By default, it shows the start of
2903    the match and the current position in the text. There are some new data escapes
2904    to vary what happens:
2905    
2906        \C+         in addition, show current contents of captured substrings
2907        \C-         do not supply a callout function
2908        \C!n        return 1 when callout number n is reached
2909        \C!n!m      return 1 when callout number n is reached for the mth time
2910    
2911    30. If pcregrep was called with the -l option and just a single file name, it
2912    output "<stdin>" if a match was found, instead of the file name.
2913    
2914    31. Improve the efficiency of the POSIX API to PCRE. If the number of capturing
2915    slots is less than POSIX_MALLOC_THRESHOLD, use a block on the stack to pass to
2916    pcre_exec(). This saves a malloc/free per call. The default value of
2917    POSIX_MALLOC_THRESHOLD is 10; it can be changed by --with-posix-malloc-threshold
2918    when configuring.
2919    
2920    32. The default maximum size of a compiled pattern is 64K. There have been a
2921    few cases of people hitting this limit. The code now uses macros to handle the
2922    storing of links as offsets within the compiled pattern. It defaults to 2-byte
2923    links, but this can be changed to 3 or 4 bytes by --with-link-size when
2924    configuring. Tests 2 and 5 work only with 2-byte links because they output
2925    debugging information about compiled patterns.
2926    
2927    33. Internal code re-arrangements:
2928    
2929    (a) Moved the debugging function for printing out a compiled regex into
2930        its own source file (printint.c) and used #include to pull it into
2931        pcretest.c and, when DEBUG is defined, into pcre.c, instead of having two
2932        separate copies.
2933    
2934    (b) Defined the list of op-code names for debugging as a macro in
2935        internal.h so that it is next to the definition of the opcodes.
2936    
2937    (c) Defined a table of op-code lengths for simpler skipping along compiled
2938        code. This is again a macro in internal.h so that it is next to the
2939        definition of the opcodes.
2940    
2941    34. Added support for recursive calls to individual subpatterns, along the
2942    lines of Robin Houston's patch (but implemented somewhat differently).
2943    
2944    35. Further mods to the Makefile to help Win32. Also, added code to pcregrep to
2945    allow it to read and process whole directories in Win32. This code was
2946    contributed by Lionel Fourquaux; it has not been tested by me.
2947    
2948    36. Added support for named subpatterns. The Python syntax (?P<name>...) is
2949    used to name a group. Names consist of alphanumerics and underscores, and must
2950    be unique. Back references use the syntax (?P=name) and recursive calls use
2951    (?P>name) which is a PCRE extension to the Python extension. Groups still have
2952    numbers. The function pcre_fullinfo() can be used after compilation to extract
2953    a name/number map. There are three relevant calls:
2954    
2955      PCRE_INFO_NAMEENTRYSIZE        yields the size of each entry in the map
2956      PCRE_INFO_NAMECOUNT            yields the number of entries
2957      PCRE_INFO_NAMETABLE            yields a pointer to the map.
2958    
2959    The map is a vector of fixed-size entries. The size of each entry depends on
2960    the length of the longest name used. The first two bytes of each entry are the
2961    group number, most significant byte first. There follows the corresponding
2962    name, zero terminated. The names are in alphabetical order.
2963    
2964    37. Make the maximum literal string in the compiled code 250 for the non-UTF-8
2965    case instead of 255. Making it the same both with and without UTF-8 support
2966    means that the same test output works with both.
2967    
2968    38. There was a case of malloc(0) in the POSIX testing code in pcretest. Avoid
2969    calling malloc() with a zero argument.
2970    
2971    39. Change 25 above had to resort to a heavy-handed test for the .* anchoring
2972    optimization. I've improved things by keeping a bitmap of backreferences with
2973    numbers 1-31 so that if .* occurs inside capturing brackets that are not in
2974    fact referenced, the optimization can be applied. It is unlikely that a
2975    relevant occurrence of .* (i.e. one which might indicate anchoring or forcing
2976    the match to follow \n) will appear inside brackets with a number greater than
2977    31, but if it does, any back reference > 31 suppresses the optimization.
2978    
2979    40. Added a new compile-time option PCRE_NO_AUTO_CAPTURE. This has the effect
2980    of disabling numbered capturing parentheses. Any opening parenthesis that is
2981    not followed by ? behaves as if it were followed by ?: but named parentheses
2982    can still be used for capturing (and they will acquire numbers in the usual
2983    way).
2984    
2985    41. Redesigned the return codes from the match() function into yes/no/error so
2986    that errors can be passed back from deep inside the nested calls. A malloc
2987    failure while inside a recursive subpattern call now causes the
2988    PCRE_ERROR_NOMEMORY return instead of quietly going wrong.
2989    
2990    42. It is now possible to set a limit on the number of times the match()
2991    function is called in a call to pcre_exec(). This facility makes it possible to
2992    limit the amount of recursion and backtracking, though not in a directly
2993    obvious way, because the match() function is used in a number of different
2994    circumstances. The count starts from zero for each position in the subject
2995    string (for non-anchored patterns). The default limit is, for compatibility, a
2996    large number, namely 10 000 000. You can change this in two ways:
2997    
2998    (a) When configuring PCRE before making, you can use --with-match-limit=n
2999        to set a default value for the compiled library.
3000    
3001    (b) For each call to pcre_exec(), you can pass a pcre_extra block in which
3002        a different value is set. See 45 below.
3003    
3004    If the limit is exceeded, pcre_exec() returns PCRE_ERROR_MATCHLIMIT.
3005    
3006    43. Added a new function pcre_config(int, void *) to enable run-time extraction
3007    of things that can be changed at compile time. The first argument specifies
3008    what is wanted and the second points to where the information is to be placed.
3009    The current list of available information is:
3010    
3011      PCRE_CONFIG_UTF8
3012    
3013    The output is an integer that is set to one if UTF-8 support is available;
3014    otherwise it is set to zero.
3015    
3016      PCRE_CONFIG_NEWLINE
3017    
3018    The output is an integer that it set to the value of the code that is used for
3019    newline. It is either LF (10) or CR (13).
3020    
3021      PCRE_CONFIG_LINK_SIZE
3022    
3023    The output is an integer that contains the number of bytes used for internal
3024    linkage in compiled expressions. The value is 2, 3, or 4. See item 32 above.
3025    
3026      PCRE_CONFIG_POSIX_MALLOC_THRESHOLD
3027    
3028    The output is an integer that contains the threshold above which the POSIX
3029    interface uses malloc() for output vectors. See item 31 above.
3030    
3031      PCRE_CONFIG_MATCH_LIMIT
3032    
3033    The output is an unsigned integer that contains the default limit of the number
3034    of match() calls in a pcre_exec() execution. See 42 above.
3035    
3036    44. pcretest has been upgraded by the addition of the -C option. This causes it
3037    to extract all the available output from the new pcre_config() function, and to
3038    output it. The program then exits immediately.
3039    
3040    45. A need has arisen to pass over additional data with calls to pcre_exec() in
3041    order to support additional features. One way would have been to define
3042    pcre_exec2() (for example) with extra arguments, but this would not have been
3043    extensible, and would also have required all calls to the original function to
3044    be mapped to the new one. Instead, I have chosen to extend the mechanism that
3045    is used for passing in "extra" data from pcre_study().
3046    
3047    The pcre_extra structure is now exposed and defined in pcre.h. It currently
3048    contains the following fields:
3049    
3050      flags         a bitmap indicating which of the following fields are set
3051      study_data    opaque data from pcre_study()
3052      match_limit   a way of specifying a limit on match() calls for a specific
3053                      call to pcre_exec()
3054      callout_data  data for callouts (see 49 below)
3055    
3056    The flag bits are also defined in pcre.h, and are
3057    
3058      PCRE_EXTRA_STUDY_DATA
3059      PCRE_EXTRA_MATCH_LIMIT
3060      PCRE_EXTRA_CALLOUT_DATA
3061    
3062    The pcre_study() function now returns one of these new pcre_extra blocks, with
3063    the actual study data pointed to by the study_data field, and the
3064    PCRE_EXTRA_STUDY_DATA flag set. This can be passed directly to pcre_exec() as
3065    before. That is, this change is entirely upwards-compatible and requires no
3066    change to existing code.
3067    
3068    If you want to pass in additional data to pcre_exec(), you can either place it
3069    in a pcre_extra block provided by pcre_study(), or create your own pcre_extra
3070    block.
3071    
3072    46. pcretest has been extended to test the PCRE_EXTRA_MATCH_LIMIT feature. If a
3073    data string contains the escape sequence \M, pcretest calls pcre_exec() several
3074    times with different match limits, until it finds the minimum value needed for
3075    pcre_exec() to complete. The value is then output. This can be instructive; for
3076    most simple matches the number is quite small, but for pathological cases it
3077    gets very large very quickly.
3078    
3079    47. There's a new option for pcre_fullinfo() called PCRE_INFO_STUDYSIZE. It
3080    returns the size of the data block pointed to by the study_data field in a
3081    pcre_extra block, that is, the value that was passed as the argument to
3082    pcre_malloc() when PCRE was getting memory in which to place the information
3083    created by pcre_study(). The fourth argument should point to a size_t variable.
3084    pcretest has been extended so that this information is shown after a successful
3085    pcre_study() call when information about the compiled regex is being displayed.
3086    
3087    48. Cosmetic change to Makefile: there's no need to have / after $(DESTDIR)
3088    because what follows is always an absolute path. (Later: it turns out that this
3089    is more than cosmetic for MinGW, because it doesn't like empty path
3090    components.)
3091    
3092    49. Some changes have been made to the callout feature (see 28 above):
3093    
3094    (i)  A callout function now has three choices for what it returns:
3095    
3096           0  =>  success, carry on matching
3097         > 0  =>  failure at this point, but backtrack if possible
3098         < 0  =>  serious error, return this value from pcre_exec()
3099    
3100         Negative values should normally be chosen from the set of PCRE_ERROR_xxx
3101         values. In particular, returning PCRE_ERROR_NOMATCH forces a standard
3102         "match failed" error. The error number PCRE_ERROR_CALLOUT is reserved for
3103         use by callout functions. It will never be used by PCRE itself.
3104    
3105    (ii) The pcre_extra structure (see 45 above) has a void * field called
3106         callout_data, with corresponding flag bit PCRE_EXTRA_CALLOUT_DATA. The
3107         pcre_callout_block structure has a field of the same name. The contents of
3108         the field passed in the pcre_extra structure are passed to the callout
3109         function in the corresponding field in the callout block. This makes it
3110         easier to use the same callout-containing regex from multiple threads. For
3111         testing, the pcretest program has a new data escape
3112    
3113           \C*n        pass the number n (may be negative) as callout_data
3114    
3115         If the callout function in pcretest receives a non-zero value as
3116         callout_data, it returns that value.
3117    
3118    50. Makefile wasn't handling CFLAGS properly when compiling dftables. Also,
3119    there were some redundant $(CFLAGS) in commands that are now specified as
3120    $(LINK), which already includes $(CFLAGS).
3121    
3122    51. Extensions to UTF-8 support are listed below. These all apply when (a) PCRE
3123    has been compiled with UTF-8 support *and* pcre_compile() has been compiled
3124    with the PCRE_UTF8 flag. Patterns that are compiled without that flag assume
3125    one-byte characters throughout. Note that case-insensitive matching applies
3126    only to characters whose values are less than 256. PCRE doesn't support the
3127    notion of cases for higher-valued characters.
3128    
3129    (i)   A character class whose characters are all within 0-255 is handled as
3130          a bit map, and the map is inverted for negative classes. Previously, a
3131          character > 255 always failed to match such a class; however it should
3132          match if the class was a negative one (e.g. [^ab]). This has been fixed.
3133    
3134    (ii)  A negated character class with a single character < 255 is coded as
3135          "not this character" (OP_NOT). This wasn't working properly when the test
3136          character was multibyte, either singly or repeated.
3137    
3138    (iii) Repeats of multibyte characters are now handled correctly in UTF-8
3139          mode, for example: \x{100}{2,3}.
3140    
3141    (iv)  The character escapes \b, \B, \d, \D, \s, \S, \w, and \W (either
3142          singly or repeated) now correctly test multibyte characters. However,
3143          PCRE doesn't recognize any characters with values greater than 255 as
3144          digits, spaces, or word characters. Such characters always match \D, \S,
3145          and \W, and never match \d, \s, or \w.
3146    
3147    (v)   Classes may now contain characters and character ranges with values
3148          greater than 255. For example: [ab\x{100}-\x{400}].
3149    
3150    (vi)  pcregrep now has a --utf-8 option (synonym -u) which makes it call
3151          PCRE in UTF-8 mode.
3152    
3153    52. The info request value PCRE_INFO_FIRSTCHAR has been renamed
3154    PCRE_INFO_FIRSTBYTE because it is a byte value. However, the old name is
3155    retained for backwards compatibility. (Note that LASTLITERAL is also a byte
3156    value.)
3157    
3158    53. The single man page has become too large. I have therefore split it up into
3159    a number of separate man pages. These also give rise to individual HTML pages;
3160    these are now put in a separate directory, and there is an index.html page that
3161    lists them all. Some hyperlinking between the pages has been installed.
3162    
3163    54. Added convenience functions for handling named capturing parentheses.
3164    
3165    55. Unknown escapes inside character classes (e.g. [\M]) and escapes that
3166    aren't interpreted therein (e.g. [\C]) are literals in Perl. This is now also
3167    true in PCRE, except when the PCRE_EXTENDED option is set, in which case they
3168    are faulted.
3169    
3170    56. Introduced HOST_CC and HOST_CFLAGS which can be set in the environment when
3171    calling configure. These values are used when compiling the dftables.c program
3172    which is run to generate the source of the default character tables. They
3173    default to the values of CC and CFLAGS. If you are cross-compiling PCRE,
3174    you will need to set these values.
3175    
3176    57. Updated the building process for Windows DLL, as provided by Fred Cox.
3177    
3178    
3179    Version 3.9 02-Jan-02
3180    ---------------------
3181    
3182    1. A bit of extraneous text had somehow crept into the pcregrep documentation.
3183    
3184    2. If --disable-static was given, the building process failed when trying to
3185    build pcretest and pcregrep. (For some reason it was using libtool to compile
3186    them, which is not right, as they aren't part of the library.)
3187    
3188    
3189  Version 3.8 18-Dec-01  Version 3.8 18-Dec-01
3190  ---------------------  ---------------------
3191    

Legend:
Removed from v.59  
changed lines
  Added in v.528

webmaster@exim.org
ViewVC Help
Powered by ViewVC 1.1.12