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

Diff of /code/trunk/ChangeLog

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

revision 15 by nigel, Sat Feb 24 21:38:25 2007 UTC revision 493 by ph10, Tue Mar 2 11:37:21 2010 UTC
# Line 1  Line 1 
1  ChangeLog for PCRE  ChangeLog for PCRE
2  ------------------  ------------------
3    
4    Version 8.02 01-Mar-2010
5    ------------------------
6    
7    1.  The Unicode data tables have been updated to Unicode 5.2.0.
8    
9    2.  Added the option --libs-cpp to pcre-config, but only when C++ support is
10        configured.
11    
12    3.  Updated the licensing terms in the pcregexp.pas file, as agreed with the
13        original author of that file, following a query about its status.
14    
15    
16    Version 8.01 19-Jan-2010
17    ------------------------
18    
19    1.  If a pattern contained a conditional subpattern with only one branch (in
20        particular, this includes all (*DEFINE) patterns), a call to pcre_study()
21        computed the wrong minimum data length (which is of course zero for such
22        subpatterns). This could cause incorrect "no match" results.
23    
24    2.  For patterns such as (?i)a(?-i)b|c where an option setting at the start of
25        the pattern is reset in the first branch, pcre_compile() failed with
26        "internal error: code overflow at offset...". This happened only when
27        the reset was to the original external option setting. (An optimization
28        abstracts leading options settings into an external setting, which was the
29        cause of this.)
30    
31    3.  A pattern such as ^(?!a(*SKIP)b) where a negative assertion contained one
32        of the verbs SKIP, PRUNE, or COMMIT, did not work correctly. When the
33        assertion pattern did not match (meaning that the assertion was true), it
34        was incorrectly treated as false if the SKIP had been reached during the
35        matching. This also applied to assertions used as conditions.
36    
37    4.  If an item that is not supported by pcre_dfa_exec() was encountered in an
38        assertion subpattern, including such a pattern used as a condition,
39        unpredictable results occurred, instead of the error return
40        PCRE_ERROR_DFA_UITEM.
41    
42    5.  The C++ GlobalReplace function was not working like Perl for the special
43        situation when an empty string is matched. It now does the fancy magic
44        stuff that is necessary.
45    
46    6.  In pcre_internal.h, obsolete includes to setjmp.h and stdarg.h have been
47        removed. (These were left over from very, very early versions of PCRE.)
48    
49    7.  Some cosmetic changes to the code to make life easier when compiling it
50        as part of something else:
51    
52        (a) Change DEBUG to PCRE_DEBUG.
53    
54        (b) In pcre_compile(), rename the member of the "branch_chain" structure
55            called "current" as "current_branch", to prevent a collision with the
56            Linux macro when compiled as a kernel module.
57    
58        (c) In pcre_study(), rename the function set_bit() as set_table_bit(), to
59            prevent a collision with the Linux macro when compiled as a kernel
60            module.
61    
62    8.  In pcre_compile() there are some checks for integer overflows that used to
63        cast potentially large values to (double). This has been changed to that
64        when building, a check for int64_t is made, and if it is found, it is used
65        instead, thus avoiding the use of floating point arithmetic. (There is no
66        other use of FP in PCRE.) If int64_t is not found, the fallback is to
67        double.
68    
69    9.  Added two casts to avoid signed/unsigned warnings from VS Studio Express
70        2005 (difference between two addresses compared to an unsigned value).
71    
72    10. Change the standard AC_CHECK_LIB test for libbz2 in configure.ac to a
73        custom one, because of the following reported problem in Windows:
74    
75          - libbz2 uses the Pascal calling convention (WINAPI) for the functions
76              under Win32.
77          - The standard autoconf AC_CHECK_LIB fails to include "bzlib.h",
78              therefore missing the function definition.
79          - The compiler thus generates a "C" signature for the test function.
80          - The linker fails to find the "C" function.
81          - PCRE fails to configure if asked to do so against libbz2.
82    
83    11. When running libtoolize from libtool-2.2.6b as part of autogen.sh, these
84        messages were output:
85    
86          Consider adding `AC_CONFIG_MACRO_DIR([m4])' to configure.ac and
87          rerunning libtoolize, to keep the correct libtool macros in-tree.
88          Consider adding `-I m4' to ACLOCAL_AMFLAGS in Makefile.am.
89    
90        I have done both of these things.
91    
92    12. Although pcre_dfa_exec() does not use nearly as much stack as pcre_exec()
93        most of the time, it *can* run out if it is given a pattern that contains a
94        runaway infinite recursion. I updated the discussion in the pcrestack man
95        page.
96    
97    13. Now that we have gone to the x.xx style of version numbers, the minor
98        version may start with zero. Using 08 or 09 is a bad idea because users
99        might check the value of PCRE_MINOR in their code, and 08 or 09 may be
100        interpreted as invalid octal numbers. I've updated the previous comment in
101        configure.ac, and also added a check that gives an error if 08 or 09 are
102        used.
103    
104    14. Change 8.00/11 was not quite complete: code had been accidentally omitted,
105        causing partial matching to fail when the end of the subject matched \W
106        in a UTF-8 pattern where \W was quantified with a minimum of 3.
107    
108    15. There were some discrepancies between the declarations in pcre_internal.h
109        of _pcre_is_newline(), _pcre_was_newline(), and _pcre_valid_utf8() and
110        their definitions. The declarations used "const uschar *" and the
111        definitions used USPTR. Even though USPTR is normally defined as "const
112        unsigned char *" (and uschar is typedeffed as "unsigned char"), it was
113        reported that: "This difference in casting confuses some C++ compilers, for
114        example, SunCC recognizes above declarations as different functions and
115        generates broken code for hbpcre." I have changed the declarations to use
116        USPTR.
117    
118    16. GNU libtool is named differently on some systems. The autogen.sh script now
119        tries several variants such as glibtoolize (MacOSX) and libtoolize1x
120        (FreeBSD).
121    
122    17. Applied Craig's patch that fixes an HP aCC compile error in pcre 8.00
123        (strtoXX undefined when compiling pcrecpp.cc). The patch contains this
124        comment: "Figure out how to create a longlong from a string: strtoll and
125        equivalent. It's not enough to call AC_CHECK_FUNCS: hpux has a strtoll, for
126        instance, but it only takes 2 args instead of 3!"
127    
128    18. A subtle bug concerned with back references has been fixed by a change of
129        specification, with a corresponding code fix. A pattern such as
130        ^(xa|=?\1a)+$ which contains a back reference inside the group to which it
131        refers, was giving matches when it shouldn't. For example, xa=xaaa would
132        match that pattern. Interestingly, Perl (at least up to 5.11.3) has the
133        same bug. Such groups have to be quantified to be useful, or contained
134        inside another quantified group. (If there's no repetition, the reference
135        can never match.) The problem arises because, having left the group and
136        moved on to the rest of the pattern, a later failure that backtracks into
137        the group uses the captured value from the final iteration of the group
138        rather than the correct earlier one. I have fixed this in PCRE by forcing
139        any group that contains a reference to itself to be an atomic group; that
140        is, there cannot be any backtracking into it once it has completed. This is
141        similar to recursive and subroutine calls.
142    
143    
144    Version 8.00 19-Oct-09
145    ----------------------
146    
147    1.  The table for translating pcre_compile() error codes into POSIX error codes
148        was out-of-date, and there was no check on the pcre_compile() error code
149        being within the table. This could lead to an OK return being given in
150        error.
151    
152    2.  Changed the call to open a subject file in pcregrep from fopen(pathname,
153        "r") to fopen(pathname, "rb"), which fixed a problem with some of the tests
154        in a Windows environment.
155    
156    3.  The pcregrep --count option prints the count for each file even when it is
157        zero, as does GNU grep. However, pcregrep was also printing all files when
158        --files-with-matches was added. Now, when both options are given, it prints
159        counts only for those files that have at least one match. (GNU grep just
160        prints the file name in this circumstance, but including the count seems
161        more useful - otherwise, why use --count?) Also ensured that the
162        combination -clh just lists non-zero counts, with no names.
163    
164    4.  The long form of the pcregrep -F option was incorrectly implemented as
165        --fixed_strings instead of --fixed-strings. This is an incompatible change,
166        but it seems right to fix it, and I didn't think it was worth preserving
167        the old behaviour.
168    
169    5.  The command line items --regex=pattern and --regexp=pattern were not
170        recognized by pcregrep, which required --regex pattern or --regexp pattern
171        (with a space rather than an '='). The man page documented the '=' forms,
172        which are compatible with GNU grep; these now work.
173    
174    6.  No libpcreposix.pc file was created for pkg-config; there was just
175        libpcre.pc and libpcrecpp.pc. The omission has been rectified.
176    
177    7.  Added #ifndef SUPPORT_UCP into the pcre_ucd.c module, to reduce its size
178        when UCP support is not needed, by modifying the Python script that
179        generates it from Unicode data files. This should not matter if the module
180        is correctly used as a library, but I received one complaint about 50K of
181        unwanted data. My guess is that the person linked everything into his
182        program rather than using a library. Anyway, it does no harm.
183    
184    8.  A pattern such as /\x{123}{2,2}+/8 was incorrectly compiled; the trigger
185        was a minimum greater than 1 for a wide character in a possessive
186        repetition. The same bug could also affect patterns like /(\x{ff}{0,2})*/8
187        which had an unlimited repeat of a nested, fixed maximum repeat of a wide
188        character. Chaos in the form of incorrect output or a compiling loop could
189        result.
190    
191    9.  The restrictions on what a pattern can contain when partial matching is
192        requested for pcre_exec() have been removed. All patterns can now be
193        partially matched by this function. In addition, if there are at least two
194        slots in the offset vector, the offset of the earliest inspected character
195        for the match and the offset of the end of the subject are set in them when
196        PCRE_ERROR_PARTIAL is returned.
197    
198    10. Partial matching has been split into two forms: PCRE_PARTIAL_SOFT, which is
199        synonymous with PCRE_PARTIAL, for backwards compatibility, and
200        PCRE_PARTIAL_HARD, which causes a partial match to supersede a full match,
201        and may be more useful for multi-segment matching.
202    
203    11. Partial matching with pcre_exec() is now more intuitive. A partial match
204        used to be given if ever the end of the subject was reached; now it is
205        given only if matching could not proceed because another character was
206        needed. This makes a difference in some odd cases such as Z(*FAIL) with the
207        string "Z", which now yields "no match" instead of "partial match". In the
208        case of pcre_dfa_exec(), "no match" is given if every matching path for the
209        final character ended with (*FAIL).
210    
211    12. Restarting a match using pcre_dfa_exec() after a partial match did not work
212        if the pattern had a "must contain" character that was already found in the
213        earlier partial match, unless partial matching was again requested. For
214        example, with the pattern /dog.(body)?/, the "must contain" character is
215        "g". If the first part-match was for the string "dog", restarting with
216        "sbody" failed. This bug has been fixed.
217    
218    13. The string returned by pcre_dfa_exec() after a partial match has been
219        changed so that it starts at the first inspected character rather than the
220        first character of the match. This makes a difference only if the pattern
221        starts with a lookbehind assertion or \b or \B (\K is not supported by
222        pcre_dfa_exec()). It's an incompatible change, but it makes the two
223        matching functions compatible, and I think it's the right thing to do.
224    
225    14. Added a pcredemo man page, created automatically from the pcredemo.c file,
226        so that the demonstration program is easily available in environments where
227        PCRE has not been installed from source.
228    
229    15. Arranged to add -DPCRE_STATIC to cflags in libpcre.pc, libpcreposix.cp,
230        libpcrecpp.pc and pcre-config when PCRE is not compiled as a shared
231        library.
232    
233    16. Added REG_UNGREEDY to the pcreposix interface, at the request of a user.
234        It maps to PCRE_UNGREEDY. It is not, of course, POSIX-compatible, but it
235        is not the first non-POSIX option to be added. Clearly some people find
236        these options useful.
237    
238    17. If a caller to the POSIX matching function regexec() passes a non-zero
239        value for nmatch with a NULL value for pmatch, the value of
240        nmatch is forced to zero.
241    
242    18. RunGrepTest did not have a test for the availability of the -u option of
243        the diff command, as RunTest does. It now checks in the same way as
244        RunTest, and also checks for the -b option.
245    
246    19. If an odd number of negated classes containing just a single character
247        interposed, within parentheses, between a forward reference to a named
248        subpattern and the definition of the subpattern, compilation crashed with
249        an internal error, complaining that it could not find the referenced
250        subpattern. An example of a crashing pattern is /(?&A)(([^m])(?<A>))/.
251        [The bug was that it was starting one character too far in when skipping
252        over the character class, thus treating the ] as data rather than
253        terminating the class. This meant it could skip too much.]
254    
255    20. Added PCRE_NOTEMPTY_ATSTART in order to be able to correctly implement the
256        /g option in pcretest when the pattern contains \K, which makes it possible
257        to have an empty string match not at the start, even when the pattern is
258        anchored. Updated pcretest and pcredemo to use this option.
259    
260    21. If the maximum number of capturing subpatterns in a recursion was greater
261        than the maximum at the outer level, the higher number was returned, but
262        with unset values at the outer level. The correct (outer level) value is
263        now given.
264    
265    22. If (*ACCEPT) appeared inside capturing parentheses, previous releases of
266        PCRE did not set those parentheses (unlike Perl). I have now found a way to
267        make it do so. The string so far is captured, making this feature
268        compatible with Perl.
269    
270    23. The tests have been re-organized, adding tests 11 and 12, to make it
271        possible to check the Perl 5.10 features against Perl 5.10.
272    
273    24. Perl 5.10 allows subroutine calls in lookbehinds, as long as the subroutine
274        pattern matches a fixed length string. PCRE did not allow this; now it
275        does. Neither allows recursion.
276    
277    25. I finally figured out how to implement a request to provide the minimum
278        length of subject string that was needed in order to match a given pattern.
279        (It was back references and recursion that I had previously got hung up
280        on.) This code has now been added to pcre_study(); it finds a lower bound
281        to the length of subject needed. It is not necessarily the greatest lower
282        bound, but using it to avoid searching strings that are too short does give
283        some useful speed-ups. The value is available to calling programs via
284        pcre_fullinfo().
285    
286    26. While implementing 25, I discovered to my embarrassment that pcretest had
287        not been passing the result of pcre_study() to pcre_dfa_exec(), so the
288        study optimizations had never been tested with that matching function.
289        Oops. What is worse, even when it was passed study data, there was a bug in
290        pcre_dfa_exec() that meant it never actually used it. Double oops. There
291        were also very few tests of studied patterns with pcre_dfa_exec().
292    
293    27. If (?| is used to create subpatterns with duplicate numbers, they are now
294        allowed to have the same name, even if PCRE_DUPNAMES is not set. However,
295        on the other side of the coin, they are no longer allowed to have different
296        names, because these cannot be distinguished in PCRE, and this has caused
297        confusion. (This is a difference from Perl.)
298    
299    28. When duplicate subpattern names are present (necessarily with different
300        numbers, as required by 27 above), and a test is made by name in a
301        conditional pattern, either for a subpattern having been matched, or for
302        recursion in such a pattern, all the associated numbered subpatterns are
303        tested, and the overall condition is true if the condition is true for any
304        one of them. This is the way Perl works, and is also more like the way
305        testing by number works.
306    
307    
308    Version 7.9 11-Apr-09
309    ---------------------
310    
311    1.  When building with support for bzlib/zlib (pcregrep) and/or readline
312        (pcretest), all targets were linked against these libraries. This included
313        libpcre, libpcreposix, and libpcrecpp, even though they do not use these
314        libraries. This caused unwanted dependencies to be created. This problem
315        has been fixed, and now only pcregrep is linked with bzlib/zlib and only
316        pcretest is linked with readline.
317    
318    2.  The "typedef int BOOL" in pcre_internal.h that was included inside the
319        "#ifndef FALSE" condition by an earlier change (probably 7.8/18) has been
320        moved outside it again, because FALSE and TRUE are already defined in AIX,
321        but BOOL is not.
322    
323    3.  The pcre_config() function was treating the PCRE_MATCH_LIMIT and
324        PCRE_MATCH_LIMIT_RECURSION values as ints, when they should be long ints.
325    
326    4.  The pcregrep documentation said spaces were inserted as well as colons (or
327        hyphens) following file names and line numbers when outputting matching
328        lines. This is not true; no spaces are inserted. I have also clarified the
329        wording for the --colour (or --color) option.
330    
331    5.  In pcregrep, when --colour was used with -o, the list of matching strings
332        was not coloured; this is different to GNU grep, so I have changed it to be
333        the same.
334    
335    6.  When --colo(u)r was used in pcregrep, only the first matching substring in
336        each matching line was coloured. Now it goes on to look for further matches
337        of any of the test patterns, which is the same behaviour as GNU grep.
338    
339    7.  A pattern that could match an empty string could cause pcregrep to loop; it
340        doesn't make sense to accept an empty string match in pcregrep, so I have
341        locked it out (using PCRE's PCRE_NOTEMPTY option). By experiment, this
342        seems to be how GNU grep behaves.
343    
344    8.  The pattern (?(?=.*b)b|^) was incorrectly compiled as "match must be at
345        start or after a newline", because the conditional assertion was not being
346        correctly handled. The rule now is that both the assertion and what follows
347        in the first alternative must satisfy the test.
348    
349    9.  If auto-callout was enabled in a pattern with a conditional group whose
350        condition was an assertion, PCRE could crash during matching, both with
351        pcre_exec() and pcre_dfa_exec().
352    
353    10. The PCRE_DOLLAR_ENDONLY option was not working when pcre_dfa_exec() was
354        used for matching.
355    
356    11. Unicode property support in character classes was not working for
357        characters (bytes) greater than 127 when not in UTF-8 mode.
358    
359    12. Added the -M command line option to pcretest.
360    
361    14. Added the non-standard REG_NOTEMPTY option to the POSIX interface.
362    
363    15. Added the PCRE_NO_START_OPTIMIZE match-time option.
364    
365    16. Added comments and documentation about mis-use of no_arg in the C++
366        wrapper.
367    
368    17. Implemented support for UTF-8 encoding in EBCDIC environments, a patch
369        from Martin Jerabek that uses macro names for all relevant character and
370        string constants.
371    
372    18. Added to pcre_internal.h two configuration checks: (a) If both EBCDIC and
373        SUPPORT_UTF8 are set, give an error; (b) If SUPPORT_UCP is set without
374        SUPPORT_UTF8, define SUPPORT_UTF8. The "configure" script handles both of
375        these, but not everybody uses configure.
376    
377    19. A conditional group that had only one branch was not being correctly
378        recognized as an item that could match an empty string. This meant that an
379        enclosing group might also not be so recognized, causing infinite looping
380        (and probably a segfault) for patterns such as ^"((?(?=[a])[^"])|b)*"$
381        with the subject "ab", where knowledge that the repeated group can match
382        nothing is needed in order to break the loop.
383    
384    20. If a pattern that was compiled with callouts was matched using pcre_dfa_
385        exec(), but without supplying a callout function, matching went wrong.
386    
387    21. If PCRE_ERROR_MATCHLIMIT occurred during a recursion, there was a memory
388        leak if the size of the offset vector was greater than 30. When the vector
389        is smaller, the saved offsets during recursion go onto a local stack
390        vector, but for larger vectors malloc() is used. It was failing to free
391        when the recursion yielded PCRE_ERROR_MATCH_LIMIT (or any other "abnormal"
392        error, in fact).
393    
394    22. There was a missing #ifdef SUPPORT_UTF8 round one of the variables in the
395        heapframe that is used only when UTF-8 support is enabled. This caused no
396        problem, but was untidy.
397    
398    23. Steven Van Ingelgem's patch to CMakeLists.txt to change the name
399        CMAKE_BINARY_DIR to PROJECT_BINARY_DIR so that it works when PCRE is
400        included within another project.
401    
402    24. Steven Van Ingelgem's patches to add more options to the CMake support,
403        slightly modified by me:
404    
405          (a) PCRE_BUILD_TESTS can be set OFF not to build the tests, including
406              not building pcregrep.
407    
408          (b) PCRE_BUILD_PCREGREP can be see OFF not to build pcregrep, but only
409              if PCRE_BUILD_TESTS is also set OFF, because the tests use pcregrep.
410    
411    25. Forward references, both numeric and by name, in patterns that made use of
412        duplicate group numbers, could behave incorrectly or give incorrect errors,
413        because when scanning forward to find the reference group, PCRE was not
414        taking into account the duplicate group numbers. A pattern such as
415        ^X(?3)(a)(?|(b)|(q))(Y) is an example.
416    
417    26. Changed a few more instances of "const unsigned char *" to USPTR, making
418        the feature of a custom pointer more persuasive (as requested by a user).
419    
420    27. Wrapped the definitions of fileno and isatty for Windows, which appear in
421        pcretest.c, inside #ifndefs, because it seems they are sometimes already
422        pre-defined.
423    
424    28. Added support for (*UTF8) at the start of a pattern.
425    
426    29. Arrange for flags added by the "release type" setting in CMake to be shown
427        in the configuration summary.
428    
429    
430    Version 7.8 05-Sep-08
431    ---------------------
432    
433    1.  Replaced UCP searching code with optimized version as implemented for Ad
434        Muncher (http://www.admuncher.com/) by Peter Kankowski. This uses a two-
435        stage table and inline lookup instead of a function, giving speed ups of 2
436        to 5 times on some simple patterns that I tested. Permission was given to
437        distribute the MultiStage2.py script that generates the tables (it's not in
438        the tarball, but is in the Subversion repository).
439    
440    2.  Updated the Unicode datatables to Unicode 5.1.0. This adds yet more
441        scripts.
442    
443    3.  Change 12 for 7.7 introduced a bug in pcre_study() when a pattern contained
444        a group with a zero qualifier. The result of the study could be incorrect,
445        or the function might crash, depending on the pattern.
446    
447    4.  Caseless matching was not working for non-ASCII characters in back
448        references. For example, /(\x{de})\1/8i was not matching \x{de}\x{fe}.
449        It now works when Unicode Property Support is available.
450    
451    5.  In pcretest, an escape such as \x{de} in the data was always generating
452        a UTF-8 string, even in non-UTF-8 mode. Now it generates a single byte in
453        non-UTF-8 mode. If the value is greater than 255, it gives a warning about
454        truncation.
455    
456    6.  Minor bugfix in pcrecpp.cc (change "" == ... to NULL == ...).
457    
458    7.  Added two (int) casts to pcregrep when printing the difference of two
459        pointers, in case they are 64-bit values.
460    
461    8.  Added comments about Mac OS X stack usage to the pcrestack man page and to
462        test 2 if it fails.
463    
464    9.  Added PCRE_CALL_CONVENTION just before the names of all exported functions,
465        and a #define of that name to empty if it is not externally set. This is to
466        allow users of MSVC to set it if necessary.
467    
468    10. The PCRE_EXP_DEFN macro which precedes exported functions was missing from
469        the convenience functions in the pcre_get.c source file.
470    
471    11. An option change at the start of a pattern that had top-level alternatives
472        could cause overwriting and/or a crash. This command provoked a crash in
473        some environments:
474    
475          printf "/(?i)[\xc3\xa9\xc3\xbd]|[\xc3\xa9\xc3\xbdA]/8\n" | pcretest
476    
477        This potential security problem was recorded as CVE-2008-2371.
478    
479    12. For a pattern where the match had to start at the beginning or immediately
480        after a newline (e.g /.*anything/ without the DOTALL flag), pcre_exec() and
481        pcre_dfa_exec() could read past the end of the passed subject if there was
482        no match. To help with detecting such bugs (e.g. with valgrind), I modified
483        pcretest so that it places the subject at the end of its malloc-ed buffer.
484    
485    13. The change to pcretest in 12 above threw up a couple more cases when pcre_
486        exec() might read past the end of the data buffer in UTF-8 mode.
487    
488    14. A similar bug to 7.3/2 existed when the PCRE_FIRSTLINE option was set and
489        the data contained the byte 0x85 as part of a UTF-8 character within its
490        first line. This applied both to normal and DFA matching.
491    
492    15. Lazy qualifiers were not working in some cases in UTF-8 mode. For example,
493        /^[^d]*?$/8 failed to match "abc".
494    
495    16. Added a missing copyright notice to pcrecpp_internal.h.
496    
497    17. Make it more clear in the documentation that values returned from
498        pcre_exec() in ovector are byte offsets, not character counts.
499    
500    18. Tidied a few places to stop certain compilers from issuing warnings.
501    
502    19. Updated the Virtual Pascal + BCC files to compile the latest v7.7, as
503        supplied by Stefan Weber. I made a further small update for 7.8 because
504        there is a change of source arrangements: the pcre_searchfuncs.c module is
505        replaced by pcre_ucd.c.
506    
507    
508    Version 7.7 07-May-08
509    ---------------------
510    
511    1.  Applied Craig's patch to sort out a long long problem: "If we can't convert
512        a string to a long long, pretend we don't even have a long long." This is
513        done by checking for the strtoq, strtoll, and _strtoi64 functions.
514    
515    2.  Applied Craig's patch to pcrecpp.cc to restore ABI compatibility with
516        pre-7.6 versions, which defined a global no_arg variable instead of putting
517        it in the RE class. (See also #8 below.)
518    
519    3.  Remove a line of dead code, identified by coverity and reported by Nuno
520        Lopes.
521    
522    4.  Fixed two related pcregrep bugs involving -r with --include or --exclude:
523    
524        (1) The include/exclude patterns were being applied to the whole pathnames
525            of files, instead of just to the final components.
526    
527        (2) If there was more than one level of directory, the subdirectories were
528            skipped unless they satisfied the include/exclude conditions. This is
529            inconsistent with GNU grep (and could even be seen as contrary to the
530            pcregrep specification - which I improved to make it absolutely clear).
531            The action now is always to scan all levels of directory, and just
532            apply the include/exclude patterns to regular files.
533    
534    5.  Added the --include_dir and --exclude_dir patterns to pcregrep, and used
535        --exclude_dir in the tests to avoid scanning .svn directories.
536    
537    6.  Applied Craig's patch to the QuoteMeta function so that it escapes the
538        NUL character as backslash + 0 rather than backslash + NUL, because PCRE
539        doesn't support NULs in patterns.
540    
541    7.  Added some missing "const"s to declarations of static tables in
542        pcre_compile.c and pcre_dfa_exec.c.
543    
544    8.  Applied Craig's patch to pcrecpp.cc to fix a problem in OS X that was
545        caused by fix #2  above. (Subsequently also a second patch to fix the
546        first patch. And a third patch - this was a messy problem.)
547    
548    9.  Applied Craig's patch to remove the use of push_back().
549    
550    10. Applied Alan Lehotsky's patch to add REG_STARTEND support to the POSIX
551        matching function regexec().
552    
553    11. Added support for the Oniguruma syntax \g<name>, \g<n>, \g'name', \g'n',
554        which, however, unlike Perl's \g{...}, are subroutine calls, not back
555        references. PCRE supports relative numbers with this syntax (I don't think
556        Oniguruma does).
557    
558    12. Previously, a group with a zero repeat such as (...){0} was completely
559        omitted from the compiled regex. However, this means that if the group
560        was called as a subroutine from elsewhere in the pattern, things went wrong
561        (an internal error was given). Such groups are now left in the compiled
562        pattern, with a new opcode that causes them to be skipped at execution
563        time.
564    
565    13. Added the PCRE_JAVASCRIPT_COMPAT option. This makes the following changes
566        to the way PCRE behaves:
567    
568        (a) A lone ] character is dis-allowed (Perl treats it as data).
569    
570        (b) A back reference to an unmatched subpattern matches an empty string
571            (Perl fails the current match path).
572    
573        (c) A data ] in a character class must be notated as \] because if the
574            first data character in a class is ], it defines an empty class. (In
575            Perl it is not possible to have an empty class.) The empty class []
576            never matches; it forces failure and is equivalent to (*FAIL) or (?!).
577            The negative empty class [^] matches any one character, independently
578            of the DOTALL setting.
579    
580    14. A pattern such as /(?2)[]a()b](abc)/ which had a forward reference to a
581        non-existent subpattern following a character class starting with ']' and
582        containing () gave an internal compiling error instead of "reference to
583        non-existent subpattern". Fortunately, when the pattern did exist, the
584        compiled code was correct. (When scanning forwards to check for the
585        existencd of the subpattern, it was treating the data ']' as terminating
586        the class, so got the count wrong. When actually compiling, the reference
587        was subsequently set up correctly.)
588    
589    15. The "always fail" assertion (?!) is optimzed to (*FAIL) by pcre_compile;
590        it was being rejected as not supported by pcre_dfa_exec(), even though
591        other assertions are supported. I have made pcre_dfa_exec() support
592        (*FAIL).
593    
594    16. The implementation of 13c above involved the invention of a new opcode,
595        OP_ALLANY, which is like OP_ANY but doesn't check the /s flag. Since /s
596        cannot be changed at match time, I realized I could make a small
597        improvement to matching performance by compiling OP_ALLANY instead of
598        OP_ANY for "." when DOTALL was set, and then removing the runtime tests
599        on the OP_ANY path.
600    
601    17. Compiling pcretest on Windows with readline support failed without the
602        following two fixes: (1) Make the unistd.h include conditional on
603        HAVE_UNISTD_H; (2) #define isatty and fileno as _isatty and _fileno.
604    
605    18. Changed CMakeLists.txt and cmake/FindReadline.cmake to arrange for the
606        ncurses library to be included for pcretest when ReadLine support is
607        requested, but also to allow for it to be overridden. This patch came from
608        Daniel Bergström.
609    
610    19. There was a typo in the file ucpinternal.h where f0_rangeflag was defined
611        as 0x00f00000 instead of 0x00800000. Luckily, this would not have caused
612        any errors with the current Unicode tables. Thanks to Peter Kankowski for
613        spotting this.
614    
615    
616    Version 7.6 28-Jan-08
617    ---------------------
618    
619    1.  A character class containing a very large number of characters with
620        codepoints greater than 255 (in UTF-8 mode, of course) caused a buffer
621        overflow.
622    
623    2.  Patch to cut out the "long long" test in pcrecpp_unittest when
624        HAVE_LONG_LONG is not defined.
625    
626    3.  Applied Christian Ehrlicher's patch to update the CMake build files to
627        bring them up to date and include new features. This patch includes:
628    
629        - Fixed PH's badly added libz and libbz2 support.
630        - Fixed a problem with static linking.
631        - Added pcredemo. [But later removed - see 7 below.]
632        - Fixed dftables problem and added an option.
633        - Added a number of HAVE_XXX tests, including HAVE_WINDOWS_H and
634            HAVE_LONG_LONG.
635        - Added readline support for pcretest.
636        - Added an listing of the option settings after cmake has run.
637    
638    4.  A user submitted a patch to Makefile that makes it easy to create
639        "pcre.dll" under mingw when using Configure/Make. I added stuff to
640        Makefile.am that cause it to include this special target, without
641        affecting anything else. Note that the same mingw target plus all
642        the other distribution libraries and programs are now supported
643        when configuring with CMake (see 6 below) instead of with
644        Configure/Make.
645    
646    5.  Applied Craig's patch that moves no_arg into the RE class in the C++ code.
647        This is an attempt to solve the reported problem "pcrecpp::no_arg is not
648        exported in the Windows port". It has not yet been confirmed that the patch
649        solves the problem, but it does no harm.
650    
651    6.  Applied Sheri's patch to CMakeLists.txt to add NON_STANDARD_LIB_PREFIX and
652        NON_STANDARD_LIB_SUFFIX for dll names built with mingw when configured
653        with CMake, and also correct the comment about stack recursion.
654    
655    7.  Remove the automatic building of pcredemo from the ./configure system and
656        from CMakeLists.txt. The whole idea of pcredemo.c is that it is an example
657        of a program that users should build themselves after PCRE is installed, so
658        building it automatically is not really right. What is more, it gave
659        trouble in some build environments.
660    
661    8.  Further tidies to CMakeLists.txt from Sheri and Christian.
662    
663    
664    Version 7.5 10-Jan-08
665    ---------------------
666    
667    1.  Applied a patch from Craig: "This patch makes it possible to 'ignore'
668        values in parens when parsing an RE using the C++ wrapper."
669    
670    2.  Negative specials like \S did not work in character classes in UTF-8 mode.
671        Characters greater than 255 were excluded from the class instead of being
672        included.
673    
674    3.  The same bug as (2) above applied to negated POSIX classes such as
675        [:^space:].
676    
677    4.  PCRECPP_STATIC was referenced in pcrecpp_internal.h, but nowhere was it
678        defined or documented. It seems to have been a typo for PCRE_STATIC, so
679        I have changed it.
680    
681    5.  The construct (?&) was not diagnosed as a syntax error (it referenced the
682        first named subpattern) and a construct such as (?&a) would reference the
683        first named subpattern whose name started with "a" (in other words, the
684        length check was missing). Both these problems are fixed. "Subpattern name
685        expected" is now given for (?&) (a zero-length name), and this patch also
686        makes it give the same error for \k'' (previously it complained that that
687        was a reference to a non-existent subpattern).
688    
689    6.  The erroneous patterns (?+-a) and (?-+a) give different error messages;
690        this is right because (?- can be followed by option settings as well as by
691        digits. I have, however, made the messages clearer.
692    
693    7.  Patterns such as (?(1)a|b) (a pattern that contains fewer subpatterns
694        than the number used in the conditional) now cause a compile-time error.
695        This is actually not compatible with Perl, which accepts such patterns, but
696        treats the conditional as always being FALSE (as PCRE used to), but it
697        seems to me that giving a diagnostic is better.
698    
699    8.  Change "alphameric" to the more common word "alphanumeric" in comments
700        and messages.
701    
702    9.  Fix two occurrences of "backslash" in comments that should have been
703        "backspace".
704    
705    10. Remove two redundant lines of code that can never be obeyed (their function
706        was moved elsewhere).
707    
708    11. The program that makes PCRE's Unicode character property table had a bug
709        which caused it to generate incorrect table entries for sequences of
710        characters that have the same character type, but are in different scripts.
711        It amalgamated them into a single range, with the script of the first of
712        them. In other words, some characters were in the wrong script. There were
713        thirteen such cases, affecting characters in the following ranges:
714    
715          U+002b0 - U+002c1
716          U+0060c - U+0060d
717          U+0061e - U+00612
718          U+0064b - U+0065e
719          U+0074d - U+0076d
720          U+01800 - U+01805
721          U+01d00 - U+01d77
722          U+01d9b - U+01dbf
723          U+0200b - U+0200f
724          U+030fc - U+030fe
725          U+03260 - U+0327f
726          U+0fb46 - U+0fbb1
727          U+10450 - U+1049d
728    
729    12. The -o option (show only the matching part of a line) for pcregrep was not
730        compatible with GNU grep in that, if there was more than one match in a
731        line, it showed only the first of them. It now behaves in the same way as
732        GNU grep.
733    
734    13. If the -o and -v options were combined for pcregrep, it printed a blank
735        line for every non-matching line. GNU grep prints nothing, and pcregrep now
736        does the same. The return code can be used to tell if there were any
737        non-matching lines.
738    
739    14. Added --file-offsets and --line-offsets to pcregrep.
740    
741    15. The pattern (?=something)(?R) was not being diagnosed as a potentially
742        infinitely looping recursion. The bug was that positive lookaheads were not
743        being skipped when checking for a possible empty match (negative lookaheads
744        and both kinds of lookbehind were skipped).
745    
746    16. Fixed two typos in the Windows-only code in pcregrep.c, and moved the
747        inclusion of <windows.h> to before rather than after the definition of
748        INVALID_FILE_ATTRIBUTES (patch from David Byron).
749    
750    17. Specifying a possessive quantifier with a specific limit for a Unicode
751        character property caused pcre_compile() to compile bad code, which led at
752        runtime to PCRE_ERROR_INTERNAL (-14). Examples of patterns that caused this
753        are: /\p{Zl}{2,3}+/8 and /\p{Cc}{2}+/8. It was the possessive "+" that
754        caused the error; without that there was no problem.
755    
756    18. Added --enable-pcregrep-libz and --enable-pcregrep-libbz2.
757    
758    19. Added --enable-pcretest-libreadline.
759    
760    20. In pcrecpp.cc, the variable 'count' was incremented twice in
761        RE::GlobalReplace(). As a result, the number of replacements returned was
762        double what it should be. I removed one of the increments, but Craig sent a
763        later patch that removed the other one (the right fix) and added unit tests
764        that check the return values (which was not done before).
765    
766    21. Several CMake things:
767    
768        (1) Arranged that, when cmake is used on Unix, the libraries end up with
769            the names libpcre and libpcreposix, not just pcre and pcreposix.
770    
771        (2) The above change means that pcretest and pcregrep are now correctly
772            linked with the newly-built libraries, not previously installed ones.
773    
774        (3) Added PCRE_SUPPORT_LIBREADLINE, PCRE_SUPPORT_LIBZ, PCRE_SUPPORT_LIBBZ2.
775    
776    22. In UTF-8 mode, with newline set to "any", a pattern such as .*a.*=.b.*
777        crashed when matching a string such as a\x{2029}b (note that \x{2029} is a
778        UTF-8 newline character). The key issue is that the pattern starts .*;
779        this means that the match must be either at the beginning, or after a
780        newline. The bug was in the code for advancing after a failed match and
781        checking that the new position followed a newline. It was not taking
782        account of UTF-8 characters correctly.
783    
784    23. PCRE was behaving differently from Perl in the way it recognized POSIX
785        character classes. PCRE was not treating the sequence [:...:] as a
786        character class unless the ... were all letters. Perl, however, seems to
787        allow any characters between [: and :], though of course it rejects as
788        unknown any "names" that contain non-letters, because all the known class
789        names consist only of letters. Thus, Perl gives an error for [[:1234:]],
790        for example, whereas PCRE did not - it did not recognize a POSIX character
791        class. This seemed a bit dangerous, so the code has been changed to be
792        closer to Perl. The behaviour is not identical to Perl, because PCRE will
793        diagnose an unknown class for, for example, [[:l\ower:]] where Perl will
794        treat it as [[:lower:]]. However, PCRE does now give "unknown" errors where
795        Perl does, and where it didn't before.
796    
797    24. Rewrite so as to remove the single use of %n from pcregrep because in some
798        Windows environments %n is disabled by default.
799    
800    
801    Version 7.4 21-Sep-07
802    ---------------------
803    
804    1.  Change 7.3/28 was implemented for classes by looking at the bitmap. This
805        means that a class such as [\s] counted as "explicit reference to CR or
806        LF". That isn't really right - the whole point of the change was to try to
807        help when there was an actual mention of one of the two characters. So now
808        the change happens only if \r or \n (or a literal CR or LF) character is
809        encountered.
810    
811    2.  The 32-bit options word was also used for 6 internal flags, but the numbers
812        of both had grown to the point where there were only 3 bits left.
813        Fortunately, there was spare space in the data structure, and so I have
814        moved the internal flags into a new 16-bit field to free up more option
815        bits.
816    
817    3.  The appearance of (?J) at the start of a pattern set the DUPNAMES option,
818        but did not set the internal JCHANGED flag - either of these is enough to
819        control the way the "get" function works - but the PCRE_INFO_JCHANGED
820        facility is supposed to tell if (?J) was ever used, so now (?J) at the
821        start sets both bits.
822    
823    4.  Added options (at build time, compile time, exec time) to change \R from
824        matching any Unicode line ending sequence to just matching CR, LF, or CRLF.
825    
826    5.  doc/pcresyntax.html was missing from the distribution.
827    
828    6.  Put back the definition of PCRE_ERROR_NULLWSLIMIT, for backward
829        compatibility, even though it is no longer used.
830    
831    7.  Added macro for snprintf to pcrecpp_unittest.cc and also for strtoll and
832        strtoull to pcrecpp.cc to select the available functions in WIN32 when the
833        windows.h file is present (where different names are used). [This was
834        reversed later after testing - see 16 below.]
835    
836    8.  Changed all #include <config.h> to #include "config.h". There were also
837        some further <pcre.h> cases that I changed to "pcre.h".
838    
839    9.  When pcregrep was used with the --colour option, it missed the line ending
840        sequence off the lines that it output.
841    
842    10. It was pointed out to me that arrays of string pointers cause lots of
843        relocations when a shared library is dynamically loaded. A technique of
844        using a single long string with a table of offsets can drastically reduce
845        these. I have refactored PCRE in four places to do this. The result is
846        dramatic:
847    
848          Originally:                          290
849          After changing UCP table:            187
850          After changing error message table:   43
851          After changing table of "verbs"       36
852          After changing table of Posix names   22
853    
854        Thanks to the folks working on Gregex for glib for this insight.
855    
856    11. --disable-stack-for-recursion caused compiling to fail unless -enable-
857        unicode-properties was also set.
858    
859    12. Updated the tests so that they work when \R is defaulted to ANYCRLF.
860    
861    13. Added checks for ANY and ANYCRLF to pcrecpp.cc where it previously
862        checked only for CRLF.
863    
864    14. Added casts to pcretest.c to avoid compiler warnings.
865    
866    15. Added Craig's patch to various pcrecpp modules to avoid compiler warnings.
867    
868    16. Added Craig's patch to remove the WINDOWS_H tests, that were not working,
869        and instead check for _strtoi64 explicitly, and avoid the use of snprintf()
870        entirely. This removes changes made in 7 above.
871    
872    17. The CMake files have been updated, and there is now more information about
873        building with CMake in the NON-UNIX-USE document.
874    
875    
876    Version 7.3 28-Aug-07
877    ---------------------
878    
879     1. In the rejigging of the build system that eventually resulted in 7.1, the
880        line "#include <pcre.h>" was included in pcre_internal.h. The use of angle
881        brackets there is not right, since it causes compilers to look for an
882        installed pcre.h, not the version that is in the source that is being
883        compiled (which of course may be different). I have changed it back to:
884    
885          #include "pcre.h"
886    
887        I have a vague recollection that the change was concerned with compiling in
888        different directories, but in the new build system, that is taken care of
889        by the VPATH setting the Makefile.
890    
891     2. The pattern .*$ when run in not-DOTALL UTF-8 mode with newline=any failed
892        when the subject happened to end in the byte 0x85 (e.g. if the last
893        character was \x{1ec5}). *Character* 0x85 is one of the "any" newline
894        characters but of course it shouldn't be taken as a newline when it is part
895        of another character. The bug was that, for an unlimited repeat of . in
896        not-DOTALL UTF-8 mode, PCRE was advancing by bytes rather than by
897        characters when looking for a newline.
898    
899     3. A small performance improvement in the DOTALL UTF-8 mode .* case.
900    
901     4. Debugging: adjusted the names of opcodes for different kinds of parentheses
902        in debug output.
903    
904     5. Arrange to use "%I64d" instead of "%lld" and "%I64u" instead of "%llu" for
905        long printing in the pcrecpp unittest when running under MinGW.
906    
907     6. ESC_K was left out of the EBCDIC table.
908    
909     7. Change 7.0/38 introduced a new limit on the number of nested non-capturing
910        parentheses; I made it 1000, which seemed large enough. Unfortunately, the
911        limit also applies to "virtual nesting" when a pattern is recursive, and in
912        this case 1000 isn't so big. I have been able to remove this limit at the
913        expense of backing off one optimization in certain circumstances. Normally,
914        when pcre_exec() would call its internal match() function recursively and
915        immediately return the result unconditionally, it uses a "tail recursion"
916        feature to save stack. However, when a subpattern that can match an empty
917        string has an unlimited repetition quantifier, it no longer makes this
918        optimization. That gives it a stack frame in which to save the data for
919        checking that an empty string has been matched. Previously this was taken
920        from the 1000-entry workspace that had been reserved. So now there is no
921        explicit limit, but more stack is used.
922    
923     8. Applied Daniel's patches to solve problems with the import/export magic
924        syntax that is required for Windows, and which was going wrong for the
925        pcreposix and pcrecpp parts of the library. These were overlooked when this
926        problem was solved for the main library.
927    
928     9. There were some crude static tests to avoid integer overflow when computing
929        the size of patterns that contain repeated groups with explicit upper
930        limits. As the maximum quantifier is 65535, the maximum group length was
931        set at 30,000 so that the product of these two numbers did not overflow a
932        32-bit integer. However, it turns out that people want to use groups that
933        are longer than 30,000 bytes (though not repeat them that many times).
934        Change 7.0/17 (the refactoring of the way the pattern size is computed) has
935        made it possible to implement the integer overflow checks in a much more
936        dynamic way, which I have now done. The artificial limitation on group
937        length has been removed - we now have only the limit on the total length of
938        the compiled pattern, which depends on the LINK_SIZE setting.
939    
940    10. Fixed a bug in the documentation for get/copy named substring when
941        duplicate names are permitted. If none of the named substrings are set, the
942        functions return PCRE_ERROR_NOSUBSTRING (7); the doc said they returned an
943        empty string.
944    
945    11. Because Perl interprets \Q...\E at a high level, and ignores orphan \E
946        instances, patterns such as [\Q\E] or [\E] or even [^\E] cause an error,
947        because the ] is interpreted as the first data character and the
948        terminating ] is not found. PCRE has been made compatible with Perl in this
949        regard. Previously, it interpreted [\Q\E] as an empty class, and [\E] could
950        cause memory overwriting.
951    
952    10. Like Perl, PCRE automatically breaks an unlimited repeat after an empty
953        string has been matched (to stop an infinite loop). It was not recognizing
954        a conditional subpattern that could match an empty string if that
955        subpattern was within another subpattern. For example, it looped when
956        trying to match  (((?(1)X|))*)  but it was OK with  ((?(1)X|)*)  where the
957        condition was not nested. This bug has been fixed.
958    
959    12. A pattern like \X?\d or \P{L}?\d in non-UTF-8 mode could cause a backtrack
960        past the start of the subject in the presence of bytes with the top bit
961        set, for example "\x8aBCD".
962    
963    13. Added Perl 5.10 experimental backtracking controls (*FAIL), (*F), (*PRUNE),
964        (*SKIP), (*THEN), (*COMMIT), and (*ACCEPT).
965    
966    14. Optimized (?!) to (*FAIL).
967    
968    15. Updated the test for a valid UTF-8 string to conform to the later RFC 3629.
969        This restricts code points to be within the range 0 to 0x10FFFF, excluding
970        the "low surrogate" sequence 0xD800 to 0xDFFF. Previously, PCRE allowed the
971        full range 0 to 0x7FFFFFFF, as defined by RFC 2279. Internally, it still
972        does: it's just the validity check that is more restrictive.
973    
974    16. Inserted checks for integer overflows during escape sequence (backslash)
975        processing, and also fixed erroneous offset values for syntax errors during
976        backslash processing.
977    
978    17. Fixed another case of looking too far back in non-UTF-8 mode (cf 12 above)
979        for patterns like [\PPP\x8a]{1,}\x80 with the subject "A\x80".
980    
981    18. An unterminated class in a pattern like (?1)\c[ with a "forward reference"
982        caused an overrun.
983    
984    19. A pattern like (?:[\PPa*]*){8,} which had an "extended class" (one with
985        something other than just ASCII characters) inside a group that had an
986        unlimited repeat caused a loop at compile time (while checking to see
987        whether the group could match an empty string).
988    
989    20. Debugging a pattern containing \p or \P could cause a crash. For example,
990        [\P{Any}] did so. (Error in the code for printing property names.)
991    
992    21. An orphan \E inside a character class could cause a crash.
993    
994    22. A repeated capturing bracket such as (A)? could cause a wild memory
995        reference during compilation.
996    
997    23. There are several functions in pcre_compile() that scan along a compiled
998        expression for various reasons (e.g. to see if it's fixed length for look
999        behind). There were bugs in these functions when a repeated \p or \P was
1000        present in the pattern. These operators have additional parameters compared
1001        with \d, etc, and these were not being taken into account when moving along
1002        the compiled data. Specifically:
1003    
1004        (a) A item such as \p{Yi}{3} in a lookbehind was not treated as fixed
1005            length.
1006    
1007        (b) An item such as \pL+ within a repeated group could cause crashes or
1008            loops.
1009    
1010        (c) A pattern such as \p{Yi}+(\P{Yi}+)(?1) could give an incorrect
1011            "reference to non-existent subpattern" error.
1012    
1013        (d) A pattern like (\P{Yi}{2}\277)? could loop at compile time.
1014    
1015    24. A repeated \S or \W in UTF-8 mode could give wrong answers when multibyte
1016        characters were involved (for example /\S{2}/8g with "A\x{a3}BC").
1017    
1018    25. Using pcregrep in multiline, inverted mode (-Mv) caused it to loop.
1019    
1020    26. Patterns such as [\P{Yi}A] which include \p or \P and just one other
1021        character were causing crashes (broken optimization).
1022    
1023    27. Patterns such as (\P{Yi}*\277)* (group with possible zero repeat containing
1024        \p or \P) caused a compile-time loop.
1025    
1026    28. More problems have arisen in unanchored patterns when CRLF is a valid line
1027        break. For example, the unstudied pattern [\r\n]A does not match the string
1028        "\r\nA" because change 7.0/46 below moves the current point on by two
1029        characters after failing to match at the start. However, the pattern \nA
1030        *does* match, because it doesn't start till \n, and if [\r\n]A is studied,
1031        the same is true. There doesn't seem any very clean way out of this, but
1032        what I have chosen to do makes the common cases work: PCRE now takes note
1033        of whether there can be an explicit match for \r or \n anywhere in the
1034        pattern, and if so, 7.0/46 no longer applies. As part of this change,
1035        there's a new PCRE_INFO_HASCRORLF option for finding out whether a compiled
1036        pattern has explicit CR or LF references.
1037    
1038    29. Added (*CR) etc for changing newline setting at start of pattern.
1039    
1040    
1041    Version 7.2 19-Jun-07
1042    ---------------------
1043    
1044     1. If the fr_FR locale cannot be found for test 3, try the "french" locale,
1045        which is apparently normally available under Windows.
1046    
1047     2. Re-jig the pcregrep tests with different newline settings in an attempt
1048        to make them independent of the local environment's newline setting.
1049    
1050     3. Add code to configure.ac to remove -g from the CFLAGS default settings.
1051    
1052     4. Some of the "internals" tests were previously cut out when the link size
1053        was not 2, because the output contained actual offsets. The recent new
1054        "Z" feature of pcretest means that these can be cut out, making the tests
1055        usable with all link sizes.
1056    
1057     5. Implemented Stan Switzer's goto replacement for longjmp() when not using
1058        stack recursion. This gives a massive performance boost under BSD, but just
1059        a small improvement under Linux. However, it saves one field in the frame
1060        in all cases.
1061    
1062     6. Added more features from the forthcoming Perl 5.10:
1063    
1064        (a) (?-n) (where n is a string of digits) is a relative subroutine or
1065            recursion call. It refers to the nth most recently opened parentheses.
1066    
1067        (b) (?+n) is also a relative subroutine call; it refers to the nth next
1068            to be opened parentheses.
1069    
1070        (c) Conditions that refer to capturing parentheses can be specified
1071            relatively, for example, (?(-2)... or (?(+3)...
1072    
1073        (d) \K resets the start of the current match so that everything before
1074            is not part of it.
1075    
1076        (e) \k{name} is synonymous with \k<name> and \k'name' (.NET compatible).
1077    
1078        (f) \g{name} is another synonym - part of Perl 5.10's unification of
1079            reference syntax.
1080    
1081        (g) (?| introduces a group in which the numbering of parentheses in each
1082            alternative starts with the same number.
1083    
1084        (h) \h, \H, \v, and \V match horizontal and vertical whitespace.
1085    
1086     7. Added two new calls to pcre_fullinfo(): PCRE_INFO_OKPARTIAL and
1087        PCRE_INFO_JCHANGED.
1088    
1089     8. A pattern such as  (.*(.)?)*  caused pcre_exec() to fail by either not
1090        terminating or by crashing. Diagnosed by Viktor Griph; it was in the code
1091        for detecting groups that can match an empty string.
1092    
1093     9. A pattern with a very large number of alternatives (more than several
1094        hundred) was running out of internal workspace during the pre-compile
1095        phase, where pcre_compile() figures out how much memory will be needed. A
1096        bit of new cunning has reduced the workspace needed for groups with
1097        alternatives. The 1000-alternative test pattern now uses 12 bytes of
1098        workspace instead of running out of the 4096 that are available.
1099    
1100    10. Inserted some missing (unsigned int) casts to get rid of compiler warnings.
1101    
1102    11. Applied patch from Google to remove an optimization that didn't quite work.
1103        The report of the bug said:
1104    
1105          pcrecpp::RE("a*").FullMatch("aaa") matches, while
1106          pcrecpp::RE("a*?").FullMatch("aaa") does not, and
1107          pcrecpp::RE("a*?\\z").FullMatch("aaa") does again.
1108    
1109    12. If \p or \P was used in non-UTF-8 mode on a character greater than 127
1110        it matched the wrong number of bytes.
1111    
1112    
1113    Version 7.1 24-Apr-07
1114    ---------------------
1115    
1116     1. Applied Bob Rossi and Daniel G's patches to convert the build system to one
1117        that is more "standard", making use of automake and other Autotools. There
1118        is some re-arrangement of the files and adjustment of comments consequent
1119        on this.
1120    
1121     2. Part of the patch fixed a problem with the pcregrep tests. The test of -r
1122        for recursive directory scanning broke on some systems because the files
1123        are not scanned in any specific order and on different systems the order
1124        was different. A call to "sort" has been inserted into RunGrepTest for the
1125        approprate test as a short-term fix. In the longer term there may be an
1126        alternative.
1127    
1128     3. I had an email from Eric Raymond about problems translating some of PCRE's
1129        man pages to HTML (despite the fact that I distribute HTML pages, some
1130        people do their own conversions for various reasons). The problems
1131        concerned the use of low-level troff macros .br and .in. I have therefore
1132        removed all such uses from the man pages (some were redundant, some could
1133        be replaced by .nf/.fi pairs). The 132html script that I use to generate
1134        HTML has been updated to handle .nf/.fi and to complain if it encounters
1135        .br or .in.
1136    
1137     4. Updated comments in configure.ac that get placed in config.h.in and also
1138        arranged for config.h to be included in the distribution, with the name
1139        config.h.generic, for the benefit of those who have to compile without
1140        Autotools (compare pcre.h, which is now distributed as pcre.h.generic).
1141    
1142     5. Updated the support (such as it is) for Virtual Pascal, thanks to Stefan
1143        Weber: (1) pcre_internal.h was missing some function renames; (2) updated
1144        makevp.bat for the current PCRE, using the additional files
1145        makevp_c.txt, makevp_l.txt, and pcregexp.pas.
1146    
1147     6. A Windows user reported a minor discrepancy with test 2, which turned out
1148        to be caused by a trailing space on an input line that had got lost in his
1149        copy. The trailing space was an accident, so I've just removed it.
1150    
1151     7. Add -Wl,-R... flags in pcre-config.in for *BSD* systems, as I'm told
1152        that is needed.
1153    
1154     8. Mark ucp_table (in ucptable.h) and ucp_gentype (in pcre_ucp_searchfuncs.c)
1155        as "const" (a) because they are and (b) because it helps the PHP
1156        maintainers who have recently made a script to detect big data structures
1157        in the php code that should be moved to the .rodata section. I remembered
1158        to update Builducptable as well, so it won't revert if ucptable.h is ever
1159        re-created.
1160    
1161     9. Added some extra #ifdef SUPPORT_UTF8 conditionals into pcretest.c,
1162        pcre_printint.src, pcre_compile.c, pcre_study.c, and pcre_tables.c, in
1163        order to be able to cut out the UTF-8 tables in the latter when UTF-8
1164        support is not required. This saves 1.5-2K of code, which is important in
1165        some applications.
1166    
1167        Later: more #ifdefs are needed in pcre_ord2utf8.c and pcre_valid_utf8.c
1168        so as not to refer to the tables, even though these functions will never be
1169        called when UTF-8 support is disabled. Otherwise there are problems with a
1170        shared library.
1171    
1172    10. Fixed two bugs in the emulated memmove() function in pcre_internal.h:
1173    
1174        (a) It was defining its arguments as char * instead of void *.
1175    
1176        (b) It was assuming that all moves were upwards in memory; this was true
1177            a long time ago when I wrote it, but is no longer the case.
1178    
1179        The emulated memove() is provided for those environments that have neither
1180        memmove() nor bcopy(). I didn't think anyone used it these days, but that
1181        is clearly not the case, as these two bugs were recently reported.
1182    
1183    11. The script PrepareRelease is now distributed: it calls 132html, CleanTxt,
1184        and Detrail to create the HTML documentation, the .txt form of the man
1185        pages, and it removes trailing spaces from listed files. It also creates
1186        pcre.h.generic and config.h.generic from pcre.h and config.h. In the latter
1187        case, it wraps all the #defines with #ifndefs. This script should be run
1188        before "make dist".
1189    
1190    12. Fixed two fairly obscure bugs concerned with quantified caseless matching
1191        with Unicode property support.
1192    
1193        (a) For a maximizing quantifier, if the two different cases of the
1194            character were of different lengths in their UTF-8 codings (there are
1195            some cases like this - I found 11), and the matching function had to
1196            back up over a mixture of the two cases, it incorrectly assumed they
1197            were both the same length.
1198    
1199        (b) When PCRE was configured to use the heap rather than the stack for
1200            recursion during matching, it was not correctly preserving the data for
1201            the other case of a UTF-8 character when checking ahead for a match
1202            while processing a minimizing repeat. If the check also involved
1203            matching a wide character, but failed, corruption could cause an
1204            erroneous result when trying to check for a repeat of the original
1205            character.
1206    
1207    13. Some tidying changes to the testing mechanism:
1208    
1209        (a) The RunTest script now detects the internal link size and whether there
1210            is UTF-8 and UCP support by running ./pcretest -C instead of relying on
1211            values substituted by "configure". (The RunGrepTest script already did
1212            this for UTF-8.) The configure.ac script no longer substitutes the
1213            relevant variables.
1214    
1215        (b) The debugging options /B and /D in pcretest show the compiled bytecode
1216            with length and offset values. This means that the output is different
1217            for different internal link sizes. Test 2 is skipped for link sizes
1218            other than 2 because of this, bypassing the problem. Unfortunately,
1219            there was also a test in test 3 (the locale tests) that used /B and
1220            failed for link sizes other than 2. Rather than cut the whole test out,
1221            I have added a new /Z option to pcretest that replaces the length and
1222            offset values with spaces. This is now used to make test 3 independent
1223            of link size. (Test 2 will be tidied up later.)
1224    
1225    14. If erroroffset was passed as NULL to pcre_compile, it provoked a
1226        segmentation fault instead of returning the appropriate error message.
1227    
1228    15. In multiline mode when the newline sequence was set to "any", the pattern
1229        ^$ would give a match between the \r and \n of a subject such as "A\r\nB".
1230        This doesn't seem right; it now treats the CRLF combination as the line
1231        ending, and so does not match in that case. It's only a pattern such as ^$
1232        that would hit this one: something like ^ABC$ would have failed after \r
1233        and then tried again after \r\n.
1234    
1235    16. Changed the comparison command for RunGrepTest from "diff -u" to "diff -ub"
1236        in an attempt to make files that differ only in their line terminators
1237        compare equal. This works on Linux.
1238    
1239    17. Under certain error circumstances pcregrep might try to free random memory
1240        as it exited. This is now fixed, thanks to valgrind.
1241    
1242    19. In pcretest, if the pattern /(?m)^$/g<any> was matched against the string
1243        "abc\r\n\r\n", it found an unwanted second match after the second \r. This
1244        was because its rules for how to advance for /g after matching an empty
1245        string at the end of a line did not allow for this case. They now check for
1246        it specially.
1247    
1248    20. pcretest is supposed to handle patterns and data of any length, by
1249        extending its buffers when necessary. It was getting this wrong when the
1250        buffer for a data line had to be extended.
1251    
1252    21. Added PCRE_NEWLINE_ANYCRLF which is like ANY, but matches only CR, LF, or
1253        CRLF as a newline sequence.
1254    
1255    22. Code for handling Unicode properties in pcre_dfa_exec() wasn't being cut
1256        out by #ifdef SUPPORT_UCP. This did no harm, as it could never be used, but
1257        I have nevertheless tidied it up.
1258    
1259    23. Added some casts to kill warnings from HP-UX ia64 compiler.
1260    
1261    24. Added a man page for pcre-config.
1262    
1263    
1264    Version 7.0 19-Dec-06
1265    ---------------------
1266    
1267     1. Fixed a signed/unsigned compiler warning in pcre_compile.c, shown up by
1268        moving to gcc 4.1.1.
1269    
1270     2. The -S option for pcretest uses setrlimit(); I had omitted to #include
1271        sys/time.h, which is documented as needed for this function. It doesn't
1272        seem to matter on Linux, but it showed up on some releases of OS X.
1273    
1274     3. It seems that there are systems where bytes whose values are greater than
1275        127 match isprint() in the "C" locale. The "C" locale should be the
1276        default when a C program starts up. In most systems, only ASCII printing
1277        characters match isprint(). This difference caused the output from pcretest
1278        to vary, making some of the tests fail. I have changed pcretest so that:
1279    
1280        (a) When it is outputting text in the compiled version of a pattern, bytes
1281            other than 32-126 are always shown as hex escapes.
1282    
1283        (b) When it is outputting text that is a matched part of a subject string,
1284            it does the same, unless a different locale has been set for the match
1285            (using the /L modifier). In this case, it uses isprint() to decide.
1286    
1287     4. Fixed a major bug that caused incorrect computation of the amount of memory
1288        required for a compiled pattern when options that changed within the
1289        pattern affected the logic of the preliminary scan that determines the
1290        length. The relevant options are -x, and -i in UTF-8 mode. The result was
1291        that the computed length was too small. The symptoms of this bug were
1292        either the PCRE error "internal error: code overflow" from pcre_compile(),
1293        or a glibc crash with a message such as "pcretest: free(): invalid next
1294        size (fast)". Examples of patterns that provoked this bug (shown in
1295        pcretest format) are:
1296    
1297          /(?-x: )/x
1298          /(?x)(?-x: \s*#\s*)/
1299          /((?i)[\x{c0}])/8
1300          /(?i:[\x{c0}])/8
1301    
1302        HOWEVER: Change 17 below makes this fix obsolete as the memory computation
1303        is now done differently.
1304    
1305     5. Applied patches from Google to: (a) add a QuoteMeta function to the C++
1306        wrapper classes; (b) implement a new function in the C++ scanner that is
1307        more efficient than the old way of doing things because it avoids levels of
1308        recursion in the regex matching; (c) add a paragraph to the documentation
1309        for the FullMatch() function.
1310    
1311     6. The escape sequence \n was being treated as whatever was defined as
1312        "newline". Not only was this contrary to the documentation, which states
1313        that \n is character 10 (hex 0A), but it also went horribly wrong when
1314        "newline" was defined as CRLF. This has been fixed.
1315    
1316     7. In pcre_dfa_exec.c the value of an unsigned integer (the variable called c)
1317        was being set to -1 for the "end of line" case (supposedly a value that no
1318        character can have). Though this value is never used (the check for end of
1319        line is "zero bytes in current character"), it caused compiler complaints.
1320        I've changed it to 0xffffffff.
1321    
1322     8. In pcre_version.c, the version string was being built by a sequence of
1323        C macros that, in the event of PCRE_PRERELEASE being defined as an empty
1324        string (as it is for production releases) called a macro with an empty
1325        argument. The C standard says the result of this is undefined. The gcc
1326        compiler treats it as an empty string (which was what was wanted) but it is
1327        reported that Visual C gives an error. The source has been hacked around to
1328        avoid this problem.
1329    
1330     9. On the advice of a Windows user, included <io.h> and <fcntl.h> in Windows
1331        builds of pcretest, and changed the call to _setmode() to use _O_BINARY
1332        instead of 0x8000. Made all the #ifdefs test both _WIN32 and WIN32 (not all
1333        of them did).
1334    
1335    10. Originally, pcretest opened its input and output without "b"; then I was
1336        told that "b" was needed in some environments, so it was added for release
1337        5.0 to both the input and output. (It makes no difference on Unix-like
1338        systems.) Later I was told that it is wrong for the input on Windows. I've
1339        now abstracted the modes into two macros, to make it easier to fiddle with
1340        them, and removed "b" from the input mode under Windows.
1341    
1342    11. Added pkgconfig support for the C++ wrapper library, libpcrecpp.
1343    
1344    12. Added -help and --help to pcretest as an official way of being reminded
1345        of the options.
1346    
1347    13. Removed some redundant semicolons after macro calls in pcrecpparg.h.in
1348        and pcrecpp.cc because they annoy compilers at high warning levels.
1349    
1350    14. A bit of tidying/refactoring in pcre_exec.c in the main bumpalong loop.
1351    
1352    15. Fixed an occurrence of == in configure.ac that should have been = (shell
1353        scripts are not C programs :-) and which was not noticed because it works
1354        on Linux.
1355    
1356    16. pcretest is supposed to handle any length of pattern and data line (as one
1357        line or as a continued sequence of lines) by extending its input buffer if
1358        necessary. This feature was broken for very long pattern lines, leading to
1359        a string of junk being passed to pcre_compile() if the pattern was longer
1360        than about 50K.
1361    
1362    17. I have done a major re-factoring of the way pcre_compile() computes the
1363        amount of memory needed for a compiled pattern. Previously, there was code
1364        that made a preliminary scan of the pattern in order to do this. That was
1365        OK when PCRE was new, but as the facilities have expanded, it has become
1366        harder and harder to keep it in step with the real compile phase, and there
1367        have been a number of bugs (see for example, 4 above). I have now found a
1368        cunning way of running the real compile function in a "fake" mode that
1369        enables it to compute how much memory it would need, while actually only
1370        ever using a few hundred bytes of working memory and without too many
1371        tests of the mode. This should make future maintenance and development
1372        easier. A side effect of this work is that the limit of 200 on the nesting
1373        depth of parentheses has been removed (though this was never a serious
1374        limitation, I suspect). However, there is a downside: pcre_compile() now
1375        runs more slowly than before (30% or more, depending on the pattern). I
1376        hope this isn't a big issue. There is no effect on runtime performance.
1377    
1378    18. Fixed a minor bug in pcretest: if a pattern line was not terminated by a
1379        newline (only possible for the last line of a file) and it was a
1380        pattern that set a locale (followed by /Lsomething), pcretest crashed.
1381    
1382    19. Added additional timing features to pcretest. (1) The -tm option now times
1383        matching only, not compiling. (2) Both -t and -tm can be followed, as a
1384        separate command line item, by a number that specifies the number of
1385        repeats to use when timing. The default is 50000; this gives better
1386        precision, but takes uncomfortably long for very large patterns.
1387    
1388    20. Extended pcre_study() to be more clever in cases where a branch of a
1389        subpattern has no definite first character. For example, (a*|b*)[cd] would
1390        previously give no result from pcre_study(). Now it recognizes that the
1391        first character must be a, b, c, or d.
1392    
1393    21. There was an incorrect error "recursive call could loop indefinitely" if
1394        a subpattern (or the entire pattern) that was being tested for matching an
1395        empty string contained only one non-empty item after a nested subpattern.
1396        For example, the pattern (?>\x{100}*)\d(?R) provoked this error
1397        incorrectly, because the \d was being skipped in the check.
1398    
1399    22. The pcretest program now has a new pattern option /B and a command line
1400        option -b, which is equivalent to adding /B to every pattern. This causes
1401        it to show the compiled bytecode, without the additional information that
1402        -d shows. The effect of -d is now the same as -b with -i (and similarly, /D
1403        is the same as /B/I).
1404    
1405    23. A new optimization is now able automatically to treat some sequences such
1406        as a*b as a*+b. More specifically, if something simple (such as a character
1407        or a simple class like \d) has an unlimited quantifier, and is followed by
1408        something that cannot possibly match the quantified thing, the quantifier
1409        is automatically "possessified".
1410    
1411    24. A recursive reference to a subpattern whose number was greater than 39
1412        went wrong under certain circumstances in UTF-8 mode. This bug could also
1413        have affected the operation of pcre_study().
1414    
1415    25. Realized that a little bit of performance could be had by replacing
1416        (c & 0xc0) == 0xc0 with c >= 0xc0 when processing UTF-8 characters.
1417    
1418    26. Timing data from pcretest is now shown to 4 decimal places instead of 3.
1419    
1420    27. Possessive quantifiers such as a++ were previously implemented by turning
1421        them into atomic groups such as ($>a+). Now they have their own opcodes,
1422        which improves performance. This includes the automatically created ones
1423        from 23 above.
1424    
1425    28. A pattern such as (?=(\w+))\1: which simulates an atomic group using a
1426        lookahead was broken if it was not anchored. PCRE was mistakenly expecting
1427        the first matched character to be a colon. This applied both to named and
1428        numbered groups.
1429    
1430    29. The ucpinternal.h header file was missing its idempotency #ifdef.
1431    
1432    30. I was sent a "project" file called libpcre.a.dev which I understand makes
1433        building PCRE on Windows easier, so I have included it in the distribution.
1434    
1435    31. There is now a check in pcretest against a ridiculously large number being
1436        returned by pcre_exec() or pcre_dfa_exec(). If this happens in a /g or /G
1437        loop, the loop is abandoned.
1438    
1439    32. Forward references to subpatterns in conditions such as (?(2)...) where
1440        subpattern 2 is defined later cause pcre_compile() to search forwards in
1441        the pattern for the relevant set of parentheses. This search went wrong
1442        when there were unescaped parentheses in a character class, parentheses
1443        escaped with \Q...\E, or parentheses in a #-comment in /x mode.
1444    
1445    33. "Subroutine" calls and backreferences were previously restricted to
1446        referencing subpatterns earlier in the regex. This restriction has now
1447        been removed.
1448    
1449    34. Added a number of extra features that are going to be in Perl 5.10. On the
1450        whole, these are just syntactic alternatives for features that PCRE had
1451        previously implemented using the Python syntax or my own invention. The
1452        other formats are all retained for compatibility.
1453    
1454        (a) Named groups can now be defined as (?<name>...) or (?'name'...) as well
1455            as (?P<name>...). The new forms, as well as being in Perl 5.10, are
1456            also .NET compatible.
1457    
1458        (b) A recursion or subroutine call to a named group can now be defined as
1459            (?&name) as well as (?P>name).
1460    
1461        (c) A backreference to a named group can now be defined as \k<name> or
1462            \k'name' as well as (?P=name). The new forms, as well as being in Perl
1463            5.10, are also .NET compatible.
1464    
1465        (d) A conditional reference to a named group can now use the syntax
1466            (?(<name>) or (?('name') as well as (?(name).
1467    
1468        (e) A "conditional group" of the form (?(DEFINE)...) can be used to define
1469            groups (named and numbered) that are never evaluated inline, but can be
1470            called as "subroutines" from elsewhere. In effect, the DEFINE condition
1471            is always false. There may be only one alternative in such a group.
1472    
1473        (f) A test for recursion can be given as (?(R1).. or (?(R&name)... as well
1474            as the simple (?(R). The condition is true only if the most recent
1475            recursion is that of the given number or name. It does not search out
1476            through the entire recursion stack.
1477    
1478        (g) The escape \gN or \g{N} has been added, where N is a positive or
1479            negative number, specifying an absolute or relative reference.
1480    
1481    35. Tidied to get rid of some further signed/unsigned compiler warnings and
1482        some "unreachable code" warnings.
1483    
1484    36. Updated the Unicode property tables to Unicode version 5.0.0. Amongst other
1485        things, this adds five new scripts.
1486    
1487    37. Perl ignores orphaned \E escapes completely. PCRE now does the same.
1488        There were also incompatibilities regarding the handling of \Q..\E inside
1489        character classes, for example with patterns like [\Qa\E-\Qz\E] where the
1490        hyphen was adjacent to \Q or \E. I hope I've cleared all this up now.
1491    
1492    38. Like Perl, PCRE detects when an indefinitely repeated parenthesized group
1493        matches an empty string, and forcibly breaks the loop. There were bugs in
1494        this code in non-simple cases. For a pattern such as  ^(a()*)*  matched
1495        against  aaaa  the result was just "a" rather than "aaaa", for example. Two
1496        separate and independent bugs (that affected different cases) have been
1497        fixed.
1498    
1499    39. Refactored the code to abolish the use of different opcodes for small
1500        capturing bracket numbers. This is a tidy that I avoided doing when I
1501        removed the limit on the number of capturing brackets for 3.5 back in 2001.
1502        The new approach is not only tidier, it makes it possible to reduce the
1503        memory needed to fix the previous bug (38).
1504    
1505    40. Implemented PCRE_NEWLINE_ANY to recognize any of the Unicode newline
1506        sequences (http://unicode.org/unicode/reports/tr18/) as "newline" when
1507        processing dot, circumflex, or dollar metacharacters, or #-comments in /x
1508        mode.
1509    
1510    41. Add \R to match any Unicode newline sequence, as suggested in the Unicode
1511        report.
1512    
1513    42. Applied patch, originally from Ari Pollak, modified by Google, to allow
1514        copy construction and assignment in the C++ wrapper.
1515    
1516    43. Updated pcregrep to support "--newline=any". In the process, I fixed a
1517        couple of bugs that could have given wrong results in the "--newline=crlf"
1518        case.
1519    
1520    44. Added a number of casts and did some reorganization of signed/unsigned int
1521        variables following suggestions from Dair Grant. Also renamed the variable
1522        "this" as "item" because it is a C++ keyword.
1523    
1524    45. Arranged for dftables to add
1525    
1526          #include "pcre_internal.h"
1527    
1528        to pcre_chartables.c because without it, gcc 4.x may remove the array
1529        definition from the final binary if PCRE is built into a static library and
1530        dead code stripping is activated.
1531    
1532    46. For an unanchored pattern, if a match attempt fails at the start of a
1533        newline sequence, and the newline setting is CRLF or ANY, and the next two
1534        characters are CRLF, advance by two characters instead of one.
1535    
1536    
1537    Version 6.7 04-Jul-06
1538    ---------------------
1539    
1540     1. In order to handle tests when input lines are enormously long, pcretest has
1541        been re-factored so that it automatically extends its buffers when
1542        necessary. The code is crude, but this _is_ just a test program. The
1543        default size has been increased from 32K to 50K.
1544    
1545     2. The code in pcre_study() was using the value of the re argument before
1546        testing it for NULL. (Of course, in any sensible call of the function, it
1547        won't be NULL.)
1548    
1549     3. The memmove() emulation function in pcre_internal.h, which is used on
1550        systems that lack both memmove() and bcopy() - that is, hardly ever -
1551        was missing a "static" storage class specifier.
1552    
1553     4. When UTF-8 mode was not set, PCRE looped when compiling certain patterns
1554        containing an extended class (one that cannot be represented by a bitmap
1555        because it contains high-valued characters or Unicode property items, e.g.
1556        [\pZ]). Almost always one would set UTF-8 mode when processing such a
1557        pattern, but PCRE should not loop if you do not (it no longer does).
1558        [Detail: two cases were found: (a) a repeated subpattern containing an
1559        extended class; (b) a recursive reference to a subpattern that followed a
1560        previous extended class. It wasn't skipping over the extended class
1561        correctly when UTF-8 mode was not set.]
1562    
1563     5. A negated single-character class was not being recognized as fixed-length
1564        in lookbehind assertions such as (?<=[^f]), leading to an incorrect
1565        compile error "lookbehind assertion is not fixed length".
1566    
1567     6. The RunPerlTest auxiliary script was showing an unexpected difference
1568        between PCRE and Perl for UTF-8 tests. It turns out that it is hard to
1569        write a Perl script that can interpret lines of an input file either as
1570        byte characters or as UTF-8, which is what "perltest" was being required to
1571        do for the non-UTF-8 and UTF-8 tests, respectively. Essentially what you
1572        can't do is switch easily at run time between having the "use utf8;" pragma
1573        or not. In the end, I fudged it by using the RunPerlTest script to insert
1574        "use utf8;" explicitly for the UTF-8 tests.
1575    
1576     7. In multiline (/m) mode, PCRE was matching ^ after a terminating newline at
1577        the end of the subject string, contrary to the documentation and to what
1578        Perl does. This was true of both matching functions. Now it matches only at
1579        the start of the subject and immediately after *internal* newlines.
1580    
1581     8. A call of pcre_fullinfo() from pcretest to get the option bits was passing
1582        a pointer to an int instead of a pointer to an unsigned long int. This
1583        caused problems on 64-bit systems.
1584    
1585     9. Applied a patch from the folks at Google to pcrecpp.cc, to fix "another
1586        instance of the 'standard' template library not being so standard".
1587    
1588    10. There was no check on the number of named subpatterns nor the maximum
1589        length of a subpattern name. The product of these values is used to compute
1590        the size of the memory block for a compiled pattern. By supplying a very
1591        long subpattern name and a large number of named subpatterns, the size
1592        computation could be caused to overflow. This is now prevented by limiting
1593        the length of names to 32 characters, and the number of named subpatterns
1594        to 10,000.
1595    
1596    11. Subpatterns that are repeated with specific counts have to be replicated in
1597        the compiled pattern. The size of memory for this was computed from the
1598        length of the subpattern and the repeat count. The latter is limited to
1599        65535, but there was no limit on the former, meaning that integer overflow
1600        could in principle occur. The compiled length of a repeated subpattern is
1601        now limited to 30,000 bytes in order to prevent this.
1602    
1603    12. Added the optional facility to have named substrings with the same name.
1604    
1605    13. Added the ability to use a named substring as a condition, using the
1606        Python syntax: (?(name)yes|no). This overloads (?(R)... and names that
1607        are numbers (not recommended). Forward references are permitted.
1608    
1609    14. Added forward references in named backreferences (if you see what I mean).
1610    
1611    15. In UTF-8 mode, with the PCRE_DOTALL option set, a quantified dot in the
1612        pattern could run off the end of the subject. For example, the pattern
1613        "(?s)(.{1,5})"8 did this with the subject "ab".
1614    
1615    16. If PCRE_DOTALL or PCRE_MULTILINE were set, pcre_dfa_exec() behaved as if
1616        PCRE_CASELESS was set when matching characters that were quantified with ?
1617        or *.
1618    
1619    17. A character class other than a single negated character that had a minimum
1620        but no maximum quantifier - for example [ab]{6,} - was not handled
1621        correctly by pce_dfa_exec(). It would match only one character.
1622    
1623    18. A valid (though odd) pattern that looked like a POSIX character
1624        class but used an invalid character after [ (for example [[,abc,]]) caused
1625        pcre_compile() to give the error "Failed: internal error: code overflow" or
1626        in some cases to crash with a glibc free() error. This could even happen if
1627        the pattern terminated after [[ but there just happened to be a sequence of
1628        letters, a binary zero, and a closing ] in the memory that followed.
1629    
1630    19. Perl's treatment of octal escapes in the range \400 to \777 has changed
1631        over the years. Originally (before any Unicode support), just the bottom 8
1632        bits were taken. Thus, for example, \500 really meant \100. Nowadays the
1633        output from "man perlunicode" includes this:
1634    
1635          The regular expression compiler produces polymorphic opcodes.  That
1636          is, the pattern adapts to the data and automatically switches to
1637          the Unicode character scheme when presented with Unicode data--or
1638          instead uses a traditional byte scheme when presented with byte
1639          data.
1640    
1641        Sadly, a wide octal escape does not cause a switch, and in a string with
1642        no other multibyte characters, these octal escapes are treated as before.
1643        Thus, in Perl, the pattern  /\500/ actually matches \100 but the pattern
1644        /\500|\x{1ff}/ matches \500 or \777 because the whole thing is treated as a
1645        Unicode string.
1646    
1647        I have not perpetrated such confusion in PCRE. Up till now, it took just
1648        the bottom 8 bits, as in old Perl. I have now made octal escapes with
1649        values greater than \377 illegal in non-UTF-8 mode. In UTF-8 mode they
1650        translate to the appropriate multibyte character.
1651    
1652    29. Applied some refactoring to reduce the number of warnings from Microsoft
1653        and Borland compilers. This has included removing the fudge introduced
1654        seven years ago for the OS/2 compiler (see 2.02/2 below) because it caused
1655        a warning about an unused variable.
1656    
1657    21. PCRE has not included VT (character 0x0b) in the set of whitespace
1658        characters since release 4.0, because Perl (from release 5.004) does not.
1659        [Or at least, is documented not to: some releases seem to be in conflict
1660        with the documentation.] However, when a pattern was studied with
1661        pcre_study() and all its branches started with \s, PCRE still included VT
1662        as a possible starting character. Of course, this did no harm; it just
1663        caused an unnecessary match attempt.
1664    
1665    22. Removed a now-redundant internal flag bit that recorded the fact that case
1666        dependency changed within the pattern. This was once needed for "required
1667        byte" processing, but is no longer used. This recovers a now-scarce options
1668        bit. Also moved the least significant internal flag bit to the most-
1669        significant bit of the word, which was not previously used (hangover from
1670        the days when it was an int rather than a uint) to free up another bit for
1671        the future.
1672    
1673    23. Added support for CRLF line endings as well as CR and LF. As well as the
1674        default being selectable at build time, it can now be changed at runtime
1675        via the PCRE_NEWLINE_xxx flags. There are now options for pcregrep to
1676        specify that it is scanning data with non-default line endings.
1677    
1678    24. Changed the definition of CXXLINK to make it agree with the definition of
1679        LINK in the Makefile, by replacing LDFLAGS to CXXFLAGS.
1680    
1681    25. Applied Ian Taylor's patches to avoid using another stack frame for tail
1682        recursions. This makes a big different to stack usage for some patterns.
1683    
1684    26. If a subpattern containing a named recursion or subroutine reference such
1685        as (?P>B) was quantified, for example (xxx(?P>B)){3}, the calculation of
1686        the space required for the compiled pattern went wrong and gave too small a
1687        value. Depending on the environment, this could lead to "Failed: internal
1688        error: code overflow at offset 49" or "glibc detected double free or
1689        corruption" errors.
1690    
1691    27. Applied patches from Google (a) to support the new newline modes and (b) to
1692        advance over multibyte UTF-8 characters in GlobalReplace.
1693    
1694    28. Change free() to pcre_free() in pcredemo.c. Apparently this makes a
1695        difference for some implementation of PCRE in some Windows version.
1696    
1697    29. Added some extra testing facilities to pcretest:
1698    
1699        \q<number>   in a data line sets the "match limit" value
1700        \Q<number>   in a data line sets the "match recursion limt" value
1701        -S <number>  sets the stack size, where <number> is in megabytes
1702    
1703        The -S option isn't available for Windows.
1704    
1705    
1706    Version 6.6 06-Feb-06
1707    ---------------------
1708    
1709     1. Change 16(a) for 6.5 broke things, because PCRE_DATA_SCOPE was not defined
1710        in pcreposix.h. I have copied the definition from pcre.h.
1711    
1712     2. Change 25 for 6.5 broke compilation in a build directory out-of-tree
1713        because pcre.h is no longer a built file.
1714    
1715     3. Added Jeff Friedl's additional debugging patches to pcregrep. These are
1716        not normally included in the compiled code.
1717    
1718    
1719    Version 6.5 01-Feb-06
1720    ---------------------
1721    
1722     1. When using the partial match feature with pcre_dfa_exec(), it was not
1723        anchoring the second and subsequent partial matches at the new starting
1724        point. This could lead to incorrect results. For example, with the pattern
1725        /1234/, partially matching against "123" and then "a4" gave a match.
1726    
1727     2. Changes to pcregrep:
1728    
1729        (a) All non-match returns from pcre_exec() were being treated as failures
1730            to match the line. Now, unless the error is PCRE_ERROR_NOMATCH, an
1731            error message is output. Some extra information is given for the
1732            PCRE_ERROR_MATCHLIMIT and PCRE_ERROR_RECURSIONLIMIT errors, which are
1733            probably the only errors that are likely to be caused by users (by
1734            specifying a regex that has nested indefinite repeats, for instance).
1735            If there are more than 20 of these errors, pcregrep is abandoned.
1736    
1737        (b) A binary zero was treated as data while matching, but terminated the
1738            output line if it was written out. This has been fixed: binary zeroes
1739            are now no different to any other data bytes.
1740    
1741        (c) Whichever of the LC_ALL or LC_CTYPE environment variables is set is
1742            used to set a locale for matching. The --locale=xxxx long option has
1743            been added (no short equivalent) to specify a locale explicitly on the
1744            pcregrep command, overriding the environment variables.
1745    
1746        (d) When -B was used with -n, some line numbers in the output were one less
1747            than they should have been.
1748    
1749        (e) Added the -o (--only-matching) option.
1750    
1751        (f) If -A or -C was used with -c (count only), some lines of context were
1752            accidentally printed for the final match.
1753    
1754        (g) Added the -H (--with-filename) option.
1755    
1756        (h) The combination of options -rh failed to suppress file names for files
1757            that were found from directory arguments.
1758    
1759        (i) Added the -D (--devices) and -d (--directories) options.
1760    
1761        (j) Added the -F (--fixed-strings) option.
1762    
1763        (k) Allow "-" to be used as a file name for -f as well as for a data file.
1764    
1765        (l) Added the --colo(u)r option.
1766    
1767        (m) Added Jeffrey Friedl's -S testing option, but within #ifdefs so that it
1768            is not present by default.
1769    
1770     3. A nasty bug was discovered in the handling of recursive patterns, that is,
1771        items such as (?R) or (?1), when the recursion could match a number of
1772        alternatives. If it matched one of the alternatives, but subsequently,
1773        outside the recursion, there was a failure, the code tried to back up into
1774        the recursion. However, because of the way PCRE is implemented, this is not
1775        possible, and the result was an incorrect result from the match.
1776    
1777        In order to prevent this happening, the specification of recursion has
1778        been changed so that all such subpatterns are automatically treated as
1779        atomic groups. Thus, for example, (?R) is treated as if it were (?>(?R)).
1780    
1781     4. I had overlooked the fact that, in some locales, there are characters for
1782        which isalpha() is true but neither isupper() nor islower() are true. In
1783        the fr_FR locale, for instance, the \xAA and \xBA characters (ordmasculine
1784        and ordfeminine) are like this. This affected the treatment of \w and \W
1785        when they appeared in character classes, but not when they appeared outside
1786        a character class. The bit map for "word" characters is now created
1787        separately from the results of isalnum() instead of just taking it from the
1788        upper, lower, and digit maps. (Plus the underscore character, of course.)
1789    
1790     5. The above bug also affected the handling of POSIX character classes such as
1791        [[:alpha:]] and [[:alnum:]]. These do not have their own bit maps in PCRE's
1792        permanent tables. Instead, the bit maps for such a class were previously
1793        created as the appropriate unions of the upper, lower, and digit bitmaps.
1794        Now they are created by subtraction from the [[:word:]] class, which has
1795        its own bitmap.
1796    
1797     6. The [[:blank:]] character class matches horizontal, but not vertical space.
1798        It is created by subtracting the vertical space characters (\x09, \x0a,
1799        \x0b, \x0c) from the [[:space:]] bitmap. Previously, however, the
1800        subtraction was done in the overall bitmap for a character class, meaning
1801        that a class such as [\x0c[:blank:]] was incorrect because \x0c would not
1802        be recognized. This bug has been fixed.
1803    
1804     7. Patches from the folks at Google:
1805    
1806          (a) pcrecpp.cc: "to handle a corner case that may or may not happen in
1807          real life, but is still worth protecting against".
1808    
1809          (b) pcrecpp.cc: "corrects a bug when negative radixes are used with
1810          regular expressions".
1811    
1812          (c) pcre_scanner.cc: avoid use of std::count() because not all systems
1813          have it.
1814    
1815          (d) Split off pcrecpparg.h from pcrecpp.h and had the former built by
1816          "configure" and the latter not, in order to fix a problem somebody had
1817          with compiling the Arg class on HP-UX.
1818    
1819          (e) Improve the error-handling of the C++ wrapper a little bit.
1820    
1821          (f) New tests for checking recursion limiting.
1822    
1823     8. The pcre_memmove() function, which is used only if the environment does not
1824        have a standard memmove() function (and is therefore rarely compiled),
1825        contained two bugs: (a) use of int instead of size_t, and (b) it was not
1826        returning a result (though PCRE never actually uses the result).
1827    
1828     9. In the POSIX regexec() interface, if nmatch is specified as a ridiculously
1829        large number - greater than INT_MAX/(3*sizeof(int)) - REG_ESPACE is
1830        returned instead of calling malloc() with an overflowing number that would
1831        most likely cause subsequent chaos.
1832    
1833    10. The debugging option of pcretest was not showing the NO_AUTO_CAPTURE flag.
1834    
1835    11. The POSIX flag REG_NOSUB is now supported. When a pattern that was compiled
1836        with this option is matched, the nmatch and pmatch options of regexec() are
1837        ignored.
1838    
1839    12. Added REG_UTF8 to the POSIX interface. This is not defined by POSIX, but is
1840        provided in case anyone wants to the the POSIX interface with UTF-8
1841        strings.
1842    
1843    13. Added CXXLDFLAGS to the Makefile parameters to provide settings only on the
1844        C++ linking (needed for some HP-UX environments).
1845    
1846    14. Avoid compiler warnings in get_ucpname() when compiled without UCP support
1847        (unused parameter) and in the pcre_printint() function (omitted "default"
1848        switch label when the default is to do nothing).
1849    
1850    15. Added some code to make it possible, when PCRE is compiled as a C++
1851        library, to replace subject pointers for pcre_exec() with a smart pointer
1852        class, thus making it possible to process discontinuous strings.
1853    
1854    16. The two macros PCRE_EXPORT and PCRE_DATA_SCOPE are confusing, and perform
1855        much the same function. They were added by different people who were trying
1856        to make PCRE easy to compile on non-Unix systems. It has been suggested
1857        that PCRE_EXPORT be abolished now that there is more automatic apparatus
1858        for compiling on Windows systems. I have therefore replaced it with
1859        PCRE_DATA_SCOPE. This is set automatically for Windows; if not set it
1860        defaults to "extern" for C or "extern C" for C++, which works fine on
1861        Unix-like systems. It is now possible to override the value of PCRE_DATA_
1862        SCOPE with something explicit in config.h. In addition:
1863    
1864        (a) pcreposix.h still had just "extern" instead of either of these macros;
1865            I have replaced it with PCRE_DATA_SCOPE.
1866    
1867        (b) Functions such as _pcre_xclass(), which are internal to the library,
1868            but external in the C sense, all had PCRE_EXPORT in their definitions.
1869            This is apparently wrong for the Windows case, so I have removed it.
1870            (It makes no difference on Unix-like systems.)
1871    
1872    17. Added a new limit, MATCH_LIMIT_RECURSION, which limits the depth of nesting
1873        of recursive calls to match(). This is different to MATCH_LIMIT because
1874        that limits the total number of calls to match(), not all of which increase
1875        the depth of recursion. Limiting the recursion depth limits the amount of
1876        stack (or heap if NO_RECURSE is set) that is used. The default can be set
1877        when PCRE is compiled, and changed at run time. A patch from Google adds
1878        this functionality to the C++ interface.
1879    
1880    18. Changes to the handling of Unicode character properties:
1881    
1882        (a) Updated the table to Unicode 4.1.0.
1883    
1884        (b) Recognize characters that are not in the table as "Cn" (undefined).
1885    
1886        (c) I revised the way the table is implemented to a much improved format
1887            which includes recognition of ranges. It now supports the ranges that
1888            are defined in UnicodeData.txt, and it also amalgamates other
1889            characters into ranges. This has reduced the number of entries in the
1890            table from around 16,000 to around 3,000, thus reducing its size
1891            considerably. I realized I did not need to use a tree structure after
1892            all - a binary chop search is just as efficient. Having reduced the
1893            number of entries, I extended their size from 6 bytes to 8 bytes to
1894            allow for more data.
1895    
1896        (d) Added support for Unicode script names via properties such as \p{Han}.
1897    
1898    19. In UTF-8 mode, a backslash followed by a non-Ascii character was not
1899        matching that character.
1900    
1901    20. When matching a repeated Unicode property with a minimum greater than zero,
1902        (for example \pL{2,}), PCRE could look past the end of the subject if it
1903        reached it while seeking the minimum number of characters. This could
1904        happen only if some of the characters were more than one byte long, because
1905        there is a check for at least the minimum number of bytes.
1906    
1907    21. Refactored the implementation of \p and \P so as to be more general, to
1908        allow for more different types of property in future. This has changed the
1909        compiled form incompatibly. Anybody with saved compiled patterns that use
1910        \p or \P will have to recompile them.
1911    
1912    22. Added "Any" and "L&" to the supported property types.
1913    
1914    23. Recognize \x{...} as a code point specifier, even when not in UTF-8 mode,
1915        but give a compile time error if the value is greater than 0xff.
1916    
1917    24. The man pages for pcrepartial, pcreprecompile, and pcre_compile2 were
1918        accidentally not being installed or uninstalled.
1919    
1920    25. The pcre.h file was built from pcre.h.in, but the only changes that were
1921        made were to insert the current release number. This seemed silly, because
1922        it made things harder for people building PCRE on systems that don't run
1923        "configure". I have turned pcre.h into a distributed file, no longer built
1924        by "configure", with the version identification directly included. There is
1925        no longer a pcre.h.in file.
1926    
1927        However, this change necessitated a change to the pcre-config script as
1928        well. It is built from pcre-config.in, and one of the substitutions was the
1929        release number. I have updated configure.ac so that ./configure now finds
1930        the release number by grepping pcre.h.
1931    
1932    26. Added the ability to run the tests under valgrind.
1933    
1934    
1935    Version 6.4 05-Sep-05
1936    ---------------------
1937    
1938     1. Change 6.0/10/(l) to pcregrep introduced a bug that caused separator lines
1939        "--" to be printed when multiple files were scanned, even when none of the
1940        -A, -B, or -C options were used. This is not compatible with Gnu grep, so I
1941        consider it to be a bug, and have restored the previous behaviour.
1942    
1943     2. A couple of code tidies to get rid of compiler warnings.
1944    
1945     3. The pcretest program used to cheat by referring to symbols in the library
1946        whose names begin with _pcre_. These are internal symbols that are not
1947        really supposed to be visible externally, and in some environments it is
1948        possible to suppress them. The cheating is now confined to including
1949        certain files from the library's source, which is a bit cleaner.
1950    
1951     4. Renamed pcre.in as pcre.h.in to go with pcrecpp.h.in; it also makes the
1952        file's purpose clearer.
1953    
1954     5. Reorganized pcre_ucp_findchar().
1955    
1956    
1957    Version 6.3 15-Aug-05
1958    ---------------------
1959    
1960     1. The file libpcre.pc.in did not have general read permission in the tarball.
1961    
1962     2. There were some problems when building without C++ support:
1963    
1964        (a) If C++ support was not built, "make install" and "make test" still
1965            tried to test it.
1966    
1967        (b) There were problems when the value of CXX was explicitly set. Some
1968            changes have been made to try to fix these, and ...
1969    
1970        (c) --disable-cpp can now be used to explicitly disable C++ support.
1971    
1972        (d) The use of @CPP_OBJ@ directly caused a blank line preceded by a
1973            backslash in a target when C++ was disabled. This confuses some
1974            versions of "make", apparently. Using an intermediate variable solves
1975            this. (Same for CPP_LOBJ.)
1976    
1977     3. $(LINK_FOR_BUILD) now includes $(CFLAGS_FOR_BUILD) and $(LINK)
1978        (non-Windows) now includes $(CFLAGS) because these flags are sometimes
1979        necessary on certain architectures.
1980    
1981     4. Added a setting of -export-symbols-regex to the link command to remove
1982        those symbols that are exported in the C sense, but actually are local
1983        within the library, and not documented. Their names all begin with
1984        "_pcre_". This is not a perfect job, because (a) we have to except some
1985        symbols that pcretest ("illegally") uses, and (b) the facility isn't always
1986        available (and never for static libraries). I have made a note to try to
1987        find a way round (a) in the future.
1988    
1989    
1990    Version 6.2 01-Aug-05
1991    ---------------------
1992    
1993     1. There was no test for integer overflow of quantifier values. A construction
1994        such as {1111111111111111} would give undefined results. What is worse, if
1995        a minimum quantifier for a parenthesized subpattern overflowed and became
1996        negative, the calculation of the memory size went wrong. This could have
1997        led to memory overwriting.
1998    
1999     2. Building PCRE using VPATH was broken. Hopefully it is now fixed.
2000    
2001     3. Added "b" to the 2nd argument of fopen() in dftables.c, for non-Unix-like
2002        operating environments where this matters.
2003    
2004     4. Applied Giuseppe Maxia's patch to add additional features for controlling
2005        PCRE options from within the C++ wrapper.
2006    
2007     5. Named capturing subpatterns were not being correctly counted when a pattern
2008        was compiled. This caused two problems: (a) If there were more than 100
2009        such subpatterns, the calculation of the memory needed for the whole
2010        compiled pattern went wrong, leading to an overflow error. (b) Numerical
2011        back references of the form \12, where the number was greater than 9, were
2012        not recognized as back references, even though there were sufficient
2013        previous subpatterns.
2014    
2015     6. Two minor patches to pcrecpp.cc in order to allow it to compile on older
2016        versions of gcc, e.g. 2.95.4.
2017    
2018    
2019    Version 6.1 21-Jun-05
2020    ---------------------
2021    
2022     1. There was one reference to the variable "posix" in pcretest.c that was not
2023        surrounded by "#if !defined NOPOSIX".
2024    
2025     2. Make it possible to compile pcretest without DFA support, UTF8 support, or
2026        the cross-check on the old pcre_info() function, for the benefit of the
2027        cut-down version of PCRE that is currently imported into Exim.
2028    
2029     3. A (silly) pattern starting with (?i)(?-i) caused an internal space
2030        allocation error. I've done the easy fix, which wastes 2 bytes for sensible
2031        patterns that start (?i) but I don't think that matters. The use of (?i) is
2032        just an example; this all applies to the other options as well.
2033    
2034     4. Since libtool seems to echo the compile commands it is issuing, the output
2035        from "make" can be reduced a bit by putting "@" in front of each libtool
2036        compile command.
2037    
2038     5. Patch from the folks at Google for configure.in to be a bit more thorough
2039        in checking for a suitable C++ installation before trying to compile the
2040        C++ stuff. This should fix a reported problem when a compiler was present,
2041        but no suitable headers.
2042    
2043     6. The man pages all had just "PCRE" as their title. I have changed them to
2044        be the relevant file name. I have also arranged that these names are
2045        retained in the file doc/pcre.txt, which is a concatenation in text format
2046        of all the man pages except the little individual ones for each function.
2047    
2048     7. The NON-UNIX-USE file had not been updated for the different set of source
2049        files that come with release 6. I also added a few comments about the C++
2050        wrapper.
2051    
2052    
2053    Version 6.0 07-Jun-05
2054    ---------------------
2055    
2056     1. Some minor internal re-organization to help with my DFA experiments.
2057    
2058     2. Some missing #ifdef SUPPORT_UCP conditionals in pcretest and printint that
2059        didn't matter for the library itself when fully configured, but did matter
2060        when compiling without UCP support, or within Exim, where the ucp files are
2061        not imported.
2062    
2063     3. Refactoring of the library code to split up the various functions into
2064        different source modules. The addition of the new DFA matching code (see
2065        below) to a single monolithic source would have made it really too
2066        unwieldy, quite apart from causing all the code to be include in a
2067        statically linked application, when only some functions are used. This is
2068        relevant even without the DFA addition now that patterns can be compiled in
2069        one application and matched in another.
2070    
2071        The downside of splitting up is that there have to be some external
2072        functions and data tables that are used internally in different modules of
2073        the library but which are not part of the API. These have all had their
2074        names changed to start with "_pcre_" so that they are unlikely to clash
2075        with other external names.
2076    
2077     4. Added an alternate matching function, pcre_dfa_exec(), which matches using
2078        a different (DFA) algorithm. Although it is slower than the original
2079        function, it does have some advantages for certain types of matching
2080        problem.
2081    
2082     5. Upgrades to pcretest in order to test the features of pcre_dfa_exec(),
2083        including restarting after a partial match.
2084    
2085     6. A patch for pcregrep that defines INVALID_FILE_ATTRIBUTES if it is not
2086        defined when compiling for Windows was sent to me. I have put it into the
2087        code, though I have no means of testing or verifying it.
2088    
2089     7. Added the pcre_refcount() auxiliary function.
2090    
2091     8. Added the PCRE_FIRSTLINE option. This constrains an unanchored pattern to
2092        match before or at the first newline in the subject string. In pcretest,
2093        the /f option on a pattern can be used to set this.
2094    
2095     9. A repeated \w when used in UTF-8 mode with characters greater than 256
2096        would behave wrongly. This has been present in PCRE since release 4.0.
2097    
2098    10. A number of changes to the pcregrep command:
2099    
2100        (a) Refactored how -x works; insert ^(...)$ instead of setting
2101            PCRE_ANCHORED and checking the length, in preparation for adding
2102            something similar for -w.
2103    
2104        (b) Added the -w (match as a word) option.
2105    
2106        (c) Refactored the way lines are read and buffered so as to have more
2107            than one at a time available.
2108    
2109        (d) Implemented a pcregrep test script.
2110    
2111        (e) Added the -M (multiline match) option. This allows patterns to match
2112            over several lines of the subject. The buffering ensures that at least
2113            8K, or the rest of the document (whichever is the shorter) is available
2114            for matching (and similarly the previous 8K for lookbehind assertions).
2115    
2116        (f) Changed the --help output so that it now says
2117    
2118              -w, --word-regex(p)
2119    
2120            instead of two lines, one with "regex" and the other with "regexp"
2121            because that confused at least one person since the short forms are the
2122            same. (This required a bit of code, as the output is generated
2123            automatically from a table. It wasn't just a text change.)
2124    
2125        (g) -- can be used to terminate pcregrep options if the next thing isn't an
2126            option but starts with a hyphen. Could be a pattern or a path name
2127            starting with a hyphen, for instance.
2128    
2129        (h) "-" can be given as a file name to represent stdin.
2130    
2131        (i) When file names are being printed, "(standard input)" is used for
2132            the standard input, for compatibility with GNU grep. Previously
2133            "<stdin>" was used.
2134    
2135        (j) The option --label=xxx can be used to supply a name to be used for
2136            stdin when file names are being printed. There is no short form.
2137    
2138        (k) Re-factored the options decoding logic because we are going to add
2139            two more options that take data. Such options can now be given in four
2140            different ways, e.g. "-fname", "-f name", "--file=name", "--file name".
2141    
2142        (l) Added the -A, -B, and -C options for requesting that lines of context
2143            around matches be printed.
2144    
2145        (m) Added the -L option to print the names of files that do not contain
2146            any matching lines, that is, the complement of -l.
2147    
2148        (n) The return code is 2 if any file cannot be opened, but pcregrep does
2149            continue to scan other files.
2150    
2151        (o) The -s option was incorrectly implemented. For compatibility with other
2152            greps, it now suppresses the error message for a non-existent or non-
2153            accessible file (but not the return code). There is a new option called
2154            -q that suppresses the output of matching lines, which was what -s was
2155            previously doing.
2156    
2157        (p) Added --include and --exclude options to specify files for inclusion
2158            and exclusion when recursing.
2159    
2160    11. The Makefile was not using the Autoconf-supported LDFLAGS macro properly.
2161        Hopefully, it now does.
2162    
2163    12. Missing cast in pcre_study().
2164    
2165    13. Added an "uninstall" target to the makefile.
2166    
2167    14. Replaced "extern" in the function prototypes in Makefile.in with
2168        "PCRE_DATA_SCOPE", which defaults to 'extern' or 'extern "C"' in the Unix
2169        world, but is set differently for Windows.
2170    
2171    15. Added a second compiling function called pcre_compile2(). The only
2172        difference is that it has an extra argument, which is a pointer to an
2173        integer error code. When there is a compile-time failure, this is set
2174        non-zero, in addition to the error test pointer being set to point to an
2175        error message. The new argument may be NULL if no error number is required
2176        (but then you may as well call pcre_compile(), which is now just a
2177        wrapper). This facility is provided because some applications need a
2178        numeric error indication, but it has also enabled me to tidy up the way
2179        compile-time errors are handled in the POSIX wrapper.
2180    
2181    16. Added VPATH=.libs to the makefile; this should help when building with one
2182        prefix path and installing with another. (Or so I'm told by someone who
2183        knows more about this stuff than I do.)
2184    
2185    17. Added a new option, REG_DOTALL, to the POSIX function regcomp(). This
2186        passes PCRE_DOTALL to the pcre_compile() function, making the "." character
2187        match everything, including newlines. This is not POSIX-compatible, but
2188        somebody wanted the feature. From pcretest it can be activated by using
2189        both the P and the s flags.
2190    
2191    18. AC_PROG_LIBTOOL appeared twice in Makefile.in. Removed one.
2192    
2193    19. libpcre.pc was being incorrectly installed as executable.
2194    
2195    20. A couple of places in pcretest check for end-of-line by looking for '\n';
2196        it now also looks for '\r' so that it will work unmodified on Windows.
2197    
2198    21. Added Google's contributed C++ wrapper to the distribution.
2199    
2200    22. Added some untidy missing memory free() calls in pcretest, to keep
2201        Electric Fence happy when testing.
2202    
2203    
2204    
2205    Version 5.0 13-Sep-04
2206    ---------------------
2207    
2208     1. Internal change: literal characters are no longer packed up into items
2209        containing multiple characters in a single byte-string. Each character
2210        is now matched using a separate opcode. However, there may be more than one
2211        byte in the character in UTF-8 mode.
2212    
2213     2. The pcre_callout_block structure has two new fields: pattern_position and
2214        next_item_length. These contain the offset in the pattern to the next match
2215        item, and its length, respectively.
2216    
2217     3. The PCRE_AUTO_CALLOUT option for pcre_compile() requests the automatic
2218        insertion of callouts before each pattern item. Added the /C option to
2219        pcretest to make use of this.
2220    
2221     4. On the advice of a Windows user, the lines
2222    
2223          #if defined(_WIN32) || defined(WIN32)
2224          _setmode( _fileno( stdout ), 0x8000 );
2225          #endif  /* defined(_WIN32) || defined(WIN32) */
2226    
2227        have been added to the source of pcretest. This apparently does useful
2228        magic in relation to line terminators.
2229    
2230     5. Changed "r" and "w" in the calls to fopen() in pcretest to "rb" and "wb"
2231        for the benefit of those environments where the "b" makes a difference.
2232    
2233     6. The icc compiler has the same options as gcc, but "configure" doesn't seem
2234        to know about it. I have put a hack into configure.in that adds in code
2235        to set GCC=yes if CC=icc. This seems to end up at a point in the
2236        generated configure script that is early enough to affect the setting of
2237        compiler options, which is what is needed, but I have no means of testing
2238        whether it really works. (The user who reported this had patched the
2239        generated configure script, which of course I cannot do.)
2240    
2241        LATER: After change 22 below (new libtool files), the configure script
2242        seems to know about icc (and also ecc). Therefore, I have commented out
2243        this hack in configure.in.
2244    
2245     7. Added support for pkg-config (2 patches were sent in).
2246    
2247     8. Negated POSIX character classes that used a combination of internal tables
2248        were completely broken. These were [[:^alpha:]], [[:^alnum:]], and
2249        [[:^ascii]]. Typically, they would match almost any characters. The other
2250        POSIX classes were not broken in this way.
2251    
2252     9. Matching the pattern "\b.*?" against "ab cd", starting at offset 1, failed
2253        to find the match, as PCRE was deluded into thinking that the match had to
2254        start at the start point or following a newline. The same bug applied to
2255        patterns with negative forward assertions or any backward assertions
2256        preceding ".*" at the start, unless the pattern required a fixed first
2257        character. This was a failing pattern: "(?!.bcd).*". The bug is now fixed.
2258    
2259    10. In UTF-8 mode, when moving forwards in the subject after a failed match
2260        starting at the last subject character, bytes beyond the end of the subject
2261        string were read.
2262    
2263    11. Renamed the variable "class" as "classbits" to make life easier for C++
2264        users. (Previously there was a macro definition, but it apparently wasn't
2265        enough.)
2266    
2267    12. Added the new field "tables" to the extra data so that tables can be passed
2268        in at exec time, or the internal tables can be re-selected. This allows
2269        a compiled regex to be saved and re-used at a later time by a different
2270        program that might have everything at different addresses.
2271    
2272    13. Modified the pcre-config script so that, when run on Solaris, it shows a
2273        -R library as well as a -L library.
2274    
2275    14. The debugging options of pcretest (-d on the command line or D on a
2276        pattern) showed incorrect output for anything following an extended class
2277        that contained multibyte characters and which was followed by a quantifier.
2278    
2279    15. Added optional support for general category Unicode character properties
2280        via the \p, \P, and \X escapes. Unicode property support implies UTF-8
2281        support. It adds about 90K to the size of the library. The meanings of the
2282        inbuilt class escapes such as \d and \s have NOT been changed.
2283    
2284    16. Updated pcredemo.c to include calls to free() to release the memory for the
2285        compiled pattern.
2286    
2287    17. The generated file chartables.c was being created in the source directory
2288        instead of in the building directory. This caused the build to fail if the
2289        source directory was different from the building directory, and was
2290        read-only.
2291    
2292    18. Added some sample Win commands from Mark Tetrode into the NON-UNIX-USE
2293        file. No doubt somebody will tell me if they don't make sense... Also added
2294        Dan Mooney's comments about building on OpenVMS.
2295    
2296    19. Added support for partial matching via the PCRE_PARTIAL option for
2297        pcre_exec() and the \P data escape in pcretest.
2298    
2299    20. Extended pcretest with 3 new pattern features:
2300    
2301        (i)   A pattern option of the form ">rest-of-line" causes pcretest to
2302              write the compiled pattern to the file whose name is "rest-of-line".
2303              This is a straight binary dump of the data, with the saved pointer to
2304              the character tables forced to be NULL. The study data, if any, is
2305              written too. After writing, pcretest reads a new pattern.
2306    
2307        (ii)  If, instead of a pattern, "<rest-of-line" is given, pcretest reads a
2308              compiled pattern from the given file. There must not be any
2309              occurrences of "<" in the file name (pretty unlikely); if there are,
2310              pcretest will instead treat the initial "<" as a pattern delimiter.
2311              After reading in the pattern, pcretest goes on to read data lines as
2312              usual.
2313    
2314        (iii) The F pattern option causes pcretest to flip the bytes in the 32-bit
2315              and 16-bit fields in a compiled pattern, to simulate a pattern that
2316              was compiled on a host of opposite endianness.
2317    
2318    21. The pcre-exec() function can now cope with patterns that were compiled on
2319        hosts of opposite endianness, with this restriction:
2320    
2321          As for any compiled expression that is saved and used later, the tables
2322          pointer field cannot be preserved; the extra_data field in the arguments
2323          to pcre_exec() should be used to pass in a tables address if a value
2324          other than the default internal tables were used at compile time.
2325    
2326    22. Calling pcre_exec() with a negative value of the "ovecsize" parameter is
2327        now diagnosed as an error. Previously, most of the time, a negative number
2328        would have been treated as zero, but if in addition "ovector" was passed as
2329        NULL, a crash could occur.
2330    
2331    23. Updated the files ltmain.sh, config.sub, config.guess, and aclocal.m4 with
2332        new versions from the libtool 1.5 distribution (the last one is a copy of
2333        a file called libtool.m4). This seems to have fixed the need to patch
2334        "configure" to support Darwin 1.3 (which I used to do). However, I still
2335        had to patch ltmain.sh to ensure that ${SED} is set (it isn't on my
2336        workstation).
2337    
2338    24. Changed the PCRE licence to be the more standard "BSD" licence.
2339    
2340    
2341    Version 4.5 01-Dec-03
2342    ---------------------
2343    
2344     1. There has been some re-arrangement of the code for the match() function so
2345        that it can be compiled in a version that does not call itself recursively.
2346        Instead, it keeps those local variables that need separate instances for
2347        each "recursion" in a frame on the heap, and gets/frees frames whenever it
2348        needs to "recurse". Keeping track of where control must go is done by means
2349        of setjmp/longjmp. The whole thing is implemented by a set of macros that
2350        hide most of the details from the main code, and operates only if
2351        NO_RECURSE is defined while compiling pcre.c. If PCRE is built using the
2352        "configure" mechanism, "--disable-stack-for-recursion" turns on this way of
2353        operating.
2354    
2355        To make it easier for callers to provide specially tailored get/free
2356        functions for this usage, two new functions, pcre_stack_malloc, and
2357        pcre_stack_free, are used. They are always called in strict stacking order,
2358        and the size of block requested is always the same.
2359    
2360        The PCRE_CONFIG_STACKRECURSE info parameter can be used to find out whether
2361        PCRE has been compiled to use the stack or the heap for recursion. The
2362        -C option of pcretest uses this to show which version is compiled.
2363    
2364        A new data escape \S, is added to pcretest; it causes the amounts of store
2365        obtained and freed by both kinds of malloc/free at match time to be added
2366        to the output.
2367    
2368     2. Changed the locale test to use "fr_FR" instead of "fr" because that's
2369        what's available on my current Linux desktop machine.
2370    
2371     3. When matching a UTF-8 string, the test for a valid string at the start has
2372        been extended. If start_offset is not zero, PCRE now checks that it points
2373        to a byte that is the start of a UTF-8 character. If not, it returns
2374        PCRE_ERROR_BADUTF8_OFFSET (-11). Note: the whole string is still checked;
2375        this is necessary because there may be backward assertions in the pattern.
2376        When matching the same subject several times, it may save resources to use
2377        PCRE_NO_UTF8_CHECK on all but the first call if the string is long.
2378    
2379     4. The code for checking the validity of UTF-8 strings has been tightened so
2380        that it rejects (a) strings containing 0xfe or 0xff bytes and (b) strings
2381        containing "overlong sequences".
2382    
2383     5. Fixed a bug (appearing twice) that I could not find any way of exploiting!
2384        I had written "if ((digitab[*p++] && chtab_digit) == 0)" where the "&&"
2385        should have been "&", but it just so happened that all the cases this let
2386        through by mistake were picked up later in the function.
2387    
2388     6. I had used a variable called "isblank" - this is a C99 function, causing
2389        some compilers to warn. To avoid this, I renamed it (as "blankclass").
2390    
2391     7. Cosmetic: (a) only output another newline at the end of pcretest if it is
2392        prompting; (b) run "./pcretest /dev/null" at the start of the test script
2393        so the version is shown; (c) stop "make test" echoing "./RunTest".
2394    
2395     8. Added patches from David Burgess to enable PCRE to run on EBCDIC systems.
2396    
2397     9. The prototype for memmove() for systems that don't have it was using
2398        size_t, but the inclusion of the header that defines size_t was later. I've
2399        moved the #includes for the C headers earlier to avoid this.
2400    
2401    10. Added some adjustments to the code to make it easier to compiler on certain
2402        special systems:
2403    
2404          (a) Some "const" qualifiers were missing.
2405          (b) Added the macro EXPORT before all exported functions; by default this
2406              is defined to be empty.
2407          (c) Changed the dftables auxiliary program (that builds chartables.c) so
2408              that it reads its output file name as an argument instead of writing
2409              to the standard output and assuming this can be redirected.
2410    
2411    11. In UTF-8 mode, if a recursive reference (e.g. (?1)) followed a character
2412        class containing characters with values greater than 255, PCRE compilation
2413        went into a loop.
2414    
2415    12. A recursive reference to a subpattern that was within another subpattern
2416        that had a minimum quantifier of zero caused PCRE to crash. For example,
2417        (x(y(?2))z)? provoked this bug with a subject that got as far as the
2418        recursion. If the recursively-called subpattern itself had a zero repeat,
2419        that was OK.
2420    
2421    13. In pcretest, the buffer for reading a data line was set at 30K, but the
2422        buffer into which it was copied (for escape processing) was still set at
2423        1024, so long lines caused crashes.
2424    
2425    14. A pattern such as /[ab]{1,3}+/ failed to compile, giving the error
2426        "internal error: code overflow...". This applied to any character class
2427        that was followed by a possessive quantifier.
2428    
2429    15. Modified the Makefile to add libpcre.la as a prerequisite for
2430        libpcreposix.la because I was told this is needed for a parallel build to
2431        work.
2432    
2433    16. If a pattern that contained .* following optional items at the start was
2434        studied, the wrong optimizing data was generated, leading to matching
2435        errors. For example, studying /[ab]*.*c/ concluded, erroneously, that any
2436        matching string must start with a or b or c. The correct conclusion for
2437        this pattern is that a match can start with any character.
2438    
2439    
2440    Version 4.4 13-Aug-03
2441    ---------------------
2442    
2443     1. In UTF-8 mode, a character class containing characters with values between
2444        127 and 255 was not handled correctly if the compiled pattern was studied.
2445        In fixing this, I have also improved the studying algorithm for such
2446        classes (slightly).
2447    
2448     2. Three internal functions had redundant arguments passed to them. Removal
2449        might give a very teeny performance improvement.
2450    
2451     3. Documentation bug: the value of the capture_top field in a callout is *one
2452        more than* the number of the hightest numbered captured substring.
2453    
2454     4. The Makefile linked pcretest and pcregrep with -lpcre, which could result
2455        in incorrectly linking with a previously installed version. They now link
2456        explicitly with libpcre.la.
2457    
2458     5. configure.in no longer needs to recognize Cygwin specially.
2459    
2460     6. A problem in pcre.in for Windows platforms is fixed.
2461    
2462     7. If a pattern was successfully studied, and the -d (or /D) flag was given to
2463        pcretest, it used to include the size of the study block as part of its
2464        output. Unfortunately, the structure contains a field that has a different
2465        size on different hardware architectures. This meant that the tests that
2466        showed this size failed. As the block is currently always of a fixed size,
2467        this information isn't actually particularly useful in pcretest output, so
2468        I have just removed it.
2469    
2470     8. Three pre-processor statements accidentally did not start in column 1.
2471        Sadly, there are *still* compilers around that complain, even though
2472        standard C has not required this for well over a decade. Sigh.
2473    
2474     9. In pcretest, the code for checking callouts passed small integers in the
2475        callout_data field, which is a void * field. However, some picky compilers
2476        complained about the casts involved for this on 64-bit systems. Now
2477        pcretest passes the address of the small integer instead, which should get
2478        rid of the warnings.
2479    
2480    10. By default, when in UTF-8 mode, PCRE now checks for valid UTF-8 strings at
2481        both compile and run time, and gives an error if an invalid UTF-8 sequence
2482        is found. There is a option for disabling this check in cases where the
2483        string is known to be correct and/or the maximum performance is wanted.
2484    
2485    11. In response to a bug report, I changed one line in Makefile.in from
2486    
2487            -Wl,--out-implib,.libs/lib@WIN_PREFIX@pcreposix.dll.a \
2488        to
2489            -Wl,--out-implib,.libs/@WIN_PREFIX@libpcreposix.dll.a \
2490    
2491        to look similar to other lines, but I have no way of telling whether this
2492        is the right thing to do, as I do not use Windows. No doubt I'll get told
2493        if it's wrong...
2494    
2495    
2496    Version 4.3 21-May-03
2497    ---------------------
2498    
2499    1. Two instances of @WIN_PREFIX@ omitted from the Windows targets in the
2500       Makefile.
2501    
2502    2. Some refactoring to improve the quality of the code:
2503    
2504       (i)   The utf8_table... variables are now declared "const".
2505    
2506       (ii)  The code for \cx, which used the "case flipping" table to upper case
2507             lower case letters, now just substracts 32. This is ASCII-specific,
2508             but the whole concept of \cx is ASCII-specific, so it seems
2509             reasonable.
2510    
2511       (iii) PCRE was using its character types table to recognize decimal and
2512             hexadecimal digits in the pattern. This is silly, because it handles
2513             only 0-9, a-f, and A-F, but the character types table is locale-
2514             specific, which means strange things might happen. A private
2515             table is now used for this - though it costs 256 bytes, a table is
2516             much faster than multiple explicit tests. Of course, the standard
2517             character types table is still used for matching digits in subject
2518             strings against \d.
2519    
2520       (iv)  Strictly, the identifier ESC_t is reserved by POSIX (all identifiers
2521             ending in _t are). So I've renamed it as ESC_tee.
2522    
2523    3. The first argument for regexec() in the POSIX wrapper should have been
2524       defined as "const".
2525    
2526    4. Changed pcretest to use malloc() for its buffers so that they can be
2527       Electric Fenced for debugging.
2528    
2529    5. There were several places in the code where, in UTF-8 mode, PCRE would try
2530       to read one or more bytes before the start of the subject string. Often this
2531       had no effect on PCRE's behaviour, but in some circumstances it could
2532       provoke a segmentation fault.
2533    
2534    6. A lookbehind at the start of a pattern in UTF-8 mode could also cause PCRE
2535       to try to read one or more bytes before the start of the subject string.
2536    
2537    7. A lookbehind in a pattern matched in non-UTF-8 mode on a PCRE compiled with
2538       UTF-8 support could misbehave in various ways if the subject string
2539       contained bytes with the 0x80 bit set and the 0x40 bit unset in a lookbehind
2540       area. (PCRE was not checking for the UTF-8 mode flag, and trying to move
2541       back over UTF-8 characters.)
2542    
2543    
2544    Version 4.2 14-Apr-03
2545    ---------------------
2546    
2547    1. Typo "#if SUPPORT_UTF8" instead of "#ifdef SUPPORT_UTF8" fixed.
2548    
2549    2. Changes to the building process, supplied by Ronald Landheer-Cieslak
2550         [ON_WINDOWS]: new variable, "#" on non-Windows platforms
2551         [NOT_ON_WINDOWS]: new variable, "#" on Windows platforms
2552         [WIN_PREFIX]: new variable, "cyg" for Cygwin
2553         * Makefile.in: use autoconf substitution for OBJEXT, EXEEXT, BUILD_OBJEXT
2554           and BUILD_EXEEXT
2555         Note: automatic setting of the BUILD variables is not yet working
2556         set CPPFLAGS and BUILD_CPPFLAGS (but don't use yet) - should be used at
2557           compile-time but not at link-time
2558         [LINK]: use for linking executables only
2559         make different versions for Windows and non-Windows
2560         [LINKLIB]: new variable, copy of UNIX-style LINK, used for linking
2561           libraries
2562         [LINK_FOR_BUILD]: new variable
2563         [OBJEXT]: use throughout
2564         [EXEEXT]: use throughout
2565         <winshared>: new target
2566         <wininstall>: new target
2567         <dftables.o>: use native compiler
2568         <dftables>: use native linker
2569         <install>: handle Windows platform correctly
2570         <clean>: ditto
2571         <check>: ditto
2572         copy DLL to top builddir before testing
2573    
2574       As part of these changes, -no-undefined was removed again. This was reported
2575       to give trouble on HP-UX 11.0, so getting rid of it seems like a good idea
2576       in any case.
2577    
2578    3. Some tidies to get rid of compiler warnings:
2579    
2580       . In the match_data structure, match_limit was an unsigned long int, whereas
2581         match_call_count was an int. I've made them both unsigned long ints.
2582    
2583       . In pcretest the fact that a const uschar * doesn't automatically cast to
2584         a void * provoked a warning.
2585    
2586       . Turning on some more compiler warnings threw up some "shadow" variables
2587         and a few more missing casts.
2588    
2589    4. If PCRE was complied with UTF-8 support, but called without the PCRE_UTF8
2590       option, a class that contained a single character with a value between 128
2591       and 255 (e.g. /[\xFF]/) caused PCRE to crash.
2592    
2593    5. If PCRE was compiled with UTF-8 support, but called without the PCRE_UTF8
2594       option, a class that contained several characters, but with at least one
2595       whose value was between 128 and 255 caused PCRE to crash.
2596    
2597    
2598    Version 4.1 12-Mar-03
2599    ---------------------
2600    
2601    1. Compiling with gcc -pedantic found a couple of places where casts were
2602    needed, and a string in dftables.c that was longer than standard compilers are
2603    required to support.
2604    
2605    2. Compiling with Sun's compiler found a few more places where the code could
2606    be tidied up in order to avoid warnings.
2607    
2608    3. The variables for cross-compiling were called HOST_CC and HOST_CFLAGS; the
2609    first of these names is deprecated in the latest Autoconf in favour of the name
2610    CC_FOR_BUILD, because "host" is typically used to mean the system on which the
2611    compiled code will be run. I can't find a reference for HOST_CFLAGS, but by
2612    analogy I have changed it to CFLAGS_FOR_BUILD.
2613    
2614    4. Added -no-undefined to the linking command in the Makefile, because this is
2615    apparently helpful for Windows. To make it work, also added "-L. -lpcre" to the
2616    linking step for the pcreposix library.
2617    
2618    5. PCRE was failing to diagnose the case of two named groups with the same
2619    name.
2620    
2621    6. A problem with one of PCRE's optimizations was discovered. PCRE remembers a
2622    literal character that is needed in the subject for a match, and scans along to
2623    ensure that it is present before embarking on the full matching process. This
2624    saves time in cases of nested unlimited repeats that are never going to match.
2625    Problem: the scan can take a lot of time if the subject is very long (e.g.
2626    megabytes), thus penalizing straightforward matches. It is now done only if the
2627    amount of subject to be scanned is less than 1000 bytes.
2628    
2629    7. A lesser problem with the same optimization is that it was recording the
2630    first character of an anchored pattern as "needed", thus provoking a search
2631    right along the subject, even when the first match of the pattern was going to
2632    fail. The "needed" character is now not set for anchored patterns, unless it
2633    follows something in the pattern that is of non-fixed length. Thus, it still
2634    fulfils its original purpose of finding quick non-matches in cases of nested
2635    unlimited repeats, but isn't used for simple anchored patterns such as /^abc/.
2636    
2637    
2638    Version 4.0 17-Feb-03
2639    ---------------------
2640    
2641    1. If a comment in an extended regex that started immediately after a meta-item
2642    extended to the end of string, PCRE compiled incorrect data. This could lead to
2643    all kinds of weird effects. Example: /#/ was bad; /()#/ was bad; /a#/ was not.
2644    
2645    2. Moved to autoconf 2.53 and libtool 1.4.2.
2646    
2647    3. Perl 5.8 no longer needs "use utf8" for doing UTF-8 things. Consequently,
2648    the special perltest8 script is no longer needed - all the tests can be run
2649    from a single perltest script.
2650    
2651    4. From 5.004, Perl has not included the VT character (0x0b) in the set defined
2652    by \s. It has now been removed in PCRE. This means it isn't recognized as
2653    whitespace in /x regexes too, which is the same as Perl. Note that the POSIX
2654    class [:space:] *does* include VT, thereby creating a mess.
2655    
2656    5. Added the class [:blank:] (a GNU extension from Perl 5.8) to match only
2657    space and tab.
2658    
2659    6. Perl 5.005 was a long time ago. It's time to amalgamate the tests that use
2660    its new features into the main test script, reducing the number of scripts.
2661    
2662    7. Perl 5.8 has changed the meaning of patterns like /a(?i)b/. Earlier versions
2663    were backward compatible, and made the (?i) apply to the whole pattern, as if
2664    /i were given. Now it behaves more logically, and applies the option setting
2665    only to what follows. PCRE has been changed to follow suit. However, if it
2666    finds options settings right at the start of the pattern, it extracts them into
2667    the global options, as before. Thus, they show up in the info data.
2668    
2669    8. Added support for the \Q...\E escape sequence. Characters in between are
2670    treated as literals. This is slightly different from Perl in that $ and @ are
2671    also handled as literals inside the quotes. In Perl, they will cause variable
2672    interpolation. Note the following examples:
2673    
2674        Pattern            PCRE matches      Perl matches
2675    
2676        \Qabc$xyz\E        abc$xyz           abc followed by the contents of $xyz
2677        \Qabc\$xyz\E       abc\$xyz          abc\$xyz
2678        \Qabc\E\$\Qxyz\E   abc$xyz           abc$xyz
2679    
2680    For compatibility with Perl, \Q...\E sequences are recognized inside character
2681    classes as well as outside them.
2682    
2683    9. Re-organized 3 code statements in pcretest to avoid "overflow in
2684    floating-point constant arithmetic" warnings from a Microsoft compiler. Added a
2685    (size_t) cast to one statement in pcretest and one in pcreposix to avoid
2686    signed/unsigned warnings.
2687    
2688    10. SunOS4 doesn't have strtoul(). This was used only for unpicking the -o
2689    option for pcretest, so I've replaced it by a simple function that does just
2690    that job.
2691    
2692    11. pcregrep was ending with code 0 instead of 2 for the commands "pcregrep" or
2693    "pcregrep -".
2694    
2695    12. Added "possessive quantifiers" ?+, *+, ++, and {,}+ which come from Sun's
2696    Java package. This provides some syntactic sugar for simple cases of what my
2697    documentation calls "once-only subpatterns". A pattern such as x*+ is the same
2698    as (?>x*). In other words, if what is inside (?>...) is just a single repeated
2699    item, you can use this simplified notation. Note that only makes sense with
2700    greedy quantifiers. Consequently, the use of the possessive quantifier forces
2701    greediness, whatever the setting of the PCRE_UNGREEDY option.
2702    
2703    13. A change of greediness default within a pattern was not taking effect at
2704    the current level for patterns like /(b+(?U)a+)/. It did apply to parenthesized
2705    subpatterns that followed. Patterns like /b+(?U)a+/ worked because the option
2706    was abstracted outside.
2707    
2708    14. PCRE now supports the \G assertion. It is true when the current matching
2709    position is at the start point of the match. This differs from \A when the
2710    starting offset is non-zero. Used with the /g option of pcretest (or similar
2711    code), it works in the same way as it does for Perl's /g option. If all
2712    alternatives of a regex begin with \G, the expression is anchored to the start
2713    match position, and the "anchored" flag is set in the compiled expression.
2714    
2715    15. Some bugs concerning the handling of certain option changes within patterns
2716    have been fixed. These applied to options other than (?ims). For example,
2717    "a(?x: b c )d" did not match "XabcdY" but did match "Xa b c dY". It should have
2718    been the other way round. Some of this was related to change 7 above.
2719    
2720    16. PCRE now gives errors for /[.x.]/ and /[=x=]/ as unsupported POSIX
2721    features, as Perl does. Previously, PCRE gave the warnings only for /[[.x.]]/
2722    and /[[=x=]]/. PCRE now also gives an error for /[:name:]/ because it supports
2723    POSIX classes only within a class (e.g. /[[:alpha:]]/).
2724    
2725    17. Added support for Perl's \C escape. This matches one byte, even in UTF8
2726    mode. Unlike ".", it always matches newline, whatever the setting of
2727    PCRE_DOTALL. However, PCRE does not permit \C to appear in lookbehind
2728    assertions. Perl allows it, but it doesn't (in general) work because it can't
2729    calculate the length of the lookbehind. At least, that's the case for Perl
2730    5.8.0 - I've been told they are going to document that it doesn't work in
2731    future.
2732    
2733    18. Added an error diagnosis for escapes that PCRE does not support: these are
2734    \L, \l, \N, \P, \p, \U, \u, and \X.
2735    
2736    19. Although correctly diagnosing a missing ']' in a character class, PCRE was
2737    reading past the end of the pattern in cases such as /[abcd/.
2738    
2739    20. PCRE was getting more memory than necessary for patterns with classes that
2740    contained both POSIX named classes and other characters, e.g. /[[:space:]abc/.
2741    
2742    21. Added some code, conditional on #ifdef VPCOMPAT, to make life easier for
2743    compiling PCRE for use with Virtual Pascal.
2744    
2745    22. Small fix to the Makefile to make it work properly if the build is done
2746    outside the source tree.
2747    
2748    23. Added a new extension: a condition to go with recursion. If a conditional
2749    subpattern starts with (?(R) the "true" branch is used if recursion has
2750    happened, whereas the "false" branch is used only at the top level.
2751    
2752    24. When there was a very long string of literal characters (over 255 bytes
2753    without UTF support, over 250 bytes with UTF support), the computation of how
2754    much memory was required could be incorrect, leading to segfaults or other
2755    strange effects.
2756    
2757    25. PCRE was incorrectly assuming anchoring (either to start of subject or to
2758    start of line for a non-DOTALL pattern) when a pattern started with (.*) and
2759    there was a subsequent back reference to those brackets. This meant that, for
2760    example, /(.*)\d+\1/ failed to match "abc123bc". Unfortunately, it isn't
2761    possible to check for precisely this case. All we can do is abandon the
2762    optimization if .* occurs inside capturing brackets when there are any back
2763    references whatsoever. (See below for a better fix that came later.)
2764    
2765    26. The handling of the optimization for finding the first character of a
2766    non-anchored pattern, and for finding a character that is required later in the
2767    match were failing in some cases. This didn't break the matching; it just
2768    failed to optimize when it could. The way this is done has been re-implemented.
2769    
2770    27. Fixed typo in error message for invalid (?R item (it said "(?p").
2771    
2772    28. Added a new feature that provides some of the functionality that Perl
2773    provides with (?{...}). The facility is termed a "callout". The way it is done
2774    in PCRE is for the caller to provide an optional function, by setting
2775    pcre_callout to its entry point. Like pcre_malloc and pcre_free, this is a
2776    global variable. By default it is unset, which disables all calling out. To get
2777    the function called, the regex must include (?C) at appropriate points. This
2778    is, in fact, equivalent to (?C0), and any number <= 255 may be given with (?C).
2779    This provides a means of identifying different callout points. When PCRE
2780    reaches such a point in the regex, if pcre_callout has been set, the external
2781    function is called. It is provided with data in a structure called
2782    pcre_callout_block, which is defined in pcre.h. If the function returns 0,
2783    matching continues; if it returns a non-zero value, the match at the current
2784    point fails. However, backtracking will occur if possible. [This was changed
2785    later and other features added - see item 49 below.]
2786    
2787    29. pcretest is upgraded to test the callout functionality. It provides a
2788    callout function that displays information. By default, it shows the start of
2789    the match and the current position in the text. There are some new data escapes
2790    to vary what happens:
2791    
2792        \C+         in addition, show current contents of captured substrings
2793        \C-         do not supply a callout function
2794        \C!n        return 1 when callout number n is reached
2795        \C!n!m      return 1 when callout number n is reached for the mth time
2796    
2797    30. If pcregrep was called with the -l option and just a single file name, it
2798    output "<stdin>" if a match was found, instead of the file name.
2799    
2800    31. Improve the efficiency of the POSIX API to PCRE. If the number of capturing
2801    slots is less than POSIX_MALLOC_THRESHOLD, use a block on the stack to pass to
2802    pcre_exec(). This saves a malloc/free per call. The default value of
2803    POSIX_MALLOC_THRESHOLD is 10; it can be changed by --with-posix-malloc-threshold
2804    when configuring.
2805    
2806    32. The default maximum size of a compiled pattern is 64K. There have been a
2807    few cases of people hitting this limit. The code now uses macros to handle the
2808    storing of links as offsets within the compiled pattern. It defaults to 2-byte
2809    links, but this can be changed to 3 or 4 bytes by --with-link-size when
2810    configuring. Tests 2 and 5 work only with 2-byte links because they output
2811    debugging information about compiled patterns.
2812    
2813    33. Internal code re-arrangements:
2814    
2815    (a) Moved the debugging function for printing out a compiled regex into
2816        its own source file (printint.c) and used #include to pull it into
2817        pcretest.c and, when DEBUG is defined, into pcre.c, instead of having two
2818        separate copies.
2819    
2820    (b) Defined the list of op-code names for debugging as a macro in
2821        internal.h so that it is next to the definition of the opcodes.
2822    
2823    (c) Defined a table of op-code lengths for simpler skipping along compiled
2824        code. This is again a macro in internal.h so that it is next to the
2825        definition of the opcodes.
2826    
2827    34. Added support for recursive calls to individual subpatterns, along the
2828    lines of Robin Houston's patch (but implemented somewhat differently).
2829    
2830    35. Further mods to the Makefile to help Win32. Also, added code to pcregrep to
2831    allow it to read and process whole directories in Win32. This code was
2832    contributed by Lionel Fourquaux; it has not been tested by me.
2833    
2834    36. Added support for named subpatterns. The Python syntax (?P<name>...) is
2835    used to name a group. Names consist of alphanumerics and underscores, and must
2836    be unique. Back references use the syntax (?P=name) and recursive calls use
2837    (?P>name) which is a PCRE extension to the Python extension. Groups still have
2838    numbers. The function pcre_fullinfo() can be used after compilation to extract
2839    a name/number map. There are three relevant calls:
2840    
2841      PCRE_INFO_NAMEENTRYSIZE        yields the size of each entry in the map
2842      PCRE_INFO_NAMECOUNT            yields the number of entries
2843      PCRE_INFO_NAMETABLE            yields a pointer to the map.
2844    
2845    The map is a vector of fixed-size entries. The size of each entry depends on
2846    the length of the longest name used. The first two bytes of each entry are the
2847    group number, most significant byte first. There follows the corresponding
2848    name, zero terminated. The names are in alphabetical order.
2849    
2850    37. Make the maximum literal string in the compiled code 250 for the non-UTF-8
2851    case instead of 255. Making it the same both with and without UTF-8 support
2852    means that the same test output works with both.
2853    
2854    38. There was a case of malloc(0) in the POSIX testing code in pcretest. Avoid
2855    calling malloc() with a zero argument.
2856    
2857    39. Change 25 above had to resort to a heavy-handed test for the .* anchoring
2858    optimization. I've improved things by keeping a bitmap of backreferences with
2859    numbers 1-31 so that if .* occurs inside capturing brackets that are not in
2860    fact referenced, the optimization can be applied. It is unlikely that a
2861    relevant occurrence of .* (i.e. one which might indicate anchoring or forcing
2862    the match to follow \n) will appear inside brackets with a number greater than
2863    31, but if it does, any back reference > 31 suppresses the optimization.
2864    
2865    40. Added a new compile-time option PCRE_NO_AUTO_CAPTURE. This has the effect
2866    of disabling numbered capturing parentheses. Any opening parenthesis that is
2867    not followed by ? behaves as if it were followed by ?: but named parentheses
2868    can still be used for capturing (and they will acquire numbers in the usual
2869    way).
2870    
2871    41. Redesigned the return codes from the match() function into yes/no/error so
2872    that errors can be passed back from deep inside the nested calls. A malloc
2873    failure while inside a recursive subpattern call now causes the
2874    PCRE_ERROR_NOMEMORY return instead of quietly going wrong.
2875    
2876    42. It is now possible to set a limit on the number of times the match()
2877    function is called in a call to pcre_exec(). This facility makes it possible to
2878    limit the amount of recursion and backtracking, though not in a directly
2879    obvious way, because the match() function is used in a number of different
2880    circumstances. The count starts from zero for each position in the subject
2881    string (for non-anchored patterns). The default limit is, for compatibility, a
2882    large number, namely 10 000 000. You can change this in two ways:
2883    
2884    (a) When configuring PCRE before making, you can use --with-match-limit=n
2885        to set a default value for the compiled library.
2886    
2887    (b) For each call to pcre_exec(), you can pass a pcre_extra block in which
2888        a different value is set. See 45 below.
2889    
2890    If the limit is exceeded, pcre_exec() returns PCRE_ERROR_MATCHLIMIT.
2891    
2892    43. Added a new function pcre_config(int, void *) to enable run-time extraction
2893    of things that can be changed at compile time. The first argument specifies
2894    what is wanted and the second points to where the information is to be placed.
2895    The current list of available information is:
2896    
2897      PCRE_CONFIG_UTF8
2898    
2899    The output is an integer that is set to one if UTF-8 support is available;
2900    otherwise it is set to zero.
2901    
2902      PCRE_CONFIG_NEWLINE
2903    
2904    The output is an integer that it set to the value of the code that is used for
2905    newline. It is either LF (10) or CR (13).
2906    
2907      PCRE_CONFIG_LINK_SIZE
2908    
2909    The output is an integer that contains the number of bytes used for internal
2910    linkage in compiled expressions. The value is 2, 3, or 4. See item 32 above.
2911    
2912      PCRE_CONFIG_POSIX_MALLOC_THRESHOLD
2913    
2914    The output is an integer that contains the threshold above which the POSIX
2915    interface uses malloc() for output vectors. See item 31 above.
2916    
2917      PCRE_CONFIG_MATCH_LIMIT
2918    
2919    The output is an unsigned integer that contains the default limit of the number
2920    of match() calls in a pcre_exec() execution. See 42 above.
2921    
2922    44. pcretest has been upgraded by the addition of the -C option. This causes it
2923    to extract all the available output from the new pcre_config() function, and to
2924    output it. The program then exits immediately.
2925    
2926    45. A need has arisen to pass over additional data with calls to pcre_exec() in
2927    order to support additional features. One way would have been to define
2928    pcre_exec2() (for example) with extra arguments, but this would not have been
2929    extensible, and would also have required all calls to the original function to
2930    be mapped to the new one. Instead, I have chosen to extend the mechanism that
2931    is used for passing in "extra" data from pcre_study().
2932    
2933    The pcre_extra structure is now exposed and defined in pcre.h. It currently
2934    contains the following fields:
2935    
2936      flags         a bitmap indicating which of the following fields are set
2937      study_data    opaque data from pcre_study()
2938      match_limit   a way of specifying a limit on match() calls for a specific
2939                      call to pcre_exec()
2940      callout_data  data for callouts (see 49 below)
2941    
2942    The flag bits are also defined in pcre.h, and are
2943    
2944      PCRE_EXTRA_STUDY_DATA
2945      PCRE_EXTRA_MATCH_LIMIT
2946      PCRE_EXTRA_CALLOUT_DATA
2947    
2948    The pcre_study() function now returns one of these new pcre_extra blocks, with
2949    the actual study data pointed to by the study_data field, and the
2950    PCRE_EXTRA_STUDY_DATA flag set. This can be passed directly to pcre_exec() as
2951    before. That is, this change is entirely upwards-compatible and requires no
2952    change to existing code.
2953    
2954    If you want to pass in additional data to pcre_exec(), you can either place it
2955    in a pcre_extra block provided by pcre_study(), or create your own pcre_extra
2956    block.
2957    
2958    46. pcretest has been extended to test the PCRE_EXTRA_MATCH_LIMIT feature. If a
2959    data string contains the escape sequence \M, pcretest calls pcre_exec() several
2960    times with different match limits, until it finds the minimum value needed for
2961    pcre_exec() to complete. The value is then output. This can be instructive; for
2962    most simple matches the number is quite small, but for pathological cases it
2963    gets very large very quickly.
2964    
2965    47. There's a new option for pcre_fullinfo() called PCRE_INFO_STUDYSIZE. It
2966    returns the size of the data block pointed to by the study_data field in a
2967    pcre_extra block, that is, the value that was passed as the argument to
2968    pcre_malloc() when PCRE was getting memory in which to place the information
2969    created by pcre_study(). The fourth argument should point to a size_t variable.
2970    pcretest has been extended so that this information is shown after a successful
2971    pcre_study() call when information about the compiled regex is being displayed.
2972    
2973    48. Cosmetic change to Makefile: there's no need to have / after $(DESTDIR)
2974    because what follows is always an absolute path. (Later: it turns out that this
2975    is more than cosmetic for MinGW, because it doesn't like empty path
2976    components.)
2977    
2978    49. Some changes have been made to the callout feature (see 28 above):
2979    
2980    (i)  A callout function now has three choices for what it returns:
2981    
2982           0  =>  success, carry on matching
2983         > 0  =>  failure at this point, but backtrack if possible
2984         < 0  =>  serious error, return this value from pcre_exec()
2985    
2986         Negative values should normally be chosen from the set of PCRE_ERROR_xxx
2987         values. In particular, returning PCRE_ERROR_NOMATCH forces a standard
2988         "match failed" error. The error number PCRE_ERROR_CALLOUT is reserved for
2989         use by callout functions. It will never be used by PCRE itself.
2990    
2991    (ii) The pcre_extra structure (see 45 above) has a void * field called
2992         callout_data, with corresponding flag bit PCRE_EXTRA_CALLOUT_DATA. The
2993         pcre_callout_block structure has a field of the same name. The contents of
2994         the field passed in the pcre_extra structure are passed to the callout
2995         function in the corresponding field in the callout block. This makes it
2996         easier to use the same callout-containing regex from multiple threads. For
2997         testing, the pcretest program has a new data escape
2998    
2999           \C*n        pass the number n (may be negative) as callout_data
3000    
3001         If the callout function in pcretest receives a non-zero value as
3002         callout_data, it returns that value.
3003    
3004    50. Makefile wasn't handling CFLAGS properly when compiling dftables. Also,
3005    there were some redundant $(CFLAGS) in commands that are now specified as
3006    $(LINK), which already includes $(CFLAGS).
3007    
3008    51. Extensions to UTF-8 support are listed below. These all apply when (a) PCRE
3009    has been compiled with UTF-8 support *and* pcre_compile() has been compiled
3010    with the PCRE_UTF8 flag. Patterns that are compiled without that flag assume
3011    one-byte characters throughout. Note that case-insensitive matching applies
3012    only to characters whose values are less than 256. PCRE doesn't support the
3013    notion of cases for higher-valued characters.
3014    
3015    (i)   A character class whose characters are all within 0-255 is handled as
3016          a bit map, and the map is inverted for negative classes. Previously, a
3017          character > 255 always failed to match such a class; however it should
3018          match if the class was a negative one (e.g. [^ab]). This has been fixed.
3019    
3020    (ii)  A negated character class with a single character < 255 is coded as
3021          "not this character" (OP_NOT). This wasn't working properly when the test
3022          character was multibyte, either singly or repeated.
3023    
3024    (iii) Repeats of multibyte characters are now handled correctly in UTF-8
3025          mode, for example: \x{100}{2,3}.
3026    
3027    (iv)  The character escapes \b, \B, \d, \D, \s, \S, \w, and \W (either
3028          singly or repeated) now correctly test multibyte characters. However,
3029          PCRE doesn't recognize any characters with values greater than 255 as
3030          digits, spaces, or word characters. Such characters always match \D, \S,
3031          and \W, and never match \d, \s, or \w.
3032    
3033    (v)   Classes may now contain characters and character ranges with values
3034          greater than 255. For example: [ab\x{100}-\x{400}].
3035    
3036    (vi)  pcregrep now has a --utf-8 option (synonym -u) which makes it call
3037          PCRE in UTF-8 mode.
3038    
3039    52. The info request value PCRE_INFO_FIRSTCHAR has been renamed
3040    PCRE_INFO_FIRSTBYTE because it is a byte value. However, the old name is
3041    retained for backwards compatibility. (Note that LASTLITERAL is also a byte
3042    value.)
3043    
3044    53. The single man page has become too large. I have therefore split it up into
3045    a number of separate man pages. These also give rise to individual HTML pages;
3046    these are now put in a separate directory, and there is an index.html page that
3047    lists them all. Some hyperlinking between the pages has been installed.
3048    
3049    54. Added convenience functions for handling named capturing parentheses.
3050    
3051    55. Unknown escapes inside character classes (e.g. [\M]) and escapes that
3052    aren't interpreted therein (e.g. [\C]) are literals in Perl. This is now also
3053    true in PCRE, except when the PCRE_EXTENDED option is set, in which case they
3054    are faulted.
3055    
3056    56. Introduced HOST_CC and HOST_CFLAGS which can be set in the environment when
3057    calling configure. These values are used when compiling the dftables.c program
3058    which is run to generate the source of the default character tables. They
3059    default to the values of CC and CFLAGS. If you are cross-compiling PCRE,
3060    you will need to set these values.
3061    
3062    57. Updated the building process for Windows DLL, as provided by Fred Cox.
3063    
3064    
3065    Version 3.9 02-Jan-02
3066    ---------------------
3067    
3068    1. A bit of extraneous text had somehow crept into the pcregrep documentation.
3069    
3070    2. If --disable-static was given, the building process failed when trying to
3071    build pcretest and pcregrep. (For some reason it was using libtool to compile
3072    them, which is not right, as they aren't part of the library.)
3073    
3074    
3075    Version 3.8 18-Dec-01
3076    ---------------------
3077    
3078    1. The experimental UTF-8 code was completely screwed up. It was packing the
3079    bytes in the wrong order. How dumb can you get?
3080    
3081    
3082    Version 3.7 29-Oct-01
3083    ---------------------
3084    
3085    1. In updating pcretest to check change 1 of version 3.6, I screwed up.
3086    This caused pcretest, when used on the test data, to segfault. Unfortunately,
3087    this didn't happen under Solaris 8, where I normally test things.
3088    
3089    2. The Makefile had to be changed to make it work on BSD systems, where 'make'
3090    doesn't seem to recognize that ./xxx and xxx are the same file. (This entry
3091    isn't in ChangeLog distributed with 3.7 because I forgot when I hastily made
3092    this fix an hour or so after the initial 3.7 release.)
3093    
3094    
3095    Version 3.6 23-Oct-01
3096    ---------------------
3097    
3098    1. Crashed with /(sens|respons)e and \1ibility/ and "sense and sensibility" if
3099    offsets passed as NULL with zero offset count.
3100    
3101    2. The config.guess and config.sub files had not been updated when I moved to
3102    the latest autoconf.
3103    
3104    
3105    Version 3.5 15-Aug-01
3106    ---------------------
3107    
3108    1. Added some missing #if !defined NOPOSIX conditionals in pcretest.c that
3109    had been forgotten.
3110    
3111    2. By using declared but undefined structures, we can avoid using "void"
3112    definitions in pcre.h while keeping the internal definitions of the structures
3113    private.
3114    
3115    3. The distribution is now built using autoconf 2.50 and libtool 1.4. From a
3116    user point of view, this means that both static and shared libraries are built
3117    by default, but this can be individually controlled. More of the work of
3118    handling this static/shared cases is now inside libtool instead of PCRE's make
3119    file.
3120    
3121    4. The pcretest utility is now installed along with pcregrep because it is
3122    useful for users (to test regexs) and by doing this, it automatically gets
3123    relinked by libtool. The documentation has been turned into a man page, so
3124    there are now .1, .txt, and .html versions in /doc.
3125    
3126    5. Upgrades to pcregrep:
3127       (i)   Added long-form option names like gnu grep.
3128       (ii)  Added --help to list all options with an explanatory phrase.
3129       (iii) Added -r, --recursive to recurse into sub-directories.
3130       (iv)  Added -f, --file to read patterns from a file.
3131    
3132    6. pcre_exec() was referring to its "code" argument before testing that
3133    argument for NULL (and giving an error if it was NULL).
3134    
3135    7. Upgraded Makefile.in to allow for compiling in a different directory from
3136    the source directory.
3137    
3138    8. Tiny buglet in pcretest: when pcre_fullinfo() was called to retrieve the
3139    options bits, the pointer it was passed was to an int instead of to an unsigned
3140    long int. This mattered only on 64-bit systems.
3141    
3142    9. Fixed typo (3.4/1) in pcre.h again. Sigh. I had changed pcre.h (which is
3143    generated) instead of pcre.in, which it its source. Also made the same change
3144    in several of the .c files.
3145    
3146    10. A new release of gcc defines printf() as a macro, which broke pcretest
3147    because it had an ifdef in the middle of a string argument for printf(). Fixed
3148    by using separate calls to printf().
3149    
3150    11. Added --enable-newline-is-cr and --enable-newline-is-lf to the configure
3151    script, to force use of CR or LF instead of \n in the source. On non-Unix
3152    systems, the value can be set in config.h.
3153    
3154    12. The limit of 200 on non-capturing parentheses is a _nesting_ limit, not an
3155    absolute limit. Changed the text of the error message to make this clear, and
3156    likewise updated the man page.
3157    
3158    13. The limit of 99 on the number of capturing subpatterns has been removed.
3159    The new limit is 65535, which I hope will not be a "real" limit.
3160    
3161    
3162    Version 3.4 22-Aug-00
3163    ---------------------
3164    
3165    1. Fixed typo in pcre.h: unsigned const char * changed to const unsigned char *.
3166    
3167    2. Diagnose condition (?(0) as an error instead of crashing on matching.
3168    
3169    
3170    Version 3.3 01-Aug-00
3171    ---------------------
3172    
3173    1. If an octal character was given, but the value was greater than \377, it
3174    was not getting masked to the least significant bits, as documented. This could
3175    lead to crashes in some systems.
3176    
3177    2. Perl 5.6 (if not earlier versions) accepts classes like [a-\d] and treats
3178    the hyphen as a literal. PCRE used to give an error; it now behaves like Perl.
3179    
3180    3. Added the functions pcre_free_substring() and pcre_free_substring_list().
3181    These just pass their arguments on to (pcre_free)(), but they are provided
3182    because some uses of PCRE bind it to non-C systems that can call its functions,
3183    but cannot call free() or pcre_free() directly.
3184    
3185    4. Add "make test" as a synonym for "make check". Corrected some comments in
3186    the Makefile.
3187    
3188    5. Add $(DESTDIR)/ in front of all the paths in the "install" target in the
3189    Makefile.
3190    
3191    6. Changed the name of pgrep to pcregrep, because Solaris has introduced a
3192    command called pgrep for grepping around the active processes.
3193    
3194    7. Added the beginnings of support for UTF-8 character strings.
3195    
3196    8. Arranged for the Makefile to pass over the settings of CC, CFLAGS, and
3197    RANLIB to ./ltconfig so that they are used by libtool. I think these are all
3198    the relevant ones. (AR is not passed because ./ltconfig does its own figuring
3199    out for the ar command.)
3200    
3201    
3202    Version 3.2 12-May-00
3203    ---------------------
3204    
3205    This is purely a bug fixing release.
3206    
3207    1. If the pattern /((Z)+|A)*/ was matched agained ZABCDEFG it matched Z instead
3208    of ZA. This was just one example of several cases that could provoke this bug,
3209    which was introduced by change 9 of version 2.00. The code for breaking
3210    infinite loops after an iteration that matches an empty string was't working
3211    correctly.
3212    
3213    2. The pcretest program was not imitating Perl correctly for the pattern /a*/g
3214    when matched against abbab (for example). After matching an empty string, it
3215    wasn't forcing anchoring when setting PCRE_NOTEMPTY for the next attempt; this
3216    caused it to match further down the string than it should.
3217    
3218    3. The code contained an inclusion of sys/types.h. It isn't clear why this
3219    was there because it doesn't seem to be needed, and it causes trouble on some
3220    systems, as it is not a Standard C header. It has been removed.
3221    
3222    4. Made 4 silly changes to the source to avoid stupid compiler warnings that
3223    were reported on the Macintosh. The changes were from
3224    
3225      while ((c = *(++ptr)) != 0 && c != '\n');
3226    to
3227      while ((c = *(++ptr)) != 0 && c != '\n') ;
3228    
3229    Totally extraordinary, but if that's what it takes...
3230    
3231    5. PCRE is being used in one environment where neither memmove() nor bcopy() is
3232    available. Added HAVE_BCOPY and an autoconf test for it; if neither
3233    HAVE_MEMMOVE nor HAVE_BCOPY is set, use a built-in emulation function which
3234    assumes the way PCRE uses memmove() (always moving upwards).
3235    
3236    6. PCRE is being used in one environment where strchr() is not available. There
3237    was only one use in pcre.c, and writing it out to avoid strchr() probably gives
3238    faster code anyway.
3239    
3240    
3241    Version 3.1 09-Feb-00
3242    ---------------------
3243    
3244    The only change in this release is the fixing of some bugs in Makefile.in for
3245    the "install" target:
3246    
3247    (1) It was failing to install pcreposix.h.
3248    
3249    (2) It was overwriting the pcre.3 man page with the pcreposix.3 man page.
3250    
3251    
3252    Version 3.0 01-Feb-00
3253    ---------------------
3254    
3255    1. Add support for the /+ modifier to perltest (to output $` like it does in
3256    pcretest).
3257    
3258    2. Add support for the /g modifier to perltest.
3259    
3260    3. Fix pcretest so that it behaves even more like Perl for /g when the pattern
3261    matches null strings.
3262    
3263    4. Fix perltest so that it doesn't do unwanted things when fed an empty
3264    pattern. Perl treats empty patterns specially - it reuses the most recent
3265    pattern, which is not what we want. Replace // by /(?#)/ in order to avoid this
3266    effect.
3267    
3268    5. The POSIX interface was broken in that it was just handing over the POSIX
3269    captured string vector to pcre_exec(), but (since release 2.00) PCRE has
3270    required a bigger vector, with some working space on the end. This means that
3271    the POSIX wrapper now has to get and free some memory, and copy the results.
3272    
3273    6. Added some simple autoconf support, placing the test data and the
3274    documentation in separate directories, re-organizing some of the
3275    information files, and making it build pcre-config (a GNU standard). Also added
3276    libtool support for building PCRE as a shared library, which is now the
3277    default.
3278    
3279    7. Got rid of the leading zero in the definition of PCRE_MINOR because 08 and
3280    09 are not valid octal constants. Single digits will be used for minor values
3281    less than 10.
3282    
3283    8. Defined REG_EXTENDED and REG_NOSUB as zero in the POSIX header, so that
3284    existing programs that set these in the POSIX interface can use PCRE without
3285    modification.
3286    
3287    9. Added a new function, pcre_fullinfo() with an extensible interface. It can
3288    return all that pcre_info() returns, plus additional data. The pcre_info()
3289    function is retained for compatibility, but is considered to be obsolete.
3290    
3291    10. Added experimental recursion feature (?R) to handle one common case that
3292    Perl 5.6 will be able to do with (?p{...}).
3293    
3294    11. Added support for POSIX character classes like [:alpha:], which Perl is
3295    adopting.
3296    
3297    
3298    Version 2.08 31-Aug-99
3299    ----------------------
3300    
3301    1. When startoffset was not zero and the pattern began with ".*", PCRE was not
3302    trying to match at the startoffset position, but instead was moving forward to
3303    the next newline as if a previous match had failed.
3304    
3305    2. pcretest was not making use of PCRE_NOTEMPTY when repeating for /g and /G,
3306    and could get into a loop if a null string was matched other than at the start
3307    of the subject.
3308    
3309    3. Added definitions of PCRE_MAJOR and PCRE_MINOR to pcre.h so the version can
3310    be distinguished at compile time, and for completeness also added PCRE_DATE.
3311    
3312    5. Added Paul Sokolovsky's minor changes to make it easy to compile a Win32 DLL
3313    in GnuWin32 environments.
3314    
3315    
3316    Version 2.07 29-Jul-99
3317    ----------------------
3318    
3319    1. The documentation is now supplied in plain text form and HTML as well as in
3320    the form of man page sources.
3321    
3322    2. C++ compilers don't like assigning (void *) values to other pointer types.
3323    In particular this affects malloc(). Although there is no problem in Standard
3324    C, I've put in casts to keep C++ compilers happy.
3325    
3326    3. Typo on pcretest.c; a cast of (unsigned char *) in the POSIX regexec() call
3327    should be (const char *).
3328    
3329    4. If NOPOSIX is defined, pcretest.c compiles without POSIX support. This may
3330    be useful for non-Unix systems who don't want to bother with the POSIX stuff.
3331    However, I haven't made this a standard facility. The documentation doesn't
3332    mention it, and the Makefile doesn't support it.
3333    
3334    5. The Makefile now contains an "install" target, with editable destinations at
3335