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

Diff of /code/trunk/ChangeLog

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

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

Legend:
Removed from v.27  
changed lines
  Added in v.364

webmaster@exim.org
ViewVC Help
Powered by ViewVC 1.1.12