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

Legend:
Removed from v.23  
changed lines
  Added in v.383

webmaster@exim.org
ViewVC Help
Powered by ViewVC 1.1.12