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

Diff of /code/trunk/ChangeLog

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

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

Legend:
Removed from v.59  
changed lines
  Added in v.475

  ViewVC Help
Powered by ViewVC 1.1.5