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

Diff of /code/trunk/ChangeLog

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

revision 17 by nigel, Sat Feb 24 21:38:29 2007 UTC revision 366 by ph10, Mon Jul 14 15:45:32 2008 UTC
# Line 1  Line 1 
1  ChangeLog for PCRE  ChangeLog for PCRE
2  ------------------  ------------------
3    
4    Version 8.0 02 Jul-08
5    ---------------------
6    
7    1.  Replaced UCP searching code with optimized version as implemented for Ad
8        Muncher (http://www.admuncher.com/) by Peter Kankowski. This uses a two-
9        stage table and inline lookup instead of a function, giving speed ups of 2
10        to 5 times on some simple patterns that I tested. Permission was given to
11        distribute the MultiStage2.py script that generates the tables (it's not in
12        the tarball, but is in the Subversion repository).
13    
14    2.  Updated the Unicode datatables to Unicode 5.1.0. This adds yet more
15        scripts.
16    
17    3.  Change 12 for 7.7 introduced a bug in pcre_study() when a pattern contained
18        a group with a zero qualifier. The result of the study could be incorrect,
19        or the function might crash, depending on the pattern.
20    
21    4.  Caseless matching was not working for non-ASCII characters in back
22        references. For example, /(\x{de})\1/8i was not matching \x{de}\x{fe}.
23        It now works when Unicode Property Support is available.
24    
25    5.  In pcretest, an escape such as \x{de} in the data was always generating
26        a UTF-8 string, even in non-UTF-8 mode. Now it generates a single byte in
27        non-UTF-8 mode. If the value is greater than 255, it gives a warning about
28        truncation.
29    
30    6.  Minor bugfix in pcrecpp.cc (change "" == ... to NULL == ...).
31    
32    7.  Added two (int) casts to pcregrep when printing the difference of two
33        pointers, in case they are 64-bit values.
34    
35    8.  Added comments about Mac OS X stack usage to the pcrestack man page and to
36        test 2 if it fails.
37    
38    9.  Added PCRE_CALL_CONVENTION just before the names of all exported functions,
39        and a #define of that name to empty if it is not externally set. This is to
40        allow users of MSVC to set it if necessary.
41    
42    10. The PCRE_EXP_DEFN macro which precedes exported functions was missing from
43        the convenience functions in the pcre_get.c source file.
44    
45    11. An option change at the start of a pattern that had top-level alternatives
46        could cause overwriting and/or a crash. This command provoked a crash in
47        some environments:
48    
49          printf "/(?i)[\xc3\xa9\xc3\xbd]|[\xc3\xa9\xc3\xbdA]/8\n" | pcretest
50    
51        This potential security problem was recorded as CVE-2008-2371.
52    
53    12. For a pattern where the match had to start at the beginning or immediately
54        after a newline (e.g /.*anything/ without the DOTALL flag), pcre_exec() and
55        pcre_dfa_exec() could read past the end of the passed subject if there was
56        no match. To help with detecting such bugs (e.g. with valgrind), I modified
57        pcretest so that it places the subject at the end of its malloc-ed buffer.
58    
59    13. The change to pcretest in 12 above threw up a couple more cases when pcre_
60        exec() might read past the end of the data buffer in UTF-8 mode.
61    
62    14. A similar bug to 7.3/2 existed when the PCRE_FIRSTLINE option was set and
63        the data contained the byte 0x85 as part of a UTF-8 character within its
64        first line. This applied both to normal and DFA matching.
65    
66    15. Lazy qualifiers were not working in some cases in UTF-8 mode. For example,
67        /^[^d]*?$/8 failed to match "abc".
68    
69    
70    Version 7.7 07-May-08
71    ---------------------
72    
73    1.  Applied Craig's patch to sort out a long long problem: "If we can't convert
74        a string to a long long, pretend we don't even have a long long." This is
75        done by checking for the strtoq, strtoll, and _strtoi64 functions.
76    
77    2.  Applied Craig's patch to pcrecpp.cc to restore ABI compatibility with
78        pre-7.6 versions, which defined a global no_arg variable instead of putting
79        it in the RE class. (See also #8 below.)
80    
81    3.  Remove a line of dead code, identified by coverity and reported by Nuno
82        Lopes.
83    
84    4.  Fixed two related pcregrep bugs involving -r with --include or --exclude:
85    
86        (1) The include/exclude patterns were being applied to the whole pathnames
87            of files, instead of just to the final components.
88    
89        (2) If there was more than one level of directory, the subdirectories were
90            skipped unless they satisfied the include/exclude conditions. This is
91            inconsistent with GNU grep (and could even be seen as contrary to the
92            pcregrep specification - which I improved to make it absolutely clear).
93            The action now is always to scan all levels of directory, and just
94            apply the include/exclude patterns to regular files.
95    
96    5.  Added the --include_dir and --exclude_dir patterns to pcregrep, and used
97        --exclude_dir in the tests to avoid scanning .svn directories.
98    
99    6.  Applied Craig's patch to the QuoteMeta function so that it escapes the
100        NUL character as backslash + 0 rather than backslash + NUL, because PCRE
101        doesn't support NULs in patterns.
102    
103    7.  Added some missing "const"s to declarations of static tables in
104        pcre_compile.c and pcre_dfa_exec.c.
105    
106    8.  Applied Craig's patch to pcrecpp.cc to fix a problem in OS X that was
107        caused by fix #2  above. (Subsequently also a second patch to fix the
108        first patch. And a third patch - this was a messy problem.)
109    
110    9.  Applied Craig's patch to remove the use of push_back().
111    
112    10. Applied Alan Lehotsky's patch to add REG_STARTEND support to the POSIX
113        matching function regexec().
114    
115    11. Added support for the Oniguruma syntax \g<name>, \g<n>, \g'name', \g'n',
116        which, however, unlike Perl's \g{...}, are subroutine calls, not back
117        references. PCRE supports relative numbers with this syntax (I don't think
118        Oniguruma does).
119    
120    12. Previously, a group with a zero repeat such as (...){0} was completely
121        omitted from the compiled regex. However, this means that if the group
122        was called as a subroutine from elsewhere in the pattern, things went wrong
123        (an internal error was given). Such groups are now left in the compiled
124        pattern, with a new opcode that causes them to be skipped at execution
125        time.
126    
127    13. Added the PCRE_JAVASCRIPT_COMPAT option. This makes the following changes
128        to the way PCRE behaves:
129    
130        (a) A lone ] character is dis-allowed (Perl treats it as data).
131    
132        (b) A back reference to an unmatched subpattern matches an empty string
133            (Perl fails the current match path).
134    
135        (c) A data ] in a character class must be notated as \] because if the
136            first data character in a class is ], it defines an empty class. (In
137            Perl it is not possible to have an empty class.) The empty class []
138            never matches; it forces failure and is equivalent to (*FAIL) or (?!).
139            The negative empty class [^] matches any one character, independently
140            of the DOTALL setting.
141    
142    14. A pattern such as /(?2)[]a()b](abc)/ which had a forward reference to a
143        non-existent subpattern following a character class starting with ']' and
144        containing () gave an internal compiling error instead of "reference to
145        non-existent subpattern". Fortunately, when the pattern did exist, the
146        compiled code was correct. (When scanning forwards to check for the
147        existencd of the subpattern, it was treating the data ']' as terminating
148        the class, so got the count wrong. When actually compiling, the reference
149        was subsequently set up correctly.)
150    
151    15. The "always fail" assertion (?!) is optimzed to (*FAIL) by pcre_compile;
152        it was being rejected as not supported by pcre_dfa_exec(), even though
153        other assertions are supported. I have made pcre_dfa_exec() support
154        (*FAIL).
155    
156    16. The implementation of 13c above involved the invention of a new opcode,
157        OP_ALLANY, which is like OP_ANY but doesn't check the /s flag. Since /s
158        cannot be changed at match time, I realized I could make a small
159        improvement to matching performance by compiling OP_ALLANY instead of
160        OP_ANY for "." when DOTALL was set, and then removing the runtime tests
161        on the OP_ANY path.
162    
163    17. Compiling pcretest on Windows with readline support failed without the
164        following two fixes: (1) Make the unistd.h include conditional on
165        HAVE_UNISTD_H; (2) #define isatty and fileno as _isatty and _fileno.
166    
167    18. Changed CMakeLists.txt and cmake/FindReadline.cmake to arrange for the
168        ncurses library to be included for pcretest when ReadLine support is
169        requested, but also to allow for it to be overridden. This patch came from
170        Daniel Bergström.
171    
172    19. There was a typo in the file ucpinternal.h where f0_rangeflag was defined
173        as 0x00f00000 instead of 0x00800000. Luckily, this would not have caused
174        any errors with the current Unicode tables. Thanks to Peter Kankowski for
175        spotting this.
176    
177    
178    Version 7.6 28-Jan-08
179    ---------------------
180    
181    1.  A character class containing a very large number of characters with
182        codepoints greater than 255 (in UTF-8 mode, of course) caused a buffer
183        overflow.
184    
185    2.  Patch to cut out the "long long" test in pcrecpp_unittest when
186        HAVE_LONG_LONG is not defined.
187    
188    3.  Applied Christian Ehrlicher's patch to update the CMake build files to
189        bring them up to date and include new features. This patch includes:
190    
191        - Fixed PH's badly added libz and libbz2 support.
192        - Fixed a problem with static linking.
193        - Added pcredemo. [But later removed - see 7 below.]
194        - Fixed dftables problem and added an option.
195        - Added a number of HAVE_XXX tests, including HAVE_WINDOWS_H and
196            HAVE_LONG_LONG.
197        - Added readline support for pcretest.
198        - Added an listing of the option settings after cmake has run.
199    
200    4.  A user submitted a patch to Makefile that makes it easy to create
201        "pcre.dll" under mingw when using Configure/Make. I added stuff to
202        Makefile.am that cause it to include this special target, without
203        affecting anything else. Note that the same mingw target plus all
204        the other distribution libraries and programs are now supported
205        when configuring with CMake (see 6 below) instead of with
206        Configure/Make.
207    
208    5.  Applied Craig's patch that moves no_arg into the RE class in the C++ code.
209        This is an attempt to solve the reported problem "pcrecpp::no_arg is not
210        exported in the Windows port". It has not yet been confirmed that the patch
211        solves the problem, but it does no harm.
212    
213    6.  Applied Sheri's patch to CMakeLists.txt to add NON_STANDARD_LIB_PREFIX and
214        NON_STANDARD_LIB_SUFFIX for dll names built with mingw when configured
215        with CMake, and also correct the comment about stack recursion.
216    
217    7.  Remove the automatic building of pcredemo from the ./configure system and
218        from CMakeLists.txt. The whole idea of pcredemo.c is that it is an example
219        of a program that users should build themselves after PCRE is installed, so
220        building it automatically is not really right. What is more, it gave
221        trouble in some build environments.
222    
223    8.  Further tidies to CMakeLists.txt from Sheri and Christian.
224    
225    
226    Version 7.5 10-Jan-08
227    ---------------------
228    
229    1.  Applied a patch from Craig: "This patch makes it possible to 'ignore'
230        values in parens when parsing an RE using the C++ wrapper."
231    
232    2.  Negative specials like \S did not work in character classes in UTF-8 mode.
233        Characters greater than 255 were excluded from the class instead of being
234        included.
235    
236    3.  The same bug as (2) above applied to negated POSIX classes such as
237        [:^space:].
238    
239    4.  PCRECPP_STATIC was referenced in pcrecpp_internal.h, but nowhere was it
240        defined or documented. It seems to have been a typo for PCRE_STATIC, so
241        I have changed it.
242    
243    5.  The construct (?&) was not diagnosed as a syntax error (it referenced the
244        first named subpattern) and a construct such as (?&a) would reference the
245        first named subpattern whose name started with "a" (in other words, the
246        length check was missing). Both these problems are fixed. "Subpattern name
247        expected" is now given for (?&) (a zero-length name), and this patch also
248        makes it give the same error for \k'' (previously it complained that that
249        was a reference to a non-existent subpattern).
250    
251    6.  The erroneous patterns (?+-a) and (?-+a) give different error messages;
252        this is right because (?- can be followed by option settings as well as by
253        digits. I have, however, made the messages clearer.
254    
255    7.  Patterns such as (?(1)a|b) (a pattern that contains fewer subpatterns
256        than the number used in the conditional) now cause a compile-time error.
257        This is actually not compatible with Perl, which accepts such patterns, but
258        treats the conditional as always being FALSE (as PCRE used to), but it
259        seems to me that giving a diagnostic is better.
260    
261    8.  Change "alphameric" to the more common word "alphanumeric" in comments
262        and messages.
263    
264    9.  Fix two occurrences of "backslash" in comments that should have been
265        "backspace".
266    
267    10. Remove two redundant lines of code that can never be obeyed (their function
268        was moved elsewhere).
269    
270    11. The program that makes PCRE's Unicode character property table had a bug
271        which caused it to generate incorrect table entries for sequences of
272        characters that have the same character type, but are in different scripts.
273        It amalgamated them into a single range, with the script of the first of
274        them. In other words, some characters were in the wrong script. There were
275        thirteen such cases, affecting characters in the following ranges:
276    
277          U+002b0 - U+002c1
278          U+0060c - U+0060d
279          U+0061e - U+00612
280          U+0064b - U+0065e
281          U+0074d - U+0076d
282          U+01800 - U+01805
283          U+01d00 - U+01d77
284          U+01d9b - U+01dbf
285          U+0200b - U+0200f
286          U+030fc - U+030fe
287          U+03260 - U+0327f
288          U+0fb46 - U+0fbb1
289          U+10450 - U+1049d
290    
291    12. The -o option (show only the matching part of a line) for pcregrep was not
292        compatible with GNU grep in that, if there was more than one match in a
293        line, it showed only the first of them. It now behaves in the same way as
294        GNU grep.
295    
296    13. If the -o and -v options were combined for pcregrep, it printed a blank
297        line for every non-matching line. GNU grep prints nothing, and pcregrep now
298        does the same. The return code can be used to tell if there were any
299        non-matching lines.
300    
301    14. Added --file-offsets and --line-offsets to pcregrep.
302    
303    15. The pattern (?=something)(?R) was not being diagnosed as a potentially
304        infinitely looping recursion. The bug was that positive lookaheads were not
305        being skipped when checking for a possible empty match (negative lookaheads
306        and both kinds of lookbehind were skipped).
307    
308    16. Fixed two typos in the Windows-only code in pcregrep.c, and moved the
309        inclusion of <windows.h> to before rather than after the definition of
310        INVALID_FILE_ATTRIBUTES (patch from David Byron).
311    
312    17. Specifying a possessive quantifier with a specific limit for a Unicode
313        character property caused pcre_compile() to compile bad code, which led at
314        runtime to PCRE_ERROR_INTERNAL (-14). Examples of patterns that caused this
315        are: /\p{Zl}{2,3}+/8 and /\p{Cc}{2}+/8. It was the possessive "+" that
316        caused the error; without that there was no problem.
317    
318    18. Added --enable-pcregrep-libz and --enable-pcregrep-libbz2.
319    
320    19. Added --enable-pcretest-libreadline.
321    
322    20. In pcrecpp.cc, the variable 'count' was incremented twice in
323        RE::GlobalReplace(). As a result, the number of replacements returned was
324        double what it should be. I removed one of the increments, but Craig sent a
325        later patch that removed the other one (the right fix) and added unit tests
326        that check the return values (which was not done before).
327    
328    21. Several CMake things:
329    
330        (1) Arranged that, when cmake is used on Unix, the libraries end up with
331            the names libpcre and libpcreposix, not just pcre and pcreposix.
332    
333        (2) The above change means that pcretest and pcregrep are now correctly
334            linked with the newly-built libraries, not previously installed ones.
335    
336        (3) Added PCRE_SUPPORT_LIBREADLINE, PCRE_SUPPORT_LIBZ, PCRE_SUPPORT_LIBBZ2.
337    
338    22. In UTF-8 mode, with newline set to "any", a pattern such as .*a.*=.b.*
339        crashed when matching a string such as a\x{2029}b (note that \x{2029} is a
340        UTF-8 newline character). The key issue is that the pattern starts .*;
341        this means that the match must be either at the beginning, or after a
342        newline. The bug was in the code for advancing after a failed match and
343        checking that the new position followed a newline. It was not taking
344        account of UTF-8 characters correctly.
345    
346    23. PCRE was behaving differently from Perl in the way it recognized POSIX
347        character classes. PCRE was not treating the sequence [:...:] as a
348        character class unless the ... were all letters. Perl, however, seems to
349        allow any characters between [: and :], though of course it rejects as
350        unknown any "names" that contain non-letters, because all the known class
351        names consist only of letters. Thus, Perl gives an error for [[:1234:]],
352        for example, whereas PCRE did not - it did not recognize a POSIX character
353        class. This seemed a bit dangerous, so the code has been changed to be
354        closer to Perl. The behaviour is not identical to Perl, because PCRE will
355        diagnose an unknown class for, for example, [[:l\ower:]] where Perl will
356        treat it as [[:lower:]]. However, PCRE does now give "unknown" errors where
357        Perl does, and where it didn't before.
358    
359    24. Rewrite so as to remove the single use of %n from pcregrep because in some
360        Windows environments %n is disabled by default.
361    
362    
363    Version 7.4 21-Sep-07
364    ---------------------
365    
366    1.  Change 7.3/28 was implemented for classes by looking at the bitmap. This
367        means that a class such as [\s] counted as "explicit reference to CR or
368        LF". That isn't really right - the whole point of the change was to try to
369        help when there was an actual mention of one of the two characters. So now
370        the change happens only if \r or \n (or a literal CR or LF) character is
371        encountered.
372    
373    2.  The 32-bit options word was also used for 6 internal flags, but the numbers
374        of both had grown to the point where there were only 3 bits left.
375        Fortunately, there was spare space in the data structure, and so I have
376        moved the internal flags into a new 16-bit field to free up more option
377        bits.
378    
379    3.  The appearance of (?J) at the start of a pattern set the DUPNAMES option,
380        but did not set the internal JCHANGED flag - either of these is enough to
381        control the way the "get" function works - but the PCRE_INFO_JCHANGED
382        facility is supposed to tell if (?J) was ever used, so now (?J) at the
383        start sets both bits.
384    
385    4.  Added options (at build time, compile time, exec time) to change \R from
386        matching any Unicode line ending sequence to just matching CR, LF, or CRLF.
387    
388    5.  doc/pcresyntax.html was missing from the distribution.
389    
390    6.  Put back the definition of PCRE_ERROR_NULLWSLIMIT, for backward
391        compatibility, even though it is no longer used.
392    
393    7.  Added macro for snprintf to pcrecpp_unittest.cc and also for strtoll and
394        strtoull to pcrecpp.cc to select the available functions in WIN32 when the
395        windows.h file is present (where different names are used). [This was
396        reversed later after testing - see 16 below.]
397    
398    8.  Changed all #include <config.h> to #include "config.h". There were also
399        some further <pcre.h> cases that I changed to "pcre.h".
400    
401    9.  When pcregrep was used with the --colour option, it missed the line ending
402        sequence off the lines that it output.
403    
404    10. It was pointed out to me that arrays of string pointers cause lots of
405        relocations when a shared library is dynamically loaded. A technique of
406        using a single long string with a table of offsets can drastically reduce
407        these. I have refactored PCRE in four places to do this. The result is
408        dramatic:
409    
410          Originally:                          290
411          After changing UCP table:            187
412          After changing error message table:   43
413          After changing table of "verbs"       36
414          After changing table of Posix names   22
415    
416        Thanks to the folks working on Gregex for glib for this insight.
417    
418    11. --disable-stack-for-recursion caused compiling to fail unless -enable-
419        unicode-properties was also set.
420    
421    12. Updated the tests so that they work when \R is defaulted to ANYCRLF.
422    
423    13. Added checks for ANY and ANYCRLF to pcrecpp.cc where it previously
424        checked only for CRLF.
425    
426    14. Added casts to pcretest.c to avoid compiler warnings.
427    
428    15. Added Craig's patch to various pcrecpp modules to avoid compiler warnings.
429    
430    16. Added Craig's patch to remove the WINDOWS_H tests, that were not working,
431        and instead check for _strtoi64 explicitly, and avoid the use of snprintf()
432        entirely. This removes changes made in 7 above.
433    
434    17. The CMake files have been updated, and there is now more information about
435        building with CMake in the NON-UNIX-USE document.
436    
437    
438    Version 7.3 28-Aug-07
439    ---------------------
440    
441     1. In the rejigging of the build system that eventually resulted in 7.1, the
442        line "#include <pcre.h>" was included in pcre_internal.h. The use of angle
443        brackets there is not right, since it causes compilers to look for an
444        installed pcre.h, not the version that is in the source that is being
445        compiled (which of course may be different). I have changed it back to:
446    
447          #include "pcre.h"
448    
449        I have a vague recollection that the change was concerned with compiling in
450        different directories, but in the new build system, that is taken care of
451        by the VPATH setting the Makefile.
452    
453     2. The pattern .*$ when run in not-DOTALL UTF-8 mode with newline=any failed
454        when the subject happened to end in the byte 0x85 (e.g. if the last
455        character was \x{1ec5}). *Character* 0x85 is one of the "any" newline
456        characters but of course it shouldn't be taken as a newline when it is part
457        of another character. The bug was that, for an unlimited repeat of . in
458        not-DOTALL UTF-8 mode, PCRE was advancing by bytes rather than by
459        characters when looking for a newline.
460    
461     3. A small performance improvement in the DOTALL UTF-8 mode .* case.
462    
463     4. Debugging: adjusted the names of opcodes for different kinds of parentheses
464        in debug output.
465    
466     5. Arrange to use "%I64d" instead of "%lld" and "%I64u" instead of "%llu" for
467        long printing in the pcrecpp unittest when running under MinGW.
468    
469     6. ESC_K was left out of the EBCDIC table.
470    
471     7. Change 7.0/38 introduced a new limit on the number of nested non-capturing
472        parentheses; I made it 1000, which seemed large enough. Unfortunately, the
473        limit also applies to "virtual nesting" when a pattern is recursive, and in
474        this case 1000 isn't so big. I have been able to remove this limit at the
475        expense of backing off one optimization in certain circumstances. Normally,
476        when pcre_exec() would call its internal match() function recursively and
477        immediately return the result unconditionally, it uses a "tail recursion"
478        feature to save stack. However, when a subpattern that can match an empty
479        string has an unlimited repetition quantifier, it no longer makes this
480        optimization. That gives it a stack frame in which to save the data for
481        checking that an empty string has been matched. Previously this was taken
482        from the 1000-entry workspace that had been reserved. So now there is no
483        explicit limit, but more stack is used.
484    
485     8. Applied Daniel's patches to solve problems with the import/export magic
486        syntax that is required for Windows, and which was going wrong for the
487        pcreposix and pcrecpp parts of the library. These were overlooked when this
488        problem was solved for the main library.
489    
490     9. There were some crude static tests to avoid integer overflow when computing
491        the size of patterns that contain repeated groups with explicit upper
492        limits. As the maximum quantifier is 65535, the maximum group length was
493        set at 30,000 so that the product of these two numbers did not overflow a
494        32-bit integer. However, it turns out that people want to use groups that
495        are longer than 30,000 bytes (though not repeat them that many times).
496        Change 7.0/17 (the refactoring of the way the pattern size is computed) has
497        made it possible to implement the integer overflow checks in a much more
498        dynamic way, which I have now done. The artificial limitation on group
499        length has been removed - we now have only the limit on the total length of
500        the compiled pattern, which depends on the LINK_SIZE setting.
501    
502    10. Fixed a bug in the documentation for get/copy named substring when
503        duplicate names are permitted. If none of the named substrings are set, the
504        functions return PCRE_ERROR_NOSUBSTRING (7); the doc said they returned an
505        empty string.
506    
507    11. Because Perl interprets \Q...\E at a high level, and ignores orphan \E
508        instances, patterns such as [\Q\E] or [\E] or even [^\E] cause an error,
509        because the ] is interpreted as the first data character and the
510        terminating ] is not found. PCRE has been made compatible with Perl in this
511        regard. Previously, it interpreted [\Q\E] as an empty class, and [\E] could
512        cause memory overwriting.
513    
514    10. Like Perl, PCRE automatically breaks an unlimited repeat after an empty
515        string has been matched (to stop an infinite loop). It was not recognizing
516        a conditional subpattern that could match an empty string if that
517        subpattern was within another subpattern. For example, it looped when
518        trying to match  (((?(1)X|))*)  but it was OK with  ((?(1)X|)*)  where the
519        condition was not nested. This bug has been fixed.
520    
521    12. A pattern like \X?\d or \P{L}?\d in non-UTF-8 mode could cause a backtrack
522        past the start of the subject in the presence of bytes with the top bit
523        set, for example "\x8aBCD".
524    
525    13. Added Perl 5.10 experimental backtracking controls (*FAIL), (*F), (*PRUNE),
526        (*SKIP), (*THEN), (*COMMIT), and (*ACCEPT).
527    
528    14. Optimized (?!) to (*FAIL).
529    
530    15. Updated the test for a valid UTF-8 string to conform to the later RFC 3629.
531        This restricts code points to be within the range 0 to 0x10FFFF, excluding
532        the "low surrogate" sequence 0xD800 to 0xDFFF. Previously, PCRE allowed the
533        full range 0 to 0x7FFFFFFF, as defined by RFC 2279. Internally, it still
534        does: it's just the validity check that is more restrictive.
535    
536    16. Inserted checks for integer overflows during escape sequence (backslash)
537        processing, and also fixed erroneous offset values for syntax errors during
538        backslash processing.
539    
540    17. Fixed another case of looking too far back in non-UTF-8 mode (cf 12 above)
541        for patterns like [\PPP\x8a]{1,}\x80 with the subject "A\x80".
542    
543    18. An unterminated class in a pattern like (?1)\c[ with a "forward reference"
544        caused an overrun.
545    
546    19. A pattern like (?:[\PPa*]*){8,} which had an "extended class" (one with
547        something other than just ASCII characters) inside a group that had an
548        unlimited repeat caused a loop at compile time (while checking to see
549        whether the group could match an empty string).
550    
551    20. Debugging a pattern containing \p or \P could cause a crash. For example,
552        [\P{Any}] did so. (Error in the code for printing property names.)
553    
554    21. An orphan \E inside a character class could cause a crash.
555    
556    22. A repeated capturing bracket such as (A)? could cause a wild memory
557        reference during compilation.
558    
559    23. There are several functions in pcre_compile() that scan along a compiled
560        expression for various reasons (e.g. to see if it's fixed length for look
561        behind). There were bugs in these functions when a repeated \p or \P was
562        present in the pattern. These operators have additional parameters compared
563        with \d, etc, and these were not being taken into account when moving along
564        the compiled data. Specifically:
565    
566        (a) A item such as \p{Yi}{3} in a lookbehind was not treated as fixed
567            length.
568    
569        (b) An item such as \pL+ within a repeated group could cause crashes or
570            loops.
571    
572        (c) A pattern such as \p{Yi}+(\P{Yi}+)(?1) could give an incorrect
573            "reference to non-existent subpattern" error.
574    
575        (d) A pattern like (\P{Yi}{2}\277)? could loop at compile time.
576    
577    24. A repeated \S or \W in UTF-8 mode could give wrong answers when multibyte
578        characters were involved (for example /\S{2}/8g with "A\x{a3}BC").
579    
580    25. Using pcregrep in multiline, inverted mode (-Mv) caused it to loop.
581    
582    26. Patterns such as [\P{Yi}A] which include \p or \P and just one other
583        character were causing crashes (broken optimization).
584    
585    27. Patterns such as (\P{Yi}*\277)* (group with possible zero repeat containing
586        \p or \P) caused a compile-time loop.
587    
588    28. More problems have arisen in unanchored patterns when CRLF is a valid line
589        break. For example, the unstudied pattern [\r\n]A does not match the string
590        "\r\nA" because change 7.0/46 below moves the current point on by two
591        characters after failing to match at the start. However, the pattern \nA
592        *does* match, because it doesn't start till \n, and if [\r\n]A is studied,
593        the same is true. There doesn't seem any very clean way out of this, but
594        what I have chosen to do makes the common cases work: PCRE now takes note
595        of whether there can be an explicit match for \r or \n anywhere in the
596        pattern, and if so, 7.0/46 no longer applies. As part of this change,
597        there's a new PCRE_INFO_HASCRORLF option for finding out whether a compiled
598        pattern has explicit CR or LF references.
599    
600    29. Added (*CR) etc for changing newline setting at start of pattern.
601    
602    
603    Version 7.2 19-Jun-07
604    ---------------------
605    
606     1. If the fr_FR locale cannot be found for test 3, try the "french" locale,
607        which is apparently normally available under Windows.
608    
609     2. Re-jig the pcregrep tests with different newline settings in an attempt
610        to make them independent of the local environment's newline setting.
611    
612     3. Add code to configure.ac to remove -g from the CFLAGS default settings.
613    
614     4. Some of the "internals" tests were previously cut out when the link size
615        was not 2, because the output contained actual offsets. The recent new
616        "Z" feature of pcretest means that these can be cut out, making the tests
617        usable with all link sizes.
618    
619     5. Implemented Stan Switzer's goto replacement for longjmp() when not using
620        stack recursion. This gives a massive performance boost under BSD, but just
621        a small improvement under Linux. However, it saves one field in the frame
622        in all cases.
623    
624     6. Added more features from the forthcoming Perl 5.10:
625    
626        (a) (?-n) (where n is a string of digits) is a relative subroutine or
627            recursion call. It refers to the nth most recently opened parentheses.
628    
629        (b) (?+n) is also a relative subroutine call; it refers to the nth next
630            to be opened parentheses.
631    
632        (c) Conditions that refer to capturing parentheses can be specified
633            relatively, for example, (?(-2)... or (?(+3)...
634    
635        (d) \K resets the start of the current match so that everything before
636            is not part of it.
637    
638        (e) \k{name} is synonymous with \k<name> and \k'name' (.NET compatible).
639    
640        (f) \g{name} is another synonym - part of Perl 5.10's unification of
641            reference syntax.
642    
643        (g) (?| introduces a group in which the numbering of parentheses in each
644            alternative starts with the same number.
645    
646        (h) \h, \H, \v, and \V match horizontal and vertical whitespace.
647    
648     7. Added two new calls to pcre_fullinfo(): PCRE_INFO_OKPARTIAL and
649        PCRE_INFO_JCHANGED.
650    
651     8. A pattern such as  (.*(.)?)*  caused pcre_exec() to fail by either not
652        terminating or by crashing. Diagnosed by Viktor Griph; it was in the code
653        for detecting groups that can match an empty string.
654    
655     9. A pattern with a very large number of alternatives (more than several
656        hundred) was running out of internal workspace during the pre-compile
657        phase, where pcre_compile() figures out how much memory will be needed. A
658        bit of new cunning has reduced the workspace needed for groups with
659        alternatives. The 1000-alternative test pattern now uses 12 bytes of
660        workspace instead of running out of the 4096 that are available.
661    
662    10. Inserted some missing (unsigned int) casts to get rid of compiler warnings.
663    
664    11. Applied patch from Google to remove an optimization that didn't quite work.
665        The report of the bug said:
666    
667          pcrecpp::RE("a*").FullMatch("aaa") matches, while
668          pcrecpp::RE("a*?").FullMatch("aaa") does not, and
669          pcrecpp::RE("a*?\\z").FullMatch("aaa") does again.
670    
671    12. If \p or \P was used in non-UTF-8 mode on a character greater than 127
672        it matched the wrong number of bytes.
673    
674    
675    Version 7.1 24-Apr-07
676    ---------------------
677    
678     1. Applied Bob Rossi and Daniel G's patches to convert the build system to one
679        that is more "standard", making use of automake and other Autotools. There
680        is some re-arrangement of the files and adjustment of comments consequent
681        on this.
682    
683     2. Part of the patch fixed a problem with the pcregrep tests. The test of -r
684        for recursive directory scanning broke on some systems because the files
685        are not scanned in any specific order and on different systems the order
686        was different. A call to "sort" has been inserted into RunGrepTest for the
687        approprate test as a short-term fix. In the longer term there may be an
688        alternative.
689    
690     3. I had an email from Eric Raymond about problems translating some of PCRE's
691        man pages to HTML (despite the fact that I distribute HTML pages, some
692        people do their own conversions for various reasons). The problems
693        concerned the use of low-level troff macros .br and .in. I have therefore
694        removed all such uses from the man pages (some were redundant, some could
695        be replaced by .nf/.fi pairs). The 132html script that I use to generate
696        HTML has been updated to handle .nf/.fi and to complain if it encounters
697        .br or .in.
698    
699     4. Updated comments in configure.ac that get placed in config.h.in and also
700        arranged for config.h to be included in the distribution, with the name
701        config.h.generic, for the benefit of those who have to compile without
702        Autotools (compare pcre.h, which is now distributed as pcre.h.generic).
703    
704     5. Updated the support (such as it is) for Virtual Pascal, thanks to Stefan
705        Weber: (1) pcre_internal.h was missing some function renames; (2) updated
706        makevp.bat for the current PCRE, using the additional files
707        makevp_c.txt, makevp_l.txt, and pcregexp.pas.
708    
709     6. A Windows user reported a minor discrepancy with test 2, which turned out
710        to be caused by a trailing space on an input line that had got lost in his
711        copy. The trailing space was an accident, so I've just removed it.
712    
713     7. Add -Wl,-R... flags in pcre-config.in for *BSD* systems, as I'm told
714        that is needed.
715    
716     8. Mark ucp_table (in ucptable.h) and ucp_gentype (in pcre_ucp_searchfuncs.c)
717        as "const" (a) because they are and (b) because it helps the PHP
718        maintainers who have recently made a script to detect big data structures
719        in the php code that should be moved to the .rodata section. I remembered
720        to update Builducptable as well, so it won't revert if ucptable.h is ever
721        re-created.
722    
723     9. Added some extra #ifdef SUPPORT_UTF8 conditionals into pcretest.c,
724        pcre_printint.src, pcre_compile.c, pcre_study.c, and pcre_tables.c, in
725        order to be able to cut out the UTF-8 tables in the latter when UTF-8
726        support is not required. This saves 1.5-2K of code, which is important in
727        some applications.
728    
729        Later: more #ifdefs are needed in pcre_ord2utf8.c and pcre_valid_utf8.c
730        so as not to refer to the tables, even though these functions will never be
731        called when UTF-8 support is disabled. Otherwise there are problems with a
732        shared library.
733    
734    10. Fixed two bugs in the emulated memmove() function in pcre_internal.h:
735    
736        (a) It was defining its arguments as char * instead of void *.
737    
738        (b) It was assuming that all moves were upwards in memory; this was true
739            a long time ago when I wrote it, but is no longer the case.
740    
741        The emulated memove() is provided for those environments that have neither
742        memmove() nor bcopy(). I didn't think anyone used it these days, but that
743        is clearly not the case, as these two bugs were recently reported.
744    
745    11. The script PrepareRelease is now distributed: it calls 132html, CleanTxt,
746        and Detrail to create the HTML documentation, the .txt form of the man
747        pages, and it removes trailing spaces from listed files. It also creates
748        pcre.h.generic and config.h.generic from pcre.h and config.h. In the latter
749        case, it wraps all the #defines with #ifndefs. This script should be run
750        before "make dist".
751    
752    12. Fixed two fairly obscure bugs concerned with quantified caseless matching
753        with Unicode property support.
754    
755        (a) For a maximizing quantifier, if the two different cases of the
756            character were of different lengths in their UTF-8 codings (there are
757            some cases like this - I found 11), and the matching function had to
758            back up over a mixture of the two cases, it incorrectly assumed they
759            were both the same length.
760    
761        (b) When PCRE was configured to use the heap rather than the stack for
762            recursion during matching, it was not correctly preserving the data for
763            the other case of a UTF-8 character when checking ahead for a match
764            while processing a minimizing repeat. If the check also involved
765            matching a wide character, but failed, corruption could cause an
766            erroneous result when trying to check for a repeat of the original
767            character.
768    
769    13. Some tidying changes to the testing mechanism:
770    
771        (a) The RunTest script now detects the internal link size and whether there
772            is UTF-8 and UCP support by running ./pcretest -C instead of relying on
773            values substituted by "configure". (The RunGrepTest script already did
774            this for UTF-8.) The configure.ac script no longer substitutes the
775            relevant variables.
776    
777        (b) The debugging options /B and /D in pcretest show the compiled bytecode
778            with length and offset values. This means that the output is different
779            for different internal link sizes. Test 2 is skipped for link sizes
780            other than 2 because of this, bypassing the problem. Unfortunately,
781            there was also a test in test 3 (the locale tests) that used /B and
782            failed for link sizes other than 2. Rather than cut the whole test out,
783            I have added a new /Z option to pcretest that replaces the length and
784            offset values with spaces. This is now used to make test 3 independent
785            of link size. (Test 2 will be tidied up later.)
786    
787    14. If erroroffset was passed as NULL to pcre_compile, it provoked a
788        segmentation fault instead of returning the appropriate error message.
789    
790    15. In multiline mode when the newline sequence was set to "any", the pattern
791        ^$ would give a match between the \r and \n of a subject such as "A\r\nB".
792        This doesn't seem right; it now treats the CRLF combination as the line
793        ending, and so does not match in that case. It's only a pattern such as ^$
794        that would hit this one: something like ^ABC$ would have failed after \r
795        and then tried again after \r\n.
796    
797    16. Changed the comparison command for RunGrepTest from "diff -u" to "diff -ub"
798        in an attempt to make files that differ only in their line terminators
799        compare equal. This works on Linux.
800    
801    17. Under certain error circumstances pcregrep might try to free random memory
802        as it exited. This is now fixed, thanks to valgrind.
803    
804    19. In pcretest, if the pattern /(?m)^$/g<any> was matched against the string
805        "abc\r\n\r\n", it found an unwanted second match after the second \r. This
806        was because its rules for how to advance for /g after matching an empty
807        string at the end of a line did not allow for this case. They now check for
808        it specially.
809    
810    20. pcretest is supposed to handle patterns and data of any length, by
811        extending its buffers when necessary. It was getting this wrong when the
812        buffer for a data line had to be extended.
813    
814    21. Added PCRE_NEWLINE_ANYCRLF which is like ANY, but matches only CR, LF, or
815        CRLF as a newline sequence.
816    
817    22. Code for handling Unicode properties in pcre_dfa_exec() wasn't being cut
818        out by #ifdef SUPPORT_UCP. This did no harm, as it could never be used, but
819        I have nevertheless tidied it up.
820    
821    23. Added some casts to kill warnings from HP-UX ia64 compiler.
822    
823    24. Added a man page for pcre-config.
824    
825    
826    Version 7.0 19-Dec-06
827    ---------------------
828    
829     1. Fixed a signed/unsigned compiler warning in pcre_compile.c, shown up by
830        moving to gcc 4.1.1.
831    
832     2. The -S option for pcretest uses setrlimit(); I had omitted to #include
833        sys/time.h, which is documented as needed for this function. It doesn't
834        seem to matter on Linux, but it showed up on some releases of OS X.
835    
836     3. It seems that there are systems where bytes whose values are greater than
837        127 match isprint() in the "C" locale. The "C" locale should be the
838        default when a C program starts up. In most systems, only ASCII printing
839        characters match isprint(). This difference caused the output from pcretest
840        to vary, making some of the tests fail. I have changed pcretest so that:
841    
842        (a) When it is outputting text in the compiled version of a pattern, bytes
843            other than 32-126 are always shown as hex escapes.
844    
845        (b) When it is outputting text that is a matched part of a subject string,
846            it does the same, unless a different locale has been set for the match
847            (using the /L modifier). In this case, it uses isprint() to decide.
848    
849     4. Fixed a major bug that caused incorrect computation of the amount of memory
850        required for a compiled pattern when options that changed within the
851        pattern affected the logic of the preliminary scan that determines the
852        length. The relevant options are -x, and -i in UTF-8 mode. The result was
853        that the computed length was too small. The symptoms of this bug were
854        either the PCRE error "internal error: code overflow" from pcre_compile(),
855        or a glibc crash with a message such as "pcretest: free(): invalid next
856        size (fast)". Examples of patterns that provoked this bug (shown in
857        pcretest format) are:
858    
859          /(?-x: )/x
860          /(?x)(?-x: \s*#\s*)/
861          /((?i)[\x{c0}])/8
862          /(?i:[\x{c0}])/8
863    
864        HOWEVER: Change 17 below makes this fix obsolete as the memory computation
865        is now done differently.
866    
867     5. Applied patches from Google to: (a) add a QuoteMeta function to the C++
868        wrapper classes; (b) implement a new function in the C++ scanner that is
869        more efficient than the old way of doing things because it avoids levels of
870        recursion in the regex matching; (c) add a paragraph to the documentation
871        for the FullMatch() function.
872    
873     6. The escape sequence \n was being treated as whatever was defined as
874        "newline". Not only was this contrary to the documentation, which states
875        that \n is character 10 (hex 0A), but it also went horribly wrong when
876        "newline" was defined as CRLF. This has been fixed.
877    
878     7. In pcre_dfa_exec.c the value of an unsigned integer (the variable called c)
879        was being set to -1 for the "end of line" case (supposedly a value that no
880        character can have). Though this value is never used (the check for end of
881        line is "zero bytes in current character"), it caused compiler complaints.
882        I've changed it to 0xffffffff.
883    
884     8. In pcre_version.c, the version string was being built by a sequence of
885        C macros that, in the event of PCRE_PRERELEASE being defined as an empty
886        string (as it is for production releases) called a macro with an empty
887        argument. The C standard says the result of this is undefined. The gcc
888        compiler treats it as an empty string (which was what was wanted) but it is
889        reported that Visual C gives an error. The source has been hacked around to
890        avoid this problem.
891    
892     9. On the advice of a Windows user, included <io.h> and <fcntl.h> in Windows
893        builds of pcretest, and changed the call to _setmode() to use _O_BINARY
894        instead of 0x8000. Made all the #ifdefs test both _WIN32 and WIN32 (not all
895        of them did).
896    
897    10. Originally, pcretest opened its input and output without "b"; then I was
898        told that "b" was needed in some environments, so it was added for release
899        5.0 to both the input and output. (It makes no difference on Unix-like
900        systems.) Later I was told that it is wrong for the input on Windows. I've
901        now abstracted the modes into two macros, to make it easier to fiddle with
902        them, and removed "b" from the input mode under Windows.
903    
904    11. Added pkgconfig support for the C++ wrapper library, libpcrecpp.
905    
906    12. Added -help and --help to pcretest as an official way of being reminded
907        of the options.
908    
909    13. Removed some redundant semicolons after macro calls in pcrecpparg.h.in
910        and pcrecpp.cc because they annoy compilers at high warning levels.
911    
912    14. A bit of tidying/refactoring in pcre_exec.c in the main bumpalong loop.
913    
914    15. Fixed an occurrence of == in configure.ac that should have been = (shell
915        scripts are not C programs :-) and which was not noticed because it works
916        on Linux.
917    
918    16. pcretest is supposed to handle any length of pattern and data line (as one
919        line or as a continued sequence of lines) by extending its input buffer if
920        necessary. This feature was broken for very long pattern lines, leading to
921        a string of junk being passed to pcre_compile() if the pattern was longer
922        than about 50K.
923    
924    17. I have done a major re-factoring of the way pcre_compile() computes the
925        amount of memory needed for a compiled pattern. Previously, there was code
926        that made a preliminary scan of the pattern in order to do this. That was
927        OK when PCRE was new, but as the facilities have expanded, it has become
928        harder and harder to keep it in step with the real compile phase, and there
929        have been a number of bugs (see for example, 4 above). I have now found a
930        cunning way of running the real compile function in a "fake" mode that
931        enables it to compute how much memory it would need, while actually only
932        ever using a few hundred bytes of working memory and without too many
933        tests of the mode. This should make future maintenance and development
934        easier. A side effect of this work is that the limit of 200 on the nesting
935        depth of parentheses has been removed (though this was never a serious
936        limitation, I suspect). However, there is a downside: pcre_compile() now
937        runs more slowly than before (30% or more, depending on the pattern). I
938        hope this isn't a big issue. There is no effect on runtime performance.
939    
940    18. Fixed a minor bug in pcretest: if a pattern line was not terminated by a
941        newline (only possible for the last line of a file) and it was a
942        pattern that set a locale (followed by /Lsomething), pcretest crashed.
943    
944    19. Added additional timing features to pcretest. (1) The -tm option now times
945        matching only, not compiling. (2) Both -t and -tm can be followed, as a
946        separate command line item, by a number that specifies the number of
947        repeats to use when timing. The default is 50000; this gives better
948        precision, but takes uncomfortably long for very large patterns.
949    
950    20. Extended pcre_study() to be more clever in cases where a branch of a
951        subpattern has no definite first character. For example, (a*|b*)[cd] would
952        previously give no result from pcre_study(). Now it recognizes that the
953        first character must be a, b, c, or d.
954    
955    21. There was an incorrect error "recursive call could loop indefinitely" if
956        a subpattern (or the entire pattern) that was being tested for matching an
957        empty string contained only one non-empty item after a nested subpattern.
958        For example, the pattern (?>\x{100}*)\d(?R) provoked this error
959        incorrectly, because the \d was being skipped in the check.
960    
961    22. The pcretest program now has a new pattern option /B and a command line
962        option -b, which is equivalent to adding /B to every pattern. This causes
963        it to show the compiled bytecode, without the additional information that
964        -d shows. The effect of -d is now the same as -b with -i (and similarly, /D
965        is the same as /B/I).
966    
967    23. A new optimization is now able automatically to treat some sequences such
968        as a*b as a*+b. More specifically, if something simple (such as a character
969        or a simple class like \d) has an unlimited quantifier, and is followed by
970        something that cannot possibly match the quantified thing, the quantifier
971        is automatically "possessified".
972    
973    24. A recursive reference to a subpattern whose number was greater than 39
974        went wrong under certain circumstances in UTF-8 mode. This bug could also
975        have affected the operation of pcre_study().
976    
977    25. Realized that a little bit of performance could be had by replacing
978        (c & 0xc0) == 0xc0 with c >= 0xc0 when processing UTF-8 characters.
979    
980    26. Timing data from pcretest is now shown to 4 decimal places instead of 3.
981    
982    27. Possessive quantifiers such as a++ were previously implemented by turning
983        them into atomic groups such as ($>a+). Now they have their own opcodes,
984        which improves performance. This includes the automatically created ones
985        from 23 above.
986    
987    28. A pattern such as (?=(\w+))\1: which simulates an atomic group using a
988        lookahead was broken if it was not anchored. PCRE was mistakenly expecting
989        the first matched character to be a colon. This applied both to named and
990        numbered groups.
991    
992    29. The ucpinternal.h header file was missing its idempotency #ifdef.
993    
994    30. I was sent a "project" file called libpcre.a.dev which I understand makes
995        building PCRE on Windows easier, so I have included it in the distribution.
996    
997    31. There is now a check in pcretest against a ridiculously large number being
998        returned by pcre_exec() or pcre_dfa_exec(). If this happens in a /g or /G
999        loop, the loop is abandoned.
1000    
1001    32. Forward references to subpatterns in conditions such as (?(2)...) where
1002        subpattern 2 is defined later cause pcre_compile() to search forwards in
1003        the pattern for the relevant set of parentheses. This search went wrong
1004        when there were unescaped parentheses in a character class, parentheses
1005        escaped with \Q...\E, or parentheses in a #-comment in /x mode.
1006    
1007    33. "Subroutine" calls and backreferences were previously restricted to
1008        referencing subpatterns earlier in the regex. This restriction has now
1009        been removed.
1010    
1011    34. Added a number of extra features that are going to be in Perl 5.10. On the
1012        whole, these are just syntactic alternatives for features that PCRE had
1013        previously implemented using the Python syntax or my own invention. The
1014        other formats are all retained for compatibility.
1015    
1016        (a) Named groups can now be defined as (?<name>...) or (?'name'...) as well
1017            as (?P<name>...). The new forms, as well as being in Perl 5.10, are
1018            also .NET compatible.
1019    
1020        (b) A recursion or subroutine call to a named group can now be defined as
1021            (?&name) as well as (?P>name).
1022    
1023        (c) A backreference to a named group can now be defined as \k<name> or
1024            \k'name' as well as (?P=name). The new forms, as well as being in Perl
1025            5.10, are also .NET compatible.
1026    
1027        (d) A conditional reference to a named group can now use the syntax
1028            (?(<name>) or (?('name') as well as (?(name).
1029    
1030        (e) A "conditional group" of the form (?(DEFINE)...) can be used to define
1031            groups (named and numbered) that are never evaluated inline, but can be
1032            called as "subroutines" from elsewhere. In effect, the DEFINE condition
1033            is always false. There may be only one alternative in such a group.
1034    
1035        (f) A test for recursion can be given as (?(R1).. or (?(R&name)... as well
1036            as the simple (?(R). The condition is true only if the most recent
1037            recursion is that of the given number or name. It does not search out
1038            through the entire recursion stack.
1039    
1040        (g) The escape \gN or \g{N} has been added, where N is a positive or
1041            negative number, specifying an absolute or relative reference.
1042    
1043    35. Tidied to get rid of some further signed/unsigned compiler warnings and
1044        some "unreachable code" warnings.
1045    
1046    36. Updated the Unicode property tables to Unicode version 5.0.0. Amongst other
1047        things, this adds five new scripts.
1048    
1049    37. Perl ignores orphaned \E escapes completely. PCRE now does the same.
1050        There were also incompatibilities regarding the handling of \Q..\E inside
1051        character classes, for example with patterns like [\Qa\E-\Qz\E] where the
1052        hyphen was adjacent to \Q or \E. I hope I've cleared all this up now.
1053    
1054    38. Like Perl, PCRE detects when an indefinitely repeated parenthesized group
1055        matches an empty string, and forcibly breaks the loop. There were bugs in
1056        this code in non-simple cases. For a pattern such as  ^(a()*)*  matched
1057        against  aaaa  the result was just "a" rather than "aaaa", for example. Two
1058        separate and independent bugs (that affected different cases) have been
1059        fixed.
1060    
1061    39. Refactored the code to abolish the use of different opcodes for small
1062        capturing bracket numbers. This is a tidy that I avoided doing when I
1063        removed the limit on the number of capturing brackets for 3.5 back in 2001.
1064        The new approach is not only tidier, it makes it possible to reduce the
1065        memory needed to fix the previous bug (38).
1066    
1067    40. Implemented PCRE_NEWLINE_ANY to recognize any of the Unicode newline
1068        sequences (http://unicode.org/unicode/reports/tr18/) as "newline" when
1069        processing dot, circumflex, or dollar metacharacters, or #-comments in /x
1070        mode.
1071    
1072    41. Add \R to match any Unicode newline sequence, as suggested in the Unicode
1073        report.
1074    
1075    42. Applied patch, originally from Ari Pollak, modified by Google, to allow
1076        copy construction and assignment in the C++ wrapper.
1077    
1078    43. Updated pcregrep to support "--newline=any". In the process, I fixed a
1079        couple of bugs that could have given wrong results in the "--newline=crlf"
1080        case.
1081    
1082    44. Added a number of casts and did some reorganization of signed/unsigned int
1083        variables following suggestions from Dair Grant. Also renamed the variable
1084        "this" as "item" because it is a C++ keyword.
1085    
1086    45. Arranged for dftables to add
1087    
1088          #include "pcre_internal.h"
1089    
1090        to pcre_chartables.c because without it, gcc 4.x may remove the array
1091        definition from the final binary if PCRE is built into a static library and
1092        dead code stripping is activated.
1093    
1094    46. For an unanchored pattern, if a match attempt fails at the start of a
1095        newline sequence, and the newline setting is CRLF or ANY, and the next two
1096        characters are CRLF, advance by two characters instead of one.
1097    
1098    
1099    Version 6.7 04-Jul-06
1100    ---------------------
1101    
1102     1. In order to handle tests when input lines are enormously long, pcretest has
1103        been re-factored so that it automatically extends its buffers when
1104        necessary. The code is crude, but this _is_ just a test program. The
1105        default size has been increased from 32K to 50K.
1106    
1107     2. The code in pcre_study() was using the value of the re argument before
1108        testing it for NULL. (Of course, in any sensible call of the function, it
1109        won't be NULL.)
1110    
1111     3. The memmove() emulation function in pcre_internal.h, which is used on
1112        systems that lack both memmove() and bcopy() - that is, hardly ever -
1113        was missing a "static" storage class specifier.
1114    
1115     4. When UTF-8 mode was not set, PCRE looped when compiling certain patterns
1116        containing an extended class (one that cannot be represented by a bitmap
1117        because it contains high-valued characters or Unicode property items, e.g.
1118        [\pZ]). Almost always one would set UTF-8 mode when processing such a
1119        pattern, but PCRE should not loop if you do not (it no longer does).
1120        [Detail: two cases were found: (a) a repeated subpattern containing an
1121        extended class; (b) a recursive reference to a subpattern that followed a
1122        previous extended class. It wasn't skipping over the extended class
1123        correctly when UTF-8 mode was not set.]
1124    
1125     5. A negated single-character class was not being recognized as fixed-length
1126        in lookbehind assertions such as (?<=[^f]), leading to an incorrect
1127        compile error "lookbehind assertion is not fixed length".
1128    
1129     6. The RunPerlTest auxiliary script was showing an unexpected difference
1130        between PCRE and Perl for UTF-8 tests. It turns out that it is hard to
1131        write a Perl script that can interpret lines of an input file either as
1132        byte characters or as UTF-8, which is what "perltest" was being required to
1133        do for the non-UTF-8 and UTF-8 tests, respectively. Essentially what you
1134        can't do is switch easily at run time between having the "use utf8;" pragma
1135        or not. In the end, I fudged it by using the RunPerlTest script to insert
1136        "use utf8;" explicitly for the UTF-8 tests.
1137    
1138     7. In multiline (/m) mode, PCRE was matching ^ after a terminating newline at
1139        the end of the subject string, contrary to the documentation and to what
1140        Perl does. This was true of both matching functions. Now it matches only at
1141        the start of the subject and immediately after *internal* newlines.
1142    
1143     8. A call of pcre_fullinfo() from pcretest to get the option bits was passing
1144        a pointer to an int instead of a pointer to an unsigned long int. This
1145        caused problems on 64-bit systems.
1146    
1147     9. Applied a patch from the folks at Google to pcrecpp.cc, to fix "another
1148        instance of the 'standard' template library not being so standard".
1149    
1150    10. There was no check on the number of named subpatterns nor the maximum
1151        length of a subpattern name. The product of these values is used to compute
1152        the size of the memory block for a compiled pattern. By supplying a very
1153        long subpattern name and a large number of named subpatterns, the size
1154        computation could be caused to overflow. This is now prevented by limiting
1155        the length of names to 32 characters, and the number of named subpatterns
1156        to 10,000.
1157    
1158    11. Subpatterns that are repeated with specific counts have to be replicated in
1159        the compiled pattern. The size of memory for this was computed from the
1160        length of the subpattern and the repeat count. The latter is limited to
1161        65535, but there was no limit on the former, meaning that integer overflow
1162        could in principle occur. The compiled length of a repeated subpattern is
1163        now limited to 30,000 bytes in order to prevent this.
1164    
1165    12. Added the optional facility to have named substrings with the same name.
1166    
1167    13. Added the ability to use a named substring as a condition, using the
1168        Python syntax: (?(name)yes|no). This overloads (?(R)... and names that
1169        are numbers (not recommended). Forward references are permitted.
1170    
1171    14. Added forward references in named backreferences (if you see what I mean).
1172    
1173    15. In UTF-8 mode, with the PCRE_DOTALL option set, a quantified dot in the
1174        pattern could run off the end of the subject. For example, the pattern
1175        "(?s)(.{1,5})"8 did this with the subject "ab".
1176    
1177    16. If PCRE_DOTALL or PCRE_MULTILINE were set, pcre_dfa_exec() behaved as if
1178        PCRE_CASELESS was set when matching characters that were quantified with ?
1179        or *.
1180    
1181    17. A character class other than a single negated character that had a minimum
1182        but no maximum quantifier - for example [ab]{6,} - was not handled
1183        correctly by pce_dfa_exec(). It would match only one character.
1184    
1185    18. A valid (though odd) pattern that looked like a POSIX character
1186        class but used an invalid character after [ (for example [[,abc,]]) caused
1187        pcre_compile() to give the error "Failed: internal error: code overflow" or
1188        in some cases to crash with a glibc free() error. This could even happen if
1189        the pattern terminated after [[ but there just happened to be a sequence of
1190        letters, a binary zero, and a closing ] in the memory that followed.
1191    
1192    19. Perl's treatment of octal escapes in the range \400 to \777 has changed
1193        over the years. Originally (before any Unicode support), just the bottom 8
1194        bits were taken. Thus, for example, \500 really meant \100. Nowadays the
1195        output from "man perlunicode" includes this:
1196    
1197          The regular expression compiler produces polymorphic opcodes.  That
1198          is, the pattern adapts to the data and automatically switches to
1199          the Unicode character scheme when presented with Unicode data--or
1200          instead uses a traditional byte scheme when presented with byte
1201          data.
1202    
1203        Sadly, a wide octal escape does not cause a switch, and in a string with
1204        no other multibyte characters, these octal escapes are treated as before.
1205        Thus, in Perl, the pattern  /\500/ actually matches \100 but the pattern
1206        /\500|\x{1ff}/ matches \500 or \777 because the whole thing is treated as a
1207        Unicode string.
1208    
1209        I have not perpetrated such confusion in PCRE. Up till now, it took just
1210        the bottom 8 bits, as in old Perl. I have now made octal escapes with
1211        values greater than \377 illegal in non-UTF-8 mode. In UTF-8 mode they
1212        translate to the appropriate multibyte character.
1213    
1214    29. Applied some refactoring to reduce the number of warnings from Microsoft
1215        and Borland compilers. This has included removing the fudge introduced
1216        seven years ago for the OS/2 compiler (see 2.02/2 below) because it caused
1217        a warning about an unused variable.
1218    
1219    21. PCRE has not included VT (character 0x0b) in the set of whitespace
1220        characters since release 4.0, because Perl (from release 5.004) does not.
1221        [Or at least, is documented not to: some releases seem to be in conflict
1222        with the documentation.] However, when a pattern was studied with
1223        pcre_study() and all its branches started with \s, PCRE still included VT
1224        as a possible starting character. Of course, this did no harm; it just
1225        caused an unnecessary match attempt.
1226    
1227    22. Removed a now-redundant internal flag bit that recorded the fact that case
1228        dependency changed within the pattern. This was once needed for "required
1229        byte" processing, but is no longer used. This recovers a now-scarce options
1230        bit. Also moved the least significant internal flag bit to the most-
1231        significant bit of the word, which was not previously used (hangover from
1232        the days when it was an int rather than a uint) to free up another bit for
1233        the future.
1234    
1235    23. Added support for CRLF line endings as well as CR and LF. As well as the
1236        default being selectable at build time, it can now be changed at runtime
1237        via the PCRE_NEWLINE_xxx flags. There are now options for pcregrep to
1238        specify that it is scanning data with non-default line endings.
1239    
1240    24. Changed the definition of CXXLINK to make it agree with the definition of
1241        LINK in the Makefile, by replacing LDFLAGS to CXXFLAGS.
1242    
1243    25. Applied Ian Taylor's patches to avoid using another stack frame for tail
1244        recursions. This makes a big different to stack usage for some patterns.
1245    
1246    26. If a subpattern containing a named recursion or subroutine reference such
1247        as (?P>B) was quantified, for example (xxx(?P>B)){3}, the calculation of
1248        the space required for the compiled pattern went wrong and gave too small a
1249        value. Depending on the environment, this could lead to "Failed: internal
1250        error: code overflow at offset 49" or "glibc detected double free or
1251        corruption" errors.
1252    
1253    27. Applied patches from Google (a) to support the new newline modes and (b) to
1254        advance over multibyte UTF-8 characters in GlobalReplace.
1255    
1256    28. Change free() to pcre_free() in pcredemo.c. Apparently this makes a
1257        difference for some implementation of PCRE in some Windows version.
1258    
1259    29. Added some extra testing facilities to pcretest:
1260    
1261        \q<number>   in a data line sets the "match limit" value
1262        \Q<number>   in a data line sets the "match recursion limt" value
1263        -S <number>  sets the stack size, where <number> is in megabytes
1264    
1265        The -S option isn't available for Windows.
1266    
1267    
1268    Version 6.6 06-Feb-06
1269    ---------------------
1270    
1271     1. Change 16(a) for 6.5 broke things, because PCRE_DATA_SCOPE was not defined
1272        in pcreposix.h. I have copied the definition from pcre.h.
1273    
1274     2. Change 25 for 6.5 broke compilation in a build directory out-of-tree
1275        because pcre.h is no longer a built file.
1276    
1277     3. Added Jeff Friedl's additional debugging patches to pcregrep. These are
1278        not normally included in the compiled code.
1279    
1280    
1281    Version 6.5 01-Feb-06
1282    ---------------------
1283    
1284     1. When using the partial match feature with pcre_dfa_exec(), it was not
1285        anchoring the second and subsequent partial matches at the new starting
1286        point. This could lead to incorrect results. For example, with the pattern
1287        /1234/, partially matching against "123" and then "a4" gave a match.
1288    
1289     2. Changes to pcregrep:
1290    
1291        (a) All non-match returns from pcre_exec() were being treated as failures
1292            to match the line. Now, unless the error is PCRE_ERROR_NOMATCH, an
1293            error message is output. Some extra information is given for the
1294            PCRE_ERROR_MATCHLIMIT and PCRE_ERROR_RECURSIONLIMIT errors, which are
1295            probably the only errors that are likely to be caused by users (by
1296            specifying a regex that has nested indefinite repeats, for instance).
1297            If there are more than 20 of these errors, pcregrep is abandoned.
1298    
1299        (b) A binary zero was treated as data while matching, but terminated the
1300            output line if it was written out. This has been fixed: binary zeroes
1301            are now no different to any other data bytes.
1302    
1303        (c) Whichever of the LC_ALL or LC_CTYPE environment variables is set is
1304            used to set a locale for matching. The --locale=xxxx long option has
1305            been added (no short equivalent) to specify a locale explicitly on the
1306            pcregrep command, overriding the environment variables.
1307    
1308        (d) When -B was used with -n, some line numbers in the output were one less
1309            than they should have been.
1310    
1311        (e) Added the -o (--only-matching) option.
1312    
1313        (f) If -A or -C was used with -c (count only), some lines of context were
1314            accidentally printed for the final match.
1315    
1316        (g) Added the -H (--with-filename) option.
1317    
1318        (h) The combination of options -rh failed to suppress file names for files
1319            that were found from directory arguments.
1320    
1321        (i) Added the -D (--devices) and -d (--directories) options.
1322    
1323        (j) Added the -F (--fixed-strings) option.
1324    
1325        (k) Allow "-" to be used as a file name for -f as well as for a data file.
1326    
1327        (l) Added the --colo(u)r option.
1328    
1329        (m) Added Jeffrey Friedl's -S testing option, but within #ifdefs so that it
1330            is not present by default.
1331    
1332     3. A nasty bug was discovered in the handling of recursive patterns, that is,
1333        items such as (?R) or (?1), when the recursion could match a number of
1334        alternatives. If it matched one of the alternatives, but subsequently,
1335        outside the recursion, there was a failure, the code tried to back up into
1336        the recursion. However, because of the way PCRE is implemented, this is not
1337        possible, and the result was an incorrect result from the match.
1338    
1339        In order to prevent this happening, the specification of recursion has
1340        been changed so that all such subpatterns are automatically treated as
1341        atomic groups. Thus, for example, (?R) is treated as if it were (?>(?R)).
1342    
1343     4. I had overlooked the fact that, in some locales, there are characters for
1344        which isalpha() is true but neither isupper() nor islower() are true. In
1345        the fr_FR locale, for instance, the \xAA and \xBA characters (ordmasculine
1346        and ordfeminine) are like this. This affected the treatment of \w and \W
1347        when they appeared in character classes, but not when they appeared outside
1348        a character class. The bit map for "word" characters is now created
1349        separately from the results of isalnum() instead of just taking it from the
1350        upper, lower, and digit maps. (Plus the underscore character, of course.)
1351    
1352     5. The above bug also affected the handling of POSIX character classes such as
1353        [[:alpha:]] and [[:alnum:]]. These do not have their own bit maps in PCRE's
1354        permanent tables. Instead, the bit maps for such a class were previously
1355        created as the appropriate unions of the upper, lower, and digit bitmaps.
1356        Now they are created by subtraction from the [[:word:]] class, which has
1357        its own bitmap.
1358    
1359     6. The [[:blank:]] character class matches horizontal, but not vertical space.
1360        It is created by subtracting the vertical space characters (\x09, \x0a,
1361        \x0b, \x0c) from the [[:space:]] bitmap. Previously, however, the
1362        subtraction was done in the overall bitmap for a character class, meaning
1363        that a class such as [\x0c[:blank:]] was incorrect because \x0c would not
1364        be recognized. This bug has been fixed.
1365    
1366     7. Patches from the folks at Google:
1367    
1368          (a) pcrecpp.cc: "to handle a corner case that may or may not happen in
1369          real life, but is still worth protecting against".
1370    
1371          (b) pcrecpp.cc: "corrects a bug when negative radixes are used with
1372          regular expressions".
1373    
1374          (c) pcre_scanner.cc: avoid use of std::count() because not all systems
1375          have it.
1376    
1377          (d) Split off pcrecpparg.h from pcrecpp.h and had the former built by
1378          "configure" and the latter not, in order to fix a problem somebody had
1379          with compiling the Arg class on HP-UX.
1380    
1381          (e) Improve the error-handling of the C++ wrapper a little bit.
1382    
1383          (f) New tests for checking recursion limiting.
1384    
1385     8. The pcre_memmove() function, which is used only if the environment does not
1386        have a standard memmove() function (and is therefore rarely compiled),
1387        contained two bugs: (a) use of int instead of size_t, and (b) it was not
1388        returning a result (though PCRE never actually uses the result).
1389    
1390     9. In the POSIX regexec() interface, if nmatch is specified as a ridiculously
1391        large number - greater than INT_MAX/(3*sizeof(int)) - REG_ESPACE is
1392        returned instead of calling malloc() with an overflowing number that would
1393        most likely cause subsequent chaos.
1394    
1395    10. The debugging option of pcretest was not showing the NO_AUTO_CAPTURE flag.
1396    
1397    11. The POSIX flag REG_NOSUB is now supported. When a pattern that was compiled
1398        with this option is matched, the nmatch and pmatch options of regexec() are
1399        ignored.
1400    
1401    12. Added REG_UTF8 to the POSIX interface. This is not defined by POSIX, but is
1402        provided in case anyone wants to the the POSIX interface with UTF-8
1403        strings.
1404    
1405    13. Added CXXLDFLAGS to the Makefile parameters to provide settings only on the
1406        C++ linking (needed for some HP-UX environments).
1407    
1408    14. Avoid compiler warnings in get_ucpname() when compiled without UCP support
1409        (unused parameter) and in the pcre_printint() function (omitted "default"
1410        switch label when the default is to do nothing).
1411    
1412    15. Added some code to make it possible, when PCRE is compiled as a C++
1413        library, to replace subject pointers for pcre_exec() with a smart pointer
1414        class, thus making it possible to process discontinuous strings.
1415    
1416    16. The two macros PCRE_EXPORT and PCRE_DATA_SCOPE are confusing, and perform
1417        much the same function. They were added by different people who were trying
1418        to make PCRE easy to compile on non-Unix systems. It has been suggested
1419        that PCRE_EXPORT be abolished now that there is more automatic apparatus
1420        for compiling on Windows systems. I have therefore replaced it with
1421        PCRE_DATA_SCOPE. This is set automatically for Windows; if not set it
1422        defaults to "extern" for C or "extern C" for C++, which works fine on
1423        Unix-like systems. It is now possible to override the value of PCRE_DATA_
1424        SCOPE with something explicit in config.h. In addition:
1425    
1426        (a) pcreposix.h still had just "extern" instead of either of these macros;
1427            I have replaced it with PCRE_DATA_SCOPE.
1428    
1429        (b) Functions such as _pcre_xclass(), which are internal to the library,
1430            but external in the C sense, all had PCRE_EXPORT in their definitions.
1431            This is apparently wrong for the Windows case, so I have removed it.
1432            (It makes no difference on Unix-like systems.)
1433    
1434    17. Added a new limit, MATCH_LIMIT_RECURSION, which limits the depth of nesting
1435        of recursive calls to match(). This is different to MATCH_LIMIT because
1436        that limits the total number of calls to match(), not all of which increase
1437        the depth of recursion. Limiting the recursion depth limits the amount of
1438        stack (or heap if NO_RECURSE is set) that is used. The default can be set
1439        when PCRE is compiled, and changed at run time. A patch from Google adds
1440        this functionality to the C++ interface.
1441    
1442    18. Changes to the handling of Unicode character properties:
1443    
1444        (a) Updated the table to Unicode 4.1.0.
1445    
1446        (b) Recognize characters that are not in the table as "Cn" (undefined).
1447    
1448        (c) I revised the way the table is implemented to a much improved format
1449            which includes recognition of ranges. It now supports the ranges that
1450            are defined in UnicodeData.txt, and it also amalgamates other
1451            characters into ranges. This has reduced the number of entries in the
1452            table from around 16,000 to around 3,000, thus reducing its size
1453            considerably. I realized I did not need to use a tree structure after
1454            all - a binary chop search is just as efficient. Having reduced the
1455            number of entries, I extended their size from 6 bytes to 8 bytes to
1456            allow for more data.
1457    
1458        (d) Added support for Unicode script names via properties such as \p{Han}.
1459    
1460    19. In UTF-8 mode, a backslash followed by a non-Ascii character was not
1461        matching that character.
1462    
1463    20. When matching a repeated Unicode property with a minimum greater than zero,
1464        (for example \pL{2,}), PCRE could look past the end of the subject if it
1465        reached it while seeking the minimum number of characters. This could
1466        happen only if some of the characters were more than one byte long, because
1467        there is a check for at least the minimum number of bytes.
1468    
1469    21. Refactored the implementation of \p and \P so as to be more general, to
1470        allow for more different types of property in future. This has changed the
1471        compiled form incompatibly. Anybody with saved compiled patterns that use
1472        \p or \P will have to recompile them.
1473    
1474    22. Added "Any" and "L&" to the supported property types.
1475    
1476    23. Recognize \x{...} as a code point specifier, even when not in UTF-8 mode,
1477        but give a compile time error if the value is greater than 0xff.
1478    
1479    24. The man pages for pcrepartial, pcreprecompile, and pcre_compile2 were
1480        accidentally not being installed or uninstalled.
1481    
1482    25. The pcre.h file was built from pcre.h.in, but the only changes that were
1483        made were to insert the current release number. This seemed silly, because
1484        it made things harder for people building PCRE on systems that don't run
1485        "configure". I have turned pcre.h into a distributed file, no longer built
1486        by "configure", with the version identification directly included. There is
1487        no longer a pcre.h.in file.
1488    
1489        However, this change necessitated a change to the pcre-config script as
1490        well. It is built from pcre-config.in, and one of the substitutions was the
1491        release number. I have updated configure.ac so that ./configure now finds
1492        the release number by grepping pcre.h.
1493    
1494    26. Added the ability to run the tests under valgrind.
1495    
1496    
1497    Version 6.4 05-Sep-05
1498    ---------------------
1499    
1500     1. Change 6.0/10/(l) to pcregrep introduced a bug that caused separator lines
1501        "--" to be printed when multiple files were scanned, even when none of the
1502        -A, -B, or -C options were used. This is not compatible with Gnu grep, so I
1503        consider it to be a bug, and have restored the previous behaviour.
1504    
1505     2. A couple of code tidies to get rid of compiler warnings.
1506    
1507     3. The pcretest program used to cheat by referring to symbols in the library
1508        whose names begin with _pcre_. These are internal symbols that are not
1509        really supposed to be visible externally, and in some environments it is
1510        possible to suppress them. The cheating is now confined to including
1511        certain files from the library's source, which is a bit cleaner.
1512    
1513     4. Renamed pcre.in as pcre.h.in to go with pcrecpp.h.in; it also makes the
1514        file's purpose clearer.
1515    
1516     5. Reorganized pcre_ucp_findchar().
1517    
1518    
1519    Version 6.3 15-Aug-05
1520    ---------------------
1521    
1522     1. The file libpcre.pc.in did not have general read permission in the tarball.
1523    
1524     2. There were some problems when building without C++ support:
1525    
1526        (a) If C++ support was not built, "make install" and "make test" still
1527            tried to test it.
1528    
1529        (b) There were problems when the value of CXX was explicitly set. Some
1530            changes have been made to try to fix these, and ...
1531    
1532        (c) --disable-cpp can now be used to explicitly disable C++ support.
1533    
1534        (d) The use of @CPP_OBJ@ directly caused a blank line preceded by a
1535            backslash in a target when C++ was disabled. This confuses some
1536            versions of "make", apparently. Using an intermediate variable solves
1537            this. (Same for CPP_LOBJ.)
1538    
1539     3. $(LINK_FOR_BUILD) now includes $(CFLAGS_FOR_BUILD) and $(LINK)
1540        (non-Windows) now includes $(CFLAGS) because these flags are sometimes
1541        necessary on certain architectures.
1542    
1543     4. Added a setting of -export-symbols-regex to the link command to remove
1544        those symbols that are exported in the C sense, but actually are local
1545        within the library, and not documented. Their names all begin with
1546        "_pcre_". This is not a perfect job, because (a) we have to except some
1547        symbols that pcretest ("illegally") uses, and (b) the facility isn't always
1548        available (and never for static libraries). I have made a note to try to
1549        find a way round (a) in the future.
1550    
1551    
1552    Version 6.2 01-Aug-05
1553    ---------------------
1554    
1555     1. There was no test for integer overflow of quantifier values. A construction
1556        such as {1111111111111111} would give undefined results. What is worse, if
1557        a minimum quantifier for a parenthesized subpattern overflowed and became
1558        negative, the calculation of the memory size went wrong. This could have
1559        led to memory overwriting.
1560    
1561     2. Building PCRE using VPATH was broken. Hopefully it is now fixed.
1562    
1563     3. Added "b" to the 2nd argument of fopen() in dftables.c, for non-Unix-like
1564        operating environments where this matters.
1565    
1566     4. Applied Giuseppe Maxia's patch to add additional features for controlling
1567        PCRE options from within the C++ wrapper.
1568    
1569     5. Named capturing subpatterns were not being correctly counted when a pattern
1570        was compiled. This caused two problems: (a) If there were more than 100
1571        such subpatterns, the calculation of the memory needed for the whole
1572        compiled pattern went wrong, leading to an overflow error. (b) Numerical
1573        back references of the form \12, where the number was greater than 9, were
1574        not recognized as back references, even though there were sufficient
1575        previous subpatterns.
1576    
1577     6. Two minor patches to pcrecpp.cc in order to allow it to compile on older
1578        versions of gcc, e.g. 2.95.4.
1579    
1580    
1581    Version 6.1 21-Jun-05
1582    ---------------------
1583    
1584     1. There was one reference to the variable "posix" in pcretest.c that was not
1585        surrounded by "#if !defined NOPOSIX".
1586    
1587     2. Make it possible to compile pcretest without DFA support, UTF8 support, or
1588        the cross-check on the old pcre_info() function, for the benefit of the
1589        cut-down version of PCRE that is currently imported into Exim.
1590    
1591     3. A (silly) pattern starting with (?i)(?-i) caused an internal space
1592        allocation error. I've done the easy fix, which wastes 2 bytes for sensible
1593        patterns that start (?i) but I don't think that matters. The use of (?i) is
1594        just an example; this all applies to the other options as well.
1595    
1596     4. Since libtool seems to echo the compile commands it is issuing, the output
1597        from "make" can be reduced a bit by putting "@" in front of each libtool
1598        compile command.
1599    
1600     5. Patch from the folks at Google for configure.in to be a bit more thorough
1601        in checking for a suitable C++ installation before trying to compile the
1602        C++ stuff. This should fix a reported problem when a compiler was present,
1603        but no suitable headers.
1604    
1605     6. The man pages all had just "PCRE" as their title. I have changed them to
1606        be the relevant file name. I have also arranged that these names are
1607        retained in the file doc/pcre.txt, which is a concatenation in text format
1608        of all the man pages except the little individual ones for each function.
1609    
1610     7. The NON-UNIX-USE file had not been updated for the different set of source
1611        files that come with release 6. I also added a few comments about the C++
1612        wrapper.
1613    
1614    
1615    Version 6.0 07-Jun-05
1616    ---------------------
1617    
1618     1. Some minor internal re-organization to help with my DFA experiments.
1619    
1620     2. Some missing #ifdef SUPPORT_UCP conditionals in pcretest and printint that
1621        didn't matter for the library itself when fully configured, but did matter
1622        when compiling without UCP support, or within Exim, where the ucp files are
1623        not imported.
1624    
1625     3. Refactoring of the library code to split up the various functions into
1626        different source modules. The addition of the new DFA matching code (see
1627        below) to a single monolithic source would have made it really too
1628        unwieldy, quite apart from causing all the code to be include in a
1629        statically linked application, when only some functions are used. This is
1630        relevant even without the DFA addition now that patterns can be compiled in
1631        one application and matched in another.
1632    
1633        The downside of splitting up is that there have to be some external
1634        functions and data tables that are used internally in different modules of
1635        the library but which are not part of the API. These have all had their
1636        names changed to start with "_pcre_" so that they are unlikely to clash
1637        with other external names.
1638    
1639     4. Added an alternate matching function, pcre_dfa_exec(), which matches using
1640        a different (DFA) algorithm. Although it is slower than the original
1641        function, it does have some advantages for certain types of matching
1642        problem.
1643    
1644     5. Upgrades to pcretest in order to test the features of pcre_dfa_exec(),
1645        including restarting after a partial match.
1646    
1647     6. A patch for pcregrep that defines INVALID_FILE_ATTRIBUTES if it is not
1648        defined when compiling for Windows was sent to me. I have put it into the
1649        code, though I have no means of testing or verifying it.
1650    
1651     7. Added the pcre_refcount() auxiliary function.
1652    
1653     8. Added the PCRE_FIRSTLINE option. This constrains an unanchored pattern to
1654        match before or at the first newline in the subject string. In pcretest,
1655        the /f option on a pattern can be used to set this.
1656    
1657     9. A repeated \w when used in UTF-8 mode with characters greater than 256
1658        would behave wrongly. This has been present in PCRE since release 4.0.
1659    
1660    10. A number of changes to the pcregrep command:
1661    
1662        (a) Refactored how -x works; insert ^(...)$ instead of setting
1663            PCRE_ANCHORED and checking the length, in preparation for adding
1664            something similar for -w.
1665    
1666        (b) Added the -w (match as a word) option.
1667    
1668        (c) Refactored the way lines are read and buffered so as to have more
1669            than one at a time available.
1670    
1671        (d) Implemented a pcregrep test script.
1672    
1673        (e) Added the -M (multiline match) option. This allows patterns to match
1674            over several lines of the subject. The buffering ensures that at least
1675            8K, or the rest of the document (whichever is the shorter) is available
1676            for matching (and similarly the previous 8K for lookbehind assertions).
1677    
1678        (f) Changed the --help output so that it now says
1679    
1680              -w, --word-regex(p)
1681    
1682            instead of two lines, one with "regex" and the other with "regexp"
1683            because that confused at least one person since the short forms are the
1684            same. (This required a bit of code, as the output is generated
1685            automatically from a table. It wasn't just a text change.)
1686    
1687        (g) -- can be used to terminate pcregrep options if the next thing isn't an
1688            option but starts with a hyphen. Could be a pattern or a path name
1689            starting with a hyphen, for instance.
1690    
1691        (h) "-" can be given as a file name to represent stdin.
1692    
1693        (i) When file names are being printed, "(standard input)" is used for
1694            the standard input, for compatibility with GNU grep. Previously
1695            "<stdin>" was used.
1696    
1697        (j) The option --label=xxx can be used to supply a name to be used for
1698            stdin when file names are being printed. There is no short form.
1699    
1700        (k) Re-factored the options decoding logic because we are going to add
1701            two more options that take data. Such options can now be given in four
1702            different ways, e.g. "-fname", "-f name", "--file=name", "--file name".
1703    
1704        (l) Added the -A, -B, and -C options for requesting that lines of context
1705            around matches be printed.
1706    
1707        (m) Added the -L option to print the names of files that do not contain
1708            any matching lines, that is, the complement of -l.
1709    
1710        (n) The return code is 2 if any file cannot be opened, but pcregrep does
1711            continue to scan other files.
1712    
1713        (o) The -s option was incorrectly implemented. For compatibility with other
1714            greps, it now suppresses the error message for a non-existent or non-
1715            accessible file (but not the return code). There is a new option called
1716            -q that suppresses the output of matching lines, which was what -s was
1717            previously doing.
1718    
1719        (p) Added --include and --exclude options to specify files for inclusion
1720            and exclusion when recursing.
1721    
1722    11. The Makefile was not using the Autoconf-supported LDFLAGS macro properly.
1723        Hopefully, it now does.
1724    
1725    12. Missing cast in pcre_study().
1726    
1727    13. Added an "uninstall" target to the makefile.
1728    
1729    14. Replaced "extern" in the function prototypes in Makefile.in with
1730        "PCRE_DATA_SCOPE", which defaults to 'extern' or 'extern "C"' in the Unix
1731        world, but is set differently for Windows.
1732    
1733    15. Added a second compiling function called pcre_compile2(). The only
1734        difference is that it has an extra argument, which is a pointer to an
1735        integer error code. When there is a compile-time failure, this is set
1736        non-zero, in addition to the error test pointer being set to point to an
1737        error message. The new argument may be NULL if no error number is required
1738        (but then you may as well call pcre_compile(), which is now just a
1739        wrapper). This facility is provided because some applications need a
1740        numeric error indication, but it has also enabled me to tidy up the way
1741        compile-time errors are handled in the POSIX wrapper.
1742    
1743    16. Added VPATH=.libs to the makefile; this should help when building with one
1744        prefix path and installing with another. (Or so I'm told by someone who
1745        knows more about this stuff than I do.)
1746    
1747    17. Added a new option, REG_DOTALL, to the POSIX function regcomp(). This
1748        passes PCRE_DOTALL to the pcre_compile() function, making the "." character
1749        match everything, including newlines. This is not POSIX-compatible, but
1750        somebody wanted the feature. From pcretest it can be activated by using
1751        both the P and the s flags.
1752    
1753    18. AC_PROG_LIBTOOL appeared twice in Makefile.in. Removed one.
1754    
1755    19. libpcre.pc was being incorrectly installed as executable.
1756    
1757    20. A couple of places in pcretest check for end-of-line by looking for '\n';
1758        it now also looks for '\r' so that it will work unmodified on Windows.
1759    
1760    21. Added Google's contributed C++ wrapper to the distribution.
1761    
1762    22. Added some untidy missing memory free() calls in pcretest, to keep
1763        Electric Fence happy when testing.
1764    
1765    
1766    
1767    Version 5.0 13-Sep-04
1768    ---------------------
1769    
1770     1. Internal change: literal characters are no longer packed up into items
1771        containing multiple characters in a single byte-string. Each character
1772        is now matched using a separate opcode. However, there may be more than one
1773        byte in the character in UTF-8 mode.
1774    
1775     2. The pcre_callout_block structure has two new fields: pattern_position and
1776        next_item_length. These contain the offset in the pattern to the next match
1777        item, and its length, respectively.
1778    
1779     3. The PCRE_AUTO_CALLOUT option for pcre_compile() requests the automatic
1780        insertion of callouts before each pattern item. Added the /C option to
1781        pcretest to make use of this.
1782    
1783     4. On the advice of a Windows user, the lines
1784    
1785          #if defined(_WIN32) || defined(WIN32)
1786          _setmode( _fileno( stdout ), 0x8000 );
1787          #endif  /* defined(_WIN32) || defined(WIN32) */
1788    
1789        have been added to the source of pcretest. This apparently does useful
1790        magic in relation to line terminators.
1791    
1792     5. Changed "r" and "w" in the calls to fopen() in pcretest to "rb" and "wb"
1793        for the benefit of those environments where the "b" makes a difference.
1794    
1795     6. The icc compiler has the same options as gcc, but "configure" doesn't seem
1796        to know about it. I have put a hack into configure.in that adds in code
1797        to set GCC=yes if CC=icc. This seems to end up at a point in the
1798        generated configure script that is early enough to affect the setting of
1799        compiler options, which is what is needed, but I have no means of testing
1800        whether it really works. (The user who reported this had patched the
1801        generated configure script, which of course I cannot do.)
1802    
1803        LATER: After change 22 below (new libtool files), the configure script
1804        seems to know about icc (and also ecc). Therefore, I have commented out
1805        this hack in configure.in.
1806    
1807     7. Added support for pkg-config (2 patches were sent in).
1808    
1809     8. Negated POSIX character classes that used a combination of internal tables
1810        were completely broken. These were [[:^alpha:]], [[:^alnum:]], and
1811        [[:^ascii]]. Typically, they would match almost any characters. The other
1812        POSIX classes were not broken in this way.
1813    
1814     9. Matching the pattern "\b.*?" against "ab cd", starting at offset 1, failed
1815        to find the match, as PCRE was deluded into thinking that the match had to
1816        start at the start point or following a newline. The same bug applied to
1817        patterns with negative forward assertions or any backward assertions
1818        preceding ".*" at the start, unless the pattern required a fixed first
1819        character. This was a failing pattern: "(?!.bcd).*". The bug is now fixed.
1820    
1821    10. In UTF-8 mode, when moving forwards in the subject after a failed match
1822        starting at the last subject character, bytes beyond the end of the subject
1823        string were read.
1824    
1825    11. Renamed the variable "class" as "classbits" to make life easier for C++
1826        users. (Previously there was a macro definition, but it apparently wasn't
1827        enough.)
1828    
1829    12. Added the new field "tables" to the extra data so that tables can be passed
1830        in at exec time, or the internal tables can be re-selected. This allows
1831        a compiled regex to be saved and re-used at a later time by a different
1832        program that might have everything at different addresses.
1833    
1834    13. Modified the pcre-config script so that, when run on Solaris, it shows a
1835        -R library as well as a -L library.
1836    
1837    14. The debugging options of pcretest (-d on the command line or D on a
1838        pattern) showed incorrect output for anything following an extended class
1839        that contained multibyte characters and which was followed by a quantifier.
1840    
1841    15. Added optional support for general category Unicode character properties
1842        via the \p, \P, and \X escapes. Unicode property support implies UTF-8
1843        support. It adds about 90K to the size of the library. The meanings of the
1844        inbuilt class escapes such as \d and \s have NOT been changed.
1845    
1846    16. Updated pcredemo.c to include calls to free() to release the memory for the
1847        compiled pattern.
1848    
1849    17. The generated file chartables.c was being created in the source directory
1850        instead of in the building directory. This caused the build to fail if the
1851        source directory was different from the building directory, and was
1852        read-only.
1853    
1854    18. Added some sample Win commands from Mark Tetrode into the NON-UNIX-USE
1855        file. No doubt somebody will tell me if they don't make sense... Also added
1856        Dan Mooney's comments about building on OpenVMS.
1857    
1858    19. Added support for partial matching via the PCRE_PARTIAL option for
1859        pcre_exec() and the \P data escape in pcretest.
1860    
1861    20. Extended pcretest with 3 new pattern features:
1862    
1863        (i)   A pattern option of the form ">rest-of-line" causes pcretest to
1864              write the compiled pattern to the file whose name is "rest-of-line".
1865              This is a straight binary dump of the data, with the saved pointer to
1866              the character tables forced to be NULL. The study data, if any, is
1867              written too. After writing, pcretest reads a new pattern.
1868    
1869        (ii)  If, instead of a pattern, "<rest-of-line" is given, pcretest reads a
1870              compiled pattern from the given file. There must not be any
1871              occurrences of "<" in the file name (pretty unlikely); if there are,
1872              pcretest will instead treat the initial "<" as a pattern delimiter.
1873              After reading in the pattern, pcretest goes on to read data lines as
1874              usual.
1875    
1876        (iii) The F pattern option causes pcretest to flip the bytes in the 32-bit
1877              and 16-bit fields in a compiled pattern, to simulate a pattern that
1878              was compiled on a host of opposite endianness.
1879    
1880    21. The pcre-exec() function can now cope with patterns that were compiled on
1881        hosts of opposite endianness, with this restriction:
1882    
1883          As for any compiled expression that is saved and used later, the tables
1884          pointer field cannot be preserved; the extra_data field in the arguments
1885          to pcre_exec() should be used to pass in a tables address if a value
1886          other than the default internal tables were used at compile time.
1887    
1888    22. Calling pcre_exec() with a negative value of the "ovecsize" parameter is
1889        now diagnosed as an error. Previously, most of the time, a negative number
1890        would have been treated as zero, but if in addition "ovector" was passed as
1891        NULL, a crash could occur.
1892    
1893    23. Updated the files ltmain.sh, config.sub, config.guess, and aclocal.m4 with
1894        new versions from the libtool 1.5 distribution (the last one is a copy of
1895        a file called libtool.m4). This seems to have fixed the need to patch
1896        "configure" to support Darwin 1.3 (which I used to do). However, I still
1897        had to patch ltmain.sh to ensure that ${SED} is set (it isn't on my
1898        workstation).
1899    
1900    24. Changed the PCRE licence to be the more standard "BSD" licence.
1901    
1902    
1903    Version 4.5 01-Dec-03
1904    ---------------------
1905    
1906     1. There has been some re-arrangement of the code for the match() function so
1907        that it can be compiled in a version that does not call itself recursively.
1908        Instead, it keeps those local variables that need separate instances for
1909        each "recursion" in a frame on the heap, and gets/frees frames whenever it
1910        needs to "recurse". Keeping track of where control must go is done by means
1911        of setjmp/longjmp. The whole thing is implemented by a set of macros that
1912        hide most of the details from the main code, and operates only if
1913        NO_RECURSE is defined while compiling pcre.c. If PCRE is built using the
1914        "configure" mechanism, "--disable-stack-for-recursion" turns on this way of
1915        operating.
1916    
1917        To make it easier for callers to provide specially tailored get/free
1918        functions for this usage, two new functions, pcre_stack_malloc, and
1919        pcre_stack_free, are used. They are always called in strict stacking order,
1920        and the size of block requested is always the same.
1921    
1922        The PCRE_CONFIG_STACKRECURSE info parameter can be used to find out whether
1923        PCRE has been compiled to use the stack or the heap for recursion. The
1924        -C option of pcretest uses this to show which version is compiled.
1925    
1926        A new data escape \S, is added to pcretest; it causes the amounts of store
1927        obtained and freed by both kinds of malloc/free at match time to be added
1928        to the output.
1929    
1930     2. Changed the locale test to use "fr_FR" instead of "fr" because that's
1931        what's available on my current Linux desktop machine.
1932    
1933     3. When matching a UTF-8 string, the test for a valid string at the start has
1934        been extended. If start_offset is not zero, PCRE now checks that it points
1935        to a byte that is the start of a UTF-8 character. If not, it returns
1936        PCRE_ERROR_BADUTF8_OFFSET (-11). Note: the whole string is still checked;
1937        this is necessary because there may be backward assertions in the pattern.
1938        When matching the same subject several times, it may save resources to use
1939        PCRE_NO_UTF8_CHECK on all but the first call if the string is long.
1940    
1941     4. The code for checking the validity of UTF-8 strings has been tightened so
1942        that it rejects (a) strings containing 0xfe or 0xff bytes and (b) strings
1943        containing "overlong sequences".
1944    
1945     5. Fixed a bug (appearing twice) that I could not find any way of exploiting!
1946        I had written "if ((digitab[*p++] && chtab_digit) == 0)" where the "&&"
1947        should have been "&", but it just so happened that all the cases this let
1948        through by mistake were picked up later in the function.
1949    
1950     6. I had used a variable called "isblank" - this is a C99 function, causing
1951        some compilers to warn. To avoid this, I renamed it (as "blankclass").
1952    
1953     7. Cosmetic: (a) only output another newline at the end of pcretest if it is
1954        prompting; (b) run "./pcretest /dev/null" at the start of the test script
1955        so the version is shown; (c) stop "make test" echoing "./RunTest".
1956    
1957     8. Added patches from David Burgess to enable PCRE to run on EBCDIC systems.
1958    
1959     9. The prototype for memmove() for systems that don't have it was using
1960        size_t, but the inclusion of the header that defines size_t was later. I've
1961        moved the #includes for the C headers earlier to avoid this.
1962    
1963    10. Added some adjustments to the code to make it easier to compiler on certain
1964        special systems:
1965    
1966          (a) Some "const" qualifiers were missing.
1967          (b) Added the macro EXPORT before all exported functions; by default this
1968              is defined to be empty.
1969          (c) Changed the dftables auxiliary program (that builds chartables.c) so
1970              that it reads its output file name as an argument instead of writing
1971              to the standard output and assuming this can be redirected.
1972    
1973    11. In UTF-8 mode, if a recursive reference (e.g. (?1)) followed a character
1974        class containing characters with values greater than 255, PCRE compilation
1975        went into a loop.
1976    
1977    12. A recursive reference to a subpattern that was within another subpattern
1978        that had a minimum quantifier of zero caused PCRE to crash. For example,
1979        (x(y(?2))z)? provoked this bug with a subject that got as far as the
1980        recursion. If the recursively-called subpattern itself had a zero repeat,
1981        that was OK.
1982    
1983    13. In pcretest, the buffer for reading a data line was set at 30K, but the
1984        buffer into which it was copied (for escape processing) was still set at
1985        1024, so long lines caused crashes.
1986    
1987    14. A pattern such as /[ab]{1,3}+/ failed to compile, giving the error
1988        "internal error: code overflow...". This applied to any character class
1989        that was followed by a possessive quantifier.
1990    
1991    15. Modified the Makefile to add libpcre.la as a prerequisite for
1992        libpcreposix.la because I was told this is needed for a parallel build to
1993        work.
1994    
1995    16. If a pattern that contained .* following optional items at the start was
1996        studied, the wrong optimizing data was generated, leading to matching
1997        errors. For example, studying /[ab]*.*c/ concluded, erroneously, that any
1998        matching string must start with a or b or c. The correct conclusion for
1999        this pattern is that a match can start with any character.
2000    
2001    
2002    Version 4.4 13-Aug-03
2003    ---------------------
2004    
2005     1. In UTF-8 mode, a character class containing characters with values between
2006        127 and 255 was not handled correctly if the compiled pattern was studied.
2007        In fixing this, I have also improved the studying algorithm for such
2008        classes (slightly).
2009    
2010     2. Three internal functions had redundant arguments passed to them. Removal
2011        might give a very teeny performance improvement.
2012    
2013     3. Documentation bug: the value of the capture_top field in a callout is *one
2014        more than* the number of the hightest numbered captured substring.
2015    
2016     4. The Makefile linked pcretest and pcregrep with -lpcre, which could result
2017        in incorrectly linking with a previously installed version. They now link
2018        explicitly with libpcre.la.
2019    
2020     5. configure.in no longer needs to recognize Cygwin specially.
2021    
2022     6. A problem in pcre.in for Windows platforms is fixed.
2023    
2024     7. If a pattern was successfully studied, and the -d (or /D) flag was given to
2025        pcretest, it used to include the size of the study block as part of its
2026        output. Unfortunately, the structure contains a field that has a different
2027        size on different hardware architectures. This meant that the tests that
2028        showed this size failed. As the block is currently always of a fixed size,
2029        this information isn't actually particularly useful in pcretest output, so
2030        I have just removed it.
2031    
2032     8. Three pre-processor statements accidentally did not start in column 1.
2033        Sadly, there are *still* compilers around that complain, even though
2034        standard C has not required this for well over a decade. Sigh.
2035    
2036     9. In pcretest, the code for checking callouts passed small integers in the
2037        callout_data field, which is a void * field. However, some picky compilers
2038        complained about the casts involved for this on 64-bit systems. Now
2039        pcretest passes the address of the small integer instead, which should get
2040        rid of the warnings.
2041    
2042    10. By default, when in UTF-8 mode, PCRE now checks for valid UTF-8 strings at
2043        both compile and run time, and gives an error if an invalid UTF-8 sequence
2044        is found. There is a option for disabling this check in cases where the
2045        string is known to be correct and/or the maximum performance is wanted.
2046    
2047    11. In response to a bug report, I changed one line in Makefile.in from
2048    
2049            -Wl,--out-implib,.libs/lib@WIN_PREFIX@pcreposix.dll.a \
2050        to
2051            -Wl,--out-implib,.libs/@WIN_PREFIX@libpcreposix.dll.a \
2052    
2053        to look similar to other lines, but I have no way of telling whether this
2054        is the right thing to do, as I do not use Windows. No doubt I'll get told
2055        if it's wrong...
2056    
2057    
2058    Version 4.3 21-May-03
2059    ---------------------
2060    
2061    1. Two instances of @WIN_PREFIX@ omitted from the Windows targets in the
2062       Makefile.
2063    
2064    2. Some refactoring to improve the quality of the code:
2065    
2066       (i)   The utf8_table... variables are now declared "const".
2067    
2068       (ii)  The code for \cx, which used the "case flipping" table to upper case
2069             lower case letters, now just substracts 32. This is ASCII-specific,
2070             but the whole concept of \cx is ASCII-specific, so it seems
2071             reasonable.
2072    
2073       (iii) PCRE was using its character types table to recognize decimal and
2074             hexadecimal digits in the pattern. This is silly, because it handles
2075             only 0-9, a-f, and A-F, but the character types table is locale-
2076             specific, which means strange things might happen. A private
2077             table is now used for this - though it costs 256 bytes, a table is
2078             much faster than multiple explicit tests. Of course, the standard
2079             character types table is still used for matching digits in subject
2080             strings against \d.
2081    
2082       (iv)  Strictly, the identifier ESC_t is reserved by POSIX (all identifiers
2083             ending in _t are). So I've renamed it as ESC_tee.
2084    
2085    3. The first argument for regexec() in the POSIX wrapper should have been
2086       defined as "const".
2087    
2088    4. Changed pcretest to use malloc() for its buffers so that they can be
2089       Electric Fenced for debugging.
2090    
2091    5. There were several places in the code where, in UTF-8 mode, PCRE would try
2092       to read one or more bytes before the start of the subject string. Often this
2093       had no effect on PCRE's behaviour, but in some circumstances it could
2094       provoke a segmentation fault.
2095    
2096    6. A lookbehind at the start of a pattern in UTF-8 mode could also cause PCRE
2097       to try to read one or more bytes before the start of the subject string.
2098    
2099    7. A lookbehind in a pattern matched in non-UTF-8 mode on a PCRE compiled with
2100       UTF-8 support could misbehave in various ways if the subject string
2101       contained bytes with the 0x80 bit set and the 0x40 bit unset in a lookbehind
2102       area. (PCRE was not checking for the UTF-8 mode flag, and trying to move
2103       back over UTF-8 characters.)
2104    
2105    
2106    Version 4.2 14-Apr-03
2107    ---------------------
2108    
2109    1. Typo "#if SUPPORT_UTF8" instead of "#ifdef SUPPORT_UTF8" fixed.
2110    
2111    2. Changes to the building process, supplied by Ronald Landheer-Cieslak
2112         [ON_WINDOWS]: new variable, "#" on non-Windows platforms
2113         [NOT_ON_WINDOWS]: new variable, "#" on Windows platforms
2114         [WIN_PREFIX]: new variable, "cyg" for Cygwin
2115         * Makefile.in: use autoconf substitution for OBJEXT, EXEEXT, BUILD_OBJEXT
2116           and BUILD_EXEEXT
2117         Note: automatic setting of the BUILD variables is not yet working
2118         set CPPFLAGS and BUILD_CPPFLAGS (but don't use yet) - should be used at
2119           compile-time but not at link-time
2120         [LINK]: use for linking executables only
2121         make different versions for Windows and non-Windows
2122         [LINKLIB]: new variable, copy of UNIX-style LINK, used for linking
2123           libraries
2124         [LINK_FOR_BUILD]: new variable
2125         [OBJEXT]: use throughout
2126         [EXEEXT]: use throughout
2127         <winshared>: new target
2128         <wininstall>: new target
2129         <dftables.o>: use native compiler
2130         <dftables>: use native linker
2131         <install>: handle Windows platform correctly
2132         <clean>: ditto
2133         <check>: ditto
2134         copy DLL to top builddir before testing
2135    
2136       As part of these changes, -no-undefined was removed again. This was reported
2137       to give trouble on HP-UX 11.0, so getting rid of it seems like a good idea
2138       in any case.
2139    
2140    3. Some tidies to get rid of compiler warnings:
2141    
2142       . In the match_data structure, match_limit was an unsigned long int, whereas
2143         match_call_count was an int. I've made them both unsigned long ints.
2144    
2145       . In pcretest the fact that a const uschar * doesn't automatically cast to
2146         a void * provoked a warning.
2147    
2148       . Turning on some more compiler warnings threw up some "shadow" variables
2149         and a few more missing casts.
2150    
2151    4. If PCRE was complied with UTF-8 support, but called without the PCRE_UTF8
2152       option, a class that contained a single character with a value between 128
2153       and 255 (e.g. /[\xFF]/) caused PCRE to crash.
2154    
2155    5. If PCRE was compiled with UTF-8 support, but called without the PCRE_UTF8
2156       option, a class that contained several characters, but with at least one
2157       whose value was between 128 and 255 caused PCRE to crash.
2158    
2159    
2160    Version 4.1 12-Mar-03
2161    ---------------------
2162    
2163    1. Compiling with gcc -pedantic found a couple of places where casts were
2164    needed, and a string in dftables.c that was longer than standard compilers are
2165    required to support.
2166    
2167    2. Compiling with Sun's compiler found a few more places where the code could
2168    be tidied up in order to avoid warnings.
2169    
2170    3. The variables for cross-compiling were called HOST_CC and HOST_CFLAGS; the
2171    first of these names is deprecated in the latest Autoconf in favour of the name
2172    CC_FOR_BUILD, because "host" is typically used to mean the system on which the
2173    compiled code will be run. I can't find a reference for HOST_CFLAGS, but by
2174    analogy I have changed it to CFLAGS_FOR_BUILD.
2175    
2176    4. Added -no-undefined to the linking command in the Makefile, because this is
2177    apparently helpful for Windows. To make it work, also added "-L. -lpcre" to the
2178    linking step for the pcreposix library.
2179    
2180    5. PCRE was failing to diagnose the case of two named groups with the same
2181    name.
2182    
2183    6. A problem with one of PCRE's optimizations was discovered. PCRE remembers a
2184    literal character that is needed in the subject for a match, and scans along to
2185    ensure that it is present before embarking on the full matching process. This
2186    saves time in cases of nested unlimited repeats that are never going to match.
2187    Problem: the scan can take a lot of time if the subject is very long (e.g.
2188    megabytes), thus penalizing straightforward matches. It is now done only if the
2189    amount of subject to be scanned is less than 1000 bytes.
2190    
2191    7. A lesser problem with the same optimization is that it was recording the
2192    first character of an anchored pattern as "needed", thus provoking a search
2193    right along the subject, even when the first match of the pattern was going to
2194    fail. The "needed" character is now not set for anchored patterns, unless it
2195    follows something in the pattern that is of non-fixed length. Thus, it still
2196    fulfils its original purpose of finding quick non-matches in cases of nested
2197    unlimited repeats, but isn't used for simple anchored patterns such as /^abc/.
2198    
2199    
2200    Version 4.0 17-Feb-03
2201    ---------------------
2202    
2203    1. If a comment in an extended regex that started immediately after a meta-item
2204    extended to the end of string, PCRE compiled incorrect data. This could lead to
2205    all kinds of weird effects. Example: /#/ was bad; /()#/ was bad; /a#/ was not.
2206    
2207    2. Moved to autoconf 2.53 and libtool 1.4.2.
2208    
2209    3. Perl 5.8 no longer needs "use utf8" for doing UTF-8 things. Consequently,
2210    the special perltest8 script is no longer needed - all the tests can be run
2211    from a single perltest script.
2212    
2213    4. From 5.004, Perl has not included the VT character (0x0b) in the set defined
2214    by \s. It has now been removed in PCRE. This means it isn't recognized as
2215    whitespace in /x regexes too, which is the same as Perl. Note that the POSIX
2216    class [:space:] *does* include VT, thereby creating a mess.
2217    
2218    5. Added the class [:blank:] (a GNU extension from Perl 5.8) to match only
2219    space and tab.
2220    
2221    6. Perl 5.005 was a long time ago. It's time to amalgamate the tests that use
2222    its new features into the main test script, reducing the number of scripts.
2223    
2224    7. Perl 5.8 has changed the meaning of patterns like /a(?i)b/. Earlier versions
2225    were backward compatible, and made the (?i) apply to the whole pattern, as if
2226    /i were given. Now it behaves more logically, and applies the option setting
2227    only to what follows. PCRE has been changed to follow suit. However, if it
2228    finds options settings right at the start of the pattern, it extracts them into
2229    the global options, as before. Thus, they show up in the info data.
2230    
2231    8. Added support for the \Q...\E escape sequence. Characters in between are
2232    treated as literals. This is slightly different from Perl in that $ and @ are
2233    also handled as literals inside the quotes. In Perl, they will cause variable
2234    interpolation. Note the following examples:
2235    
2236        Pattern            PCRE matches      Perl matches
2237    
2238        \Qabc$xyz\E        abc$xyz           abc followed by the contents of $xyz
2239        \Qabc\$xyz\E       abc\$xyz          abc\$xyz
2240        \Qabc\E\$\Qxyz\E   abc$xyz           abc$xyz
2241    
2242    For compatibility with Perl, \Q...\E sequences are recognized inside character
2243    classes as well as outside them.
2244    
2245    9. Re-organized 3 code statements in pcretest to avoid "overflow in
2246    floating-point constant arithmetic" warnings from a Microsoft compiler. Added a
2247    (size_t) cast to one statement in pcretest and one in pcreposix to avoid
2248    signed/unsigned warnings.
2249    
2250    10. SunOS4 doesn't have strtoul(). This was used only for unpicking the -o
2251    option for pcretest, so I've replaced it by a simple function that does just
2252    that job.
2253    
2254    11. pcregrep was ending with code 0 instead of 2 for the commands "pcregrep" or
2255    "pcregrep -".
2256    
2257    12. Added "possessive quantifiers" ?+, *+, ++, and {,}+ which come from Sun's
2258    Java package. This provides some syntactic sugar for simple cases of what my
2259    documentation calls "once-only subpatterns". A pattern such as x*+ is the same
2260    as (?>x*). In other words, if what is inside (?>...) is just a single repeated
2261    item, you can use this simplified notation. Note that only makes sense with
2262    greedy quantifiers. Consequently, the use of the possessive quantifier forces
2263    greediness, whatever the setting of the PCRE_UNGREEDY option.
2264    
2265    13. A change of greediness default within a pattern was not taking effect at
2266    the current level for patterns like /(b+(?U)a+)/. It did apply to parenthesized
2267    subpatterns that followed. Patterns like /b+(?U)a+/ worked because the option
2268    was abstracted outside.
2269    
2270    14. PCRE now supports the \G assertion. It is true when the current matching
2271    position is at the start point of the match. This differs from \A when the
2272    starting offset is non-zero. Used with the /g option of pcretest (or similar
2273    code), it works in the same way as it does for Perl's /g option. If all
2274    alternatives of a regex begin with \G, the expression is anchored to the start
2275    match position, and the "anchored" flag is set in the compiled expression.
2276    
2277    15. Some bugs concerning the handling of certain option changes within patterns
2278    have been fixed. These applied to options other than (?ims). For example,
2279    "a(?x: b c )d" did not match "XabcdY" but did match "Xa b c dY". It should have
2280    been the other way round. Some of this was related to change 7 above.
2281    
2282    16. PCRE now gives errors for /[.x.]/ and /[=x=]/ as unsupported POSIX
2283    features, as Perl does. Previously, PCRE gave the warnings only for /[[.x.]]/
2284    and /[[=x=]]/. PCRE now also gives an error for /[:name:]/ because it supports
2285    POSIX classes only within a class (e.g. /[[:alpha:]]/).
2286    
2287    17. Added support for Perl's \C escape. This matches one byte, even in UTF8
2288    mode. Unlike ".", it always matches newline, whatever the setting of
2289    PCRE_DOTALL. However, PCRE does not permit \C to appear in lookbehind
2290    assertions. Perl allows it, but it doesn't (in general) work because it can't
2291    calculate the length of the lookbehind. At least, that's the case for Perl
2292    5.8.0 - I've been told they are going to document that it doesn't work in
2293    future.
2294    
2295    18. Added an error diagnosis for escapes that PCRE does not support: these are
2296    \L, \l, \N, \P, \p, \U, \u, and \X.
2297    
2298    19. Although correctly diagnosing a missing ']' in a character class, PCRE was
2299    reading past the end of the pattern in cases such as /[abcd/.
2300    
2301    20. PCRE was getting more memory than necessary for patterns with classes that
2302    contained both POSIX named classes and other characters, e.g. /[[:space:]abc/.
2303    
2304    21. Added some code, conditional on #ifdef VPCOMPAT, to make life easier for
2305    compiling PCRE for use with Virtual Pascal.
2306    
2307    22. Small fix to the Makefile to make it work properly if the build is done
2308    outside the source tree.
2309    
2310    23. Added a new extension: a condition to go with recursion. If a conditional
2311    subpattern starts with (?(R) the "true" branch is used if recursion has
2312    happened, whereas the "false" branch is used only at the top level.
2313    
2314    24. When there was a very long string of literal characters (over 255 bytes
2315    without UTF support, over 250 bytes with UTF support), the computation of how
2316    much memory was required could be incorrect, leading to segfaults or other
2317    strange effects.
2318    
2319    25. PCRE was incorrectly assuming anchoring (either to start of subject or to
2320    start of line for a non-DOTALL pattern) when a pattern started with (.*) and
2321    there was a subsequent back reference to those brackets. This meant that, for
2322    example, /(.*)\d+\1/ failed to match "abc123bc". Unfortunately, it isn't
2323    possible to check for precisely this case. All we can do is abandon the
2324    optimization if .* occurs inside capturing brackets when there are any back
2325    references whatsoever. (See below for a better fix that came later.)
2326    
2327    26. The handling of the optimization for finding the first character of a
2328    non-anchored pattern, and for finding a character that is required later in the
2329    match were failing in some cases. This didn't break the matching; it just
2330    failed to optimize when it could. The way this is done has been re-implemented.
2331    
2332    27. Fixed typo in error message for invalid (?R item (it said "(?p").
2333    
2334    28. Added a new feature that provides some of the functionality that Perl
2335    provides with (?{...}). The facility is termed a "callout". The way it is done
2336    in PCRE is for the caller to provide an optional function, by setting
2337    pcre_callout to its entry point. Like pcre_malloc and pcre_free, this is a
2338    global variable. By default it is unset, which disables all calling out. To get
2339    the function called, the regex must include (?C) at appropriate points. This
2340    is, in fact, equivalent to (?C0), and any number <= 255 may be given with (?C).
2341    This provides a means of identifying different callout points. When PCRE
2342    reaches such a point in the regex, if pcre_callout has been set, the external
2343    function is called. It is provided with data in a structure called
2344    pcre_callout_block, which is defined in pcre.h. If the function returns 0,
2345    matching continues; if it returns a non-zero value, the match at the current
2346    point fails. However, backtracking will occur if possible. [This was changed
2347    later and other features added - see item 49 below.]
2348    
2349    29. pcretest is upgraded to test the callout functionality. It provides a
2350    callout function that displays information. By default, it shows the start of
2351    the match and the current position in the text. There are some new data escapes
2352    to vary what happens:
2353    
2354        \C+         in addition, show current contents of captured substrings
2355        \C-         do not supply a callout function
2356        \C!n        return 1 when callout number n is reached
2357        \C!n!m      return 1 when callout number n is reached for the mth time
2358    
2359    30. If pcregrep was called with the -l option and just a single file name, it
2360    output "<stdin>" if a match was found, instead of the file name.
2361    
2362    31. Improve the efficiency of the POSIX API to PCRE. If the number of capturing
2363    slots is less than POSIX_MALLOC_THRESHOLD, use a block on the stack to pass to
2364    pcre_exec(). This saves a malloc/free per call. The default value of
2365    POSIX_MALLOC_THRESHOLD is 10; it can be changed by --with-posix-malloc-threshold
2366    when configuring.
2367    
2368    32. The default maximum size of a compiled pattern is 64K. There have been a
2369    few cases of people hitting this limit. The code now uses macros to handle the
2370    storing of links as offsets within the compiled pattern. It defaults to 2-byte
2371    links, but this can be changed to 3 or 4 bytes by --with-link-size when
2372    configuring. Tests 2 and 5 work only with 2-byte links because they output
2373    debugging information about compiled patterns.
2374    
2375    33. Internal code re-arrangements:
2376    
2377    (a) Moved the debugging function for printing out a compiled regex into
2378        its own source file (printint.c) and used #include to pull it into
2379        pcretest.c and, when DEBUG is defined, into pcre.c, instead of having two
2380        separate copies.
2381    
2382    (b) Defined the list of op-code names for debugging as a macro in
2383        internal.h so that it is next to the definition of the opcodes.
2384    
2385    (c) Defined a table of op-code lengths for simpler skipping along compiled
2386        code. This is again a macro in internal.h so that it is next to the
2387        definition of the opcodes.
2388    
2389    34. Added support for recursive calls to individual subpatterns, along the
2390    lines of Robin Houston's patch (but implemented somewhat differently).
2391    
2392    35. Further mods to the Makefile to help Win32. Also, added code to pcregrep to
2393    allow it to read and process whole directories in Win32. This code was
2394    contributed by Lionel Fourquaux; it has not been tested by me.
2395    
2396    36. Added support for named subpatterns. The Python syntax (?P<name>...) is
2397    used to name a group. Names consist of alphanumerics and underscores, and must
2398    be unique. Back references use the syntax (?P=name) and recursive calls use
2399    (?P>name) which is a PCRE extension to the Python extension. Groups still have
2400    numbers. The function pcre_fullinfo() can be used after compilation to extract
2401    a name/number map. There are three relevant calls:
2402    
2403      PCRE_INFO_NAMEENTRYSIZE        yields the size of each entry in the map
2404      PCRE_INFO_NAMECOUNT            yields the number of entries
2405      PCRE_INFO_NAMETABLE            yields a pointer to the map.
2406    
2407    The map is a vector of fixed-size entries. The size of each entry depends on
2408    the length of the longest name used. The first two bytes of each entry are the
2409    group number, most significant byte first. There follows the corresponding
2410    name, zero terminated. The names are in alphabetical order.
2411    
2412    37. Make the maximum literal string in the compiled code 250 for the non-UTF-8
2413    case instead of 255. Making it the same both with and without UTF-8 support
2414    means that the same test output works with both.
2415    
2416    38. There was a case of malloc(0) in the POSIX testing code in pcretest. Avoid
2417    calling malloc() with a zero argument.
2418    
2419    39. Change 25 above had to resort to a heavy-handed test for the .* anchoring
2420    optimization. I've improved things by keeping a bitmap of backreferences with
2421    numbers 1-31 so that if .* occurs inside capturing brackets that are not in
2422    fact referenced, the optimization can be applied. It is unlikely that a
2423    relevant occurrence of .* (i.e. one which might indicate anchoring or forcing
2424    the match to follow \n) will appear inside brackets with a number greater than
2425    31, but if it does, any back reference > 31 suppresses the optimization.
2426    
2427    40. Added a new compile-time option PCRE_NO_AUTO_CAPTURE. This has the effect
2428    of disabling numbered capturing parentheses. Any opening parenthesis that is
2429    not followed by ? behaves as if it were followed by ?: but named parentheses
2430    can still be used for capturing (and they will acquire numbers in the usual
2431    way).
2432    
2433    41. Redesigned the return codes from the match() function into yes/no/error so
2434    that errors can be passed back from deep inside the nested calls. A malloc
2435    failure while inside a recursive subpattern call now causes the
2436    PCRE_ERROR_NOMEMORY return instead of quietly going wrong.
2437    
2438    42. It is now possible to set a limit on the number of times the match()
2439    function is called in a call to pcre_exec(). This facility makes it possible to
2440    limit the amount of recursion and backtracking, though not in a directly
2441    obvious way, because the match() function is used in a number of different
2442    circumstances. The count starts from zero for each position in the subject
2443    string (for non-anchored patterns). The default limit is, for compatibility, a
2444    large number, namely 10 000 000. You can change this in two ways:
2445    
2446    (a) When configuring PCRE before making, you can use --with-match-limit=n
2447        to set a default value for the compiled library.
2448    
2449    (b) For each call to pcre_exec(), you can pass a pcre_extra block in which
2450        a different value is set. See 45 below.
2451    
2452    If the limit is exceeded, pcre_exec() returns PCRE_ERROR_MATCHLIMIT.
2453    
2454    43. Added a new function pcre_config(int, void *) to enable run-time extraction
2455    of things that can be changed at compile time. The first argument specifies
2456    what is wanted and the second points to where the information is to be placed.
2457    The current list of available information is:
2458    
2459      PCRE_CONFIG_UTF8
2460    
2461    The output is an integer that is set to one if UTF-8 support is available;
2462    otherwise it is set to zero.
2463    
2464      PCRE_CONFIG_NEWLINE
2465    
2466    The output is an integer that it set to the value of the code that is used for
2467    newline. It is either LF (10) or CR (13).
2468    
2469      PCRE_CONFIG_LINK_SIZE
2470    
2471    The output is an integer that contains the number of bytes used for internal
2472    linkage in compiled expressions. The value is 2, 3, or 4. See item 32 above.
2473    
2474      PCRE_CONFIG_POSIX_MALLOC_THRESHOLD
2475    
2476    The output is an integer that contains the threshold above which the POSIX
2477    interface uses malloc() for output vectors. See item 31 above.
2478    
2479      PCRE_CONFIG_MATCH_LIMIT
2480    
2481    The output is an unsigned integer that contains the default limit of the number
2482    of match() calls in a pcre_exec() execution. See 42 above.
2483    
2484    44. pcretest has been upgraded by the addition of the -C option. This causes it
2485    to extract all the available output from the new pcre_config() function, and to
2486    output it. The program then exits immediately.
2487    
2488    45. A need has arisen to pass over additional data with calls to pcre_exec() in
2489    order to support additional features. One way would have been to define
2490    pcre_exec2() (for example) with extra arguments, but this would not have been
2491    extensible, and would also have required all calls to the original function to
2492    be mapped to the new one. Instead, I have chosen to extend the mechanism that
2493    is used for passing in "extra" data from pcre_study().
2494    
2495    The pcre_extra structure is now exposed and defined in pcre.h. It currently
2496    contains the following fields:
2497    
2498      flags         a bitmap indicating which of the following fields are set
2499      study_data    opaque data from pcre_study()
2500      match_limit   a way of specifying a limit on match() calls for a specific
2501                      call to pcre_exec()
2502      callout_data  data for callouts (see 49 below)
2503    
2504    The flag bits are also defined in pcre.h, and are
2505    
2506      PCRE_EXTRA_STUDY_DATA
2507      PCRE_EXTRA_MATCH_LIMIT
2508      PCRE_EXTRA_CALLOUT_DATA
2509    
2510    The pcre_study() function now returns one of these new pcre_extra blocks, with
2511    the actual study data pointed to by the study_data field, and the
2512    PCRE_EXTRA_STUDY_DATA flag set. This can be passed directly to pcre_exec() as
2513    before. That is, this change is entirely upwards-compatible and requires no
2514    change to existing code.
2515    
2516    If you want to pass in additional data to pcre_exec(), you can either place it
2517    in a pcre_extra block provided by pcre_study(), or create your own pcre_extra
2518    block.
2519    
2520    46. pcretest has been extended to test the PCRE_EXTRA_MATCH_LIMIT feature. If a
2521    data string contains the escape sequence \M, pcretest calls pcre_exec() several
2522    times with different match limits, until it finds the minimum value needed for
2523    pcre_exec() to complete. The value is then output. This can be instructive; for
2524    most simple matches the number is quite small, but for pathological cases it
2525    gets very large very quickly.
2526    
2527    47. There's a new option for pcre_fullinfo() called PCRE_INFO_STUDYSIZE. It
2528    returns the size of the data block pointed to by the study_data field in a
2529    pcre_extra block, that is, the value that was passed as the argument to
2530    pcre_malloc() when PCRE was getting memory in which to place the information
2531    created by pcre_study(). The fourth argument should point to a size_t variable.
2532    pcretest has been extended so that this information is shown after a successful
2533    pcre_study() call when information about the compiled regex is being displayed.
2534    
2535    48. Cosmetic change to Makefile: there's no need to have / after $(DESTDIR)
2536    because what follows is always an absolute path. (Later: it turns out that this
2537    is more than cosmetic for MinGW, because it doesn't like empty path
2538    components.)
2539    
2540    49. Some changes have been made to the callout feature (see 28 above):
2541    
2542    (i)  A callout function now has three choices for what it returns:
2543    
2544           0  =>  success, carry on matching
2545         > 0  =>  failure at this point, but backtrack if possible
2546         < 0  =>  serious error, return this value from pcre_exec()
2547    
2548         Negative values should normally be chosen from the set of PCRE_ERROR_xxx
2549         values. In particular, returning PCRE_ERROR_NOMATCH forces a standard
2550         "match failed" error. The error number PCRE_ERROR_CALLOUT is reserved for
2551         use by callout functions. It will never be used by PCRE itself.
2552    
2553    (ii) The pcre_extra structure (see 45 above) has a void * field called
2554         callout_data, with corresponding flag bit PCRE_EXTRA_CALLOUT_DATA. The
2555         pcre_callout_block structure has a field of the same name. The contents of
2556         the field passed in the pcre_extra structure are passed to the callout
2557         function in the corresponding field in the callout block. This makes it
2558         easier to use the same callout-containing regex from multiple threads. For
2559         testing, the pcretest program has a new data escape
2560    
2561           \C*n        pass the number n (may be negative) as callout_data
2562    
2563         If the callout function in pcretest receives a non-zero value as
2564         callout_data, it returns that value.
2565    
2566    50. Makefile wasn't handling CFLAGS properly when compiling dftables. Also,
2567    there were some redundant $(CFLAGS) in commands that are now specified as
2568    $(LINK), which already includes $(CFLAGS).
2569    
2570    51. Extensions to UTF-8 support are listed below. These all apply when (a) PCRE
2571    has been compiled with UTF-8 support *and* pcre_compile() has been compiled
2572    with the PCRE_UTF8 flag. Patterns that are compiled without that flag assume
2573    one-byte characters throughout. Note that case-insensitive matching applies
2574    only to characters whose values are less than 256. PCRE doesn't support the
2575    notion of cases for higher-valued characters.
2576    
2577    (i)   A character class whose characters are all within 0-255 is handled as
2578          a bit map, and the map is inverted for negative classes. Previously, a
2579          character > 255 always failed to match such a class; however it should
2580          match if the class was a negative one (e.g. [^ab]). This has been fixed.
2581    
2582    (ii)  A negated character class with a single character < 255 is coded as
2583          "not this character" (OP_NOT). This wasn't working properly when the test
2584          character was multibyte, either singly or repeated.
2585    
2586    (iii) Repeats of multibyte characters are now handled correctly in UTF-8
2587          mode, for example: \x{100}{2,3}.
2588    
2589    (iv)  The character escapes \b, \B, \d, \D, \s, \S, \w, and \W (either
2590          singly or repeated) now correctly test multibyte characters. However,
2591          PCRE doesn't recognize any characters with values greater than 255 as
2592          digits, spaces, or word characters. Such characters always match \D, \S,
2593          and \W, and never match \d, \s, or \w.
2594    
2595    (v)   Classes may now contain characters and character ranges with values
2596          greater than 255. For example: [ab\x{100}-\x{400}].
2597    
2598    (vi)  pcregrep now has a --utf-8 option (synonym -u) which makes it call
2599          PCRE in UTF-8 mode.
2600    
2601    52. The info request value PCRE_INFO_FIRSTCHAR has been renamed
2602    PCRE_INFO_FIRSTBYTE because it is a byte value. However, the old name is
2603    retained for backwards compatibility. (Note that LASTLITERAL is also a byte
2604    value.)
2605    
2606    53. The single man page has become too large. I have therefore split it up into
2607    a number of separate man pages. These also give rise to individual HTML pages;
2608    these are now put in a separate directory, and there is an index.html page that
2609    lists them all. Some hyperlinking between the pages has been installed.
2610    
2611    54. Added convenience functions for handling named capturing parentheses.
2612    
2613    55. Unknown escapes inside character classes (e.g. [\M]) and escapes that
2614    aren't interpreted therein (e.g. [\C]) are literals in Perl. This is now also
2615    true in PCRE, except when the PCRE_EXTENDED option is set, in which case they
2616    are faulted.
2617    
2618    56. Introduced HOST_CC and HOST_CFLAGS which can be set in the environment when
2619    calling configure. These values are used when compiling the dftables.c program
2620    which is run to generate the source of the default character tables. They
2621    default to the values of CC and CFLAGS. If you are cross-compiling PCRE,
2622    you will need to set these values.
2623    
2624    57. Updated the building process for Windows DLL, as provided by Fred Cox.
2625    
2626    
2627    Version 3.9 02-Jan-02
2628    ---------------------
2629    
2630    1. A bit of extraneous text had somehow crept into the pcregrep documentation.
2631    
2632    2. If --disable-static was given, the building process failed when trying to
2633    build pcretest and pcregrep. (For some reason it was using libtool to compile
2634    them, which is not right, as they aren't part of the library.)
2635    
2636    
2637    Version 3.8 18-Dec-01
2638    ---------------------
2639    
2640    1. The experimental UTF-8 code was completely screwed up. It was packing the
2641    bytes in the wrong order. How dumb can you get?
2642    
2643    
2644    Version 3.7 29-Oct-01
2645    ---------------------
2646    
2647    1. In updating pcretest to check change 1 of version 3.6, I screwed up.
2648    This caused pcretest, when used on the test data, to segfault. Unfortunately,
2649    this didn't happen under Solaris 8, where I normally test things.
2650    
2651    2. The Makefile had to be changed to make it work on BSD systems, where 'make'
2652    doesn't seem to recognize that ./xxx and xxx are the same file. (This entry
2653    isn't in ChangeLog distributed with 3.7 because I forgot when I hastily made
2654    this fix an hour or so after the initial 3.7 release.)
2655    
2656    
2657    Version 3.6 23-Oct-01
2658    ---------------------
2659    
2660    1. Crashed with /(sens|respons)e and \1ibility/ and "sense and sensibility" if
2661    offsets passed as NULL with zero offset count.
2662    
2663    2. The config.guess and config.sub files had not been updated when I moved to
2664    the latest autoconf.
2665    
2666    
2667    Version 3.5 15-Aug-01
2668    ---------------------
2669    
2670    1. Added some missing #if !defined NOPOSIX conditionals in pcretest.c that
2671    had been forgotten.
2672    
2673    2. By using declared but undefined structures, we can avoid using "void"
2674    definitions in pcre.h while keeping the internal definitions of the structures
2675    private.
2676    
2677    3. The distribution is now built using autoconf 2.50 and libtool 1.4. From a
2678    user point of view, this means that both static and shared libraries are built
2679    by default, but this can be individually controlled. More of the work of
2680    handling this static/shared cases is now inside libtool instead of PCRE's make
2681    file.
2682    
2683    4. The pcretest utility is now installed along with pcregrep because it is
2684    useful for users (to test regexs) and by doing this, it automatically gets
2685    relinked by libtool. The documentation has been turned into a man page, so
2686    there are now .1, .txt, and .html versions in /doc.
2687    
2688    5. Upgrades to pcregrep:
2689       (i)   Added long-form option names like gnu grep.
2690       (ii)  Added --help to list all options with an explanatory phrase.
2691       (iii) Added -r, --recursive to recurse into sub-directories.
2692       (iv)  Added -f, --file to read patterns from a file.
2693    
2694    6. pcre_exec() was referring to its "code" argument before testing that
2695    argument for NULL (and giving an error if it was NULL).
2696    
2697    7. Upgraded Makefile.in to allow for compiling in a different directory from
2698    the source directory.
2699    
2700    8. Tiny buglet in pcretest: when pcre_fullinfo() was called to retrieve the
2701    options bits, the pointer it was passed was to an int instead of to an unsigned
2702    long int. This mattered only on 64-bit systems.
2703    
2704    9. Fixed typo (3.4/1) in pcre.h again. Sigh. I had changed pcre.h (which is
2705    generated) instead of pcre.in, which it its source. Also made the same change
2706    in several of the .c files.
2707    
2708    10. A new release of gcc defines printf() as a macro, which broke pcretest
2709    because it had an ifdef in the middle of a string argument for printf(). Fixed
2710    by using separate calls to printf().
2711    
2712    11. Added --enable-newline-is-cr and --enable-newline-is-lf to the configure
2713    script, to force use of CR or LF instead of \n in the source. On non-Unix
2714    systems, the value can be set in config.h.
2715    
2716    12. The limit of 200 on non-capturing parentheses is a _nesting_ limit, not an
2717    absolute limit. Changed the text of the error message to make this clear, and
2718    likewise updated the man page.
2719    
2720    13. The limit of 99 on the number of capturing subpatterns has been removed.
2721    The new limit is 65535, which I hope will not be a "real" limit.
2722    
2723    
2724    Version 3.4 22-Aug-00
2725    ---------------------
2726    
2727    1. Fixed typo in pcre.h: unsigned const char * changed to const unsigned char *.
2728    
2729    2. Diagnose condition (?(0) as an error instead of crashing on matching.
2730    
2731    
2732    Version 3.3 01-Aug-00
2733    ---------------------
2734    
2735    1. If an octal character was given, but the value was greater than \377, it
2736    was not getting masked to the least significant bits, as documented. This could
2737    lead to crashes in some systems.
2738    
2739    2. Perl 5.6 (if not earlier versions) accepts classes like [a-\d] and treats
2740    the hyphen as a literal. PCRE used to give an error; it now behaves like Perl.
2741    
2742    3. Added the functions pcre_free_substring() and pcre_free_substring_list().
2743    These just pass their arguments on to (pcre_free)(), but they are provided
2744    because some uses of PCRE bind it to non-C systems that can call its functions,
2745    but cannot call free() or pcre_free() directly.
2746    
2747    4. Add "make test" as a synonym for "make check". Corrected some comments in
2748    the Makefile.
2749    
2750    5. Add $(DESTDIR)/ in front of all the paths in the "install" target in the
2751    Makefile.
2752    
2753    6. Changed the name of pgrep to pcregrep, because Solaris has introduced a
2754    command called pgrep for grepping around the active processes.
2755    
2756    7. Added the beginnings of support for UTF-8 character strings.
2757    
2758    8. Arranged for the Makefile to pass over the settings of CC, CFLAGS, and
2759    RANLIB to ./ltconfig so that they are used by libtool. I think these are all
2760    the relevant ones. (AR is not passed because ./ltconfig does its own figuring
2761    out for the ar command.)
2762    
2763    
2764    Version 3.2 12-May-00
2765    ---------------------
2766    
2767    This is purely a bug fixing release.
2768    
2769    1. If the pattern /((Z)+|A)*/ was matched agained ZABCDEFG it matched Z instead
2770    of ZA. This was just one example of several cases that could provoke this bug,
2771    which was introduced by change 9 of version 2.00. The code for breaking
2772    infinite loops after an iteration that matches an empty string was't working
2773    correctly.
2774    
2775    2. The pcretest program was not imitating Perl correctly for the pattern /a*/g
2776    when matched against abbab (for example). After matching an empty string, it
2777    wasn't forcing anchoring when setting PCRE_NOTEMPTY for the next attempt; this
2778    caused it to match further down the string than it should.
2779    
2780    3. The code contained an inclusion of sys/types.h. It isn't clear why this
2781    was there because it doesn't seem to be needed, and it causes trouble on some
2782    systems, as it is not a Standard C header. It has been removed.
2783    
2784    4. Made 4 silly changes to the source to avoid stupid compiler warnings that
2785    were reported on the Macintosh. The changes were from
2786    
2787      while ((c = *(++ptr)) != 0 && c != '\n');
2788    to
2789      while ((c = *(++ptr)) != 0 && c != '\n') ;
2790    
2791    Totally extraordinary, but if that's what it takes...
2792    
2793    5. PCRE is being used in one environment where neither memmove() nor bcopy() is
2794    available. Added HAVE_BCOPY and an autoconf test for it; if neither
2795    HAVE_MEMMOVE nor HAVE_BCOPY is set, use a built-in emulation function which
2796    assumes the way PCRE uses memmove() (always moving upwards).
2797    
2798    6. PCRE is being used in one environment where strchr() is not available. There
2799    was only one use in pcre.c, and writing it out to avoid strchr() probably gives
2800    faster code anyway.
2801    
2802    
2803    Version 3.1 09-Feb-00
2804    ---------------------
2805    
2806    The only change in this release is the fixing of some bugs in Makefile.in for
2807    the "install" target:
2808    
2809    (1) It was failing to install pcreposix.h.
2810    
2811    (2) It was overwriting the pcre.3 man page with the pcreposix.3 man page.
2812    
2813    
2814    Version 3.0 01-Feb-00
2815    ---------------------
2816    
2817    1. Add support for the /+ modifier to perltest (to output $` like it does in
2818    pcretest).
2819    
2820    2. Add support for the /g modifier to perltest.
2821    
2822    3. Fix pcretest so that it behaves even more like Perl for /g when the pattern
2823    matches null strings.
2824    
2825    4. Fix perltest so that it doesn't do unwanted things when fed an empty
2826    pattern. Perl treats empty patterns specially - it reuses the most recent
2827    pattern, which is not what we want. Replace // by /(?#)/ in order to avoid this
2828    effect.
2829    
2830    5. The POSIX interface was broken in that it was just handing over the POSIX
2831    captured string vector to pcre_exec(), but (since release 2.00) PCRE has
2832    required a bigger vector, with some working space on the end. This means that
2833    the POSIX wrapper now has to get and free some memory, and copy the results.
2834    
2835    6. Added some simple autoconf support, placing the test data and the
2836    documentation in separate directories, re-organizing some of the
2837    information files, and making it build pcre-config (a GNU standard). Also added
2838    libtool support for building PCRE as a shared library, which is now the
2839    default.
2840    
2841    7. Got rid of the leading zero in the definition of PCRE_MINOR because 08 and
2842    09 are not valid octal constants. Single digits will be used for minor values
2843    less than 10.
2844    
2845    8. Defined REG_EXTENDED and REG_NOSUB as zero in the POSIX header, so that
2846    existing programs that set these in the POSIX interface can use PCRE without
2847    modification.
2848    
2849    9. Added a new function, pcre_fullinfo() with an extensible interface. It can
2850    return all that pcre_info() returns, plus additional data. The pcre_info()
2851    function is retained for compatibility, but is considered to be obsolete.
2852    
2853    10. Added experimental recursion feature (?R) to handle one common case that
2854    Perl 5.6 will be able to do with (?p{...}).
2855    
2856    11. Added support for POSIX character classes like [:alpha:], which Perl is
2857    adopting.
2858    
2859    
2860    Version 2.08 31-Aug-99
2861    ----------------------
2862    
2863    1. When startoffset was not zero and the pattern began with ".*", PCRE was not
2864    trying to match at the startoffset position, but instead was moving forward to
2865    the next newline as if a previous match had failed.
2866    
2867    2. pcretest was not making use of PCRE_NOTEMPTY when repeating for /g and /G,
2868    and could get into a loop if a null string was matched other than at the start
2869    of the subject.
2870    
2871    3. Added definitions of PCRE_MAJOR and PCRE_MINOR to pcre.h so the version can
2872    be distinguished at compile time, and for completeness also added PCRE_DATE.
2873    
2874    5. Added Paul Sokolovsky's minor changes to make it easy to compile a Win32 DLL
2875    in GnuWin32 environments.
2876    
2877    
2878    Version 2.07 29-Jul-99
2879    ----------------------
2880    
2881    1. The documentation is now supplied in plain text form and HTML as well as in
2882    the form of man page sources.
2883    
2884    2. C++ compilers don't like assigning (void *) values to other pointer types.
2885    In particular this affects malloc(). Although there is no problem in Standard
2886    C, I've put in casts to keep C++ compilers happy.
2887    
2888    3. Typo on pcretest.c; a cast of (unsigned char *) in the POSIX regexec() call
2889    should be (const char *).
2890    
2891    4. If NOPOSIX is defined, pcretest.c compiles without POSIX support. This may
2892    be useful for non-Unix systems who don't want to bother with the POSIX stuff.
2893    However, I haven't made this a standard facility. The documentation doesn't
2894    mention it, and the Makefile doesn't support it.
2895    
2896    5. The Makefile now contains an "install" target, with editable destinations at
2897    the top of the file. The pcretest program is not installed.
2898    
2899    6. pgrep -V now gives the PCRE version number and date.
2900    
2901    7. Fixed bug: a zero repetition after a literal string (e.g. /abcde{0}/) was
2902    causing the entire string to be ignored, instead of just the last character.
2903    
2904    8. If a pattern like /"([^\\"]+|\\.)*"/ is applied in the normal way to a
2905    non-matching string, it can take a very, very long time, even for strings of
2906    quite modest length, because of the nested recursion. PCRE now does better in
2907    some of these cases. It does this by remembering the last required literal
2908    character in the pattern, and pre-searching the subject to ensure it is present
2909    before running the real match. In other words, it applies a heuristic to detect
2910    some types of certain failure quickly, and in the above example, if presented
2911    with a string that has no trailing " it gives "no match" very quickly.
2912    
2913    9. A new runtime option PCRE_NOTEMPTY causes null string matches to be ignored;
2914    other alternatives are tried instead.
2915    
2916    
2917    Version 2.06 09-Jun-99
2918    ----------------------
2919    
2920    1. Change pcretest's output for amount of store used to show just the code
2921    space, because the remainder (the data block) varies in size between 32-bit and
2922    64-bit systems.
2923    
2924    2. Added an extra argument to pcre_exec() to supply an offset in the subject to
2925    start matching at. This allows lookbehinds to work when searching for multiple
2926    occurrences in a string.
2927    
2928    3. Added additional options to pcretest for testing multiple occurrences:
2929    
2930       /+   outputs the rest of the string that follows a match
2931       /g   loops for multiple occurrences, using the new startoffset argument
2932       /G   loops for multiple occurrences by passing an incremented pointer
2933    
2934    4. PCRE wasn't doing the "first character" optimization for patterns starting
2935    with \b or \B, though it was doing it for other lookbehind assertions. That is,
2936    it wasn't noticing that a match for a pattern such as /\bxyz/ has to start with
2937    the letter 'x'. On long subject strings, this gives a significant speed-up.
2938    
2939    
2940    Version 2.05 21-Apr-99
2941    ----------------------
2942    
2943    1. Changed the type of magic_number from int to long int so that it works
2944    properly on 16-bit systems.
2945    
2946    2. Fixed a bug which caused patterns starting with .* not to work correctly
2947    when the subject string contained newline characters. PCRE was assuming
2948    anchoring for such patterns in all cases, which is not correct because .* will
2949    not pass a newline unless PCRE_DOTALL is set. It now assumes anchoring only if
2950    DOTALL is set at top level; otherwise it knows that patterns starting with .*
2951    must be retried after every newline in the subject.
2952    
2953    
2954    Version 2.04 18-Feb-99
2955    ----------------------
2956    
2957    1. For parenthesized subpatterns with repeats whose minimum was zero, the
2958    computation of the store needed to hold the pattern was incorrect (too large).
2959    If such patterns were nested a few deep, this could multiply and become a real
2960    problem.
2961    
2962    2. Added /M option to pcretest to show the memory requirement of a specific
2963    pattern. Made -m a synonym of -s (which does this globally) for compatibility.
2964    
2965    3. Subpatterns of the form (regex){n,m} (i.e. limited maximum) were being
2966    compiled in such a way that the backtracking after subsequent failure was
2967    pessimal. Something like (a){0,3} was compiled as (a)?(a)?(a)? instead of
2968    ((a)((a)(a)?)?)? with disastrous performance if the maximum was of any size.
2969    
2970    
2971    Version 2.03 02-Feb-99
2972    ----------------------
2973    
2974    1. Fixed typo and small mistake in man page.
2975    
2976    2. Added 4th condition (GPL supersedes if conflict) and created separate
2977    LICENCE file containing the conditions.
2978    
2979    3. Updated pcretest so that patterns such as /abc\/def/ work like they do in
2980    Perl, that is the internal \ allows the delimiter to be included in the
2981    pattern. Locked out the use of \ as a delimiter. If \ immediately follows
2982    the final delimiter, add \ to the end of the pattern (to test the error).
2983    
2984    4. Added the convenience functions for extracting substrings after a successful
2985    match. Updated pcretest to make it able to test these functions.
2986    
2987    
2988    Version 2.02 14-Jan-99
2989    ----------------------
2990    
2991    1. Initialized the working variables associated with each extraction so that
2992    their saving and restoring doesn't refer to uninitialized store.
2993    
2994    2. Put dummy code into study.c in order to trick the optimizer of the IBM C
2995    compiler for OS/2 into generating correct code. Apparently IBM isn't going to
2996    fix the problem.
2997    
2998    3. Pcretest: the timing code wasn't using LOOPREPEAT for timing execution
2999    calls, and wasn't printing the correct value for compiling calls. Increased the
3000    default value of LOOPREPEAT, and the number of significant figures in the
3001    times.
3002    
3003    4. Changed "/bin/rm" in the Makefile to "-rm" so it works on Windows NT.
3004    
3005    5. Renamed "deftables" as "dftables" to get it down to 8 characters, to avoid
3006    a building problem on Windows NT with a FAT file system.
3007    
3008    
3009    Version 2.01 21-Oct-98
3010    ----------------------
3011    
3012    1. Changed the API for pcre_compile() to allow for the provision of a pointer
3013    to character tables built by pcre_maketables() in the current locale. If NULL
3014    is passed, the default tables are used.
3015    
3016    
3017    Version 2.00 24-Sep-98
3018    ----------------------
3019    
3020    1. Since the (>?) facility is in Perl 5.005, don't require PCRE_EXTRA to enable
3021    it any more.
3022    
3023    2. Allow quantification of (?>) groups, and make it work correctly.
3024    
3025    3. The first character computation wasn't working for (?>) groups.
3026    
3027    4. Correct the implementation of \Z (it is permitted to match on the \n at the
3028    end of the subject) and add 5.005's \z, which really does match only at the
3029    very end of the subject.
3030    
3031    5. Remove the \X "cut" facility; Perl doesn't have it, and (?> is neater.
3032    
3033    6. Remove the ability to specify CASELESS, MULTILINE, DOTALL, and
3034    DOLLAR_END_ONLY at runtime, to make it possible to implement the Perl 5.005
3035    localized options. All options to pcre_study() were also removed.
3036    
3037    7. Add other new features from 5.005:
3038    
3039       $(?<=           positive lookbehind
3040       $(?<!           negative lookbehind
3041       (?imsx-imsx)    added the unsetting capability
3042                       such a setting is global if at outer level; local otherwise
3043       (?imsx-imsx:)   non-capturing groups with option setting
3044       (?(cond)re|re)  conditional pattern matching
3045    
3046       A backreference to itself in a repeated group matches the previous
3047       captured string.
3048    
3049    8. General tidying up of studying (both automatic and via "study")
3050    consequential on the addition of new assertions.
3051    
3052    9. As in 5.005, unlimited repeated groups that could match an empty substring
3053    are no longer faulted at compile time. Instead, the loop is forcibly broken at
3054    runtime if any iteration does actually match an empty substring.
3055    
3056    10. Include the RunTest script in the distribution.
3057    
3058    11. Added tests from the Perl 5.005_02 distribution. This showed up a few
3059    discrepancies, some of which were old and were also with respect to 5.004. They
3060    have now been fixed.
3061    
3062    
3063    Version 1.09 28-Apr-98
3064    ----------------------
3065    
3066    1. A negated single character class followed by a quantifier with a minimum
3067    value of one (e.g.  [^x]{1,6}  ) was not compiled correctly. This could lead to
3068    program crashes, or just wrong answers. This did not apply to negated classes
3069    containing more than one character, or to minima other than one.
3070    
3071    
3072    Version 1.08 27-Mar-98
3073    ----------------------
3074    
3075    1. Add PCRE_UNGREEDY to invert the greediness of quantifiers.
3076    
3077    2. Add (?U) and (?X) to set PCRE_UNGREEDY and PCRE_EXTRA respectively. The
3078    latter must appear before anything that relies on it in the pattern.
3079    
3080    
3081  Version 1.07 16-Feb-98  Version 1.07 16-Feb-98
3082  ----------------------  ----------------------

Legend:
Removed from v.17  
changed lines
  Added in v.366

webmaster@exim.org
ViewVC Help
Powered by ViewVC 1.1.12