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

Diff of /code/trunk/ChangeLog

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

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

Legend:
Removed from v.23  
changed lines
  Added in v.379

  ViewVC Help
Powered by ViewVC 1.1.5