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

Diff of /code/trunk/ChangeLog

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

revision 21 by nigel, Sat Feb 24 21:38:37 2007 UTC revision 378 by ph10, Sun Mar 1 14:13:34 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  Version 1.09 28 Apr-98  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    
36    Version 7.8 05-Sep-08
37    ---------------------
38    
39    1.  Replaced UCP searching code with optimized version as implemented for Ad
40        Muncher (http://www.admuncher.com/) by Peter Kankowski. This uses a two-
41        stage table and inline lookup instead of a function, giving speed ups of 2
42        to 5 times on some simple patterns that I tested. Permission was given to
43        distribute the MultiStage2.py script that generates the tables (it's not in
44        the tarball, but is in the Subversion repository).
45    
46    2.  Updated the Unicode datatables to Unicode 5.1.0. This adds yet more
47        scripts.
48    
49    3.  Change 12 for 7.7 introduced a bug in pcre_study() when a pattern contained
50        a group with a zero qualifier. The result of the study could be incorrect,
51        or the function might crash, depending on the pattern.
52    
53    4.  Caseless matching was not working for non-ASCII characters in back
54        references. For example, /(\x{de})\1/8i was not matching \x{de}\x{fe}.
55        It now works when Unicode Property Support is available.
56    
57    5.  In pcretest, an escape such as \x{de} in the data was always generating
58        a UTF-8 string, even in non-UTF-8 mode. Now it generates a single byte in
59        non-UTF-8 mode. If the value is greater than 255, it gives a warning about
60        truncation.
61    
62    6.  Minor bugfix in pcrecpp.cc (change "" == ... to NULL == ...).
63    
64    7.  Added two (int) casts to pcregrep when printing the difference of two
65        pointers, in case they are 64-bit values.
66    
67    8.  Added comments about Mac OS X stack usage to the pcrestack man page and to
68        test 2 if it fails.
69    
70    9.  Added PCRE_CALL_CONVENTION just before the names of all exported functions,
71        and a #define of that name to empty if it is not externally set. This is to
72        allow users of MSVC to set it if necessary.
73    
74    10. The PCRE_EXP_DEFN macro which precedes exported functions was missing from
75        the convenience functions in the pcre_get.c source file.
76    
77    11. An option change at the start of a pattern that had top-level alternatives
78        could cause overwriting and/or a crash. This command provoked a crash in
79        some environments:
80    
81          printf "/(?i)[\xc3\xa9\xc3\xbd]|[\xc3\xa9\xc3\xbdA]/8\n" | pcretest
82    
83        This potential security problem was recorded as CVE-2008-2371.
84    
85    12. For a pattern where the match had to start at the beginning or immediately
86        after a newline (e.g /.*anything/ without the DOTALL flag), pcre_exec() and
87        pcre_dfa_exec() could read past the end of the passed subject if there was
88        no match. To help with detecting such bugs (e.g. with valgrind), I modified
89        pcretest so that it places the subject at the end of its malloc-ed buffer.
90    
91    13. The change to pcretest in 12 above threw up a couple more cases when pcre_
92        exec() might read past the end of the data buffer in UTF-8 mode.
93    
94    14. A similar bug to 7.3/2 existed when the PCRE_FIRSTLINE option was set and
95        the data contained the byte 0x85 as part of a UTF-8 character within its
96        first line. This applied both to normal and DFA matching.
97    
98    15. Lazy qualifiers were not working in some cases in UTF-8 mode. For example,
99        /^[^d]*?$/8 failed to match "abc".
100    
101    16. Added a missing copyright notice to pcrecpp_internal.h.
102    
103    17. Make it more clear in the documentation that values returned from
104        pcre_exec() in ovector are byte offsets, not character counts.
105    
106    18. Tidied a few places to stop certain compilers from issuing warnings.
107    
108    19. Updated the Virtual Pascal + BCC files to compile the latest v7.7, as
109        supplied by Stefan Weber. I made a further small update for 7.8 because
110        there is a change of source arrangements: the pcre_searchfuncs.c module is
111        replaced by pcre_ucd.c.
112    
113    
114    Version 7.7 07-May-08
115    ---------------------
116    
117    1.  Applied Craig's patch to sort out a long long problem: "If we can't convert
118        a string to a long long, pretend we don't even have a long long." This is
119        done by checking for the strtoq, strtoll, and _strtoi64 functions.
120    
121    2.  Applied Craig's patch to pcrecpp.cc to restore ABI compatibility with
122        pre-7.6 versions, which defined a global no_arg variable instead of putting
123        it in the RE class. (See also #8 below.)
124    
125    3.  Remove a line of dead code, identified by coverity and reported by Nuno
126        Lopes.
127    
128    4.  Fixed two related pcregrep bugs involving -r with --include or --exclude:
129    
130        (1) The include/exclude patterns were being applied to the whole pathnames
131            of files, instead of just to the final components.
132    
133        (2) If there was more than one level of directory, the subdirectories were
134            skipped unless they satisfied the include/exclude conditions. This is
135            inconsistent with GNU grep (and could even be seen as contrary to the
136            pcregrep specification - which I improved to make it absolutely clear).
137            The action now is always to scan all levels of directory, and just
138            apply the include/exclude patterns to regular files.
139    
140    5.  Added the --include_dir and --exclude_dir patterns to pcregrep, and used
141        --exclude_dir in the tests to avoid scanning .svn directories.
142    
143    6.  Applied Craig's patch to the QuoteMeta function so that it escapes the
144        NUL character as backslash + 0 rather than backslash + NUL, because PCRE
145        doesn't support NULs in patterns.
146    
147    7.  Added some missing "const"s to declarations of static tables in
148        pcre_compile.c and pcre_dfa_exec.c.
149    
150    8.  Applied Craig's patch to pcrecpp.cc to fix a problem in OS X that was
151        caused by fix #2  above. (Subsequently also a second patch to fix the
152        first patch. And a third patch - this was a messy problem.)
153    
154    9.  Applied Craig's patch to remove the use of push_back().
155    
156    10. Applied Alan Lehotsky's patch to add REG_STARTEND support to the POSIX
157        matching function regexec().
158    
159    11. Added support for the Oniguruma syntax \g<name>, \g<n>, \g'name', \g'n',
160        which, however, unlike Perl's \g{...}, are subroutine calls, not back
161        references. PCRE supports relative numbers with this syntax (I don't think
162        Oniguruma does).
163    
164    12. Previously, a group with a zero repeat such as (...){0} was completely
165        omitted from the compiled regex. However, this means that if the group
166        was called as a subroutine from elsewhere in the pattern, things went wrong
167        (an internal error was given). Such groups are now left in the compiled
168        pattern, with a new opcode that causes them to be skipped at execution
169        time.
170    
171    13. Added the PCRE_JAVASCRIPT_COMPAT option. This makes the following changes
172        to the way PCRE behaves:
173    
174        (a) A lone ] character is dis-allowed (Perl treats it as data).
175    
176        (b) A back reference to an unmatched subpattern matches an empty string
177            (Perl fails the current match path).
178    
179        (c) A data ] in a character class must be notated as \] because if the
180            first data character in a class is ], it defines an empty class. (In
181            Perl it is not possible to have an empty class.) The empty class []
182            never matches; it forces failure and is equivalent to (*FAIL) or (?!).
183            The negative empty class [^] matches any one character, independently
184            of the DOTALL setting.
185    
186    14. A pattern such as /(?2)[]a()b](abc)/ which had a forward reference to a
187        non-existent subpattern following a character class starting with ']' and
188        containing () gave an internal compiling error instead of "reference to
189        non-existent subpattern". Fortunately, when the pattern did exist, the
190        compiled code was correct. (When scanning forwards to check for the
191        existencd of the subpattern, it was treating the data ']' as terminating
192        the class, so got the count wrong. When actually compiling, the reference
193        was subsequently set up correctly.)
194    
195    15. The "always fail" assertion (?!) is optimzed to (*FAIL) by pcre_compile;
196        it was being rejected as not supported by pcre_dfa_exec(), even though
197        other assertions are supported. I have made pcre_dfa_exec() support
198        (*FAIL).
199    
200    16. The implementation of 13c above involved the invention of a new opcode,
201        OP_ALLANY, which is like OP_ANY but doesn't check the /s flag. Since /s
202        cannot be changed at match time, I realized I could make a small
203        improvement to matching performance by compiling OP_ALLANY instead of
204        OP_ANY for "." when DOTALL was set, and then removing the runtime tests
205        on the OP_ANY path.
206    
207    17. Compiling pcretest on Windows with readline support failed without the
208        following two fixes: (1) Make the unistd.h include conditional on
209        HAVE_UNISTD_H; (2) #define isatty and fileno as _isatty and _fileno.
210    
211    18. Changed CMakeLists.txt and cmake/FindReadline.cmake to arrange for the
212        ncurses library to be included for pcretest when ReadLine support is
213        requested, but also to allow for it to be overridden. This patch came from
214        Daniel Bergström.
215    
216    19. There was a typo in the file ucpinternal.h where f0_rangeflag was defined
217        as 0x00f00000 instead of 0x00800000. Luckily, this would not have caused
218        any errors with the current Unicode tables. Thanks to Peter Kankowski for
219        spotting this.
220    
221    
222    Version 7.6 28-Jan-08
223    ---------------------
224    
225    1.  A character class containing a very large number of characters with
226        codepoints greater than 255 (in UTF-8 mode, of course) caused a buffer
227        overflow.
228    
229    2.  Patch to cut out the "long long" test in pcrecpp_unittest when
230        HAVE_LONG_LONG is not defined.
231    
232    3.  Applied Christian Ehrlicher's patch to update the CMake build files to
233        bring them up to date and include new features. This patch includes:
234    
235        - Fixed PH's badly added libz and libbz2 support.
236        - Fixed a problem with static linking.
237        - Added pcredemo. [But later removed - see 7 below.]
238        - Fixed dftables problem and added an option.
239        - Added a number of HAVE_XXX tests, including HAVE_WINDOWS_H and
240            HAVE_LONG_LONG.
241        - Added readline support for pcretest.
242        - Added an listing of the option settings after cmake has run.
243    
244    4.  A user submitted a patch to Makefile that makes it easy to create
245        "pcre.dll" under mingw when using Configure/Make. I added stuff to
246        Makefile.am that cause it to include this special target, without
247        affecting anything else. Note that the same mingw target plus all
248        the other distribution libraries and programs are now supported
249        when configuring with CMake (see 6 below) instead of with
250        Configure/Make.
251    
252    5.  Applied Craig's patch that moves no_arg into the RE class in the C++ code.
253        This is an attempt to solve the reported problem "pcrecpp::no_arg is not
254        exported in the Windows port". It has not yet been confirmed that the patch
255        solves the problem, but it does no harm.
256    
257    6.  Applied Sheri's patch to CMakeLists.txt to add NON_STANDARD_LIB_PREFIX and
258        NON_STANDARD_LIB_SUFFIX for dll names built with mingw when configured
259        with CMake, and also correct the comment about stack recursion.
260    
261    7.  Remove the automatic building of pcredemo from the ./configure system and
262        from CMakeLists.txt. The whole idea of pcredemo.c is that it is an example
263        of a program that users should build themselves after PCRE is installed, so
264        building it automatically is not really right. What is more, it gave
265        trouble in some build environments.
266    
267    8.  Further tidies to CMakeLists.txt from Sheri and Christian.
268    
269    
270    Version 7.5 10-Jan-08
271    ---------------------
272    
273    1.  Applied a patch from Craig: "This patch makes it possible to 'ignore'
274        values in parens when parsing an RE using the C++ wrapper."
275    
276    2.  Negative specials like \S did not work in character classes in UTF-8 mode.
277        Characters greater than 255 were excluded from the class instead of being
278        included.
279    
280    3.  The same bug as (2) above applied to negated POSIX classes such as
281        [:^space:].
282    
283    4.  PCRECPP_STATIC was referenced in pcrecpp_internal.h, but nowhere was it
284        defined or documented. It seems to have been a typo for PCRE_STATIC, so
285        I have changed it.
286    
287    5.  The construct (?&) was not diagnosed as a syntax error (it referenced the
288        first named subpattern) and a construct such as (?&a) would reference the
289        first named subpattern whose name started with "a" (in other words, the
290        length check was missing). Both these problems are fixed. "Subpattern name
291        expected" is now given for (?&) (a zero-length name), and this patch also
292        makes it give the same error for \k'' (previously it complained that that
293        was a reference to a non-existent subpattern).
294    
295    6.  The erroneous patterns (?+-a) and (?-+a) give different error messages;
296        this is right because (?- can be followed by option settings as well as by
297        digits. I have, however, made the messages clearer.
298    
299    7.  Patterns such as (?(1)a|b) (a pattern that contains fewer subpatterns
300        than the number used in the conditional) now cause a compile-time error.
301        This is actually not compatible with Perl, which accepts such patterns, but
302        treats the conditional as always being FALSE (as PCRE used to), but it
303        seems to me that giving a diagnostic is better.
304    
305    8.  Change "alphameric" to the more common word "alphanumeric" in comments
306        and messages.
307    
308    9.  Fix two occurrences of "backslash" in comments that should have been
309        "backspace".
310    
311    10. Remove two redundant lines of code that can never be obeyed (their function
312        was moved elsewhere).
313    
314    11. The program that makes PCRE's Unicode character property table had a bug
315        which caused it to generate incorrect table entries for sequences of
316        characters that have the same character type, but are in different scripts.
317        It amalgamated them into a single range, with the script of the first of
318        them. In other words, some characters were in the wrong script. There were
319        thirteen such cases, affecting characters in the following ranges:
320    
321          U+002b0 - U+002c1
322          U+0060c - U+0060d
323          U+0061e - U+00612
324          U+0064b - U+0065e
325          U+0074d - U+0076d
326          U+01800 - U+01805
327          U+01d00 - U+01d77
328          U+01d9b - U+01dbf
329          U+0200b - U+0200f
330          U+030fc - U+030fe
331          U+03260 - U+0327f
332          U+0fb46 - U+0fbb1
333          U+10450 - U+1049d
334    
335    12. The -o option (show only the matching part of a line) for pcregrep was not
336        compatible with GNU grep in that, if there was more than one match in a
337        line, it showed only the first of them. It now behaves in the same way as
338        GNU grep.
339    
340    13. If the -o and -v options were combined for pcregrep, it printed a blank
341        line for every non-matching line. GNU grep prints nothing, and pcregrep now
342        does the same. The return code can be used to tell if there were any
343        non-matching lines.
344    
345    14. Added --file-offsets and --line-offsets to pcregrep.
346    
347    15. The pattern (?=something)(?R) was not being diagnosed as a potentially
348        infinitely looping recursion. The bug was that positive lookaheads were not
349        being skipped when checking for a possible empty match (negative lookaheads
350        and both kinds of lookbehind were skipped).
351    
352    16. Fixed two typos in the Windows-only code in pcregrep.c, and moved the
353        inclusion of <windows.h> to before rather than after the definition of
354        INVALID_FILE_ATTRIBUTES (patch from David Byron).
355    
356    17. Specifying a possessive quantifier with a specific limit for a Unicode
357        character property caused pcre_compile() to compile bad code, which led at
358        runtime to PCRE_ERROR_INTERNAL (-14). Examples of patterns that caused this
359        are: /\p{Zl}{2,3}+/8 and /\p{Cc}{2}+/8. It was the possessive "+" that
360        caused the error; without that there was no problem.
361    
362    18. Added --enable-pcregrep-libz and --enable-pcregrep-libbz2.
363    
364    19. Added --enable-pcretest-libreadline.
365    
366    20. In pcrecpp.cc, the variable 'count' was incremented twice in
367        RE::GlobalReplace(). As a result, the number of replacements returned was
368        double what it should be. I removed one of the increments, but Craig sent a
369        later patch that removed the other one (the right fix) and added unit tests
370        that check the return values (which was not done before).
371    
372    21. Several CMake things:
373    
374        (1) Arranged that, when cmake is used on Unix, the libraries end up with
375            the names libpcre and libpcreposix, not just pcre and pcreposix.
376    
377        (2) The above change means that pcretest and pcregrep are now correctly
378            linked with the newly-built libraries, not previously installed ones.
379    
380        (3) Added PCRE_SUPPORT_LIBREADLINE, PCRE_SUPPORT_LIBZ, PCRE_SUPPORT_LIBBZ2.
381    
382    22. In UTF-8 mode, with newline set to "any", a pattern such as .*a.*=.b.*
383        crashed when matching a string such as a\x{2029}b (note that \x{2029} is a
384        UTF-8 newline character). The key issue is that the pattern starts .*;
385        this means that the match must be either at the beginning, or after a
386        newline. The bug was in the code for advancing after a failed match and
387        checking that the new position followed a newline. It was not taking
388        account of UTF-8 characters correctly.
389    
390    23. PCRE was behaving differently from Perl in the way it recognized POSIX
391        character classes. PCRE was not treating the sequence [:...:] as a
392        character class unless the ... were all letters. Perl, however, seems to
393        allow any characters between [: and :], though of course it rejects as
394        unknown any "names" that contain non-letters, because all the known class
395        names consist only of letters. Thus, Perl gives an error for [[:1234:]],
396        for example, whereas PCRE did not - it did not recognize a POSIX character
397        class. This seemed a bit dangerous, so the code has been changed to be
398        closer to Perl. The behaviour is not identical to Perl, because PCRE will
399        diagnose an unknown class for, for example, [[:l\ower:]] where Perl will
400        treat it as [[:lower:]]. However, PCRE does now give "unknown" errors where
401        Perl does, and where it didn't before.
402    
403    24. Rewrite so as to remove the single use of %n from pcregrep because in some
404        Windows environments %n is disabled by default.
405    
406    
407    Version 7.4 21-Sep-07
408    ---------------------
409    
410    1.  Change 7.3/28 was implemented for classes by looking at the bitmap. This
411        means that a class such as [\s] counted as "explicit reference to CR or
412        LF". That isn't really right - the whole point of the change was to try to
413        help when there was an actual mention of one of the two characters. So now
414        the change happens only if \r or \n (or a literal CR or LF) character is
415        encountered.
416    
417    2.  The 32-bit options word was also used for 6 internal flags, but the numbers
418        of both had grown to the point where there were only 3 bits left.
419        Fortunately, there was spare space in the data structure, and so I have
420        moved the internal flags into a new 16-bit field to free up more option
421        bits.
422    
423    3.  The appearance of (?J) at the start of a pattern set the DUPNAMES option,
424        but did not set the internal JCHANGED flag - either of these is enough to
425        control the way the "get" function works - but the PCRE_INFO_JCHANGED
426        facility is supposed to tell if (?J) was ever used, so now (?J) at the
427        start sets both bits.
428    
429    4.  Added options (at build time, compile time, exec time) to change \R from
430        matching any Unicode line ending sequence to just matching CR, LF, or CRLF.
431    
432    5.  doc/pcresyntax.html was missing from the distribution.
433    
434    6.  Put back the definition of PCRE_ERROR_NULLWSLIMIT, for backward
435        compatibility, even though it is no longer used.
436    
437    7.  Added macro for snprintf to pcrecpp_unittest.cc and also for strtoll and
438        strtoull to pcrecpp.cc to select the available functions in WIN32 when the
439        windows.h file is present (where different names are used). [This was
440        reversed later after testing - see 16 below.]
441    
442    8.  Changed all #include <config.h> to #include "config.h". There were also
443        some further <pcre.h> cases that I changed to "pcre.h".
444    
445    9.  When pcregrep was used with the --colour option, it missed the line ending
446        sequence off the lines that it output.
447    
448    10. It was pointed out to me that arrays of string pointers cause lots of
449        relocations when a shared library is dynamically loaded. A technique of
450        using a single long string with a table of offsets can drastically reduce
451        these. I have refactored PCRE in four places to do this. The result is
452        dramatic:
453    
454          Originally:                          290
455          After changing UCP table:            187
456          After changing error message table:   43
457          After changing table of "verbs"       36
458          After changing table of Posix names   22
459    
460        Thanks to the folks working on Gregex for glib for this insight.
461    
462    11. --disable-stack-for-recursion caused compiling to fail unless -enable-
463        unicode-properties was also set.
464    
465    12. Updated the tests so that they work when \R is defaulted to ANYCRLF.
466    
467    13. Added checks for ANY and ANYCRLF to pcrecpp.cc where it previously
468        checked only for CRLF.
469    
470    14. Added casts to pcretest.c to avoid compiler warnings.
471    
472    15. Added Craig's patch to various pcrecpp modules to avoid compiler warnings.
473    
474    16. Added Craig's patch to remove the WINDOWS_H tests, that were not working,
475        and instead check for _strtoi64 explicitly, and avoid the use of snprintf()
476        entirely. This removes changes made in 7 above.
477    
478    17. The CMake files have been updated, and there is now more information about
479        building with CMake in the NON-UNIX-USE document.
480    
481    
482    Version 7.3 28-Aug-07
483    ---------------------
484    
485     1. In the rejigging of the build system that eventually resulted in 7.1, the
486        line "#include <pcre.h>" was included in pcre_internal.h. The use of angle
487        brackets there is not right, since it causes compilers to look for an
488        installed pcre.h, not the version that is in the source that is being
489        compiled (which of course may be different). I have changed it back to:
490    
491          #include "pcre.h"
492    
493        I have a vague recollection that the change was concerned with compiling in
494        different directories, but in the new build system, that is taken care of
495        by the VPATH setting the Makefile.
496    
497     2. The pattern .*$ when run in not-DOTALL UTF-8 mode with newline=any failed
498        when the subject happened to end in the byte 0x85 (e.g. if the last
499        character was \x{1ec5}). *Character* 0x85 is one of the "any" newline
500        characters but of course it shouldn't be taken as a newline when it is part
501        of another character. The bug was that, for an unlimited repeat of . in
502        not-DOTALL UTF-8 mode, PCRE was advancing by bytes rather than by
503        characters when looking for a newline.
504    
505     3. A small performance improvement in the DOTALL UTF-8 mode .* case.
506    
507     4. Debugging: adjusted the names of opcodes for different kinds of parentheses
508        in debug output.
509    
510     5. Arrange to use "%I64d" instead of "%lld" and "%I64u" instead of "%llu" for
511        long printing in the pcrecpp unittest when running under MinGW.
512    
513     6. ESC_K was left out of the EBCDIC table.
514    
515     7. Change 7.0/38 introduced a new limit on the number of nested non-capturing
516        parentheses; I made it 1000, which seemed large enough. Unfortunately, the
517        limit also applies to "virtual nesting" when a pattern is recursive, and in
518        this case 1000 isn't so big. I have been able to remove this limit at the
519        expense of backing off one optimization in certain circumstances. Normally,
520        when pcre_exec() would call its internal match() function recursively and
521        immediately return the result unconditionally, it uses a "tail recursion"
522        feature to save stack. However, when a subpattern that can match an empty
523        string has an unlimited repetition quantifier, it no longer makes this
524        optimization. That gives it a stack frame in which to save the data for
525        checking that an empty string has been matched. Previously this was taken
526        from the 1000-entry workspace that had been reserved. So now there is no
527        explicit limit, but more stack is used.
528    
529     8. Applied Daniel's patches to solve problems with the import/export magic
530        syntax that is required for Windows, and which was going wrong for the
531        pcreposix and pcrecpp parts of the library. These were overlooked when this
532        problem was solved for the main library.
533    
534     9. There were some crude static tests to avoid integer overflow when computing
535        the size of patterns that contain repeated groups with explicit upper
536        limits. As the maximum quantifier is 65535, the maximum group length was
537        set at 30,000 so that the product of these two numbers did not overflow a
538        32-bit integer. However, it turns out that people want to use groups that
539        are longer than 30,000 bytes (though not repeat them that many times).
540        Change 7.0/17 (the refactoring of the way the pattern size is computed) has
541        made it possible to implement the integer overflow checks in a much more
542        dynamic way, which I have now done. The artificial limitation on group
543        length has been removed - we now have only the limit on the total length of
544        the compiled pattern, which depends on the LINK_SIZE setting.
545    
546    10. Fixed a bug in the documentation for get/copy named substring when
547        duplicate names are permitted. If none of the named substrings are set, the
548        functions return PCRE_ERROR_NOSUBSTRING (7); the doc said they returned an
549        empty string.
550    
551    11. Because Perl interprets \Q...\E at a high level, and ignores orphan \E
552        instances, patterns such as [\Q\E] or [\E] or even [^\E] cause an error,
553        because the ] is interpreted as the first data character and the
554        terminating ] is not found. PCRE has been made compatible with Perl in this
555        regard. Previously, it interpreted [\Q\E] as an empty class, and [\E] could
556        cause memory overwriting.
557    
558    10. Like Perl, PCRE automatically breaks an unlimited repeat after an empty
559        string has been matched (to stop an infinite loop). It was not recognizing
560        a conditional subpattern that could match an empty string if that
561        subpattern was within another subpattern. For example, it looped when
562        trying to match  (((?(1)X|))*)  but it was OK with  ((?(1)X|)*)  where the
563        condition was not nested. This bug has been fixed.
564    
565    12. A pattern like \X?\d or \P{L}?\d in non-UTF-8 mode could cause a backtrack
566        past the start of the subject in the presence of bytes with the top bit
567        set, for example "\x8aBCD".
568    
569    13. Added Perl 5.10 experimental backtracking controls (*FAIL), (*F), (*PRUNE),
570        (*SKIP), (*THEN), (*COMMIT), and (*ACCEPT).
571    
572    14. Optimized (?!) to (*FAIL).
573    
574    15. Updated the test for a valid UTF-8 string to conform to the later RFC 3629.
575        This restricts code points to be within the range 0 to 0x10FFFF, excluding
576        the "low surrogate" sequence 0xD800 to 0xDFFF. Previously, PCRE allowed the
577        full range 0 to 0x7FFFFFFF, as defined by RFC 2279. Internally, it still
578        does: it's just the validity check that is more restrictive.
579    
580    16. Inserted checks for integer overflows during escape sequence (backslash)
581        processing, and also fixed erroneous offset values for syntax errors during
582        backslash processing.
583    
584    17. Fixed another case of looking too far back in non-UTF-8 mode (cf 12 above)
585        for patterns like [\PPP\x8a]{1,}\x80 with the subject "A\x80".
586    
587    18. An unterminated class in a pattern like (?1)\c[ with a "forward reference"
588        caused an overrun.
589    
590    19. A pattern like (?:[\PPa*]*){8,} which had an "extended class" (one with
591        something other than just ASCII characters) inside a group that had an
592        unlimited repeat caused a loop at compile time (while checking to see
593        whether the group could match an empty string).
594    
595    20. Debugging a pattern containing \p or \P could cause a crash. For example,
596        [\P{Any}] did so. (Error in the code for printing property names.)
597    
598    21. An orphan \E inside a character class could cause a crash.
599    
600    22. A repeated capturing bracket such as (A)? could cause a wild memory
601        reference during compilation.
602    
603    23. There are several functions in pcre_compile() that scan along a compiled
604        expression for various reasons (e.g. to see if it's fixed length for look
605        behind). There were bugs in these functions when a repeated \p or \P was
606        present in the pattern. These operators have additional parameters compared
607        with \d, etc, and these were not being taken into account when moving along
608        the compiled data. Specifically:
609    
610        (a) A item such as \p{Yi}{3} in a lookbehind was not treated as fixed
611            length.
612    
613        (b) An item such as \pL+ within a repeated group could cause crashes or
614            loops.
615    
616        (c) A pattern such as \p{Yi}+(\P{Yi}+)(?1) could give an incorrect
617            "reference to non-existent subpattern" error.
618    
619        (d) A pattern like (\P{Yi}{2}\277)? could loop at compile time.
620    
621    24. A repeated \S or \W in UTF-8 mode could give wrong answers when multibyte
622        characters were involved (for example /\S{2}/8g with "A\x{a3}BC").
623    
624    25. Using pcregrep in multiline, inverted mode (-Mv) caused it to loop.
625    
626    26. Patterns such as [\P{Yi}A] which include \p or \P and just one other
627        character were causing crashes (broken optimization).
628    
629    27. Patterns such as (\P{Yi}*\277)* (group with possible zero repeat containing
630        \p or \P) caused a compile-time loop.
631    
632    28. More problems have arisen in unanchored patterns when CRLF is a valid line
633        break. For example, the unstudied pattern [\r\n]A does not match the string
634        "\r\nA" because change 7.0/46 below moves the current point on by two
635        characters after failing to match at the start. However, the pattern \nA
636        *does* match, because it doesn't start till \n, and if [\r\n]A is studied,
637        the same is true. There doesn't seem any very clean way out of this, but
638        what I have chosen to do makes the common cases work: PCRE now takes note
639        of whether there can be an explicit match for \r or \n anywhere in the
640        pattern, and if so, 7.0/46 no longer applies. As part of this change,
641        there's a new PCRE_INFO_HASCRORLF option for finding out whether a compiled
642        pattern has explicit CR or LF references.
643    
644    29. Added (*CR) etc for changing newline setting at start of pattern.
645    
646    
647    Version 7.2 19-Jun-07
648    ---------------------
649    
650     1. If the fr_FR locale cannot be found for test 3, try the "french" locale,
651        which is apparently normally available under Windows.
652    
653     2. Re-jig the pcregrep tests with different newline settings in an attempt
654        to make them independent of the local environment's newline setting.
655    
656     3. Add code to configure.ac to remove -g from the CFLAGS default settings.
657    
658     4. Some of the "internals" tests were previously cut out when the link size
659        was not 2, because the output contained actual offsets. The recent new
660        "Z" feature of pcretest means that these can be cut out, making the tests
661        usable with all link sizes.
662    
663     5. Implemented Stan Switzer's goto replacement for longjmp() when not using
664        stack recursion. This gives a massive performance boost under BSD, but just
665        a small improvement under Linux. However, it saves one field in the frame
666        in all cases.
667    
668     6. Added more features from the forthcoming Perl 5.10:
669    
670        (a) (?-n) (where n is a string of digits) is a relative subroutine or
671            recursion call. It refers to the nth most recently opened parentheses.
672    
673        (b) (?+n) is also a relative subroutine call; it refers to the nth next
674            to be opened parentheses.
675    
676        (c) Conditions that refer to capturing parentheses can be specified
677            relatively, for example, (?(-2)... or (?(+3)...
678    
679        (d) \K resets the start of the current match so that everything before
680            is not part of it.
681    
682        (e) \k{name} is synonymous with \k<name> and \k'name' (.NET compatible).
683    
684        (f) \g{name} is another synonym - part of Perl 5.10's unification of
685            reference syntax.
686    
687        (g) (?| introduces a group in which the numbering of parentheses in each
688            alternative starts with the same number.
689    
690        (h) \h, \H, \v, and \V match horizontal and vertical whitespace.
691    
692     7. Added two new calls to pcre_fullinfo(): PCRE_INFO_OKPARTIAL and
693        PCRE_INFO_JCHANGED.
694    
695     8. A pattern such as  (.*(.)?)*  caused pcre_exec() to fail by either not
696        terminating or by crashing. Diagnosed by Viktor Griph; it was in the code
697        for detecting groups that can match an empty string.
698    
699     9. A pattern with a very large number of alternatives (more than several
700        hundred) was running out of internal workspace during the pre-compile
701        phase, where pcre_compile() figures out how much memory will be needed. A
702        bit of new cunning has reduced the workspace needed for groups with
703        alternatives. The 1000-alternative test pattern now uses 12 bytes of
704        workspace instead of running out of the 4096 that are available.
705    
706    10. Inserted some missing (unsigned int) casts to get rid of compiler warnings.
707    
708    11. Applied patch from Google to remove an optimization that didn't quite work.
709        The report of the bug said:
710    
711          pcrecpp::RE("a*").FullMatch("aaa") matches, while
712          pcrecpp::RE("a*?").FullMatch("aaa") does not, and
713          pcrecpp::RE("a*?\\z").FullMatch("aaa") does again.
714    
715    12. If \p or \P was used in non-UTF-8 mode on a character greater than 127
716        it matched the wrong number of bytes.
717    
718    
719    Version 7.1 24-Apr-07
720    ---------------------
721    
722     1. Applied Bob Rossi and Daniel G's patches to convert the build system to one
723        that is more "standard", making use of automake and other Autotools. There
724        is some re-arrangement of the files and adjustment of comments consequent
725        on this.
726    
727     2. Part of the patch fixed a problem with the pcregrep tests. The test of -r
728        for recursive directory scanning broke on some systems because the files
729        are not scanned in any specific order and on different systems the order
730        was different. A call to "sort" has been inserted into RunGrepTest for the
731        approprate test as a short-term fix. In the longer term there may be an
732        alternative.
733    
734     3. I had an email from Eric Raymond about problems translating some of PCRE's
735        man pages to HTML (despite the fact that I distribute HTML pages, some
736        people do their own conversions for various reasons). The problems
737        concerned the use of low-level troff macros .br and .in. I have therefore
738        removed all such uses from the man pages (some were redundant, some could
739        be replaced by .nf/.fi pairs). The 132html script that I use to generate
740        HTML has been updated to handle .nf/.fi and to complain if it encounters
741        .br or .in.
742    
743     4. Updated comments in configure.ac that get placed in config.h.in and also
744        arranged for config.h to be included in the distribution, with the name
745        config.h.generic, for the benefit of those who have to compile without
746        Autotools (compare pcre.h, which is now distributed as pcre.h.generic).
747    
748     5. Updated the support (such as it is) for Virtual Pascal, thanks to Stefan
749        Weber: (1) pcre_internal.h was missing some function renames; (2) updated
750        makevp.bat for the current PCRE, using the additional files
751        makevp_c.txt, makevp_l.txt, and pcregexp.pas.
752    
753     6. A Windows user reported a minor discrepancy with test 2, which turned out
754        to be caused by a trailing space on an input line that had got lost in his
755        copy. The trailing space was an accident, so I've just removed it.
756    
757     7. Add -Wl,-R... flags in pcre-config.in for *BSD* systems, as I'm told
758        that is needed.
759    
760     8. Mark ucp_table (in ucptable.h) and ucp_gentype (in pcre_ucp_searchfuncs.c)
761        as "const" (a) because they are and (b) because it helps the PHP
762        maintainers who have recently made a script to detect big data structures
763        in the php code that should be moved to the .rodata section. I remembered
764        to update Builducptable as well, so it won't revert if ucptable.h is ever
765        re-created.
766    
767     9. Added some extra #ifdef SUPPORT_UTF8 conditionals into pcretest.c,
768        pcre_printint.src, pcre_compile.c, pcre_study.c, and pcre_tables.c, in
769        order to be able to cut out the UTF-8 tables in the latter when UTF-8
770        support is not required. This saves 1.5-2K of code, which is important in
771        some applications.
772    
773        Later: more #ifdefs are needed in pcre_ord2utf8.c and pcre_valid_utf8.c
774        so as not to refer to the tables, even though these functions will never be
775        called when UTF-8 support is disabled. Otherwise there are problems with a
776        shared library.
777    
778    10. Fixed two bugs in the emulated memmove() function in pcre_internal.h:
779    
780        (a) It was defining its arguments as char * instead of void *.
781    
782        (b) It was assuming that all moves were upwards in memory; this was true
783            a long time ago when I wrote it, but is no longer the case.
784    
785        The emulated memove() is provided for those environments that have neither
786        memmove() nor bcopy(). I didn't think anyone used it these days, but that
787        is clearly not the case, as these two bugs were recently reported.
788    
789    11. The script PrepareRelease is now distributed: it calls 132html, CleanTxt,
790        and Detrail to create the HTML documentation, the .txt form of the man
791        pages, and it removes trailing spaces from listed files. It also creates
792        pcre.h.generic and config.h.generic from pcre.h and config.h. In the latter
793        case, it wraps all the #defines with #ifndefs. This script should be run
794        before "make dist".
795    
796    12. Fixed two fairly obscure bugs concerned with quantified caseless matching
797        with Unicode property support.
798    
799        (a) For a maximizing quantifier, if the two different cases of the
800            character were of different lengths in their UTF-8 codings (there are
801            some cases like this - I found 11), and the matching function had to
802            back up over a mixture of the two cases, it incorrectly assumed they
803            were both the same length.
804    
805        (b) When PCRE was configured to use the heap rather than the stack for
806            recursion during matching, it was not correctly preserving the data for
807            the other case of a UTF-8 character when checking ahead for a match
808            while processing a minimizing repeat. If the check also involved
809            matching a wide character, but failed, corruption could cause an
810            erroneous result when trying to check for a repeat of the original
811            character.
812    
813    13. Some tidying changes to the testing mechanism:
814    
815        (a) The RunTest script now detects the internal link size and whether there
816            is UTF-8 and UCP support by running ./pcretest -C instead of relying on
817            values substituted by "configure". (The RunGrepTest script already did
818            this for UTF-8.) The configure.ac script no longer substitutes the
819            relevant variables.
820    
821        (b) The debugging options /B and /D in pcretest show the compiled bytecode
822            with length and offset values. This means that the output is different
823            for different internal link sizes. Test 2 is skipped for link sizes
824            other than 2 because of this, bypassing the problem. Unfortunately,
825            there was also a test in test 3 (the locale tests) that used /B and
826            failed for link sizes other than 2. Rather than cut the whole test out,
827            I have added a new /Z option to pcretest that replaces the length and
828            offset values with spaces. This is now used to make test 3 independent
829            of link size. (Test 2 will be tidied up later.)
830    
831    14. If erroroffset was passed as NULL to pcre_compile, it provoked a
832        segmentation fault instead of returning the appropriate error message.
833    
834    15. In multiline mode when the newline sequence was set to "any", the pattern
835        ^$ would give a match between the \r and \n of a subject such as "A\r\nB".
836        This doesn't seem right; it now treats the CRLF combination as the line
837        ending, and so does not match in that case. It's only a pattern such as ^$
838        that would hit this one: something like ^ABC$ would have failed after \r
839        and then tried again after \r\n.
840    
841    16. Changed the comparison command for RunGrepTest from "diff -u" to "diff -ub"
842        in an attempt to make files that differ only in their line terminators
843        compare equal. This works on Linux.
844    
845    17. Under certain error circumstances pcregrep might try to free random memory
846        as it exited. This is now fixed, thanks to valgrind.
847    
848    19. In pcretest, if the pattern /(?m)^$/g<any> was matched against the string
849        "abc\r\n\r\n", it found an unwanted second match after the second \r. This
850        was because its rules for how to advance for /g after matching an empty
851        string at the end of a line did not allow for this case. They now check for
852        it specially.
853    
854    20. pcretest is supposed to handle patterns and data of any length, by
855        extending its buffers when necessary. It was getting this wrong when the
856        buffer for a data line had to be extended.
857    
858    21. Added PCRE_NEWLINE_ANYCRLF which is like ANY, but matches only CR, LF, or
859        CRLF as a newline sequence.
860    
861    22. Code for handling Unicode properties in pcre_dfa_exec() wasn't being cut
862        out by #ifdef SUPPORT_UCP. This did no harm, as it could never be used, but
863        I have nevertheless tidied it up.
864    
865    23. Added some casts to kill warnings from HP-UX ia64 compiler.
866    
867    24. Added a man page for pcre-config.
868    
869    
870    Version 7.0 19-Dec-06
871    ---------------------
872    
873     1. Fixed a signed/unsigned compiler warning in pcre_compile.c, shown up by
874        moving to gcc 4.1.1.
875    
876     2. The -S option for pcretest uses setrlimit(); I had omitted to #include
877        sys/time.h, which is documented as needed for this function. It doesn't
878        seem to matter on Linux, but it showed up on some releases of OS X.
879    
880     3. It seems that there are systems where bytes whose values are greater than
881        127 match isprint() in the "C" locale. The "C" locale should be the
882        default when a C program starts up. In most systems, only ASCII printing
883        characters match isprint(). This difference caused the output from pcretest
884        to vary, making some of the tests fail. I have changed pcretest so that:
885    
886        (a) When it is outputting text in the compiled version of a pattern, bytes
887            other than 32-126 are always shown as hex escapes.
888    
889        (b) When it is outputting text that is a matched part of a subject string,
890            it does the same, unless a different locale has been set for the match
891            (using the /L modifier). In this case, it uses isprint() to decide.
892    
893     4. Fixed a major bug that caused incorrect computation of the amount of memory
894        required for a compiled pattern when options that changed within the
895        pattern affected the logic of the preliminary scan that determines the
896        length. The relevant options are -x, and -i in UTF-8 mode. The result was
897        that the computed length was too small. The symptoms of this bug were
898        either the PCRE error "internal error: code overflow" from pcre_compile(),
899        or a glibc crash with a message such as "pcretest: free(): invalid next
900        size (fast)". Examples of patterns that provoked this bug (shown in
901        pcretest format) are:
902    
903          /(?-x: )/x
904          /(?x)(?-x: \s*#\s*)/
905          /((?i)[\x{c0}])/8
906          /(?i:[\x{c0}])/8
907    
908        HOWEVER: Change 17 below makes this fix obsolete as the memory computation
909        is now done differently.
910    
911     5. Applied patches from Google to: (a) add a QuoteMeta function to the C++
912        wrapper classes; (b) implement a new function in the C++ scanner that is
913        more efficient than the old way of doing things because it avoids levels of
914        recursion in the regex matching; (c) add a paragraph to the documentation
915        for the FullMatch() function.
916    
917     6. The escape sequence \n was being treated as whatever was defined as
918        "newline". Not only was this contrary to the documentation, which states
919        that \n is character 10 (hex 0A), but it also went horribly wrong when
920        "newline" was defined as CRLF. This has been fixed.
921    
922     7. In pcre_dfa_exec.c the value of an unsigned integer (the variable called c)
923        was being set to -1 for the "end of line" case (supposedly a value that no
924        character can have). Though this value is never used (the check for end of
925        line is "zero bytes in current character"), it caused compiler complaints.
926        I've changed it to 0xffffffff.
927    
928     8. In pcre_version.c, the version string was being built by a sequence of
929        C macros that, in the event of PCRE_PRERELEASE being defined as an empty
930        string (as it is for production releases) called a macro with an empty
931        argument. The C standard says the result of this is undefined. The gcc
932        compiler treats it as an empty string (which was what was wanted) but it is
933        reported that Visual C gives an error. The source has been hacked around to
934        avoid this problem.
935    
936     9. On the advice of a Windows user, included <io.h> and <fcntl.h> in Windows
937        builds of pcretest, and changed the call to _setmode() to use _O_BINARY
938        instead of 0x8000. Made all the #ifdefs test both _WIN32 and WIN32 (not all
939        of them did).
940    
941    10. Originally, pcretest opened its input and output without "b"; then I was
942        told that "b" was needed in some environments, so it was added for release
943        5.0 to both the input and output. (It makes no difference on Unix-like
944        systems.) Later I was told that it is wrong for the input on Windows. I've
945        now abstracted the modes into two macros, to make it easier to fiddle with
946        them, and removed "b" from the input mode under Windows.
947    
948    11. Added pkgconfig support for the C++ wrapper library, libpcrecpp.
949    
950    12. Added -help and --help to pcretest as an official way of being reminded
951        of the options.
952    
953    13. Removed some redundant semicolons after macro calls in pcrecpparg.h.in
954        and pcrecpp.cc because they annoy compilers at high warning levels.
955    
956    14. A bit of tidying/refactoring in pcre_exec.c in the main bumpalong loop.
957    
958    15. Fixed an occurrence of == in configure.ac that should have been = (shell
959        scripts are not C programs :-) and which was not noticed because it works
960        on Linux.
961    
962    16. pcretest is supposed to handle any length of pattern and data line (as one
963        line or as a continued sequence of lines) by extending its input buffer if
964        necessary. This feature was broken for very long pattern lines, leading to
965        a string of junk being passed to pcre_compile() if the pattern was longer
966        than about 50K.
967    
968    17. I have done a major re-factoring of the way pcre_compile() computes the
969        amount of memory needed for a compiled pattern. Previously, there was code
970        that made a preliminary scan of the pattern in order to do this. That was
971        OK when PCRE was new, but as the facilities have expanded, it has become
972        harder and harder to keep it in step with the real compile phase, and there
973        have been a number of bugs (see for example, 4 above). I have now found a
974        cunning way of running the real compile function in a "fake" mode that
975        enables it to compute how much memory it would need, while actually only
976        ever using a few hundred bytes of working memory and without too many
977        tests of the mode. This should make future maintenance and development
978        easier. A side effect of this work is that the limit of 200 on the nesting
979        depth of parentheses has been removed (though this was never a serious
980        limitation, I suspect). However, there is a downside: pcre_compile() now
981        runs more slowly than before (30% or more, depending on the pattern). I
982        hope this isn't a big issue. There is no effect on runtime performance.
983    
984    18. Fixed a minor bug in pcretest: if a pattern line was not terminated by a
985        newline (only possible for the last line of a file) and it was a
986        pattern that set a locale (followed by /Lsomething), pcretest crashed.
987    
988    19. Added additional timing features to pcretest. (1) The -tm option now times
989        matching only, not compiling. (2) Both -t and -tm can be followed, as a
990        separate command line item, by a number that specifies the number of
991        repeats to use when timing. The default is 50000; this gives better
992        precision, but takes uncomfortably long for very large patterns.
993    
994    20. Extended pcre_study() to be more clever in cases where a branch of a
995        subpattern has no definite first character. For example, (a*|b*)[cd] would
996        previously give no result from pcre_study(). Now it recognizes that the
997        first character must be a, b, c, or d.
998    
999    21. There was an incorrect error "recursive call could loop indefinitely" if
1000        a subpattern (or the entire pattern) that was being tested for matching an
1001        empty string contained only one non-empty item after a nested subpattern.
1002        For example, the pattern (?>\x{100}*)\d(?R) provoked this error
1003        incorrectly, because the \d was being skipped in the check.
1004    
1005    22. The pcretest program now has a new pattern option /B and a command line
1006        option -b, which is equivalent to adding /B to every pattern. This causes
1007        it to show the compiled bytecode, without the additional information that
1008        -d shows. The effect of -d is now the same as -b with -i (and similarly, /D
1009        is the same as /B/I).
1010    
1011    23. A new optimization is now able automatically to treat some sequences such
1012        as a*b as a*+b. More specifically, if something simple (such as a character
1013        or a simple class like \d) has an unlimited quantifier, and is followed by
1014        something that cannot possibly match the quantified thing, the quantifier
1015        is automatically "possessified".
1016    
1017    24. A recursive reference to a subpattern whose number was greater than 39
1018        went wrong under certain circumstances in UTF-8 mode. This bug could also
1019        have affected the operation of pcre_study().
1020    
1021    25. Realized that a little bit of performance could be had by replacing
1022        (c & 0xc0) == 0xc0 with c >= 0xc0 when processing UTF-8 characters.
1023    
1024    26. Timing data from pcretest is now shown to 4 decimal places instead of 3.
1025    
1026    27. Possessive quantifiers such as a++ were previously implemented by turning
1027        them into atomic groups such as ($>a+). Now they have their own opcodes,
1028        which improves performance. This includes the automatically created ones
1029        from 23 above.
1030    
1031    28. A pattern such as (?=(\w+))\1: which simulates an atomic group using a
1032        lookahead was broken if it was not anchored. PCRE was mistakenly expecting
1033        the first matched character to be a colon. This applied both to named and
1034        numbered groups.
1035    
1036    29. The ucpinternal.h header file was missing its idempotency #ifdef.
1037    
1038    30. I was sent a "project" file called libpcre.a.dev which I understand makes
1039        building PCRE on Windows easier, so I have included it in the distribution.
1040    
1041    31. There is now a check in pcretest against a ridiculously large number being
1042        returned by pcre_exec() or pcre_dfa_exec(). If this happens in a /g or /G
1043        loop, the loop is abandoned.
1044    
1045    32. Forward references to subpatterns in conditions such as (?(2)...) where
1046        subpattern 2 is defined later cause pcre_compile() to search forwards in
1047        the pattern for the relevant set of parentheses. This search went wrong
1048        when there were unescaped parentheses in a character class, parentheses
1049        escaped with \Q...\E, or parentheses in a #-comment in /x mode.
1050    
1051    33. "Subroutine" calls and backreferences were previously restricted to
1052        referencing subpatterns earlier in the regex. This restriction has now
1053        been removed.
1054    
1055    34. Added a number of extra features that are going to be in Perl 5.10. On the
1056        whole, these are just syntactic alternatives for features that PCRE had
1057        previously implemented using the Python syntax or my own invention. The
1058        other formats are all retained for compatibility.
1059    
1060        (a) Named groups can now be defined as (?<name>...) or (?'name'...) as well
1061            as (?P<name>...). The new forms, as well as being in Perl 5.10, are
1062            also .NET compatible.
1063    
1064        (b) A recursion or subroutine call to a named group can now be defined as
1065            (?&name) as well as (?P>name).
1066    
1067        (c) A backreference to a named group can now be defined as \k<name> or
1068            \k'name' as well as (?P=name). The new forms, as well as being in Perl
1069            5.10, are also .NET compatible.
1070    
1071        (d) A conditional reference to a named group can now use the syntax
1072            (?(<name>) or (?('name') as well as (?(name).
1073    
1074        (e) A "conditional group" of the form (?(DEFINE)...) can be used to define
1075            groups (named and numbered) that are never evaluated inline, but can be
1076            called as "subroutines" from elsewhere. In effect, the DEFINE condition
1077            is always false. There may be only one alternative in such a group.
1078    
1079        (f) A test for recursion can be given as (?(R1).. or (?(R&name)... as well
1080            as the simple (?(R). The condition is true only if the most recent
1081            recursion is that of the given number or name. It does not search out
1082            through the entire recursion stack.
1083    
1084        (g) The escape \gN or \g{N} has been added, where N is a positive or
1085            negative number, specifying an absolute or relative reference.
1086    
1087    35. Tidied to get rid of some further signed/unsigned compiler warnings and
1088        some "unreachable code" warnings.
1089    
1090    36. Updated the Unicode property tables to Unicode version 5.0.0. Amongst other
1091        things, this adds five new scripts.
1092    
1093    37. Perl ignores orphaned \E escapes completely. PCRE now does the same.
1094        There were also incompatibilities regarding the handling of \Q..\E inside
1095        character classes, for example with patterns like [\Qa\E-\Qz\E] where the
1096        hyphen was adjacent to \Q or \E. I hope I've cleared all this up now.
1097    
1098    38. Like Perl, PCRE detects when an indefinitely repeated parenthesized group
1099        matches an empty string, and forcibly breaks the loop. There were bugs in
1100        this code in non-simple cases. For a pattern such as  ^(a()*)*  matched
1101        against  aaaa  the result was just "a" rather than "aaaa", for example. Two
1102        separate and independent bugs (that affected different cases) have been
1103        fixed.
1104    
1105    39. Refactored the code to abolish the use of different opcodes for small
1106        capturing bracket numbers. This is a tidy that I avoided doing when I
1107        removed the limit on the number of capturing brackets for 3.5 back in 2001.
1108        The new approach is not only tidier, it makes it possible to reduce the
1109        memory needed to fix the previous bug (38).
1110    
1111    40. Implemented PCRE_NEWLINE_ANY to recognize any of the Unicode newline
1112        sequences (http://unicode.org/unicode/reports/tr18/) as "newline" when
1113        processing dot, circumflex, or dollar metacharacters, or #-comments in /x
1114        mode.
1115    
1116    41. Add \R to match any Unicode newline sequence, as suggested in the Unicode
1117        report.
1118    
1119    42. Applied patch, originally from Ari Pollak, modified by Google, to allow
1120        copy construction and assignment in the C++ wrapper.
1121    
1122    43. Updated pcregrep to support "--newline=any". In the process, I fixed a
1123        couple of bugs that could have given wrong results in the "--newline=crlf"
1124        case.
1125    
1126    44. Added a number of casts and did some reorganization of signed/unsigned int
1127        variables following suggestions from Dair Grant. Also renamed the variable
1128        "this" as "item" because it is a C++ keyword.
1129    
1130    45. Arranged for dftables to add
1131    
1132          #include "pcre_internal.h"
1133    
1134        to pcre_chartables.c because without it, gcc 4.x may remove the array
1135        definition from the final binary if PCRE is built into a static library and
1136        dead code stripping is activated.
1137    
1138    46. For an unanchored pattern, if a match attempt fails at the start of a
1139        newline sequence, and the newline setting is CRLF or ANY, and the next two
1140        characters are CRLF, advance by two characters instead of one.
1141    
1142    
1143    Version 6.7 04-Jul-06
1144    ---------------------
1145    
1146     1. In order to handle tests when input lines are enormously long, pcretest has
1147        been re-factored so that it automatically extends its buffers when
1148        necessary. The code is crude, but this _is_ just a test program. The
1149        default size has been increased from 32K to 50K.
1150    
1151     2. The code in pcre_study() was using the value of the re argument before
1152        testing it for NULL. (Of course, in any sensible call of the function, it
1153        won't be NULL.)
1154    
1155     3. The memmove() emulation function in pcre_internal.h, which is used on
1156        systems that lack both memmove() and bcopy() - that is, hardly ever -
1157        was missing a "static" storage class specifier.
1158    
1159     4. When UTF-8 mode was not set, PCRE looped when compiling certain patterns
1160        containing an extended class (one that cannot be represented by a bitmap
1161        because it contains high-valued characters or Unicode property items, e.g.
1162        [\pZ]). Almost always one would set UTF-8 mode when processing such a
1163        pattern, but PCRE should not loop if you do not (it no longer does).
1164        [Detail: two cases were found: (a) a repeated subpattern containing an
1165        extended class; (b) a recursive reference to a subpattern that followed a
1166        previous extended class. It wasn't skipping over the extended class
1167        correctly when UTF-8 mode was not set.]
1168    
1169     5. A negated single-character class was not being recognized as fixed-length
1170        in lookbehind assertions such as (?<=[^f]), leading to an incorrect
1171        compile error "lookbehind assertion is not fixed length".
1172    
1173     6. The RunPerlTest auxiliary script was showing an unexpected difference
1174        between PCRE and Perl for UTF-8 tests. It turns out that it is hard to
1175        write a Perl script that can interpret lines of an input file either as
1176        byte characters or as UTF-8, which is what "perltest" was being required to
1177        do for the non-UTF-8 and UTF-8 tests, respectively. Essentially what you
1178        can't do is switch easily at run time between having the "use utf8;" pragma
1179        or not. In the end, I fudged it by using the RunPerlTest script to insert
1180        "use utf8;" explicitly for the UTF-8 tests.
1181    
1182     7. In multiline (/m) mode, PCRE was matching ^ after a terminating newline at
1183        the end of the subject string, contrary to the documentation and to what
1184        Perl does. This was true of both matching functions. Now it matches only at
1185        the start of the subject and immediately after *internal* newlines.
1186    
1187     8. A call of pcre_fullinfo() from pcretest to get the option bits was passing
1188        a pointer to an int instead of a pointer to an unsigned long int. This
1189        caused problems on 64-bit systems.
1190    
1191     9. Applied a patch from the folks at Google to pcrecpp.cc, to fix "another
1192        instance of the 'standard' template library not being so standard".
1193    
1194    10. There was no check on the number of named subpatterns nor the maximum
1195        length of a subpattern name. The product of these values is used to compute
1196        the size of the memory block for a compiled pattern. By supplying a very
1197        long subpattern name and a large number of named subpatterns, the size
1198        computation could be caused to overflow. This is now prevented by limiting
1199        the length of names to 32 characters, and the number of named subpatterns
1200        to 10,000.
1201    
1202    11. Subpatterns that are repeated with specific counts have to be replicated in
1203        the compiled pattern. The size of memory for this was computed from the
1204        length of the subpattern and the repeat count. The latter is limited to
1205        65535, but there was no limit on the former, meaning that integer overflow
1206        could in principle occur. The compiled length of a repeated subpattern is
1207        now limited to 30,000 bytes in order to prevent this.
1208    
1209    12. Added the optional facility to have named substrings with the same name.
1210    
1211    13. Added the ability to use a named substring as a condition, using the
1212        Python syntax: (?(name)yes|no). This overloads (?(R)... and names that
1213        are numbers (not recommended). Forward references are permitted.
1214    
1215    14. Added forward references in named backreferences (if you see what I mean).
1216    
1217    15. In UTF-8 mode, with the PCRE_DOTALL option set, a quantified dot in the
1218        pattern could run off the end of the subject. For example, the pattern
1219        "(?s)(.{1,5})"8 did this with the subject "ab".
1220    
1221    16. If PCRE_DOTALL or PCRE_MULTILINE were set, pcre_dfa_exec() behaved as if
1222        PCRE_CASELESS was set when matching characters that were quantified with ?
1223        or *.
1224    
1225    17. A character class other than a single negated character that had a minimum
1226        but no maximum quantifier - for example [ab]{6,} - was not handled
1227        correctly by pce_dfa_exec(). It would match only one character.
1228    
1229    18. A valid (though odd) pattern that looked like a POSIX character
1230        class but used an invalid character after [ (for example [[,abc,]]) caused
1231        pcre_compile() to give the error "Failed: internal error: code overflow" or
1232        in some cases to crash with a glibc free() error. This could even happen if
1233        the pattern terminated after [[ but there just happened to be a sequence of
1234        letters, a binary zero, and a closing ] in the memory that followed.
1235    
1236    19. Perl's treatment of octal escapes in the range \400 to \777 has changed
1237        over the years. Originally (before any Unicode support), just the bottom 8
1238        bits were taken. Thus, for example, \500 really meant \100. Nowadays the
1239        output from "man perlunicode" includes this:
1240    
1241          The regular expression compiler produces polymorphic opcodes.  That
1242          is, the pattern adapts to the data and automatically switches to
1243          the Unicode character scheme when presented with Unicode data--or
1244          instead uses a traditional byte scheme when presented with byte
1245          data.
1246    
1247        Sadly, a wide octal escape does not cause a switch, and in a string with
1248        no other multibyte characters, these octal escapes are treated as before.
1249        Thus, in Perl, the pattern  /\500/ actually matches \100 but the pattern
1250        /\500|\x{1ff}/ matches \500 or \777 because the whole thing is treated as a
1251        Unicode string.
1252    
1253        I have not perpetrated such confusion in PCRE. Up till now, it took just
1254        the bottom 8 bits, as in old Perl. I have now made octal escapes with
1255        values greater than \377 illegal in non-UTF-8 mode. In UTF-8 mode they
1256        translate to the appropriate multibyte character.
1257    
1258    29. Applied some refactoring to reduce the number of warnings from Microsoft
1259        and Borland compilers. This has included removing the fudge introduced
1260        seven years ago for the OS/2 compiler (see 2.02/2 below) because it caused
1261        a warning about an unused variable.
1262    
1263    21. PCRE has not included VT (character 0x0b) in the set of whitespace
1264        characters since release 4.0, because Perl (from release 5.004) does not.
1265        [Or at least, is documented not to: some releases seem to be in conflict
1266        with the documentation.] However, when a pattern was studied with
1267        pcre_study() and all its branches started with \s, PCRE still included VT
1268        as a possible starting character. Of course, this did no harm; it just
1269        caused an unnecessary match attempt.
1270    
1271    22. Removed a now-redundant internal flag bit that recorded the fact that case
1272        dependency changed within the pattern. This was once needed for "required
1273        byte" processing, but is no longer used. This recovers a now-scarce options
1274        bit. Also moved the least significant internal flag bit to the most-
1275        significant bit of the word, which was not previously used (hangover from
1276        the days when it was an int rather than a uint) to free up another bit for
1277        the future.
1278    
1279    23. Added support for CRLF line endings as well as CR and LF. As well as the
1280        default being selectable at build time, it can now be changed at runtime
1281        via the PCRE_NEWLINE_xxx flags. There are now options for pcregrep to
1282        specify that it is scanning data with non-default line endings.
1283    
1284    24. Changed the definition of CXXLINK to make it agree with the definition of
1285        LINK in the Makefile, by replacing LDFLAGS to CXXFLAGS.
1286    
1287    25. Applied Ian Taylor's patches to avoid using another stack frame for tail
1288        recursions. This makes a big different to stack usage for some patterns.
1289    
1290    26. If a subpattern containing a named recursion or subroutine reference such
1291        as (?P>B) was quantified, for example (xxx(?P>B)){3}, the calculation of
1292        the space required for the compiled pattern went wrong and gave too small a
1293        value. Depending on the environment, this could lead to "Failed: internal
1294        error: code overflow at offset 49" or "glibc detected double free or
1295        corruption" errors.
1296    
1297    27. Applied patches from Google (a) to support the new newline modes and (b) to
1298        advance over multibyte UTF-8 characters in GlobalReplace.
1299    
1300    28. Change free() to pcre_free() in pcredemo.c. Apparently this makes a
1301        difference for some implementation of PCRE in some Windows version.
1302    
1303    29. Added some extra testing facilities to pcretest:
1304    
1305        \q<number>   in a data line sets the "match limit" value
1306        \Q<number>   in a data line sets the "match recursion limt" value
1307        -S <number>  sets the stack size, where <number> is in megabytes
1308    
1309        The -S option isn't available for Windows.
1310    
1311    
1312    Version 6.6 06-Feb-06
1313    ---------------------
1314    
1315     1. Change 16(a) for 6.5 broke things, because PCRE_DATA_SCOPE was not defined
1316        in pcreposix.h. I have copied the definition from pcre.h.
1317    
1318     2. Change 25 for 6.5 broke compilation in a build directory out-of-tree
1319        because pcre.h is no longer a built file.
1320    
1321     3. Added Jeff Friedl's additional debugging patches to pcregrep. These are
1322        not normally included in the compiled code.
1323    
1324    
1325    Version 6.5 01-Feb-06
1326    ---------------------
1327    
1328     1. When using the partial match feature with pcre_dfa_exec(), it was not
1329        anchoring the second and subsequent partial matches at the new starting
1330        point. This could lead to incorrect results. For example, with the pattern
1331        /1234/, partially matching against "123" and then "a4" gave a match.
1332    
1333     2. Changes to pcregrep:
1334    
1335        (a) All non-match returns from pcre_exec() were being treated as failures
1336            to match the line. Now, unless the error is PCRE_ERROR_NOMATCH, an
1337            error message is output. Some extra information is given for the
1338            PCRE_ERROR_MATCHLIMIT and PCRE_ERROR_RECURSIONLIMIT errors, which are
1339            probably the only errors that are likely to be caused by users (by
1340            specifying a regex that has nested indefinite repeats, for instance).
1341            If there are more than 20 of these errors, pcregrep is abandoned.
1342    
1343        (b) A binary zero was treated as data while matching, but terminated the
1344            output line if it was written out. This has been fixed: binary zeroes
1345            are now no different to any other data bytes.
1346    
1347        (c) Whichever of the LC_ALL or LC_CTYPE environment variables is set is
1348            used to set a locale for matching. The --locale=xxxx long option has
1349            been added (no short equivalent) to specify a locale explicitly on the
1350            pcregrep command, overriding the environment variables.
1351    
1352        (d) When -B was used with -n, some line numbers in the output were one less
1353            than they should have been.
1354    
1355        (e) Added the -o (--only-matching) option.
1356    
1357        (f) If -A or -C was used with -c (count only), some lines of context were
1358            accidentally printed for the final match.
1359    
1360        (g) Added the -H (--with-filename) option.
1361    
1362        (h) The combination of options -rh failed to suppress file names for files
1363            that were found from directory arguments.
1364    
1365        (i) Added the -D (--devices) and -d (--directories) options.
1366    
1367        (j) Added the -F (--fixed-strings) option.
1368    
1369        (k) Allow "-" to be used as a file name for -f as well as for a data file.
1370    
1371        (l) Added the --colo(u)r option.
1372    
1373        (m) Added Jeffrey Friedl's -S testing option, but within #ifdefs so that it
1374            is not present by default.
1375    
1376     3. A nasty bug was discovered in the handling of recursive patterns, that is,
1377        items such as (?R) or (?1), when the recursion could match a number of
1378        alternatives. If it matched one of the alternatives, but subsequently,
1379        outside the recursion, there was a failure, the code tried to back up into
1380        the recursion. However, because of the way PCRE is implemented, this is not
1381        possible, and the result was an incorrect result from the match.
1382    
1383        In order to prevent this happening, the specification of recursion has
1384        been changed so that all such subpatterns are automatically treated as
1385        atomic groups. Thus, for example, (?R) is treated as if it were (?>(?R)).
1386    
1387     4. I had overlooked the fact that, in some locales, there are characters for
1388        which isalpha() is true but neither isupper() nor islower() are true. In
1389        the fr_FR locale, for instance, the \xAA and \xBA characters (ordmasculine
1390        and ordfeminine) are like this. This affected the treatment of \w and \W
1391        when they appeared in character classes, but not when they appeared outside
1392        a character class. The bit map for "word" characters is now created
1393        separately from the results of isalnum() instead of just taking it from the
1394        upper, lower, and digit maps. (Plus the underscore character, of course.)
1395    
1396     5. The above bug also affected the handling of POSIX character classes such as
1397        [[:alpha:]] and [[:alnum:]]. These do not have their own bit maps in PCRE's
1398        permanent tables. Instead, the bit maps for such a class were previously
1399        created as the appropriate unions of the upper, lower, and digit bitmaps.
1400        Now they are created by subtraction from the [[:word:]] class, which has
1401        its own bitmap.
1402    
1403     6. The [[:blank:]] character class matches horizontal, but not vertical space.
1404        It is created by subtracting the vertical space characters (\x09, \x0a,
1405        \x0b, \x0c) from the [[:space:]] bitmap. Previously, however, the
1406        subtraction was done in the overall bitmap for a character class, meaning
1407        that a class such as [\x0c[:blank:]] was incorrect because \x0c would not
1408        be recognized. This bug has been fixed.
1409    
1410     7. Patches from the folks at Google:
1411    
1412          (a) pcrecpp.cc: "to handle a corner case that may or may not happen in
1413          real life, but is still worth protecting against".
1414    
1415          (b) pcrecpp.cc: "corrects a bug when negative radixes are used with
1416          regular expressions".
1417    
1418          (c) pcre_scanner.cc: avoid use of std::count() because not all systems
1419          have it.
1420    
1421          (d) Split off pcrecpparg.h from pcrecpp.h and had the former built by
1422          "configure" and the latter not, in order to fix a problem somebody had
1423          with compiling the Arg class on HP-UX.
1424    
1425          (e) Improve the error-handling of the C++ wrapper a little bit.
1426    
1427          (f) New tests for checking recursion limiting.
1428    
1429     8. The pcre_memmove() function, which is used only if the environment does not
1430        have a standard memmove() function (and is therefore rarely compiled),
1431        contained two bugs: (a) use of int instead of size_t, and (b) it was not
1432        returning a result (though PCRE never actually uses the result).
1433    
1434     9. In the POSIX regexec() interface, if nmatch is specified as a ridiculously
1435        large number - greater than INT_MAX/(3*sizeof(int)) - REG_ESPACE is
1436        returned instead of calling malloc() with an overflowing number that would
1437        most likely cause subsequent chaos.
1438    
1439    10. The debugging option of pcretest was not showing the NO_AUTO_CAPTURE flag.
1440    
1441    11. The POSIX flag REG_NOSUB is now supported. When a pattern that was compiled
1442        with this option is matched, the nmatch and pmatch options of regexec() are
1443        ignored.
1444    
1445    12. Added REG_UTF8 to the POSIX interface. This is not defined by POSIX, but is
1446        provided in case anyone wants to the the POSIX interface with UTF-8
1447        strings.
1448    
1449    13. Added CXXLDFLAGS to the Makefile parameters to provide settings only on the
1450        C++ linking (needed for some HP-UX environments).
1451    
1452    14. Avoid compiler warnings in get_ucpname() when compiled without UCP support
1453        (unused parameter) and in the pcre_printint() function (omitted "default"
1454        switch label when the default is to do nothing).
1455    
1456    15. Added some code to make it possible, when PCRE is compiled as a C++
1457        library, to replace subject pointers for pcre_exec() with a smart pointer
1458        class, thus making it possible to process discontinuous strings.
1459    
1460    16. The two macros PCRE_EXPORT and PCRE_DATA_SCOPE are confusing, and perform
1461        much the same function. They were added by different people who were trying
1462        to make PCRE easy to compile on non-Unix systems. It has been suggested
1463        that PCRE_EXPORT be abolished now that there is more automatic apparatus
1464        for compiling on Windows systems. I have therefore replaced it with
1465        PCRE_DATA_SCOPE. This is set automatically for Windows; if not set it
1466        defaults to "extern" for C or "extern C" for C++, which works fine on
1467        Unix-like systems. It is now possible to override the value of PCRE_DATA_
1468        SCOPE with something explicit in config.h. In addition:
1469    
1470        (a) pcreposix.h still had just "extern" instead of either of these macros;
1471            I have replaced it with PCRE_DATA_SCOPE.
1472    
1473        (b) Functions such as _pcre_xclass(), which are internal to the library,
1474            but external in the C sense, all had PCRE_EXPORT in their definitions.
1475            This is apparently wrong for the Windows case, so I have removed it.
1476            (It makes no difference on Unix-like systems.)
1477    
1478    17. Added a new limit, MATCH_LIMIT_RECURSION, which limits the depth of nesting
1479        of recursive calls to match(). This is different to MATCH_LIMIT because
1480        that limits the total number of calls to match(), not all of which increase
1481        the depth of recursion. Limiting the recursion depth limits the amount of
1482        stack (or heap if NO_RECURSE is set) that is used. The default can be set
1483        when PCRE is compiled, and changed at run time. A patch from Google adds
1484        this functionality to the C++ interface.
1485    
1486    18. Changes to the handling of Unicode character properties:
1487    
1488        (a) Updated the table to Unicode 4.1.0.
1489    
1490        (b) Recognize characters that are not in the table as "Cn" (undefined).
1491    
1492        (c) I revised the way the table is implemented to a much improved format
1493            which includes recognition of ranges. It now supports the ranges that
1494            are defined in UnicodeData.txt, and it also amalgamates other
1495            characters into ranges. This has reduced the number of entries in the
1496            table from around 16,000 to around 3,000, thus reducing its size
1497            considerably. I realized I did not need to use a tree structure after
1498            all - a binary chop search is just as efficient. Having reduced the
1499            number of entries, I extended their size from 6 bytes to 8 bytes to
1500            allow for more data.
1501    
1502        (d) Added support for Unicode script names via properties such as \p{Han}.
1503    
1504    19. In UTF-8 mode, a backslash followed by a non-Ascii character was not
1505        matching that character.
1506    
1507    20. When matching a repeated Unicode property with a minimum greater than zero,
1508        (for example \pL{2,}), PCRE could look past the end of the subject if it
1509        reached it while seeking the minimum number of characters. This could
1510        happen only if some of the characters were more than one byte long, because
1511        there is a check for at least the minimum number of bytes.
1512    
1513    21. Refactored the implementation of \p and \P so as to be more general, to
1514        allow for more different types of property in future. This has changed the
1515        compiled form incompatibly. Anybody with saved compiled patterns that use
1516        \p or \P will have to recompile them.
1517    
1518    22. Added "Any" and "L&" to the supported property types.
1519    
1520    23. Recognize \x{...} as a code point specifier, even when not in UTF-8 mode,
1521        but give a compile time error if the value is greater than 0xff.
1522    
1523    24. The man pages for pcrepartial, pcreprecompile, and pcre_compile2 were
1524        accidentally not being installed or uninstalled.
1525    
1526    25. The pcre.h file was built from pcre.h.in, but the only changes that were
1527        made were to insert the current release number. This seemed silly, because
1528        it made things harder for people building PCRE on systems that don't run
1529        "configure". I have turned pcre.h into a distributed file, no longer built
1530        by "configure", with the version identification directly included. There is
1531        no longer a pcre.h.in file.
1532    
1533        However, this change necessitated a change to the pcre-config script as
1534        well. It is built from pcre-config.in, and one of the substitutions was the
1535        release number. I have updated configure.ac so that ./configure now finds
1536        the release number by grepping pcre.h.
1537    
1538    26. Added the ability to run the tests under valgrind.
1539    
1540    
1541    Version 6.4 05-Sep-05
1542    ---------------------
1543    
1544     1. Change 6.0/10/(l) to pcregrep introduced a bug that caused separator lines
1545        "--" to be printed when multiple files were scanned, even when none of the
1546        -A, -B, or -C options were used. This is not compatible with Gnu grep, so I
1547        consider it to be a bug, and have restored the previous behaviour.
1548    
1549     2. A couple of code tidies to get rid of compiler warnings.
1550    
1551     3. The pcretest program used to cheat by referring to symbols in the library
1552        whose names begin with _pcre_. These are internal symbols that are not
1553        really supposed to be visible externally, and in some environments it is
1554        possible to suppress them. The cheating is now confined to including
1555        certain files from the library's source, which is a bit cleaner.
1556    
1557     4. Renamed pcre.in as pcre.h.in to go with pcrecpp.h.in; it also makes the
1558        file's purpose clearer.
1559    
1560     5. Reorganized pcre_ucp_findchar().
1561    
1562    
1563    Version 6.3 15-Aug-05
1564    ---------------------
1565    
1566     1. The file libpcre.pc.in did not have general read permission in the tarball.
1567    
1568     2. There were some problems when building without C++ support:
1569    
1570        (a) If C++ support was not built, "make install" and "make test" still
1571            tried to test it.
1572    
1573        (b) There were problems when the value of CXX was explicitly set. Some
1574            changes have been made to try to fix these, and ...
1575    
1576        (c) --disable-cpp can now be used to explicitly disable C++ support.
1577    
1578        (d) The use of @CPP_OBJ@ directly caused a blank line preceded by a
1579            backslash in a target when C++ was disabled. This confuses some
1580            versions of "make", apparently. Using an intermediate variable solves
1581            this. (Same for CPP_LOBJ.)
1582    
1583     3. $(LINK_FOR_BUILD) now includes $(CFLAGS_FOR_BUILD) and $(LINK)
1584        (non-Windows) now includes $(CFLAGS) because these flags are sometimes
1585        necessary on certain architectures.
1586    
1587     4. Added a setting of -export-symbols-regex to the link command to remove
1588        those symbols that are exported in the C sense, but actually are local
1589        within the library, and not documented. Their names all begin with
1590        "_pcre_". This is not a perfect job, because (a) we have to except some
1591        symbols that pcretest ("illegally") uses, and (b) the facility isn't always
1592        available (and never for static libraries). I have made a note to try to
1593        find a way round (a) in the future.
1594    
1595    
1596    Version 6.2 01-Aug-05
1597    ---------------------
1598    
1599     1. There was no test for integer overflow of quantifier values. A construction
1600        such as {1111111111111111} would give undefined results. What is worse, if
1601        a minimum quantifier for a parenthesized subpattern overflowed and became
1602        negative, the calculation of the memory size went wrong. This could have
1603        led to memory overwriting.
1604    
1605     2. Building PCRE using VPATH was broken. Hopefully it is now fixed.
1606    
1607     3. Added "b" to the 2nd argument of fopen() in dftables.c, for non-Unix-like
1608        operating environments where this matters.
1609    
1610     4. Applied Giuseppe Maxia's patch to add additional features for controlling
1611        PCRE options from within the C++ wrapper.
1612    
1613     5. Named capturing subpatterns were not being correctly counted when a pattern
1614        was compiled. This caused two problems: (a) If there were more than 100
1615        such subpatterns, the calculation of the memory needed for the whole
1616        compiled pattern went wrong, leading to an overflow error. (b) Numerical
1617        back references of the form \12, where the number was greater than 9, were
1618        not recognized as back references, even though there were sufficient
1619        previous subpatterns.
1620    
1621     6. Two minor patches to pcrecpp.cc in order to allow it to compile on older
1622        versions of gcc, e.g. 2.95.4.
1623    
1624    
1625    Version 6.1 21-Jun-05
1626    ---------------------
1627    
1628     1. There was one reference to the variable "posix" in pcretest.c that was not
1629        surrounded by "#if !defined NOPOSIX".
1630    
1631     2. Make it possible to compile pcretest without DFA support, UTF8 support, or
1632        the cross-check on the old pcre_info() function, for the benefit of the
1633        cut-down version of PCRE that is currently imported into Exim.
1634    
1635     3. A (silly) pattern starting with (?i)(?-i) caused an internal space
1636        allocation error. I've done the easy fix, which wastes 2 bytes for sensible
1637        patterns that start (?i) but I don't think that matters. The use of (?i) is
1638        just an example; this all applies to the other options as well.
1639    
1640     4. Since libtool seems to echo the compile commands it is issuing, the output
1641        from "make" can be reduced a bit by putting "@" in front of each libtool
1642        compile command.
1643    
1644     5. Patch from the folks at Google for configure.in to be a bit more thorough
1645        in checking for a suitable C++ installation before trying to compile the
1646        C++ stuff. This should fix a reported problem when a compiler was present,
1647        but no suitable headers.
1648    
1649     6. The man pages all had just "PCRE" as their title. I have changed them to
1650        be the relevant file name. I have also arranged that these names are
1651        retained in the file doc/pcre.txt, which is a concatenation in text format
1652        of all the man pages except the little individual ones for each function.
1653    
1654     7. The NON-UNIX-USE file had not been updated for the different set of source
1655        files that come with release 6. I also added a few comments about the C++
1656        wrapper.
1657    
1658    
1659    Version 6.0 07-Jun-05
1660    ---------------------
1661    
1662     1. Some minor internal re-organization to help with my DFA experiments.
1663    
1664     2. Some missing #ifdef SUPPORT_UCP conditionals in pcretest and printint that
1665        didn't matter for the library itself when fully configured, but did matter
1666        when compiling without UCP support, or within Exim, where the ucp files are
1667        not imported.
1668    
1669     3. Refactoring of the library code to split up the various functions into
1670        different source modules. The addition of the new DFA matching code (see
1671        below) to a single monolithic source would have made it really too
1672        unwieldy, quite apart from causing all the code to be include in a
1673        statically linked application, when only some functions are used. This is
1674        relevant even without the DFA addition now that patterns can be compiled in
1675        one application and matched in another.
1676    
1677        The downside of splitting up is that there have to be some external
1678        functions and data tables that are used internally in different modules of
1679        the library but which are not part of the API. These have all had their
1680        names changed to start with "_pcre_" so that they are unlikely to clash
1681        with other external names.
1682    
1683     4. Added an alternate matching function, pcre_dfa_exec(), which matches using
1684        a different (DFA) algorithm. Although it is slower than the original
1685        function, it does have some advantages for certain types of matching
1686        problem.
1687    
1688     5. Upgrades to pcretest in order to test the features of pcre_dfa_exec(),
1689        including restarting after a partial match.
1690    
1691     6. A patch for pcregrep that defines INVALID_FILE_ATTRIBUTES if it is not
1692        defined when compiling for Windows was sent to me. I have put it into the
1693        code, though I have no means of testing or verifying it.
1694    
1695     7. Added the pcre_refcount() auxiliary function.
1696    
1697     8. Added the PCRE_FIRSTLINE option. This constrains an unanchored pattern to
1698        match before or at the first newline in the subject string. In pcretest,
1699        the /f option on a pattern can be used to set this.
1700    
1701     9. A repeated \w when used in UTF-8 mode with characters greater than 256
1702        would behave wrongly. This has been present in PCRE since release 4.0.
1703    
1704    10. A number of changes to the pcregrep command:
1705    
1706        (a) Refactored how -x works; insert ^(...)$ instead of setting
1707            PCRE_ANCHORED and checking the length, in preparation for adding
1708            something similar for -w.
1709    
1710        (b) Added the -w (match as a word) option.
1711    
1712        (c) Refactored the way lines are read and buffered so as to have more
1713            than one at a time available.
1714    
1715        (d) Implemented a pcregrep test script.
1716    
1717        (e) Added the -M (multiline match) option. This allows patterns to match
1718            over several lines of the subject. The buffering ensures that at least
1719            8K, or the rest of the document (whichever is the shorter) is available
1720            for matching (and similarly the previous 8K for lookbehind assertions).
1721    
1722        (f) Changed the --help output so that it now says
1723    
1724              -w, --word-regex(p)
1725    
1726            instead of two lines, one with "regex" and the other with "regexp"
1727            because that confused at least one person since the short forms are the
1728            same. (This required a bit of code, as the output is generated
1729            automatically from a table. It wasn't just a text change.)
1730    
1731        (g) -- can be used to terminate pcregrep options if the next thing isn't an
1732            option but starts with a hyphen. Could be a pattern or a path name
1733            starting with a hyphen, for instance.
1734    
1735        (h) "-" can be given as a file name to represent stdin.
1736    
1737        (i) When file names are being printed, "(standard input)" is used for
1738            the standard input, for compatibility with GNU grep. Previously
1739            "<stdin>" was used.
1740    
1741        (j) The option --label=xxx can be used to supply a name to be used for
1742            stdin when file names are being printed. There is no short form.
1743    
1744        (k) Re-factored the options decoding logic because we are going to add
1745            two more options that take data. Such options can now be given in four
1746            different ways, e.g. "-fname", "-f name", "--file=name", "--file name".
1747    
1748        (l) Added the -A, -B, and -C options for requesting that lines of context
1749            around matches be printed.
1750    
1751        (m) Added the -L option to print the names of files that do not contain
1752            any matching lines, that is, the complement of -l.
1753    
1754        (n) The return code is 2 if any file cannot be opened, but pcregrep does
1755            continue to scan other files.
1756    
1757        (o) The -s option was incorrectly implemented. For compatibility with other
1758            greps, it now suppresses the error message for a non-existent or non-
1759            accessible file (but not the return code). There is a new option called
1760            -q that suppresses the output of matching lines, which was what -s was
1761            previously doing.
1762    
1763        (p) Added --include and --exclude options to specify files for inclusion
1764            and exclusion when recursing.
1765    
1766    11. The Makefile was not using the Autoconf-supported LDFLAGS macro properly.
1767        Hopefully, it now does.
1768    
1769    12. Missing cast in pcre_study().
1770    
1771    13. Added an "uninstall" target to the makefile.
1772    
1773    14. Replaced "extern" in the function prototypes in Makefile.in with
1774        "PCRE_DATA_SCOPE", which defaults to 'extern' or 'extern "C"' in the Unix
1775        world, but is set differently for Windows.
1776    
1777    15. Added a second compiling function called pcre_compile2(). The only
1778        difference is that it has an extra argument, which is a pointer to an
1779        integer error code. When there is a compile-time failure, this is set
1780        non-zero, in addition to the error test pointer being set to point to an
1781        error message. The new argument may be NULL if no error number is required
1782        (but then you may as well call pcre_compile(), which is now just a
1783        wrapper). This facility is provided because some applications need a
1784        numeric error indication, but it has also enabled me to tidy up the way
1785        compile-time errors are handled in the POSIX wrapper.
1786    
1787    16. Added VPATH=.libs to the makefile; this should help when building with one
1788        prefix path and installing with another. (Or so I'm told by someone who
1789        knows more about this stuff than I do.)
1790    
1791    17. Added a new option, REG_DOTALL, to the POSIX function regcomp(). This
1792        passes PCRE_DOTALL to the pcre_compile() function, making the "." character
1793        match everything, including newlines. This is not POSIX-compatible, but
1794        somebody wanted the feature. From pcretest it can be activated by using
1795        both the P and the s flags.
1796    
1797    18. AC_PROG_LIBTOOL appeared twice in Makefile.in. Removed one.
1798    
1799    19. libpcre.pc was being incorrectly installed as executable.
1800    
1801    20. A couple of places in pcretest check for end-of-line by looking for '\n';
1802        it now also looks for '\r' so that it will work unmodified on Windows.
1803    
1804    21. Added Google's contributed C++ wrapper to the distribution.
1805    
1806    22. Added some untidy missing memory free() calls in pcretest, to keep
1807        Electric Fence happy when testing.
1808    
1809    
1810    
1811    Version 5.0 13-Sep-04
1812    ---------------------
1813    
1814     1. Internal change: literal characters are no longer packed up into items
1815        containing multiple characters in a single byte-string. Each character
1816        is now matched using a separate opcode. However, there may be more than one
1817        byte in the character in UTF-8 mode.
1818    
1819     2. The pcre_callout_block structure has two new fields: pattern_position and
1820        next_item_length. These contain the offset in the pattern to the next match
1821        item, and its length, respectively.
1822    
1823     3. The PCRE_AUTO_CALLOUT option for pcre_compile() requests the automatic
1824        insertion of callouts before each pattern item. Added the /C option to
1825        pcretest to make use of this.
1826    
1827     4. On the advice of a Windows user, the lines
1828    
1829          #if defined(_WIN32) || defined(WIN32)
1830          _setmode( _fileno( stdout ), 0x8000 );
1831          #endif  /* defined(_WIN32) || defined(WIN32) */
1832    
1833        have been added to the source of pcretest. This apparently does useful
1834        magic in relation to line terminators.
1835    
1836     5. Changed "r" and "w" in the calls to fopen() in pcretest to "rb" and "wb"
1837        for the benefit of those environments where the "b" makes a difference.
1838    
1839     6. The icc compiler has the same options as gcc, but "configure" doesn't seem
1840        to know about it. I have put a hack into configure.in that adds in code
1841        to set GCC=yes if CC=icc. This seems to end up at a point in the
1842        generated configure script that is early enough to affect the setting of
1843        compiler options, which is what is needed, but I have no means of testing
1844        whether it really works. (The user who reported this had patched the
1845        generated configure script, which of course I cannot do.)
1846    
1847        LATER: After change 22 below (new libtool files), the configure script
1848        seems to know about icc (and also ecc). Therefore, I have commented out
1849        this hack in configure.in.
1850    
1851     7. Added support for pkg-config (2 patches were sent in).
1852    
1853     8. Negated POSIX character classes that used a combination of internal tables
1854        were completely broken. These were [[:^alpha:]], [[:^alnum:]], and
1855        [[:^ascii]]. Typically, they would match almost any characters. The other
1856        POSIX classes were not broken in this way.
1857    
1858     9. Matching the pattern "\b.*?" against "ab cd", starting at offset 1, failed
1859        to find the match, as PCRE was deluded into thinking that the match had to
1860        start at the start point or following a newline. The same bug applied to
1861        patterns with negative forward assertions or any backward assertions
1862        preceding ".*" at the start, unless the pattern required a fixed first
1863        character. This was a failing pattern: "(?!.bcd).*". The bug is now fixed.
1864    
1865    10. In UTF-8 mode, when moving forwards in the subject after a failed match
1866        starting at the last subject character, bytes beyond the end of the subject
1867        string were read.
1868    
1869    11. Renamed the variable "class" as "classbits" to make life easier for C++
1870        users. (Previously there was a macro definition, but it apparently wasn't
1871        enough.)
1872    
1873    12. Added the new field "tables" to the extra data so that tables can be passed
1874        in at exec time, or the internal tables can be re-selected. This allows
1875        a compiled regex to be saved and re-used at a later time by a different
1876        program that might have everything at different addresses.
1877    
1878    13. Modified the pcre-config script so that, when run on Solaris, it shows a
1879        -R library as well as a -L library.
1880    
1881    14. The debugging options of pcretest (-d on the command line or D on a
1882        pattern) showed incorrect output for anything following an extended class
1883        that contained multibyte characters and which was followed by a quantifier.
1884    
1885    15. Added optional support for general category Unicode character properties
1886        via the \p, \P, and \X escapes. Unicode property support implies UTF-8
1887        support. It adds about 90K to the size of the library. The meanings of the
1888        inbuilt class escapes such as \d and \s have NOT been changed.
1889    
1890    16. Updated pcredemo.c to include calls to free() to release the memory for the
1891        compiled pattern.
1892    
1893    17. The generated file chartables.c was being created in the source directory
1894        instead of in the building directory. This caused the build to fail if the
1895        source directory was different from the building directory, and was
1896        read-only.
1897    
1898    18. Added some sample Win commands from Mark Tetrode into the NON-UNIX-USE
1899        file. No doubt somebody will tell me if they don't make sense... Also added
1900        Dan Mooney's comments about building on OpenVMS.
1901    
1902    19. Added support for partial matching via the PCRE_PARTIAL option for
1903        pcre_exec() and the \P data escape in pcretest.
1904    
1905    20. Extended pcretest with 3 new pattern features:
1906    
1907        (i)   A pattern option of the form ">rest-of-line" causes pcretest to
1908              write the compiled pattern to the file whose name is "rest-of-line".
1909              This is a straight binary dump of the data, with the saved pointer to
1910              the character tables forced to be NULL. The study data, if any, is
1911              written too. After writing, pcretest reads a new pattern.
1912    
1913        (ii)  If, instead of a pattern, "<rest-of-line" is given, pcretest reads a
1914              compiled pattern from the given file. There must not be any
1915              occurrences of "<" in the file name (pretty unlikely); if there are,
1916              pcretest will instead treat the initial "<" as a pattern delimiter.
1917              After reading in the pattern, pcretest goes on to read data lines as
1918              usual.
1919    
1920        (iii) The F pattern option causes pcretest to flip the bytes in the 32-bit
1921              and 16-bit fields in a compiled pattern, to simulate a pattern that
1922              was compiled on a host of opposite endianness.
1923    
1924    21. The pcre-exec() function can now cope with patterns that were compiled on
1925        hosts of opposite endianness, with this restriction:
1926    
1927          As for any compiled expression that is saved and used later, the tables
1928          pointer field cannot be preserved; the extra_data field in the arguments
1929          to pcre_exec() should be used to pass in a tables address if a value
1930          other than the default internal tables were used at compile time.
1931    
1932    22. Calling pcre_exec() with a negative value of the "ovecsize" parameter is
1933        now diagnosed as an error. Previously, most of the time, a negative number
1934        would have been treated as zero, but if in addition "ovector" was passed as
1935        NULL, a crash could occur.
1936    
1937    23. Updated the files ltmain.sh, config.sub, config.guess, and aclocal.m4 with
1938        new versions from the libtool 1.5 distribution (the last one is a copy of
1939        a file called libtool.m4). This seems to have fixed the need to patch
1940        "configure" to support Darwin 1.3 (which I used to do). However, I still
1941        had to patch ltmain.sh to ensure that ${SED} is set (it isn't on my
1942        workstation).
1943    
1944    24. Changed the PCRE licence to be the more standard "BSD" licence.
1945    
1946    
1947    Version 4.5 01-Dec-03
1948    ---------------------
1949    
1950     1. There has been some re-arrangement of the code for the match() function so
1951        that it can be compiled in a version that does not call itself recursively.
1952        Instead, it keeps those local variables that need separate instances for
1953        each "recursion" in a frame on the heap, and gets/frees frames whenever it
1954        needs to "recurse". Keeping track of where control must go is done by means
1955        of setjmp/longjmp. The whole thing is implemented by a set of macros that
1956        hide most of the details from the main code, and operates only if
1957        NO_RECURSE is defined while compiling pcre.c. If PCRE is built using the
1958        "configure" mechanism, "--disable-stack-for-recursion" turns on this way of
1959        operating.
1960    
1961        To make it easier for callers to provide specially tailored get/free
1962        functions for this usage, two new functions, pcre_stack_malloc, and
1963        pcre_stack_free, are used. They are always called in strict stacking order,
1964        and the size of block requested is always the same.
1965    
1966        The PCRE_CONFIG_STACKRECURSE info parameter can be used to find out whether
1967        PCRE has been compiled to use the stack or the heap for recursion. The
1968        -C option of pcretest uses this to show which version is compiled.
1969    
1970        A new data escape \S, is added to pcretest; it causes the amounts of store
1971        obtained and freed by both kinds of malloc/free at match time to be added
1972        to the output.
1973    
1974     2. Changed the locale test to use "fr_FR" instead of "fr" because that's
1975        what's available on my current Linux desktop machine.
1976    
1977     3. When matching a UTF-8 string, the test for a valid string at the start has
1978        been extended. If start_offset is not zero, PCRE now checks that it points
1979        to a byte that is the start of a UTF-8 character. If not, it returns
1980        PCRE_ERROR_BADUTF8_OFFSET (-11). Note: the whole string is still checked;
1981        this is necessary because there may be backward assertions in the pattern.
1982        When matching the same subject several times, it may save resources to use
1983        PCRE_NO_UTF8_CHECK on all but the first call if the string is long.
1984    
1985     4. The code for checking the validity of UTF-8 strings has been tightened so
1986        that it rejects (a) strings containing 0xfe or 0xff bytes and (b) strings
1987        containing "overlong sequences".
1988    
1989     5. Fixed a bug (appearing twice) that I could not find any way of exploiting!
1990        I had written "if ((digitab[*p++] && chtab_digit) == 0)" where the "&&"
1991        should have been "&", but it just so happened that all the cases this let
1992        through by mistake were picked up later in the function.
1993    
1994     6. I had used a variable called "isblank" - this is a C99 function, causing
1995        some compilers to warn. To avoid this, I renamed it (as "blankclass").
1996    
1997     7. Cosmetic: (a) only output another newline at the end of pcretest if it is
1998        prompting; (b) run "./pcretest /dev/null" at the start of the test script
1999        so the version is shown; (c) stop "make test" echoing "./RunTest".
2000    
2001     8. Added patches from David Burgess to enable PCRE to run on EBCDIC systems.
2002    
2003     9. The prototype for memmove() for systems that don't have it was using
2004        size_t, but the inclusion of the header that defines size_t was later. I've
2005        moved the #includes for the C headers earlier to avoid this.
2006    
2007    10. Added some adjustments to the code to make it easier to compiler on certain
2008        special systems:
2009    
2010          (a) Some "const" qualifiers were missing.
2011          (b) Added the macro EXPORT before all exported functions; by default this
2012              is defined to be empty.
2013          (c) Changed the dftables auxiliary program (that builds chartables.c) so
2014              that it reads its output file name as an argument instead of writing
2015              to the standard output and assuming this can be redirected.
2016    
2017    11. In UTF-8 mode, if a recursive reference (e.g. (?1)) followed a character
2018        class containing characters with values greater than 255, PCRE compilation
2019        went into a loop.
2020    
2021    12. A recursive reference to a subpattern that was within another subpattern
2022        that had a minimum quantifier of zero caused PCRE to crash. For example,
2023        (x(y(?2))z)? provoked this bug with a subject that got as far as the
2024        recursion. If the recursively-called subpattern itself had a zero repeat,
2025        that was OK.
2026    
2027    13. In pcretest, the buffer for reading a data line was set at 30K, but the
2028        buffer into which it was copied (for escape processing) was still set at
2029        1024, so long lines caused crashes.
2030    
2031    14. A pattern such as /[ab]{1,3}+/ failed to compile, giving the error
2032        "internal error: code overflow...". This applied to any character class
2033        that was followed by a possessive quantifier.
2034    
2035    15. Modified the Makefile to add libpcre.la as a prerequisite for
2036        libpcreposix.la because I was told this is needed for a parallel build to
2037        work.
2038    
2039    16. If a pattern that contained .* following optional items at the start was
2040        studied, the wrong optimizing data was generated, leading to matching
2041        errors. For example, studying /[ab]*.*c/ concluded, erroneously, that any
2042        matching string must start with a or b or c. The correct conclusion for
2043        this pattern is that a match can start with any character.
2044    
2045    
2046    Version 4.4 13-Aug-03
2047    ---------------------
2048    
2049     1. In UTF-8 mode, a character class containing characters with values between
2050        127 and 255 was not handled correctly if the compiled pattern was studied.
2051        In fixing this, I have also improved the studying algorithm for such
2052        classes (slightly).
2053    
2054     2. Three internal functions had redundant arguments passed to them. Removal
2055        might give a very teeny performance improvement.
2056    
2057     3. Documentation bug: the value of the capture_top field in a callout is *one
2058        more than* the number of the hightest numbered captured substring.
2059    
2060     4. The Makefile linked pcretest and pcregrep with -lpcre, which could result
2061        in incorrectly linking with a previously installed version. They now link
2062        explicitly with libpcre.la.
2063    
2064     5. configure.in no longer needs to recognize Cygwin specially.
2065    
2066     6. A problem in pcre.in for Windows platforms is fixed.
2067    
2068     7. If a pattern was successfully studied, and the -d (or /D) flag was given to
2069        pcretest, it used to include the size of the study block as part of its
2070        output. Unfortunately, the structure contains a field that has a different
2071        size on different hardware architectures. This meant that the tests that
2072        showed this size failed. As the block is currently always of a fixed size,
2073        this information isn't actually particularly useful in pcretest output, so
2074        I have just removed it.
2075    
2076     8. Three pre-processor statements accidentally did not start in column 1.
2077        Sadly, there are *still* compilers around that complain, even though
2078        standard C has not required this for well over a decade. Sigh.
2079    
2080     9. In pcretest, the code for checking callouts passed small integers in the
2081        callout_data field, which is a void * field. However, some picky compilers
2082        complained about the casts involved for this on 64-bit systems. Now
2083        pcretest passes the address of the small integer instead, which should get
2084        rid of the warnings.
2085    
2086    10. By default, when in UTF-8 mode, PCRE now checks for valid UTF-8 strings at
2087        both compile and run time, and gives an error if an invalid UTF-8 sequence
2088        is found. There is a option for disabling this check in cases where the
2089        string is known to be correct and/or the maximum performance is wanted.
2090    
2091    11. In response to a bug report, I changed one line in Makefile.in from
2092    
2093            -Wl,--out-implib,.libs/lib@WIN_PREFIX@pcreposix.dll.a \
2094        to
2095            -Wl,--out-implib,.libs/@WIN_PREFIX@libpcreposix.dll.a \
2096    
2097        to look similar to other lines, but I have no way of telling whether this
2098        is the right thing to do, as I do not use Windows. No doubt I'll get told
2099        if it's wrong...
2100    
2101    
2102    Version 4.3 21-May-03
2103    ---------------------
2104    
2105    1. Two instances of @WIN_PREFIX@ omitted from the Windows targets in the
2106       Makefile.
2107    
2108    2. Some refactoring to improve the quality of the code:
2109    
2110       (i)   The utf8_table... variables are now declared "const".
2111    
2112       (ii)  The code for \cx, which used the "case flipping" table to upper case
2113             lower case letters, now just substracts 32. This is ASCII-specific,
2114             but the whole concept of \cx is ASCII-specific, so it seems
2115             reasonable.
2116    
2117       (iii) PCRE was using its character types table to recognize decimal and
2118             hexadecimal digits in the pattern. This is silly, because it handles
2119             only 0-9, a-f, and A-F, but the character types table is locale-
2120             specific, which means strange things might happen. A private
2121             table is now used for this - though it costs 256 bytes, a table is
2122             much faster than multiple explicit tests. Of course, the standard
2123             character types table is still used for matching digits in subject
2124             strings against \d.
2125    
2126       (iv)  Strictly, the identifier ESC_t is reserved by POSIX (all identifiers
2127             ending in _t are). So I've renamed it as ESC_tee.
2128    
2129    3. The first argument for regexec() in the POSIX wrapper should have been
2130       defined as "const".
2131    
2132    4. Changed pcretest to use malloc() for its buffers so that they can be
2133       Electric Fenced for debugging.
2134    
2135    5. There were several places in the code where, in UTF-8 mode, PCRE would try
2136       to read one or more bytes before the start of the subject string. Often this
2137       had no effect on PCRE's behaviour, but in some circumstances it could
2138       provoke a segmentation fault.
2139    
2140    6. A lookbehind at the start of a pattern in UTF-8 mode could also cause PCRE
2141       to try to read one or more bytes before the start of the subject string.
2142    
2143    7. A lookbehind in a pattern matched in non-UTF-8 mode on a PCRE compiled with
2144       UTF-8 support could misbehave in various ways if the subject string
2145       contained bytes with the 0x80 bit set and the 0x40 bit unset in a lookbehind
2146       area. (PCRE was not checking for the UTF-8 mode flag, and trying to move
2147       back over UTF-8 characters.)
2148    
2149    
2150    Version 4.2 14-Apr-03
2151    ---------------------
2152    
2153    1. Typo "#if SUPPORT_UTF8" instead of "#ifdef SUPPORT_UTF8" fixed.
2154    
2155    2. Changes to the building process, supplied by Ronald Landheer-Cieslak
2156         [ON_WINDOWS]: new variable, "#" on non-Windows platforms
2157         [NOT_ON_WINDOWS]: new variable, "#" on Windows platforms
2158         [WIN_PREFIX]: new variable, "cyg" for Cygwin
2159         * Makefile.in: use autoconf substitution for OBJEXT, EXEEXT, BUILD_OBJEXT
2160           and BUILD_EXEEXT
2161         Note: automatic setting of the BUILD variables is not yet working
2162         set CPPFLAGS and BUILD_CPPFLAGS (but don't use yet) - should be used at
2163           compile-time but not at link-time
2164         [LINK]: use for linking executables only
2165         make different versions for Windows and non-Windows
2166         [LINKLIB]: new variable, copy of UNIX-style LINK, used for linking
2167           libraries
2168         [LINK_FOR_BUILD]: new variable
2169         [OBJEXT]: use throughout
2170         [EXEEXT]: use throughout
2171         <winshared>: new target
2172         <wininstall>: new target
2173         <dftables.o>: use native compiler
2174         <dftables>: use native linker
2175         <install>: handle Windows platform correctly
2176         <clean>: ditto
2177         <check>: ditto
2178         copy DLL to top builddir before testing
2179    
2180       As part of these changes, -no-undefined was removed again. This was reported
2181       to give trouble on HP-UX 11.0, so getting rid of it seems like a good idea
2182       in any case.
2183    
2184    3. Some tidies to get rid of compiler warnings:
2185    
2186       . In the match_data structure, match_limit was an unsigned long int, whereas
2187         match_call_count was an int. I've made them both unsigned long ints.
2188    
2189       . In pcretest the fact that a const uschar * doesn't automatically cast to
2190         a void * provoked a warning.
2191    
2192       . Turning on some more compiler warnings threw up some "shadow" variables
2193         and a few more missing casts.
2194    
2195    4. If PCRE was complied with UTF-8 support, but called without the PCRE_UTF8
2196       option, a class that contained a single character with a value between 128
2197       and 255 (e.g. /[\xFF]/) caused PCRE to crash.
2198    
2199    5. If PCRE was compiled with UTF-8 support, but called without the PCRE_UTF8
2200       option, a class that contained several characters, but with at least one
2201       whose value was between 128 and 255 caused PCRE to crash.
2202    
2203    
2204    Version 4.1 12-Mar-03
2205    ---------------------
2206    
2207    1. Compiling with gcc -pedantic found a couple of places where casts were
2208    needed, and a string in dftables.c that was longer than standard compilers are
2209    required to support.
2210    
2211    2. Compiling with Sun's compiler found a few more places where the code could
2212    be tidied up in order to avoid warnings.
2213    
2214    3. The variables for cross-compiling were called HOST_CC and HOST_CFLAGS; the
2215    first of these names is deprecated in the latest Autoconf in favour of the name
2216    CC_FOR_BUILD, because "host" is typically used to mean the system on which the
2217    compiled code will be run. I can't find a reference for HOST_CFLAGS, but by
2218    analogy I have changed it to CFLAGS_FOR_BUILD.
2219    
2220    4. Added -no-undefined to the linking command in the Makefile, because this is
2221    apparently helpful for Windows. To make it work, also added "-L. -lpcre" to the
2222    linking step for the pcreposix library.
2223    
2224    5. PCRE was failing to diagnose the case of two named groups with the same
2225    name.
2226    
2227    6. A problem with one of PCRE's optimizations was discovered. PCRE remembers a
2228    literal character that is needed in the subject for a match, and scans along to
2229    ensure that it is present before embarking on the full matching process. This
2230    saves time in cases of nested unlimited repeats that are never going to match.
2231    Problem: the scan can take a lot of time if the subject is very long (e.g.
2232    megabytes), thus penalizing straightforward matches. It is now done only if the
2233    amount of subject to be scanned is less than 1000 bytes.
2234    
2235    7. A lesser problem with the same optimization is that it was recording the
2236    first character of an anchored pattern as "needed", thus provoking a search
2237    right along the subject, even when the first match of the pattern was going to
2238    fail. The "needed" character is now not set for anchored patterns, unless it
2239    follows something in the pattern that is of non-fixed length. Thus, it still
2240    fulfils its original purpose of finding quick non-matches in cases of nested
2241    unlimited repeats, but isn't used for simple anchored patterns such as /^abc/.
2242    
2243    
2244    Version 4.0 17-Feb-03
2245    ---------------------
2246    
2247    1. If a comment in an extended regex that started immediately after a meta-item
2248    extended to the end of string, PCRE compiled incorrect data. This could lead to
2249    all kinds of weird effects. Example: /#/ was bad; /()#/ was bad; /a#/ was not.
2250    
2251    2. Moved to autoconf 2.53 and libtool 1.4.2.
2252    
2253    3. Perl 5.8 no longer needs "use utf8" for doing UTF-8 things. Consequently,
2254    the special perltest8 script is no longer needed - all the tests can be run
2255    from a single perltest script.
2256    
2257    4. From 5.004, Perl has not included the VT character (0x0b) in the set defined
2258    by \s. It has now been removed in PCRE. This means it isn't recognized as
2259    whitespace in /x regexes too, which is the same as Perl. Note that the POSIX
2260    class [:space:] *does* include VT, thereby creating a mess.
2261    
2262    5. Added the class [:blank:] (a GNU extension from Perl 5.8) to match only
2263    space and tab.
2264    
2265    6. Perl 5.005 was a long time ago. It's time to amalgamate the tests that use
2266    its new features into the main test script, reducing the number of scripts.
2267    
2268    7. Perl 5.8 has changed the meaning of patterns like /a(?i)b/. Earlier versions
2269    were backward compatible, and made the (?i) apply to the whole pattern, as if
2270    /i were given. Now it behaves more logically, and applies the option setting
2271    only to what follows. PCRE has been changed to follow suit. However, if it
2272    finds options settings right at the start of the pattern, it extracts them into
2273    the global options, as before. Thus, they show up in the info data.
2274    
2275    8. Added support for the \Q...\E escape sequence. Characters in between are
2276    treated as literals. This is slightly different from Perl in that $ and @ are
2277    also handled as literals inside the quotes. In Perl, they will cause variable
2278    interpolation. Note the following examples:
2279    
2280        Pattern            PCRE matches      Perl matches
2281    
2282        \Qabc$xyz\E        abc$xyz           abc followed by the contents of $xyz
2283        \Qabc\$xyz\E       abc\$xyz          abc\$xyz
2284        \Qabc\E\$\Qxyz\E   abc$xyz           abc$xyz
2285    
2286    For compatibility with Perl, \Q...\E sequences are recognized inside character
2287    classes as well as outside them.
2288    
2289    9. Re-organized 3 code statements in pcretest to avoid "overflow in
2290    floating-point constant arithmetic" warnings from a Microsoft compiler. Added a
2291    (size_t) cast to one statement in pcretest and one in pcreposix to avoid
2292    signed/unsigned warnings.
2293    
2294    10. SunOS4 doesn't have strtoul(). This was used only for unpicking the -o
2295    option for pcretest, so I've replaced it by a simple function that does just
2296    that job.
2297    
2298    11. pcregrep was ending with code 0 instead of 2 for the commands "pcregrep" or
2299    "pcregrep -".
2300    
2301    12. Added "possessive quantifiers" ?+, *+, ++, and {,}+ which come from Sun's
2302    Java package. This provides some syntactic sugar for simple cases of what my
2303    documentation calls "once-only subpatterns". A pattern such as x*+ is the same
2304    as (?>x*). In other words, if what is inside (?>...) is just a single repeated
2305    item, you can use this simplified notation. Note that only makes sense with
2306    greedy quantifiers. Consequently, the use of the possessive quantifier forces
2307    greediness, whatever the setting of the PCRE_UNGREEDY option.
2308    
2309    13. A change of greediness default within a pattern was not taking effect at
2310    the current level for patterns like /(b+(?U)a+)/. It did apply to parenthesized
2311    subpatterns that followed. Patterns like /b+(?U)a+/ worked because the option
2312    was abstracted outside.
2313    
2314    14. PCRE now supports the \G assertion. It is true when the current matching
2315    position is at the start point of the match. This differs from \A when the
2316    starting offset is non-zero. Used with the /g option of pcretest (or similar
2317    code), it works in the same way as it does for Perl's /g option. If all
2318    alternatives of a regex begin with \G, the expression is anchored to the start
2319    match position, and the "anchored" flag is set in the compiled expression.
2320    
2321    15. Some bugs concerning the handling of certain option changes within patterns
2322    have been fixed. These applied to options other than (?ims). For example,
2323    "a(?x: b c )d" did not match "XabcdY" but did match "Xa b c dY". It should have
2324    been the other way round. Some of this was related to change 7 above.
2325    
2326    16. PCRE now gives errors for /[.x.]/ and /[=x=]/ as unsupported POSIX
2327    features, as Perl does. Previously, PCRE gave the warnings only for /[[.x.]]/
2328    and /[[=x=]]/. PCRE now also gives an error for /[:name:]/ because it supports
2329    POSIX classes only within a class (e.g. /[[:alpha:]]/).
2330    
2331    17. Added support for Perl's \C escape. This matches one byte, even in UTF8
2332    mode. Unlike ".", it always matches newline, whatever the setting of
2333    PCRE_DOTALL. However, PCRE does not permit \C to appear in lookbehind
2334    assertions. Perl allows it, but it doesn't (in general) work because it can't
2335    calculate the length of the lookbehind. At least, that's the case for Perl
2336    5.8.0 - I've been told they are going to document that it doesn't work in
2337    future.
2338    
2339    18. Added an error diagnosis for escapes that PCRE does not support: these are
2340    \L, \l, \N, \P, \p, \U, \u, and \X.
2341    
2342    19. Although correctly diagnosing a missing ']' in a character class, PCRE was
2343    reading past the end of the pattern in cases such as /[abcd/.
2344    
2345    20. PCRE was getting more memory than necessary for patterns with classes that
2346    contained both POSIX named classes and other characters, e.g. /[[:space:]abc/.
2347    
2348    21. Added some code, conditional on #ifdef VPCOMPAT, to make life easier for
2349    compiling PCRE for use with Virtual Pascal.
2350    
2351    22. Small fix to the Makefile to make it work properly if the build is done
2352    outside the source tree.
2353    
2354    23. Added a new extension: a condition to go with recursion. If a conditional
2355    subpattern starts with (?(R) the "true" branch is used if recursion has
2356    happened, whereas the "false" branch is used only at the top level.
2357    
2358    24. When there was a very long string of literal characters (over 255 bytes
2359    without UTF support, over 250 bytes with UTF support), the computation of how
2360    much memory was required could be incorrect, leading to segfaults or other
2361    strange effects.
2362    
2363    25. PCRE was incorrectly assuming anchoring (either to start of subject or to
2364    start of line for a non-DOTALL pattern) when a pattern started with (.*) and
2365    there was a subsequent back reference to those brackets. This meant that, for
2366    example, /(.*)\d+\1/ failed to match "abc123bc". Unfortunately, it isn't
2367    possible to check for precisely this case. All we can do is abandon the
2368    optimization if .* occurs inside capturing brackets when there are any back
2369    references whatsoever. (See below for a better fix that came later.)
2370    
2371    26. The handling of the optimization for finding the first character of a
2372    non-anchored pattern, and for finding a character that is required later in the
2373    match were failing in some cases. This didn't break the matching; it just
2374    failed to optimize when it could. The way this is done has been re-implemented.
2375    
2376    27. Fixed typo in error message for invalid (?R item (it said "(?p").
2377    
2378    28. Added a new feature that provides some of the functionality that Perl
2379    provides with (?{...}). The facility is termed a "callout". The way it is done
2380    in PCRE is for the caller to provide an optional function, by setting
2381    pcre_callout to its entry point. Like pcre_malloc and pcre_free, this is a
2382    global variable. By default it is unset, which disables all calling out. To get
2383    the function called, the regex must include (?C) at appropriate points. This
2384    is, in fact, equivalent to (?C0), and any number <= 255 may be given with (?C).
2385    This provides a means of identifying different callout points. When PCRE
2386    reaches such a point in the regex, if pcre_callout has been set, the external
2387    function is called. It is provided with data in a structure called
2388    pcre_callout_block, which is defined in pcre.h. If the function returns 0,
2389    matching continues; if it returns a non-zero value, the match at the current
2390    point fails. However, backtracking will occur if possible. [This was changed
2391    later and other features added - see item 49 below.]
2392    
2393    29. pcretest is upgraded to test the callout functionality. It provides a
2394    callout function that displays information. By default, it shows the start of
2395    the match and the current position in the text. There are some new data escapes
2396    to vary what happens:
2397    
2398        \C+         in addition, show current contents of captured substrings
2399        \C-         do not supply a callout function
2400        \C!n        return 1 when callout number n is reached
2401        \C!n!m      return 1 when callout number n is reached for the mth time
2402    
2403    30. If pcregrep was called with the -l option and just a single file name, it
2404    output "<stdin>" if a match was found, instead of the file name.
2405    
2406    31. Improve the efficiency of the POSIX API to PCRE. If the number of capturing
2407    slots is less than POSIX_MALLOC_THRESHOLD, use a block on the stack to pass to
2408    pcre_exec(). This saves a malloc/free per call. The default value of
2409    POSIX_MALLOC_THRESHOLD is 10; it can be changed by --with-posix-malloc-threshold
2410    when configuring.
2411    
2412    32. The default maximum size of a compiled pattern is 64K. There have been a
2413    few cases of people hitting this limit. The code now uses macros to handle the
2414    storing of links as offsets within the compiled pattern. It defaults to 2-byte
2415    links, but this can be changed to 3 or 4 bytes by --with-link-size when
2416    configuring. Tests 2 and 5 work only with 2-byte links because they output
2417    debugging information about compiled patterns.
2418    
2419    33. Internal code re-arrangements:
2420    
2421    (a) Moved the debugging function for printing out a compiled regex into
2422        its own source file (printint.c) and used #include to pull it into
2423        pcretest.c and, when DEBUG is defined, into pcre.c, instead of having two
2424        separate copies.
2425    
2426    (b) Defined the list of op-code names for debugging as a macro in
2427        internal.h so that it is next to the definition of the opcodes.
2428    
2429    (c) Defined a table of op-code lengths for simpler skipping along compiled
2430        code. This is again a macro in internal.h so that it is next to the
2431        definition of the opcodes.
2432    
2433    34. Added support for recursive calls to individual subpatterns, along the
2434    lines of Robin Houston's patch (but implemented somewhat differently).
2435    
2436    35. Further mods to the Makefile to help Win32. Also, added code to pcregrep to
2437    allow it to read and process whole directories in Win32. This code was
2438    contributed by Lionel Fourquaux; it has not been tested by me.
2439    
2440    36. Added support for named subpatterns. The Python syntax (?P<name>...) is
2441    used to name a group. Names consist of alphanumerics and underscores, and must
2442    be unique. Back references use the syntax (?P=name) and recursive calls use
2443    (?P>name) which is a PCRE extension to the Python extension. Groups still have
2444    numbers. The function pcre_fullinfo() can be used after compilation to extract
2445    a name/number map. There are three relevant calls:
2446    
2447      PCRE_INFO_NAMEENTRYSIZE        yields the size of each entry in the map
2448      PCRE_INFO_NAMECOUNT            yields the number of entries
2449      PCRE_INFO_NAMETABLE            yields a pointer to the map.
2450    
2451    The map is a vector of fixed-size entries. The size of each entry depends on
2452    the length of the longest name used. The first two bytes of each entry are the
2453    group number, most significant byte first. There follows the corresponding
2454    name, zero terminated. The names are in alphabetical order.
2455    
2456    37. Make the maximum literal string in the compiled code 250 for the non-UTF-8
2457    case instead of 255. Making it the same both with and without UTF-8 support
2458    means that the same test output works with both.
2459    
2460    38. There was a case of malloc(0) in the POSIX testing code in pcretest. Avoid
2461    calling malloc() with a zero argument.
2462    
2463    39. Change 25 above had to resort to a heavy-handed test for the .* anchoring
2464    optimization. I've improved things by keeping a bitmap of backreferences with
2465    numbers 1-31 so that if .* occurs inside capturing brackets that are not in
2466    fact referenced, the optimization can be applied. It is unlikely that a
2467    relevant occurrence of .* (i.e. one which might indicate anchoring or forcing
2468    the match to follow \n) will appear inside brackets with a number greater than
2469    31, but if it does, any back reference > 31 suppresses the optimization.
2470    
2471    40. Added a new compile-time option PCRE_NO_AUTO_CAPTURE. This has the effect
2472    of disabling numbered capturing parentheses. Any opening parenthesis that is
2473    not followed by ? behaves as if it were followed by ?: but named parentheses
2474    can still be used for capturing (and they will acquire numbers in the usual
2475    way).
2476    
2477    41. Redesigned the return codes from the match() function into yes/no/error so
2478    that errors can be passed back from deep inside the nested calls. A malloc
2479    failure while inside a recursive subpattern call now causes the
2480    PCRE_ERROR_NOMEMORY return instead of quietly going wrong.
2481    
2482    42. It is now possible to set a limit on the number of times the match()
2483    function is called in a call to pcre_exec(). This facility makes it possible to
2484    limit the amount of recursion and backtracking, though not in a directly
2485    obvious way, because the match() function is used in a number of different
2486    circumstances. The count starts from zero for each position in the subject
2487    string (for non-anchored patterns). The default limit is, for compatibility, a
2488    large number, namely 10 000 000. You can change this in two ways:
2489    
2490    (a) When configuring PCRE before making, you can use --with-match-limit=n
2491        to set a default value for the compiled library.
2492    
2493    (b) For each call to pcre_exec(), you can pass a pcre_extra block in which
2494        a different value is set. See 45 below.
2495    
2496    If the limit is exceeded, pcre_exec() returns PCRE_ERROR_MATCHLIMIT.
2497    
2498    43. Added a new function pcre_config(int, void *) to enable run-time extraction
2499    of things that can be changed at compile time. The first argument specifies
2500    what is wanted and the second points to where the information is to be placed.
2501    The current list of available information is:
2502    
2503      PCRE_CONFIG_UTF8
2504    
2505    The output is an integer that is set to one if UTF-8 support is available;
2506    otherwise it is set to zero.
2507    
2508      PCRE_CONFIG_NEWLINE
2509    
2510    The output is an integer that it set to the value of the code that is used for
2511    newline. It is either LF (10) or CR (13).
2512    
2513      PCRE_CONFIG_LINK_SIZE
2514    
2515    The output is an integer that contains the number of bytes used for internal
2516    linkage in compiled expressions. The value is 2, 3, or 4. See item 32 above.
2517    
2518      PCRE_CONFIG_POSIX_MALLOC_THRESHOLD
2519    
2520    The output is an integer that contains the threshold above which the POSIX
2521    interface uses malloc() for output vectors. See item 31 above.
2522    
2523      PCRE_CONFIG_MATCH_LIMIT
2524    
2525    The output is an unsigned integer that contains the default limit of the number
2526    of match() calls in a pcre_exec() execution. See 42 above.
2527    
2528    44. pcretest has been upgraded by the addition of the -C option. This causes it
2529    to extract all the available output from the new pcre_config() function, and to
2530    output it. The program then exits immediately.
2531    
2532    45. A need has arisen to pass over additional data with calls to pcre_exec() in
2533    order to support additional features. One way would have been to define
2534    pcre_exec2() (for example) with extra arguments, but this would not have been
2535    extensible, and would also have required all calls to the original function to
2536    be mapped to the new one. Instead, I have chosen to extend the mechanism that
2537    is used for passing in "extra" data from pcre_study().
2538    
2539    The pcre_extra structure is now exposed and defined in pcre.h. It currently
2540    contains the following fields:
2541    
2542      flags         a bitmap indicating which of the following fields are set
2543      study_data    opaque data from pcre_study()
2544      match_limit   a way of specifying a limit on match() calls for a specific
2545                      call to pcre_exec()
2546      callout_data  data for callouts (see 49 below)
2547    
2548    The flag bits are also defined in pcre.h, and are
2549    
2550      PCRE_EXTRA_STUDY_DATA
2551      PCRE_EXTRA_MATCH_LIMIT
2552      PCRE_EXTRA_CALLOUT_DATA
2553    
2554    The pcre_study() function now returns one of these new pcre_extra blocks, with
2555    the actual study data pointed to by the study_data field, and the
2556    PCRE_EXTRA_STUDY_DATA flag set. This can be passed directly to pcre_exec() as
2557    before. That is, this change is entirely upwards-compatible and requires no
2558    change to existing code.
2559    
2560    If you want to pass in additional data to pcre_exec(), you can either place it
2561    in a pcre_extra block provided by pcre_study(), or create your own pcre_extra
2562    block.
2563    
2564    46. pcretest has been extended to test the PCRE_EXTRA_MATCH_LIMIT feature. If a
2565    data string contains the escape sequence \M, pcretest calls pcre_exec() several
2566    times with different match limits, until it finds the minimum value needed for
2567    pcre_exec() to complete. The value is then output. This can be instructive; for
2568    most simple matches the number is quite small, but for pathological cases it
2569    gets very large very quickly.
2570    
2571    47. There's a new option for pcre_fullinfo() called PCRE_INFO_STUDYSIZE. It
2572    returns the size of the data block pointed to by the study_data field in a
2573    pcre_extra block, that is, the value that was passed as the argument to
2574    pcre_malloc() when PCRE was getting memory in which to place the information
2575    created by pcre_study(). The fourth argument should point to a size_t variable.
2576    pcretest has been extended so that this information is shown after a successful
2577    pcre_study() call when information about the compiled regex is being displayed.
2578    
2579    48. Cosmetic change to Makefile: there's no need to have / after $(DESTDIR)
2580    because what follows is always an absolute path. (Later: it turns out that this
2581    is more than cosmetic for MinGW, because it doesn't like empty path
2582    components.)
2583    
2584    49. Some changes have been made to the callout feature (see 28 above):
2585    
2586    (i)  A callout function now has three choices for what it returns:
2587    
2588           0  =>  success, carry on matching
2589         > 0  =>  failure at this point, but backtrack if possible
2590         < 0  =>  serious error, return this value from pcre_exec()
2591    
2592         Negative values should normally be chosen from the set of PCRE_ERROR_xxx
2593         values. In particular, returning PCRE_ERROR_NOMATCH forces a standard
2594         "match failed" error. The error number PCRE_ERROR_CALLOUT is reserved for
2595         use by callout functions. It will never be used by PCRE itself.
2596    
2597    (ii) The pcre_extra structure (see 45 above) has a void * field called
2598         callout_data, with corresponding flag bit PCRE_EXTRA_CALLOUT_DATA. The
2599         pcre_callout_block structure has a field of the same name. The contents of
2600         the field passed in the pcre_extra structure are passed to the callout
2601         function in the corresponding field in the callout block. This makes it
2602         easier to use the same callout-containing regex from multiple threads. For
2603         testing, the pcretest program has a new data escape
2604    
2605           \C*n        pass the number n (may be negative) as callout_data
2606    
2607         If the callout function in pcretest receives a non-zero value as
2608         callout_data, it returns that value.
2609    
2610    50. Makefile wasn't handling CFLAGS properly when compiling dftables. Also,
2611    there were some redundant $(CFLAGS) in commands that are now specified as
2612    $(LINK), which already includes $(CFLAGS).
2613    
2614    51. Extensions to UTF-8 support are listed below. These all apply when (a) PCRE
2615    has been compiled with UTF-8 support *and* pcre_compile() has been compiled
2616    with the PCRE_UTF8 flag. Patterns that are compiled without that flag assume
2617    one-byte characters throughout. Note that case-insensitive matching applies
2618    only to characters whose values are less than 256. PCRE doesn't support the
2619    notion of cases for higher-valued characters.
2620    
2621    (i)   A character class whose characters are all within 0-255 is handled as
2622          a bit map, and the map is inverted for negative classes. Previously, a
2623          character > 255 always failed to match such a class; however it should
2624          match if the class was a negative one (e.g. [^ab]). This has been fixed.
2625    
2626    (ii)  A negated character class with a single character < 255 is coded as
2627          "not this character" (OP_NOT). This wasn't working properly when the test
2628          character was multibyte, either singly or repeated.
2629    
2630    (iii) Repeats of multibyte characters are now handled correctly in UTF-8
2631          mode, for example: \x{100}{2,3}.
2632    
2633    (iv)  The character escapes \b, \B, \d, \D, \s, \S, \w, and \W (either
2634          singly or repeated) now correctly test multibyte characters. However,
2635          PCRE doesn't recognize any characters with values greater than 255 as
2636          digits, spaces, or word characters. Such characters always match \D, \S,
2637          and \W, and never match \d, \s, or \w.
2638    
2639    (v)   Classes may now contain characters and character ranges with values
2640          greater than 255. For example: [ab\x{100}-\x{400}].
2641    
2642    (vi)  pcregrep now has a --utf-8 option (synonym -u) which makes it call
2643          PCRE in UTF-8 mode.
2644    
2645    52. The info request value PCRE_INFO_FIRSTCHAR has been renamed
2646    PCRE_INFO_FIRSTBYTE because it is a byte value. However, the old name is
2647    retained for backwards compatibility. (Note that LASTLITERAL is also a byte
2648    value.)
2649    
2650    53. The single man page has become too large. I have therefore split it up into
2651    a number of separate man pages. These also give rise to individual HTML pages;
2652    these are now put in a separate directory, and there is an index.html page that
2653    lists them all. Some hyperlinking between the pages has been installed.
2654    
2655    54. Added convenience functions for handling named capturing parentheses.
2656    
2657    55. Unknown escapes inside character classes (e.g. [\M]) and escapes that
2658    aren't interpreted therein (e.g. [\C]) are literals in Perl. This is now also
2659    true in PCRE, except when the PCRE_EXTENDED option is set, in which case they
2660    are faulted.
2661    
2662    56. Introduced HOST_CC and HOST_CFLAGS which can be set in the environment when
2663    calling configure. These values are used when compiling the dftables.c program
2664    which is run to generate the source of the default character tables. They
2665    default to the values of CC and CFLAGS. If you are cross-compiling PCRE,
2666    you will need to set these values.
2667    
2668    57. Updated the building process for Windows DLL, as provided by Fred Cox.
2669    
2670    
2671    Version 3.9 02-Jan-02
2672    ---------------------
2673    
2674    1. A bit of extraneous text had somehow crept into the pcregrep documentation.
2675    
2676    2. If --disable-static was given, the building process failed when trying to
2677    build pcretest and pcregrep. (For some reason it was using libtool to compile
2678    them, which is not right, as they aren't part of the library.)
2679    
2680    
2681    Version 3.8 18-Dec-01
2682    ---------------------
2683    
2684    1. The experimental UTF-8 code was completely screwed up. It was packing the
2685    bytes in the wrong order. How dumb can you get?
2686    
2687    
2688    Version 3.7 29-Oct-01
2689    ---------------------
2690    
2691    1. In updating pcretest to check change 1 of version 3.6, I screwed up.
2692    This caused pcretest, when used on the test data, to segfault. Unfortunately,
2693    this didn't happen under Solaris 8, where I normally test things.
2694    
2695    2. The Makefile had to be changed to make it work on BSD systems, where 'make'
2696    doesn't seem to recognize that ./xxx and xxx are the same file. (This entry
2697    isn't in ChangeLog distributed with 3.7 because I forgot when I hastily made
2698    this fix an hour or so after the initial 3.7 release.)
2699    
2700    
2701    Version 3.6 23-Oct-01
2702    ---------------------
2703    
2704    1. Crashed with /(sens|respons)e and \1ibility/ and "sense and sensibility" if
2705    offsets passed as NULL with zero offset count.
2706    
2707    2. The config.guess and config.sub files had not been updated when I moved to
2708    the latest autoconf.
2709    
2710    
2711    Version 3.5 15-Aug-01
2712    ---------------------
2713    
2714    1. Added some missing #if !defined NOPOSIX conditionals in pcretest.c that
2715    had been forgotten.
2716    
2717    2. By using declared but undefined structures, we can avoid using "void"
2718    definitions in pcre.h while keeping the internal definitions of the structures
2719    private.
2720    
2721    3. The distribution is now built using autoconf 2.50 and libtool 1.4. From a
2722    user point of view, this means that both static and shared libraries are built
2723    by default, but this can be individually controlled. More of the work of
2724    handling this static/shared cases is now inside libtool instead of PCRE's make
2725    file.
2726    
2727    4. The pcretest utility is now installed along with pcregrep because it is
2728    useful for users (to test regexs) and by doing this, it automatically gets
2729    relinked by libtool. The documentation has been turned into a man page, so
2730    there are now .1, .txt, and .html versions in /doc.
2731    
2732    5. Upgrades to pcregrep:
2733       (i)   Added long-form option names like gnu grep.
2734       (ii)  Added --help to list all options with an explanatory phrase.
2735       (iii) Added -r, --recursive to recurse into sub-directories.
2736       (iv)  Added -f, --file to read patterns from a file.
2737    
2738    6. pcre_exec() was referring to its "code" argument before testing that
2739    argument for NULL (and giving an error if it was NULL).
2740    
2741    7. Upgraded Makefile.in to allow for compiling in a different directory from
2742    the source directory.
2743    
2744    8. Tiny buglet in pcretest: when pcre_fullinfo() was called to retrieve the
2745    options bits, the pointer it was passed was to an int instead of to an unsigned
2746    long int. This mattered only on 64-bit systems.
2747    
2748    9. Fixed typo (3.4/1) in pcre.h again. Sigh. I had changed pcre.h (which is
2749    generated) instead of pcre.in, which it its source. Also made the same change
2750    in several of the .c files.
2751    
2752    10. A new release of gcc defines printf() as a macro, which broke pcretest
2753    because it had an ifdef in the middle of a string argument for printf(). Fixed
2754    by using separate calls to printf().
2755    
2756    11. Added --enable-newline-is-cr and --enable-newline-is-lf to the configure
2757    script, to force use of CR or LF instead of \n in the source. On non-Unix
2758    systems, the value can be set in config.h.
2759    
2760    12. The limit of 200 on non-capturing parentheses is a _nesting_ limit, not an
2761    absolute limit. Changed the text of the error message to make this clear, and
2762    likewise updated the man page.
2763    
2764    13. The limit of 99 on the number of capturing subpatterns has been removed.
2765    The new limit is 65535, which I hope will not be a "real" limit.
2766    
2767    
2768    Version 3.4 22-Aug-00
2769    ---------------------
2770    
2771    1. Fixed typo in pcre.h: unsigned const char * changed to const unsigned char *.
2772    
2773    2. Diagnose condition (?(0) as an error instead of crashing on matching.
2774    
2775    
2776    Version 3.3 01-Aug-00
2777    ---------------------
2778    
2779    1. If an octal character was given, but the value was greater than \377, it
2780    was not getting masked to the least significant bits, as documented. This could
2781    lead to crashes in some systems.
2782    
2783    2. Perl 5.6 (if not earlier versions) accepts classes like [a-\d] and treats
2784    the hyphen as a literal. PCRE used to give an error; it now behaves like Perl.
2785    
2786    3. Added the functions pcre_free_substring() and pcre_free_substring_list().
2787    These just pass their arguments on to (pcre_free)(), but they are provided
2788    because some uses of PCRE bind it to non-C systems that can call its functions,
2789    but cannot call free() or pcre_free() directly.
2790    
2791    4. Add "make test" as a synonym for "make check". Corrected some comments in
2792    the Makefile.
2793    
2794    5. Add $(DESTDIR)/ in front of all the paths in the "install" target in the
2795    Makefile.
2796    
2797    6. Changed the name of pgrep to pcregrep, because Solaris has introduced a
2798    command called pgrep for grepping around the active processes.
2799    
2800    7. Added the beginnings of support for UTF-8 character strings.
2801    
2802    8. Arranged for the Makefile to pass over the settings of CC, CFLAGS, and
2803    RANLIB to ./ltconfig so that they are used by libtool. I think these are all
2804    the relevant ones. (AR is not passed because ./ltconfig does its own figuring
2805    out for the ar command.)
2806    
2807    
2808    Version 3.2 12-May-00
2809    ---------------------
2810    
2811    This is purely a bug fixing release.
2812    
2813    1. If the pattern /((Z)+|A)*/ was matched agained ZABCDEFG it matched Z instead
2814    of ZA. This was just one example of several cases that could provoke this bug,
2815    which was introduced by change 9 of version 2.00. The code for breaking
2816    infinite loops after an iteration that matches an empty string was't working
2817    correctly.
2818    
2819    2. The pcretest program was not imitating Perl correctly for the pattern /a*/g
2820    when matched against abbab (for example). After matching an empty string, it
2821    wasn't forcing anchoring when setting PCRE_NOTEMPTY for the next attempt; this
2822    caused it to match further down the string than it should.
2823    
2824    3. The code contained an inclusion of sys/types.h. It isn't clear why this
2825    was there because it doesn't seem to be needed, and it causes trouble on some
2826    systems, as it is not a Standard C header. It has been removed.
2827    
2828    4. Made 4 silly changes to the source to avoid stupid compiler warnings that
2829    were reported on the Macintosh. The changes were from
2830    
2831      while ((c = *(++ptr)) != 0 && c != '\n');
2832    to
2833      while ((c = *(++ptr)) != 0 && c != '\n') ;
2834    
2835    Totally extraordinary, but if that's what it takes...
2836    
2837    5. PCRE is being used in one environment where neither memmove() nor bcopy() is
2838    available. Added HAVE_BCOPY and an autoconf test for it; if neither
2839    HAVE_MEMMOVE nor HAVE_BCOPY is set, use a built-in emulation function which
2840    assumes the way PCRE uses memmove() (always moving upwards).
2841    
2842    6. PCRE is being used in one environment where strchr() is not available. There
2843    was only one use in pcre.c, and writing it out to avoid strchr() probably gives
2844    faster code anyway.
2845    
2846    
2847    Version 3.1 09-Feb-00
2848    ---------------------
2849    
2850    The only change in this release is the fixing of some bugs in Makefile.in for
2851    the "install" target:
2852    
2853    (1) It was failing to install pcreposix.h.
2854    
2855    (2) It was overwriting the pcre.3 man page with the pcreposix.3 man page.
2856    
2857    
2858    Version 3.0 01-Feb-00
2859    ---------------------
2860    
2861    1. Add support for the /+ modifier to perltest (to output $` like it does in
2862    pcretest).
2863    
2864    2. Add support for the /g modifier to perltest.
2865    
2866    3. Fix pcretest so that it behaves even more like Perl for /g when the pattern
2867    matches null strings.
2868    
2869    4. Fix perltest so that it doesn't do unwanted things when fed an empty
2870    pattern. Perl treats empty patterns specially - it reuses the most recent
2871    pattern, which is not what we want. Replace // by /(?#)/ in order to avoid this
2872    effect.
2873    
2874    5. The POSIX interface was broken in that it was just handing over the POSIX
2875    captured string vector to pcre_exec(), but (since release 2.00) PCRE has
2876    required a bigger vector, with some working space on the end. This means that
2877    the POSIX wrapper now has to get and free some memory, and copy the results.
2878    
2879    6. Added some simple autoconf support, placing the test data and the
2880    documentation in separate directories, re-organizing some of the
2881    information files, and making it build pcre-config (a GNU standard). Also added
2882    libtool support for building PCRE as a shared library, which is now the
2883    default.
2884    
2885    7. Got rid of the leading zero in the definition of PCRE_MINOR because 08 and
2886    09 are not valid octal constants. Single digits will be used for minor values
2887    less than 10.
2888    
2889    8. Defined REG_EXTENDED and REG_NOSUB as zero in the POSIX header, so that
2890    existing programs that set these in the POSIX interface can use PCRE without
2891    modification.
2892    
2893    9. Added a new function, pcre_fullinfo() with an extensible interface. It can
2894    return all that pcre_info() returns, plus additional data. The pcre_info()
2895    function is retained for compatibility, but is considered to be obsolete.
2896    
2897    10. Added experimental recursion feature (?R) to handle one common case that
2898    Perl 5.6 will be able to do with (?p{...}).
2899    
2900    11. Added support for POSIX character classes like [:alpha:], which Perl is
2901    adopting.
2902    
2903    
2904    Version 2.08 31-Aug-99
2905    ----------------------
2906    
2907    1. When startoffset was not zero and the pattern began with ".*", PCRE was not
2908    trying to match at the startoffset position, but instead was moving forward to
2909    the next newline as if a previous match had failed.
2910    
2911    2. pcretest was not making use of PCRE_NOTEMPTY when repeating for /g and /G,
2912    and could get into a loop if a null string was matched other than at the start
2913    of the subject.
2914    
2915    3. Added definitions of PCRE_MAJOR and PCRE_MINOR to pcre.h so the version can
2916    be distinguished at compile time, and for completeness also added PCRE_DATE.
2917    
2918    5. Added Paul Sokolovsky's minor changes to make it easy to compile a Win32 DLL
2919    in GnuWin32 environments.
2920    
2921    
2922    Version 2.07 29-Jul-99
2923    ----------------------
2924    
2925    1. The documentation is now supplied in plain text form and HTML as well as in
2926    the form of man page sources.
2927    
2928    2. C++ compilers don't like assigning (void *) values to other pointer types.
2929    In particular this affects malloc(). Although there is no problem in Standard
2930    C, I've put in casts to keep C++ compilers happy.
2931    
2932    3. Typo on pcretest.c; a cast of (unsigned char *) in the POSIX regexec() call
2933    should be (const char *).
2934    
2935    4. If NOPOSIX is defined, pcretest.c compiles without POSIX support. This may
2936    be useful for non-Unix systems who don't want to bother with the POSIX stuff.
2937    However, I haven't made this a standard facility. The documentation doesn't
2938    mention it, and the Makefile doesn't support it.
2939    
2940    5. The Makefile now contains an "install" target, with editable destinations at
2941    the top of the file. The pcretest program is not installed.
2942    
2943    6. pgrep -V now gives the PCRE version number and date.
2944    
2945    7. Fixed bug: a zero repetition after a literal string (e.g. /abcde{0}/) was
2946    causing the entire string to be ignored, instead of just the last character.
2947    
2948    8. If a pattern like /"([^\\"]+|\\.)*"/ is applied in the normal way to a
2949    non-matching string, it can take a very, very long time, even for strings of
2950    quite modest length, because of the nested recursion. PCRE now does better in
2951    some of these cases. It does this by remembering the last required literal
2952    character in the pattern, and pre-searching the subject to ensure it is present
2953    before running the real match. In other words, it applies a heuristic to detect
2954    some types of certain failure quickly, and in the above example, if presented
2955    with a string that has no trailing " it gives "no match" very quickly.
2956    
2957    9. A new runtime option PCRE_NOTEMPTY causes null string matches to be ignored;
2958    other alternatives are tried instead.
2959    
2960    
2961    Version 2.06 09-Jun-99
2962    ----------------------
2963    
2964    1. Change pcretest's output for amount of store used to show just the code
2965    space, because the remainder (the data block) varies in size between 32-bit and
2966    64-bit systems.
2967    
2968    2. Added an extra argument to pcre_exec() to supply an offset in the subject to
2969    start matching at. This allows lookbehinds to work when searching for multiple
2970    occurrences in a string.
2971    
2972    3. Added additional options to pcretest for testing multiple occurrences:
2973    
2974       /+   outputs the rest of the string that follows a match
2975       /g   loops for multiple occurrences, using the new startoffset argument
2976       /G   loops for multiple occurrences by passing an incremented pointer
2977    
2978    4. PCRE wasn't doing the "first character" optimization for patterns starting
2979    with \b or \B, though it was doing it for other lookbehind assertions. That is,
2980    it wasn't noticing that a match for a pattern such as /\bxyz/ has to start with
2981    the letter 'x'. On long subject strings, this gives a significant speed-up.
2982    
2983    
2984    Version 2.05 21-Apr-99
2985    ----------------------
2986    
2987    1. Changed the type of magic_number from int to long int so that it works
2988    properly on 16-bit systems.
2989    
2990    2. Fixed a bug which caused patterns starting with .* not to work correctly
2991    when the subject string contained newline characters. PCRE was assuming
2992    anchoring for such patterns in all cases, which is not correct because .* will
2993    not pass a newline unless PCRE_DOTALL is set. It now assumes anchoring only if
2994    DOTALL is set at top level; otherwise it knows that patterns starting with .*
2995    must be retried after every newline in the subject.
2996    
2997    
2998    Version 2.04 18-Feb-99
2999    ----------------------
3000    
3001    1. For parenthesized subpatterns with repeats whose minimum was zero, the
3002    computation of the store needed to hold the pattern was incorrect (too large).
3003    If such patterns were nested a few deep, this could multiply and become a real
3004    problem.
3005    
3006    2. Added /M option to pcretest to show the memory requirement of a specific
3007    pattern. Made -m a synonym of -s (which does this globally) for compatibility.
3008    
3009    3. Subpatterns of the form (regex){n,m} (i.e. limited maximum) were being
3010    compiled in such a way that the backtracking after subsequent failure was
3011    pessimal. Something like (a){0,3} was compiled as (a)?(a)?(a)? instead of
3012    ((a)((a)(a)?)?)? with disastrous performance if the maximum was of any size.
3013    
3014    
3015    Version 2.03 02-Feb-99
3016    ----------------------
3017    
3018    1. Fixed typo and small mistake in man page.
3019    
3020    2. Added 4th condition (GPL supersedes if conflict) and created separate
3021    LICENCE file containing the conditions.
3022    
3023    3. Updated pcretest so that patterns such as /abc\/def/ work like they do in
3024    Perl, that is the internal \ allows the delimiter to be included in the
3025    pattern. Locked out the use of \ as a delimiter. If \ immediately follows
3026    the final delimiter, add \ to the end of the pattern (to test the error).
3027    
3028    4. Added the convenience functions for extracting substrings after a successful
3029    match. Updated pcretest to make it able to test these functions.
3030    
3031    
3032    Version 2.02 14-Jan-99
3033    ----------------------
3034    
3035    1. Initialized the working variables associated with each extraction so that
3036    their saving and restoring doesn't refer to uninitialized store.
3037    
3038    2. Put dummy code into study.c in order to trick the optimizer of the IBM C
3039    compiler for OS/2 into generating correct code. Apparently IBM isn't going to
3040    fix the problem.
3041    
3042    3. Pcretest: the timing code wasn't using LOOPREPEAT for timing execution
3043    calls, and wasn't printing the correct value for compiling calls. Increased the
3044    default value of LOOPREPEAT, and the number of significant figures in the
3045    times.
3046    
3047    4. Changed "/bin/rm" in the Makefile to "-rm" so it works on Windows NT.
3048    
3049    5. Renamed "deftables" as "dftables" to get it down to 8 characters, to avoid
3050    a building problem on Windows NT with a FAT file system.
3051    
3052    
3053    Version 2.01 21-Oct-98
3054    ----------------------
3055    
3056    1. Changed the API for pcre_compile() to allow for the provision of a pointer
3057    to character tables built by pcre_maketables() in the current locale. If NULL
3058    is passed, the default tables are used.
3059    
3060    
3061    Version 2.00 24-Sep-98
3062    ----------------------
3063    
3064    1. Since the (>?) facility is in Perl 5.005, don't require PCRE_EXTRA to enable
3065    it any more.
3066    
3067    2. Allow quantification of (?>) groups, and make it work correctly.
3068    
3069    3. The first character computation wasn't working for (?>) groups.
3070    
3071    4. Correct the implementation of \Z (it is permitted to match on the \n at the
3072    end of the subject) and add 5.005's \z, which really does match only at the
3073    very end of the subject.
3074    
3075    5. Remove the \X "cut" facility; Perl doesn't have it, and (?> is neater.
3076    
3077    6. Remove the ability to specify CASELESS, MULTILINE, DOTALL, and
3078    DOLLAR_END_ONLY at runtime, to make it possible to implement the Perl 5.005
3079    localized options. All options to pcre_study() were also removed.
3080    
3081    7. Add other new features from 5.005:
3082    
3083       $(?<=           positive lookbehind
3084       $(?<!           negative lookbehind
3085       (?imsx-imsx)    added the unsetting capability
3086                       such a setting is global if at outer level; local otherwise
3087       (?imsx-imsx:)   non-capturing groups with option setting
3088       (?(cond)re|re)  conditional pattern matching
3089    
3090       A backreference to itself in a repeated group matches the previous
3091       captured string.
3092    
3093    8. General tidying up of studying (both automatic and via "study")
3094    consequential on the addition of new assertions.
3095    
3096    9. As in 5.005, unlimited repeated groups that could match an empty substring
3097    are no longer faulted at compile time. Instead, the loop is forcibly broken at
3098    runtime if any iteration does actually match an empty substring.
3099    
3100    10. Include the RunTest script in the distribution.
3101    
3102    11. Added tests from the Perl 5.005_02 distribution. This showed up a few
3103    discrepancies, some of which were old and were also with respect to 5.004. They
3104    have now been fixed.
3105    
3106    
3107    Version 1.09 28-Apr-98
3108  ----------------------  ----------------------
3109    
3110  1. A negated single character class followed by a quantifier with a minimum  1. A negated single character class followed by a quantifier with a minimum

Legend:
Removed from v.21  
changed lines
  Added in v.378

  ViewVC Help
Powered by ViewVC 1.1.5