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

Diff of /code/trunk/ChangeLog

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

revision 87 by nigel, Sat Feb 24 21:41:21 2007 UTC revision 386 by ph10, Tue Mar 10 11:48:33 2009 UTC
# Line 1  Line 1 
1  ChangeLog for PCRE  ChangeLog for PCRE
2  ------------------  ------------------
3    
4    Version 7.9 xx-xxx-09
5    ---------------------
6    
7    1.  When building with support for bzlib/zlib (pcregrep) and/or readline
8        (pcretest), all targets were linked against these libraries. This included
9        libpcre, libpcreposix, and libpcrecpp, even though they do not use these
10        libraries. This caused unwanted dependencies to be created. This problem
11        has been fixed, and now only pcregrep is linked with bzlib/zlib and only
12        pcretest is linked with readline.
13    
14    2.  The "typedef int BOOL" in pcre_internal.h that was included inside the
15        "#ifndef FALSE" condition by an earlier change (probably 7.8/18) has been
16        moved outside it again, because FALSE and TRUE are already defined in AIX,
17        but BOOL is not.
18    
19    3.  The pcre_config() function was treating the PCRE_MATCH_LIMIT and
20        PCRE_MATCH_LIMIT_RETURSION values as ints, when they should be long ints.
21    
22    4.  The pcregrep documentation said spaces were inserted as well as colons (or
23        hyphens) following file names and line numbers when outputting matching
24        lines. This is not true; no spaces are inserted. I have also clarified the
25        wording for the --colour (or --color) option.
26    
27    5.  In pcregrep, when --colour was used with -o, the list of matching strings
28        was not coloured; this is different to GNU grep, so I have changed it to be
29        the same.
30    
31    6.  When --colo(u)r was used in pcregrep, only the first matching substring in
32        each matching line was coloured. Now it goes on to look for further matches
33        of any of the test patterns, which is the same behaviour as GNU grep.
34    
35    7.  A pattern that could match an empty string could cause pcregrep to loop; it
36        doesn't make sense to accept an empty string match in pcregrep, so I have
37        locked it out (using PCRE's PCRE_NOTEMPTY option). By experiment, this
38        seems to be how GNU grep behaves.
39    
40    8.  The pattern (?(?=.*b)b|^) was incorrectly compiled as "match must be at
41        start or after a newline", because the conditional assertion was not being
42        correctly handled. The rule now is that both the assertion and what follows
43        in the first alternative must satisfy the test.
44    
45    9.  If auto-callout was enabled in a pattern with a conditional group, PCRE
46        could crash during matching.
47    
48    10. The PCRE_DOLLAR_ENDONLY option was not working when pcre_dfa_exec() was
49        used for matching.
50    
51    11. Unicode property support in character classes was not working for
52        characters (bytes) greater than 127 when not in UTF-8 mode.
53    
54    12. Added the -M command line option to pcretest.
55    
56    
57    Version 7.8 05-Sep-08
58    ---------------------
59    
60    1.  Replaced UCP searching code with optimized version as implemented for Ad
61        Muncher (http://www.admuncher.com/) by Peter Kankowski. This uses a two-
62        stage table and inline lookup instead of a function, giving speed ups of 2
63        to 5 times on some simple patterns that I tested. Permission was given to
64        distribute the MultiStage2.py script that generates the tables (it's not in
65        the tarball, but is in the Subversion repository).
66    
67    2.  Updated the Unicode datatables to Unicode 5.1.0. This adds yet more
68        scripts.
69    
70    3.  Change 12 for 7.7 introduced a bug in pcre_study() when a pattern contained
71        a group with a zero qualifier. The result of the study could be incorrect,
72        or the function might crash, depending on the pattern.
73    
74    4.  Caseless matching was not working for non-ASCII characters in back
75        references. For example, /(\x{de})\1/8i was not matching \x{de}\x{fe}.
76        It now works when Unicode Property Support is available.
77    
78    5.  In pcretest, an escape such as \x{de} in the data was always generating
79        a UTF-8 string, even in non-UTF-8 mode. Now it generates a single byte in
80        non-UTF-8 mode. If the value is greater than 255, it gives a warning about
81        truncation.
82    
83    6.  Minor bugfix in pcrecpp.cc (change "" == ... to NULL == ...).
84    
85    7.  Added two (int) casts to pcregrep when printing the difference of two
86        pointers, in case they are 64-bit values.
87    
88    8.  Added comments about Mac OS X stack usage to the pcrestack man page and to
89        test 2 if it fails.
90    
91    9.  Added PCRE_CALL_CONVENTION just before the names of all exported functions,
92        and a #define of that name to empty if it is not externally set. This is to
93        allow users of MSVC to set it if necessary.
94    
95    10. The PCRE_EXP_DEFN macro which precedes exported functions was missing from
96        the convenience functions in the pcre_get.c source file.
97    
98    11. An option change at the start of a pattern that had top-level alternatives
99        could cause overwriting and/or a crash. This command provoked a crash in
100        some environments:
101    
102          printf "/(?i)[\xc3\xa9\xc3\xbd]|[\xc3\xa9\xc3\xbdA]/8\n" | pcretest
103    
104        This potential security problem was recorded as CVE-2008-2371.
105    
106    12. For a pattern where the match had to start at the beginning or immediately
107        after a newline (e.g /.*anything/ without the DOTALL flag), pcre_exec() and
108        pcre_dfa_exec() could read past the end of the passed subject if there was
109        no match. To help with detecting such bugs (e.g. with valgrind), I modified
110        pcretest so that it places the subject at the end of its malloc-ed buffer.
111    
112    13. The change to pcretest in 12 above threw up a couple more cases when pcre_
113        exec() might read past the end of the data buffer in UTF-8 mode.
114    
115    14. A similar bug to 7.3/2 existed when the PCRE_FIRSTLINE option was set and
116        the data contained the byte 0x85 as part of a UTF-8 character within its
117        first line. This applied both to normal and DFA matching.
118    
119    15. Lazy qualifiers were not working in some cases in UTF-8 mode. For example,
120        /^[^d]*?$/8 failed to match "abc".
121    
122    16. Added a missing copyright notice to pcrecpp_internal.h.
123    
124    17. Make it more clear in the documentation that values returned from
125        pcre_exec() in ovector are byte offsets, not character counts.
126    
127    18. Tidied a few places to stop certain compilers from issuing warnings.
128    
129    19. Updated the Virtual Pascal + BCC files to compile the latest v7.7, as
130        supplied by Stefan Weber. I made a further small update for 7.8 because
131        there is a change of source arrangements: the pcre_searchfuncs.c module is
132        replaced by pcre_ucd.c.
133    
134    
135    Version 7.7 07-May-08
136    ---------------------
137    
138    1.  Applied Craig's patch to sort out a long long problem: "If we can't convert
139        a string to a long long, pretend we don't even have a long long." This is
140        done by checking for the strtoq, strtoll, and _strtoi64 functions.
141    
142    2.  Applied Craig's patch to pcrecpp.cc to restore ABI compatibility with
143        pre-7.6 versions, which defined a global no_arg variable instead of putting
144        it in the RE class. (See also #8 below.)
145    
146    3.  Remove a line of dead code, identified by coverity and reported by Nuno
147        Lopes.
148    
149    4.  Fixed two related pcregrep bugs involving -r with --include or --exclude:
150    
151        (1) The include/exclude patterns were being applied to the whole pathnames
152            of files, instead of just to the final components.
153    
154        (2) If there was more than one level of directory, the subdirectories were
155            skipped unless they satisfied the include/exclude conditions. This is
156            inconsistent with GNU grep (and could even be seen as contrary to the
157            pcregrep specification - which I improved to make it absolutely clear).
158            The action now is always to scan all levels of directory, and just
159            apply the include/exclude patterns to regular files.
160    
161    5.  Added the --include_dir and --exclude_dir patterns to pcregrep, and used
162        --exclude_dir in the tests to avoid scanning .svn directories.
163    
164    6.  Applied Craig's patch to the QuoteMeta function so that it escapes the
165        NUL character as backslash + 0 rather than backslash + NUL, because PCRE
166        doesn't support NULs in patterns.
167    
168    7.  Added some missing "const"s to declarations of static tables in
169        pcre_compile.c and pcre_dfa_exec.c.
170    
171    8.  Applied Craig's patch to pcrecpp.cc to fix a problem in OS X that was
172        caused by fix #2  above. (Subsequently also a second patch to fix the
173        first patch. And a third patch - this was a messy problem.)
174    
175    9.  Applied Craig's patch to remove the use of push_back().
176    
177    10. Applied Alan Lehotsky's patch to add REG_STARTEND support to the POSIX
178        matching function regexec().
179    
180    11. Added support for the Oniguruma syntax \g<name>, \g<n>, \g'name', \g'n',
181        which, however, unlike Perl's \g{...}, are subroutine calls, not back
182        references. PCRE supports relative numbers with this syntax (I don't think
183        Oniguruma does).
184    
185    12. Previously, a group with a zero repeat such as (...){0} was completely
186        omitted from the compiled regex. However, this means that if the group
187        was called as a subroutine from elsewhere in the pattern, things went wrong
188        (an internal error was given). Such groups are now left in the compiled
189        pattern, with a new opcode that causes them to be skipped at execution
190        time.
191    
192    13. Added the PCRE_JAVASCRIPT_COMPAT option. This makes the following changes
193        to the way PCRE behaves:
194    
195        (a) A lone ] character is dis-allowed (Perl treats it as data).
196    
197        (b) A back reference to an unmatched subpattern matches an empty string
198            (Perl fails the current match path).
199    
200        (c) A data ] in a character class must be notated as \] because if the
201            first data character in a class is ], it defines an empty class. (In
202            Perl it is not possible to have an empty class.) The empty class []
203            never matches; it forces failure and is equivalent to (*FAIL) or (?!).
204            The negative empty class [^] matches any one character, independently
205            of the DOTALL setting.
206    
207    14. A pattern such as /(?2)[]a()b](abc)/ which had a forward reference to a
208        non-existent subpattern following a character class starting with ']' and
209        containing () gave an internal compiling error instead of "reference to
210        non-existent subpattern". Fortunately, when the pattern did exist, the
211        compiled code was correct. (When scanning forwards to check for the
212        existencd of the subpattern, it was treating the data ']' as terminating
213        the class, so got the count wrong. When actually compiling, the reference
214        was subsequently set up correctly.)
215    
216    15. The "always fail" assertion (?!) is optimzed to (*FAIL) by pcre_compile;
217        it was being rejected as not supported by pcre_dfa_exec(), even though
218        other assertions are supported. I have made pcre_dfa_exec() support
219        (*FAIL).
220    
221    16. The implementation of 13c above involved the invention of a new opcode,
222        OP_ALLANY, which is like OP_ANY but doesn't check the /s flag. Since /s
223        cannot be changed at match time, I realized I could make a small
224        improvement to matching performance by compiling OP_ALLANY instead of
225        OP_ANY for "." when DOTALL was set, and then removing the runtime tests
226        on the OP_ANY path.
227    
228    17. Compiling pcretest on Windows with readline support failed without the
229        following two fixes: (1) Make the unistd.h include conditional on
230        HAVE_UNISTD_H; (2) #define isatty and fileno as _isatty and _fileno.
231    
232    18. Changed CMakeLists.txt and cmake/FindReadline.cmake to arrange for the
233        ncurses library to be included for pcretest when ReadLine support is
234        requested, but also to allow for it to be overridden. This patch came from
235        Daniel Bergström.
236    
237    19. There was a typo in the file ucpinternal.h where f0_rangeflag was defined
238        as 0x00f00000 instead of 0x00800000. Luckily, this would not have caused
239        any errors with the current Unicode tables. Thanks to Peter Kankowski for
240        spotting this.
241    
242    
243    Version 7.6 28-Jan-08
244    ---------------------
245    
246    1.  A character class containing a very large number of characters with
247        codepoints greater than 255 (in UTF-8 mode, of course) caused a buffer
248        overflow.
249    
250    2.  Patch to cut out the "long long" test in pcrecpp_unittest when
251        HAVE_LONG_LONG is not defined.
252    
253    3.  Applied Christian Ehrlicher's patch to update the CMake build files to
254        bring them up to date and include new features. This patch includes:
255    
256        - Fixed PH's badly added libz and libbz2 support.
257        - Fixed a problem with static linking.
258        - Added pcredemo. [But later removed - see 7 below.]
259        - Fixed dftables problem and added an option.
260        - Added a number of HAVE_XXX tests, including HAVE_WINDOWS_H and
261            HAVE_LONG_LONG.
262        - Added readline support for pcretest.
263        - Added an listing of the option settings after cmake has run.
264    
265    4.  A user submitted a patch to Makefile that makes it easy to create
266        "pcre.dll" under mingw when using Configure/Make. I added stuff to
267        Makefile.am that cause it to include this special target, without
268        affecting anything else. Note that the same mingw target plus all
269        the other distribution libraries and programs are now supported
270        when configuring with CMake (see 6 below) instead of with
271        Configure/Make.
272    
273    5.  Applied Craig's patch that moves no_arg into the RE class in the C++ code.
274        This is an attempt to solve the reported problem "pcrecpp::no_arg is not
275        exported in the Windows port". It has not yet been confirmed that the patch
276        solves the problem, but it does no harm.
277    
278    6.  Applied Sheri's patch to CMakeLists.txt to add NON_STANDARD_LIB_PREFIX and
279        NON_STANDARD_LIB_SUFFIX for dll names built with mingw when configured
280        with CMake, and also correct the comment about stack recursion.
281    
282    7.  Remove the automatic building of pcredemo from the ./configure system and
283        from CMakeLists.txt. The whole idea of pcredemo.c is that it is an example
284        of a program that users should build themselves after PCRE is installed, so
285        building it automatically is not really right. What is more, it gave
286        trouble in some build environments.
287    
288    8.  Further tidies to CMakeLists.txt from Sheri and Christian.
289    
290    
291    Version 7.5 10-Jan-08
292    ---------------------
293    
294    1.  Applied a patch from Craig: "This patch makes it possible to 'ignore'
295        values in parens when parsing an RE using the C++ wrapper."
296    
297    2.  Negative specials like \S did not work in character classes in UTF-8 mode.
298        Characters greater than 255 were excluded from the class instead of being
299        included.
300    
301    3.  The same bug as (2) above applied to negated POSIX classes such as
302        [:^space:].
303    
304    4.  PCRECPP_STATIC was referenced in pcrecpp_internal.h, but nowhere was it
305        defined or documented. It seems to have been a typo for PCRE_STATIC, so
306        I have changed it.
307    
308    5.  The construct (?&) was not diagnosed as a syntax error (it referenced the
309        first named subpattern) and a construct such as (?&a) would reference the
310        first named subpattern whose name started with "a" (in other words, the
311        length check was missing). Both these problems are fixed. "Subpattern name
312        expected" is now given for (?&) (a zero-length name), and this patch also
313        makes it give the same error for \k'' (previously it complained that that
314        was a reference to a non-existent subpattern).
315    
316    6.  The erroneous patterns (?+-a) and (?-+a) give different error messages;
317        this is right because (?- can be followed by option settings as well as by
318        digits. I have, however, made the messages clearer.
319    
320    7.  Patterns such as (?(1)a|b) (a pattern that contains fewer subpatterns
321        than the number used in the conditional) now cause a compile-time error.
322        This is actually not compatible with Perl, which accepts such patterns, but
323        treats the conditional as always being FALSE (as PCRE used to), but it
324        seems to me that giving a diagnostic is better.
325    
326    8.  Change "alphameric" to the more common word "alphanumeric" in comments
327        and messages.
328    
329    9.  Fix two occurrences of "backslash" in comments that should have been
330        "backspace".
331    
332    10. Remove two redundant lines of code that can never be obeyed (their function
333        was moved elsewhere).
334    
335    11. The program that makes PCRE's Unicode character property table had a bug
336        which caused it to generate incorrect table entries for sequences of
337        characters that have the same character type, but are in different scripts.
338        It amalgamated them into a single range, with the script of the first of
339        them. In other words, some characters were in the wrong script. There were
340        thirteen such cases, affecting characters in the following ranges:
341    
342          U+002b0 - U+002c1
343          U+0060c - U+0060d
344          U+0061e - U+00612
345          U+0064b - U+0065e
346          U+0074d - U+0076d
347          U+01800 - U+01805
348          U+01d00 - U+01d77
349          U+01d9b - U+01dbf
350          U+0200b - U+0200f
351          U+030fc - U+030fe
352          U+03260 - U+0327f
353          U+0fb46 - U+0fbb1
354          U+10450 - U+1049d
355    
356    12. The -o option (show only the matching part of a line) for pcregrep was not
357        compatible with GNU grep in that, if there was more than one match in a
358        line, it showed only the first of them. It now behaves in the same way as
359        GNU grep.
360    
361    13. If the -o and -v options were combined for pcregrep, it printed a blank
362        line for every non-matching line. GNU grep prints nothing, and pcregrep now
363        does the same. The return code can be used to tell if there were any
364        non-matching lines.
365    
366    14. Added --file-offsets and --line-offsets to pcregrep.
367    
368    15. The pattern (?=something)(?R) was not being diagnosed as a potentially
369        infinitely looping recursion. The bug was that positive lookaheads were not
370        being skipped when checking for a possible empty match (negative lookaheads
371        and both kinds of lookbehind were skipped).
372    
373    16. Fixed two typos in the Windows-only code in pcregrep.c, and moved the
374        inclusion of <windows.h> to before rather than after the definition of
375        INVALID_FILE_ATTRIBUTES (patch from David Byron).
376    
377    17. Specifying a possessive quantifier with a specific limit for a Unicode
378        character property caused pcre_compile() to compile bad code, which led at
379        runtime to PCRE_ERROR_INTERNAL (-14). Examples of patterns that caused this
380        are: /\p{Zl}{2,3}+/8 and /\p{Cc}{2}+/8. It was the possessive "+" that
381        caused the error; without that there was no problem.
382    
383    18. Added --enable-pcregrep-libz and --enable-pcregrep-libbz2.
384    
385    19. Added --enable-pcretest-libreadline.
386    
387    20. In pcrecpp.cc, the variable 'count' was incremented twice in
388        RE::GlobalReplace(). As a result, the number of replacements returned was
389        double what it should be. I removed one of the increments, but Craig sent a
390        later patch that removed the other one (the right fix) and added unit tests
391        that check the return values (which was not done before).
392    
393    21. Several CMake things:
394    
395        (1) Arranged that, when cmake is used on Unix, the libraries end up with
396            the names libpcre and libpcreposix, not just pcre and pcreposix.
397    
398        (2) The above change means that pcretest and pcregrep are now correctly
399            linked with the newly-built libraries, not previously installed ones.
400    
401        (3) Added PCRE_SUPPORT_LIBREADLINE, PCRE_SUPPORT_LIBZ, PCRE_SUPPORT_LIBBZ2.
402    
403    22. In UTF-8 mode, with newline set to "any", a pattern such as .*a.*=.b.*
404        crashed when matching a string such as a\x{2029}b (note that \x{2029} is a
405        UTF-8 newline character). The key issue is that the pattern starts .*;
406        this means that the match must be either at the beginning, or after a
407        newline. The bug was in the code for advancing after a failed match and
408        checking that the new position followed a newline. It was not taking
409        account of UTF-8 characters correctly.
410    
411    23. PCRE was behaving differently from Perl in the way it recognized POSIX
412        character classes. PCRE was not treating the sequence [:...:] as a
413        character class unless the ... were all letters. Perl, however, seems to
414        allow any characters between [: and :], though of course it rejects as
415        unknown any "names" that contain non-letters, because all the known class
416        names consist only of letters. Thus, Perl gives an error for [[:1234:]],
417        for example, whereas PCRE did not - it did not recognize a POSIX character
418        class. This seemed a bit dangerous, so the code has been changed to be
419        closer to Perl. The behaviour is not identical to Perl, because PCRE will
420        diagnose an unknown class for, for example, [[:l\ower:]] where Perl will
421        treat it as [[:lower:]]. However, PCRE does now give "unknown" errors where
422        Perl does, and where it didn't before.
423    
424    24. Rewrite so as to remove the single use of %n from pcregrep because in some
425        Windows environments %n is disabled by default.
426    
427    
428    Version 7.4 21-Sep-07
429    ---------------------
430    
431    1.  Change 7.3/28 was implemented for classes by looking at the bitmap. This
432        means that a class such as [\s] counted as "explicit reference to CR or
433        LF". That isn't really right - the whole point of the change was to try to
434        help when there was an actual mention of one of the two characters. So now
435        the change happens only if \r or \n (or a literal CR or LF) character is
436        encountered.
437    
438    2.  The 32-bit options word was also used for 6 internal flags, but the numbers
439        of both had grown to the point where there were only 3 bits left.
440        Fortunately, there was spare space in the data structure, and so I have
441        moved the internal flags into a new 16-bit field to free up more option
442        bits.
443    
444    3.  The appearance of (?J) at the start of a pattern set the DUPNAMES option,
445        but did not set the internal JCHANGED flag - either of these is enough to
446        control the way the "get" function works - but the PCRE_INFO_JCHANGED
447        facility is supposed to tell if (?J) was ever used, so now (?J) at the
448        start sets both bits.
449    
450    4.  Added options (at build time, compile time, exec time) to change \R from
451        matching any Unicode line ending sequence to just matching CR, LF, or CRLF.
452    
453    5.  doc/pcresyntax.html was missing from the distribution.
454    
455    6.  Put back the definition of PCRE_ERROR_NULLWSLIMIT, for backward
456        compatibility, even though it is no longer used.
457    
458    7.  Added macro for snprintf to pcrecpp_unittest.cc and also for strtoll and
459        strtoull to pcrecpp.cc to select the available functions in WIN32 when the
460        windows.h file is present (where different names are used). [This was
461        reversed later after testing - see 16 below.]
462    
463    8.  Changed all #include <config.h> to #include "config.h". There were also
464        some further <pcre.h> cases that I changed to "pcre.h".
465    
466    9.  When pcregrep was used with the --colour option, it missed the line ending
467        sequence off the lines that it output.
468    
469    10. It was pointed out to me that arrays of string pointers cause lots of
470        relocations when a shared library is dynamically loaded. A technique of
471        using a single long string with a table of offsets can drastically reduce
472        these. I have refactored PCRE in four places to do this. The result is
473        dramatic:
474    
475          Originally:                          290
476          After changing UCP table:            187
477          After changing error message table:   43
478          After changing table of "verbs"       36
479          After changing table of Posix names   22
480    
481        Thanks to the folks working on Gregex for glib for this insight.
482    
483    11. --disable-stack-for-recursion caused compiling to fail unless -enable-
484        unicode-properties was also set.
485    
486    12. Updated the tests so that they work when \R is defaulted to ANYCRLF.
487    
488    13. Added checks for ANY and ANYCRLF to pcrecpp.cc where it previously
489        checked only for CRLF.
490    
491    14. Added casts to pcretest.c to avoid compiler warnings.
492    
493    15. Added Craig's patch to various pcrecpp modules to avoid compiler warnings.
494    
495    16. Added Craig's patch to remove the WINDOWS_H tests, that were not working,
496        and instead check for _strtoi64 explicitly, and avoid the use of snprintf()
497        entirely. This removes changes made in 7 above.
498    
499    17. The CMake files have been updated, and there is now more information about
500        building with CMake in the NON-UNIX-USE document.
501    
502    
503    Version 7.3 28-Aug-07
504    ---------------------
505    
506     1. In the rejigging of the build system that eventually resulted in 7.1, the
507        line "#include <pcre.h>" was included in pcre_internal.h. The use of angle
508        brackets there is not right, since it causes compilers to look for an
509        installed pcre.h, not the version that is in the source that is being
510        compiled (which of course may be different). I have changed it back to:
511    
512          #include "pcre.h"
513    
514        I have a vague recollection that the change was concerned with compiling in
515        different directories, but in the new build system, that is taken care of
516        by the VPATH setting the Makefile.
517    
518     2. The pattern .*$ when run in not-DOTALL UTF-8 mode with newline=any failed
519        when the subject happened to end in the byte 0x85 (e.g. if the last
520        character was \x{1ec5}). *Character* 0x85 is one of the "any" newline
521        characters but of course it shouldn't be taken as a newline when it is part
522        of another character. The bug was that, for an unlimited repeat of . in
523        not-DOTALL UTF-8 mode, PCRE was advancing by bytes rather than by
524        characters when looking for a newline.
525    
526     3. A small performance improvement in the DOTALL UTF-8 mode .* case.
527    
528     4. Debugging: adjusted the names of opcodes for different kinds of parentheses
529        in debug output.
530    
531     5. Arrange to use "%I64d" instead of "%lld" and "%I64u" instead of "%llu" for
532        long printing in the pcrecpp unittest when running under MinGW.
533    
534     6. ESC_K was left out of the EBCDIC table.
535    
536     7. Change 7.0/38 introduced a new limit on the number of nested non-capturing
537        parentheses; I made it 1000, which seemed large enough. Unfortunately, the
538        limit also applies to "virtual nesting" when a pattern is recursive, and in
539        this case 1000 isn't so big. I have been able to remove this limit at the
540        expense of backing off one optimization in certain circumstances. Normally,
541        when pcre_exec() would call its internal match() function recursively and
542        immediately return the result unconditionally, it uses a "tail recursion"
543        feature to save stack. However, when a subpattern that can match an empty
544        string has an unlimited repetition quantifier, it no longer makes this
545        optimization. That gives it a stack frame in which to save the data for
546        checking that an empty string has been matched. Previously this was taken
547        from the 1000-entry workspace that had been reserved. So now there is no
548        explicit limit, but more stack is used.
549    
550     8. Applied Daniel's patches to solve problems with the import/export magic
551        syntax that is required for Windows, and which was going wrong for the
552        pcreposix and pcrecpp parts of the library. These were overlooked when this
553        problem was solved for the main library.
554    
555     9. There were some crude static tests to avoid integer overflow when computing
556        the size of patterns that contain repeated groups with explicit upper
557        limits. As the maximum quantifier is 65535, the maximum group length was
558        set at 30,000 so that the product of these two numbers did not overflow a
559        32-bit integer. However, it turns out that people want to use groups that
560        are longer than 30,000 bytes (though not repeat them that many times).
561        Change 7.0/17 (the refactoring of the way the pattern size is computed) has
562        made it possible to implement the integer overflow checks in a much more
563        dynamic way, which I have now done. The artificial limitation on group
564        length has been removed - we now have only the limit on the total length of
565        the compiled pattern, which depends on the LINK_SIZE setting.
566    
567    10. Fixed a bug in the documentation for get/copy named substring when
568        duplicate names are permitted. If none of the named substrings are set, the
569        functions return PCRE_ERROR_NOSUBSTRING (7); the doc said they returned an
570        empty string.
571    
572    11. Because Perl interprets \Q...\E at a high level, and ignores orphan \E
573        instances, patterns such as [\Q\E] or [\E] or even [^\E] cause an error,
574        because the ] is interpreted as the first data character and the
575        terminating ] is not found. PCRE has been made compatible with Perl in this
576        regard. Previously, it interpreted [\Q\E] as an empty class, and [\E] could
577        cause memory overwriting.
578    
579    10. Like Perl, PCRE automatically breaks an unlimited repeat after an empty
580        string has been matched (to stop an infinite loop). It was not recognizing
581        a conditional subpattern that could match an empty string if that
582        subpattern was within another subpattern. For example, it looped when
583        trying to match  (((?(1)X|))*)  but it was OK with  ((?(1)X|)*)  where the
584        condition was not nested. This bug has been fixed.
585    
586    12. A pattern like \X?\d or \P{L}?\d in non-UTF-8 mode could cause a backtrack
587        past the start of the subject in the presence of bytes with the top bit
588        set, for example "\x8aBCD".
589    
590    13. Added Perl 5.10 experimental backtracking controls (*FAIL), (*F), (*PRUNE),
591        (*SKIP), (*THEN), (*COMMIT), and (*ACCEPT).
592    
593    14. Optimized (?!) to (*FAIL).
594    
595    15. Updated the test for a valid UTF-8 string to conform to the later RFC 3629.
596        This restricts code points to be within the range 0 to 0x10FFFF, excluding
597        the "low surrogate" sequence 0xD800 to 0xDFFF. Previously, PCRE allowed the
598        full range 0 to 0x7FFFFFFF, as defined by RFC 2279. Internally, it still
599        does: it's just the validity check that is more restrictive.
600    
601    16. Inserted checks for integer overflows during escape sequence (backslash)
602        processing, and also fixed erroneous offset values for syntax errors during
603        backslash processing.
604    
605    17. Fixed another case of looking too far back in non-UTF-8 mode (cf 12 above)
606        for patterns like [\PPP\x8a]{1,}\x80 with the subject "A\x80".
607    
608    18. An unterminated class in a pattern like (?1)\c[ with a "forward reference"
609        caused an overrun.
610    
611    19. A pattern like (?:[\PPa*]*){8,} which had an "extended class" (one with
612        something other than just ASCII characters) inside a group that had an
613        unlimited repeat caused a loop at compile time (while checking to see
614        whether the group could match an empty string).
615    
616    20. Debugging a pattern containing \p or \P could cause a crash. For example,
617        [\P{Any}] did so. (Error in the code for printing property names.)
618    
619    21. An orphan \E inside a character class could cause a crash.
620    
621    22. A repeated capturing bracket such as (A)? could cause a wild memory
622        reference during compilation.
623    
624    23. There are several functions in pcre_compile() that scan along a compiled
625        expression for various reasons (e.g. to see if it's fixed length for look
626        behind). There were bugs in these functions when a repeated \p or \P was
627        present in the pattern. These operators have additional parameters compared
628        with \d, etc, and these were not being taken into account when moving along
629        the compiled data. Specifically:
630    
631        (a) A item such as \p{Yi}{3} in a lookbehind was not treated as fixed
632            length.
633    
634        (b) An item such as \pL+ within a repeated group could cause crashes or
635            loops.
636    
637        (c) A pattern such as \p{Yi}+(\P{Yi}+)(?1) could give an incorrect
638            "reference to non-existent subpattern" error.
639    
640        (d) A pattern like (\P{Yi}{2}\277)? could loop at compile time.
641    
642    24. A repeated \S or \W in UTF-8 mode could give wrong answers when multibyte
643        characters were involved (for example /\S{2}/8g with "A\x{a3}BC").
644    
645    25. Using pcregrep in multiline, inverted mode (-Mv) caused it to loop.
646    
647    26. Patterns such as [\P{Yi}A] which include \p or \P and just one other
648        character were causing crashes (broken optimization).
649    
650    27. Patterns such as (\P{Yi}*\277)* (group with possible zero repeat containing
651        \p or \P) caused a compile-time loop.
652    
653    28. More problems have arisen in unanchored patterns when CRLF is a valid line
654        break. For example, the unstudied pattern [\r\n]A does not match the string
655        "\r\nA" because change 7.0/46 below moves the current point on by two
656        characters after failing to match at the start. However, the pattern \nA
657        *does* match, because it doesn't start till \n, and if [\r\n]A is studied,
658        the same is true. There doesn't seem any very clean way out of this, but
659        what I have chosen to do makes the common cases work: PCRE now takes note
660        of whether there can be an explicit match for \r or \n anywhere in the
661        pattern, and if so, 7.0/46 no longer applies. As part of this change,
662        there's a new PCRE_INFO_HASCRORLF option for finding out whether a compiled
663        pattern has explicit CR or LF references.
664    
665    29. Added (*CR) etc for changing newline setting at start of pattern.
666    
667    
668    Version 7.2 19-Jun-07
669    ---------------------
670    
671     1. If the fr_FR locale cannot be found for test 3, try the "french" locale,
672        which is apparently normally available under Windows.
673    
674     2. Re-jig the pcregrep tests with different newline settings in an attempt
675        to make them independent of the local environment's newline setting.
676    
677     3. Add code to configure.ac to remove -g from the CFLAGS default settings.
678    
679     4. Some of the "internals" tests were previously cut out when the link size
680        was not 2, because the output contained actual offsets. The recent new
681        "Z" feature of pcretest means that these can be cut out, making the tests
682        usable with all link sizes.
683    
684     5. Implemented Stan Switzer's goto replacement for longjmp() when not using
685        stack recursion. This gives a massive performance boost under BSD, but just
686        a small improvement under Linux. However, it saves one field in the frame
687        in all cases.
688    
689     6. Added more features from the forthcoming Perl 5.10:
690    
691        (a) (?-n) (where n is a string of digits) is a relative subroutine or
692            recursion call. It refers to the nth most recently opened parentheses.
693    
694        (b) (?+n) is also a relative subroutine call; it refers to the nth next
695            to be opened parentheses.
696    
697        (c) Conditions that refer to capturing parentheses can be specified
698            relatively, for example, (?(-2)... or (?(+3)...
699    
700        (d) \K resets the start of the current match so that everything before
701            is not part of it.
702    
703        (e) \k{name} is synonymous with \k<name> and \k'name' (.NET compatible).
704    
705        (f) \g{name} is another synonym - part of Perl 5.10's unification of
706            reference syntax.
707    
708        (g) (?| introduces a group in which the numbering of parentheses in each
709            alternative starts with the same number.
710    
711        (h) \h, \H, \v, and \V match horizontal and vertical whitespace.
712    
713     7. Added two new calls to pcre_fullinfo(): PCRE_INFO_OKPARTIAL and
714        PCRE_INFO_JCHANGED.
715    
716     8. A pattern such as  (.*(.)?)*  caused pcre_exec() to fail by either not
717        terminating or by crashing. Diagnosed by Viktor Griph; it was in the code
718        for detecting groups that can match an empty string.
719    
720     9. A pattern with a very large number of alternatives (more than several
721        hundred) was running out of internal workspace during the pre-compile
722        phase, where pcre_compile() figures out how much memory will be needed. A
723        bit of new cunning has reduced the workspace needed for groups with
724        alternatives. The 1000-alternative test pattern now uses 12 bytes of
725        workspace instead of running out of the 4096 that are available.
726    
727    10. Inserted some missing (unsigned int) casts to get rid of compiler warnings.
728    
729    11. Applied patch from Google to remove an optimization that didn't quite work.
730        The report of the bug said:
731    
732          pcrecpp::RE("a*").FullMatch("aaa") matches, while
733          pcrecpp::RE("a*?").FullMatch("aaa") does not, and
734          pcrecpp::RE("a*?\\z").FullMatch("aaa") does again.
735    
736    12. If \p or \P was used in non-UTF-8 mode on a character greater than 127
737        it matched the wrong number of bytes.
738    
739    
740    Version 7.1 24-Apr-07
741    ---------------------
742    
743     1. Applied Bob Rossi and Daniel G's patches to convert the build system to one
744        that is more "standard", making use of automake and other Autotools. There
745        is some re-arrangement of the files and adjustment of comments consequent
746        on this.
747    
748     2. Part of the patch fixed a problem with the pcregrep tests. The test of -r
749        for recursive directory scanning broke on some systems because the files
750        are not scanned in any specific order and on different systems the order
751        was different. A call to "sort" has been inserted into RunGrepTest for the
752        approprate test as a short-term fix. In the longer term there may be an
753        alternative.
754    
755     3. I had an email from Eric Raymond about problems translating some of PCRE's
756        man pages to HTML (despite the fact that I distribute HTML pages, some
757        people do their own conversions for various reasons). The problems
758        concerned the use of low-level troff macros .br and .in. I have therefore
759        removed all such uses from the man pages (some were redundant, some could
760        be replaced by .nf/.fi pairs). The 132html script that I use to generate
761        HTML has been updated to handle .nf/.fi and to complain if it encounters
762        .br or .in.
763    
764     4. Updated comments in configure.ac that get placed in config.h.in and also
765        arranged for config.h to be included in the distribution, with the name
766        config.h.generic, for the benefit of those who have to compile without
767        Autotools (compare pcre.h, which is now distributed as pcre.h.generic).
768    
769     5. Updated the support (such as it is) for Virtual Pascal, thanks to Stefan
770        Weber: (1) pcre_internal.h was missing some function renames; (2) updated
771        makevp.bat for the current PCRE, using the additional files
772        makevp_c.txt, makevp_l.txt, and pcregexp.pas.
773    
774     6. A Windows user reported a minor discrepancy with test 2, which turned out
775        to be caused by a trailing space on an input line that had got lost in his
776        copy. The trailing space was an accident, so I've just removed it.
777    
778     7. Add -Wl,-R... flags in pcre-config.in for *BSD* systems, as I'm told
779        that is needed.
780    
781     8. Mark ucp_table (in ucptable.h) and ucp_gentype (in pcre_ucp_searchfuncs.c)
782        as "const" (a) because they are and (b) because it helps the PHP
783        maintainers who have recently made a script to detect big data structures
784        in the php code that should be moved to the .rodata section. I remembered
785        to update Builducptable as well, so it won't revert if ucptable.h is ever
786        re-created.
787    
788     9. Added some extra #ifdef SUPPORT_UTF8 conditionals into pcretest.c,
789        pcre_printint.src, pcre_compile.c, pcre_study.c, and pcre_tables.c, in
790        order to be able to cut out the UTF-8 tables in the latter when UTF-8
791        support is not required. This saves 1.5-2K of code, which is important in
792        some applications.
793    
794        Later: more #ifdefs are needed in pcre_ord2utf8.c and pcre_valid_utf8.c
795        so as not to refer to the tables, even though these functions will never be
796        called when UTF-8 support is disabled. Otherwise there are problems with a
797        shared library.
798    
799    10. Fixed two bugs in the emulated memmove() function in pcre_internal.h:
800    
801        (a) It was defining its arguments as char * instead of void *.
802    
803        (b) It was assuming that all moves were upwards in memory; this was true
804            a long time ago when I wrote it, but is no longer the case.
805    
806        The emulated memove() is provided for those environments that have neither
807        memmove() nor bcopy(). I didn't think anyone used it these days, but that
808        is clearly not the case, as these two bugs were recently reported.
809    
810    11. The script PrepareRelease is now distributed: it calls 132html, CleanTxt,
811        and Detrail to create the HTML documentation, the .txt form of the man
812        pages, and it removes trailing spaces from listed files. It also creates
813        pcre.h.generic and config.h.generic from pcre.h and config.h. In the latter
814        case, it wraps all the #defines with #ifndefs. This script should be run
815        before "make dist".
816    
817    12. Fixed two fairly obscure bugs concerned with quantified caseless matching
818        with Unicode property support.
819    
820        (a) For a maximizing quantifier, if the two different cases of the
821            character were of different lengths in their UTF-8 codings (there are
822            some cases like this - I found 11), and the matching function had to
823            back up over a mixture of the two cases, it incorrectly assumed they
824            were both the same length.
825    
826        (b) When PCRE was configured to use the heap rather than the stack for
827            recursion during matching, it was not correctly preserving the data for
828            the other case of a UTF-8 character when checking ahead for a match
829            while processing a minimizing repeat. If the check also involved
830            matching a wide character, but failed, corruption could cause an
831            erroneous result when trying to check for a repeat of the original
832            character.
833    
834    13. Some tidying changes to the testing mechanism:
835    
836        (a) The RunTest script now detects the internal link size and whether there
837            is UTF-8 and UCP support by running ./pcretest -C instead of relying on
838            values substituted by "configure". (The RunGrepTest script already did
839            this for UTF-8.) The configure.ac script no longer substitutes the
840            relevant variables.
841    
842        (b) The debugging options /B and /D in pcretest show the compiled bytecode
843            with length and offset values. This means that the output is different
844            for different internal link sizes. Test 2 is skipped for link sizes
845            other than 2 because of this, bypassing the problem. Unfortunately,
846            there was also a test in test 3 (the locale tests) that used /B and
847            failed for link sizes other than 2. Rather than cut the whole test out,
848            I have added a new /Z option to pcretest that replaces the length and
849            offset values with spaces. This is now used to make test 3 independent
850            of link size. (Test 2 will be tidied up later.)
851    
852    14. If erroroffset was passed as NULL to pcre_compile, it provoked a
853        segmentation fault instead of returning the appropriate error message.
854    
855    15. In multiline mode when the newline sequence was set to "any", the pattern
856        ^$ would give a match between the \r and \n of a subject such as "A\r\nB".
857        This doesn't seem right; it now treats the CRLF combination as the line
858        ending, and so does not match in that case. It's only a pattern such as ^$
859        that would hit this one: something like ^ABC$ would have failed after \r
860        and then tried again after \r\n.
861    
862    16. Changed the comparison command for RunGrepTest from "diff -u" to "diff -ub"
863        in an attempt to make files that differ only in their line terminators
864        compare equal. This works on Linux.
865    
866    17. Under certain error circumstances pcregrep might try to free random memory
867        as it exited. This is now fixed, thanks to valgrind.
868    
869    19. In pcretest, if the pattern /(?m)^$/g<any> was matched against the string
870        "abc\r\n\r\n", it found an unwanted second match after the second \r. This
871        was because its rules for how to advance for /g after matching an empty
872        string at the end of a line did not allow for this case. They now check for
873        it specially.
874    
875    20. pcretest is supposed to handle patterns and data of any length, by
876        extending its buffers when necessary. It was getting this wrong when the
877        buffer for a data line had to be extended.
878    
879    21. Added PCRE_NEWLINE_ANYCRLF which is like ANY, but matches only CR, LF, or
880        CRLF as a newline sequence.
881    
882    22. Code for handling Unicode properties in pcre_dfa_exec() wasn't being cut
883        out by #ifdef SUPPORT_UCP. This did no harm, as it could never be used, but
884        I have nevertheless tidied it up.
885    
886    23. Added some casts to kill warnings from HP-UX ia64 compiler.
887    
888    24. Added a man page for pcre-config.
889    
890    
891    Version 7.0 19-Dec-06
892    ---------------------
893    
894     1. Fixed a signed/unsigned compiler warning in pcre_compile.c, shown up by
895        moving to gcc 4.1.1.
896    
897     2. The -S option for pcretest uses setrlimit(); I had omitted to #include
898        sys/time.h, which is documented as needed for this function. It doesn't
899        seem to matter on Linux, but it showed up on some releases of OS X.
900    
901     3. It seems that there are systems where bytes whose values are greater than
902        127 match isprint() in the "C" locale. The "C" locale should be the
903        default when a C program starts up. In most systems, only ASCII printing
904        characters match isprint(). This difference caused the output from pcretest
905        to vary, making some of the tests fail. I have changed pcretest so that:
906    
907        (a) When it is outputting text in the compiled version of a pattern, bytes
908            other than 32-126 are always shown as hex escapes.
909    
910        (b) When it is outputting text that is a matched part of a subject string,
911            it does the same, unless a different locale has been set for the match
912            (using the /L modifier). In this case, it uses isprint() to decide.
913    
914     4. Fixed a major bug that caused incorrect computation of the amount of memory
915        required for a compiled pattern when options that changed within the
916        pattern affected the logic of the preliminary scan that determines the
917        length. The relevant options are -x, and -i in UTF-8 mode. The result was
918        that the computed length was too small. The symptoms of this bug were
919        either the PCRE error "internal error: code overflow" from pcre_compile(),
920        or a glibc crash with a message such as "pcretest: free(): invalid next
921        size (fast)". Examples of patterns that provoked this bug (shown in
922        pcretest format) are:
923    
924          /(?-x: )/x
925          /(?x)(?-x: \s*#\s*)/
926          /((?i)[\x{c0}])/8
927          /(?i:[\x{c0}])/8
928    
929        HOWEVER: Change 17 below makes this fix obsolete as the memory computation
930        is now done differently.
931    
932     5. Applied patches from Google to: (a) add a QuoteMeta function to the C++
933        wrapper classes; (b) implement a new function in the C++ scanner that is
934        more efficient than the old way of doing things because it avoids levels of
935        recursion in the regex matching; (c) add a paragraph to the documentation
936        for the FullMatch() function.
937    
938     6. The escape sequence \n was being treated as whatever was defined as
939        "newline". Not only was this contrary to the documentation, which states
940        that \n is character 10 (hex 0A), but it also went horribly wrong when
941        "newline" was defined as CRLF. This has been fixed.
942    
943     7. In pcre_dfa_exec.c the value of an unsigned integer (the variable called c)
944        was being set to -1 for the "end of line" case (supposedly a value that no
945        character can have). Though this value is never used (the check for end of
946        line is "zero bytes in current character"), it caused compiler complaints.
947        I've changed it to 0xffffffff.
948    
949     8. In pcre_version.c, the version string was being built by a sequence of
950        C macros that, in the event of PCRE_PRERELEASE being defined as an empty
951        string (as it is for production releases) called a macro with an empty
952        argument. The C standard says the result of this is undefined. The gcc
953        compiler treats it as an empty string (which was what was wanted) but it is
954        reported that Visual C gives an error. The source has been hacked around to
955        avoid this problem.
956    
957     9. On the advice of a Windows user, included <io.h> and <fcntl.h> in Windows
958        builds of pcretest, and changed the call to _setmode() to use _O_BINARY
959        instead of 0x8000. Made all the #ifdefs test both _WIN32 and WIN32 (not all
960        of them did).
961    
962    10. Originally, pcretest opened its input and output without "b"; then I was
963        told that "b" was needed in some environments, so it was added for release
964        5.0 to both the input and output. (It makes no difference on Unix-like
965        systems.) Later I was told that it is wrong for the input on Windows. I've
966        now abstracted the modes into two macros, to make it easier to fiddle with
967        them, and removed "b" from the input mode under Windows.
968    
969    11. Added pkgconfig support for the C++ wrapper library, libpcrecpp.
970    
971    12. Added -help and --help to pcretest as an official way of being reminded
972        of the options.
973    
974    13. Removed some redundant semicolons after macro calls in pcrecpparg.h.in
975        and pcrecpp.cc because they annoy compilers at high warning levels.
976    
977    14. A bit of tidying/refactoring in pcre_exec.c in the main bumpalong loop.
978    
979    15. Fixed an occurrence of == in configure.ac that should have been = (shell
980        scripts are not C programs :-) and which was not noticed because it works
981        on Linux.
982    
983    16. pcretest is supposed to handle any length of pattern and data line (as one
984        line or as a continued sequence of lines) by extending its input buffer if
985        necessary. This feature was broken for very long pattern lines, leading to
986        a string of junk being passed to pcre_compile() if the pattern was longer
987        than about 50K.
988    
989    17. I have done a major re-factoring of the way pcre_compile() computes the
990        amount of memory needed for a compiled pattern. Previously, there was code
991        that made a preliminary scan of the pattern in order to do this. That was
992        OK when PCRE was new, but as the facilities have expanded, it has become
993        harder and harder to keep it in step with the real compile phase, and there
994        have been a number of bugs (see for example, 4 above). I have now found a
995        cunning way of running the real compile function in a "fake" mode that
996        enables it to compute how much memory it would need, while actually only
997        ever using a few hundred bytes of working memory and without too many
998        tests of the mode. This should make future maintenance and development
999        easier. A side effect of this work is that the limit of 200 on the nesting
1000        depth of parentheses has been removed (though this was never a serious
1001        limitation, I suspect). However, there is a downside: pcre_compile() now
1002        runs more slowly than before (30% or more, depending on the pattern). I
1003        hope this isn't a big issue. There is no effect on runtime performance.
1004    
1005    18. Fixed a minor bug in pcretest: if a pattern line was not terminated by a
1006        newline (only possible for the last line of a file) and it was a
1007        pattern that set a locale (followed by /Lsomething), pcretest crashed.
1008    
1009    19. Added additional timing features to pcretest. (1) The -tm option now times
1010        matching only, not compiling. (2) Both -t and -tm can be followed, as a
1011        separate command line item, by a number that specifies the number of
1012        repeats to use when timing. The default is 50000; this gives better
1013        precision, but takes uncomfortably long for very large patterns.
1014    
1015    20. Extended pcre_study() to be more clever in cases where a branch of a
1016        subpattern has no definite first character. For example, (a*|b*)[cd] would
1017        previously give no result from pcre_study(). Now it recognizes that the
1018        first character must be a, b, c, or d.
1019    
1020    21. There was an incorrect error "recursive call could loop indefinitely" if
1021        a subpattern (or the entire pattern) that was being tested for matching an
1022        empty string contained only one non-empty item after a nested subpattern.
1023        For example, the pattern (?>\x{100}*)\d(?R) provoked this error
1024        incorrectly, because the \d was being skipped in the check.
1025    
1026    22. The pcretest program now has a new pattern option /B and a command line
1027        option -b, which is equivalent to adding /B to every pattern. This causes
1028        it to show the compiled bytecode, without the additional information that
1029        -d shows. The effect of -d is now the same as -b with -i (and similarly, /D
1030        is the same as /B/I).
1031    
1032    23. A new optimization is now able automatically to treat some sequences such
1033        as a*b as a*+b. More specifically, if something simple (such as a character
1034        or a simple class like \d) has an unlimited quantifier, and is followed by
1035        something that cannot possibly match the quantified thing, the quantifier
1036        is automatically "possessified".
1037    
1038    24. A recursive reference to a subpattern whose number was greater than 39
1039        went wrong under certain circumstances in UTF-8 mode. This bug could also
1040        have affected the operation of pcre_study().
1041    
1042    25. Realized that a little bit of performance could be had by replacing
1043        (c & 0xc0) == 0xc0 with c >= 0xc0 when processing UTF-8 characters.
1044    
1045    26. Timing data from pcretest is now shown to 4 decimal places instead of 3.
1046    
1047    27. Possessive quantifiers such as a++ were previously implemented by turning
1048        them into atomic groups such as ($>a+). Now they have their own opcodes,
1049        which improves performance. This includes the automatically created ones
1050        from 23 above.
1051    
1052    28. A pattern such as (?=(\w+))\1: which simulates an atomic group using a
1053        lookahead was broken if it was not anchored. PCRE was mistakenly expecting
1054        the first matched character to be a colon. This applied both to named and
1055        numbered groups.
1056    
1057    29. The ucpinternal.h header file was missing its idempotency #ifdef.
1058    
1059    30. I was sent a "project" file called libpcre.a.dev which I understand makes
1060        building PCRE on Windows easier, so I have included it in the distribution.
1061    
1062    31. There is now a check in pcretest against a ridiculously large number being
1063        returned by pcre_exec() or pcre_dfa_exec(). If this happens in a /g or /G
1064        loop, the loop is abandoned.
1065    
1066    32. Forward references to subpatterns in conditions such as (?(2)...) where
1067        subpattern 2 is defined later cause pcre_compile() to search forwards in
1068        the pattern for the relevant set of parentheses. This search went wrong
1069        when there were unescaped parentheses in a character class, parentheses
1070        escaped with \Q...\E, or parentheses in a #-comment in /x mode.
1071    
1072    33. "Subroutine" calls and backreferences were previously restricted to
1073        referencing subpatterns earlier in the regex. This restriction has now
1074        been removed.
1075    
1076    34. Added a number of extra features that are going to be in Perl 5.10. On the
1077        whole, these are just syntactic alternatives for features that PCRE had
1078        previously implemented using the Python syntax or my own invention. The
1079        other formats are all retained for compatibility.
1080    
1081        (a) Named groups can now be defined as (?<name>...) or (?'name'...) as well
1082            as (?P<name>...). The new forms, as well as being in Perl 5.10, are
1083            also .NET compatible.
1084    
1085        (b) A recursion or subroutine call to a named group can now be defined as
1086            (?&name) as well as (?P>name).
1087    
1088        (c) A backreference to a named group can now be defined as \k<name> or
1089            \k'name' as well as (?P=name). The new forms, as well as being in Perl
1090            5.10, are also .NET compatible.
1091    
1092        (d) A conditional reference to a named group can now use the syntax
1093            (?(<name>) or (?('name') as well as (?(name).
1094    
1095        (e) A "conditional group" of the form (?(DEFINE)...) can be used to define
1096            groups (named and numbered) that are never evaluated inline, but can be
1097            called as "subroutines" from elsewhere. In effect, the DEFINE condition
1098            is always false. There may be only one alternative in such a group.
1099    
1100        (f) A test for recursion can be given as (?(R1).. or (?(R&name)... as well
1101            as the simple (?(R). The condition is true only if the most recent
1102            recursion is that of the given number or name. It does not search out
1103            through the entire recursion stack.
1104    
1105        (g) The escape \gN or \g{N} has been added, where N is a positive or
1106            negative number, specifying an absolute or relative reference.
1107    
1108    35. Tidied to get rid of some further signed/unsigned compiler warnings and
1109        some "unreachable code" warnings.
1110    
1111    36. Updated the Unicode property tables to Unicode version 5.0.0. Amongst other
1112        things, this adds five new scripts.
1113    
1114    37. Perl ignores orphaned \E escapes completely. PCRE now does the same.
1115        There were also incompatibilities regarding the handling of \Q..\E inside
1116        character classes, for example with patterns like [\Qa\E-\Qz\E] where the
1117        hyphen was adjacent to \Q or \E. I hope I've cleared all this up now.
1118    
1119    38. Like Perl, PCRE detects when an indefinitely repeated parenthesized group
1120        matches an empty string, and forcibly breaks the loop. There were bugs in
1121        this code in non-simple cases. For a pattern such as  ^(a()*)*  matched
1122        against  aaaa  the result was just "a" rather than "aaaa", for example. Two
1123        separate and independent bugs (that affected different cases) have been
1124        fixed.
1125    
1126    39. Refactored the code to abolish the use of different opcodes for small
1127        capturing bracket numbers. This is a tidy that I avoided doing when I
1128        removed the limit on the number of capturing brackets for 3.5 back in 2001.
1129        The new approach is not only tidier, it makes it possible to reduce the
1130        memory needed to fix the previous bug (38).
1131    
1132    40. Implemented PCRE_NEWLINE_ANY to recognize any of the Unicode newline
1133        sequences (http://unicode.org/unicode/reports/tr18/) as "newline" when
1134        processing dot, circumflex, or dollar metacharacters, or #-comments in /x
1135        mode.
1136    
1137    41. Add \R to match any Unicode newline sequence, as suggested in the Unicode
1138        report.
1139    
1140    42. Applied patch, originally from Ari Pollak, modified by Google, to allow
1141        copy construction and assignment in the C++ wrapper.
1142    
1143    43. Updated pcregrep to support "--newline=any". In the process, I fixed a
1144        couple of bugs that could have given wrong results in the "--newline=crlf"
1145        case.
1146    
1147    44. Added a number of casts and did some reorganization of signed/unsigned int
1148        variables following suggestions from Dair Grant. Also renamed the variable
1149        "this" as "item" because it is a C++ keyword.
1150    
1151    45. Arranged for dftables to add
1152    
1153          #include "pcre_internal.h"
1154    
1155        to pcre_chartables.c because without it, gcc 4.x may remove the array
1156        definition from the final binary if PCRE is built into a static library and
1157        dead code stripping is activated.
1158    
1159    46. For an unanchored pattern, if a match attempt fails at the start of a
1160        newline sequence, and the newline setting is CRLF or ANY, and the next two
1161        characters are CRLF, advance by two characters instead of one.
1162    
1163    
1164    Version 6.7 04-Jul-06
1165    ---------------------
1166    
1167     1. In order to handle tests when input lines are enormously long, pcretest has
1168        been re-factored so that it automatically extends its buffers when
1169        necessary. The code is crude, but this _is_ just a test program. The
1170        default size has been increased from 32K to 50K.
1171    
1172     2. The code in pcre_study() was using the value of the re argument before
1173        testing it for NULL. (Of course, in any sensible call of the function, it
1174        won't be NULL.)
1175    
1176     3. The memmove() emulation function in pcre_internal.h, which is used on
1177        systems that lack both memmove() and bcopy() - that is, hardly ever -
1178        was missing a "static" storage class specifier.
1179    
1180     4. When UTF-8 mode was not set, PCRE looped when compiling certain patterns
1181        containing an extended class (one that cannot be represented by a bitmap
1182        because it contains high-valued characters or Unicode property items, e.g.
1183        [\pZ]). Almost always one would set UTF-8 mode when processing such a
1184        pattern, but PCRE should not loop if you do not (it no longer does).
1185        [Detail: two cases were found: (a) a repeated subpattern containing an
1186        extended class; (b) a recursive reference to a subpattern that followed a
1187        previous extended class. It wasn't skipping over the extended class
1188        correctly when UTF-8 mode was not set.]
1189    
1190     5. A negated single-character class was not being recognized as fixed-length
1191        in lookbehind assertions such as (?<=[^f]), leading to an incorrect
1192        compile error "lookbehind assertion is not fixed length".
1193    
1194     6. The RunPerlTest auxiliary script was showing an unexpected difference
1195        between PCRE and Perl for UTF-8 tests. It turns out that it is hard to
1196        write a Perl script that can interpret lines of an input file either as
1197        byte characters or as UTF-8, which is what "perltest" was being required to
1198        do for the non-UTF-8 and UTF-8 tests, respectively. Essentially what you
1199        can't do is switch easily at run time between having the "use utf8;" pragma
1200        or not. In the end, I fudged it by using the RunPerlTest script to insert
1201        "use utf8;" explicitly for the UTF-8 tests.
1202    
1203     7. In multiline (/m) mode, PCRE was matching ^ after a terminating newline at
1204        the end of the subject string, contrary to the documentation and to what
1205        Perl does. This was true of both matching functions. Now it matches only at
1206        the start of the subject and immediately after *internal* newlines.
1207    
1208     8. A call of pcre_fullinfo() from pcretest to get the option bits was passing
1209        a pointer to an int instead of a pointer to an unsigned long int. This
1210        caused problems on 64-bit systems.
1211    
1212     9. Applied a patch from the folks at Google to pcrecpp.cc, to fix "another
1213        instance of the 'standard' template library not being so standard".
1214    
1215    10. There was no check on the number of named subpatterns nor the maximum
1216        length of a subpattern name. The product of these values is used to compute
1217        the size of the memory block for a compiled pattern. By supplying a very
1218        long subpattern name and a large number of named subpatterns, the size
1219        computation could be caused to overflow. This is now prevented by limiting
1220        the length of names to 32 characters, and the number of named subpatterns
1221        to 10,000.
1222    
1223    11. Subpatterns that are repeated with specific counts have to be replicated in
1224        the compiled pattern. The size of memory for this was computed from the
1225        length of the subpattern and the repeat count. The latter is limited to
1226        65535, but there was no limit on the former, meaning that integer overflow
1227        could in principle occur. The compiled length of a repeated subpattern is
1228        now limited to 30,000 bytes in order to prevent this.
1229    
1230    12. Added the optional facility to have named substrings with the same name.
1231    
1232    13. Added the ability to use a named substring as a condition, using the
1233        Python syntax: (?(name)yes|no). This overloads (?(R)... and names that
1234        are numbers (not recommended). Forward references are permitted.
1235    
1236    14. Added forward references in named backreferences (if you see what I mean).
1237    
1238    15. In UTF-8 mode, with the PCRE_DOTALL option set, a quantified dot in the
1239        pattern could run off the end of the subject. For example, the pattern
1240        "(?s)(.{1,5})"8 did this with the subject "ab".
1241    
1242    16. If PCRE_DOTALL or PCRE_MULTILINE were set, pcre_dfa_exec() behaved as if
1243        PCRE_CASELESS was set when matching characters that were quantified with ?
1244        or *.
1245    
1246    17. A character class other than a single negated character that had a minimum
1247        but no maximum quantifier - for example [ab]{6,} - was not handled
1248        correctly by pce_dfa_exec(). It would match only one character.
1249    
1250    18. A valid (though odd) pattern that looked like a POSIX character
1251        class but used an invalid character after [ (for example [[,abc,]]) caused
1252        pcre_compile() to give the error "Failed: internal error: code overflow" or
1253        in some cases to crash with a glibc free() error. This could even happen if
1254        the pattern terminated after [[ but there just happened to be a sequence of
1255        letters, a binary zero, and a closing ] in the memory that followed.
1256    
1257    19. Perl's treatment of octal escapes in the range \400 to \777 has changed
1258        over the years. Originally (before any Unicode support), just the bottom 8
1259        bits were taken. Thus, for example, \500 really meant \100. Nowadays the
1260        output from "man perlunicode" includes this:
1261    
1262          The regular expression compiler produces polymorphic opcodes.  That
1263          is, the pattern adapts to the data and automatically switches to
1264          the Unicode character scheme when presented with Unicode data--or
1265          instead uses a traditional byte scheme when presented with byte
1266          data.
1267    
1268        Sadly, a wide octal escape does not cause a switch, and in a string with
1269        no other multibyte characters, these octal escapes are treated as before.
1270        Thus, in Perl, the pattern  /\500/ actually matches \100 but the pattern
1271        /\500|\x{1ff}/ matches \500 or \777 because the whole thing is treated as a
1272        Unicode string.
1273    
1274        I have not perpetrated such confusion in PCRE. Up till now, it took just
1275        the bottom 8 bits, as in old Perl. I have now made octal escapes with
1276        values greater than \377 illegal in non-UTF-8 mode. In UTF-8 mode they
1277        translate to the appropriate multibyte character.
1278    
1279    29. Applied some refactoring to reduce the number of warnings from Microsoft
1280        and Borland compilers. This has included removing the fudge introduced
1281        seven years ago for the OS/2 compiler (see 2.02/2 below) because it caused
1282        a warning about an unused variable.
1283    
1284    21. PCRE has not included VT (character 0x0b) in the set of whitespace
1285        characters since release 4.0, because Perl (from release 5.004) does not.
1286        [Or at least, is documented not to: some releases seem to be in conflict
1287        with the documentation.] However, when a pattern was studied with
1288        pcre_study() and all its branches started with \s, PCRE still included VT
1289        as a possible starting character. Of course, this did no harm; it just
1290        caused an unnecessary match attempt.
1291    
1292    22. Removed a now-redundant internal flag bit that recorded the fact that case
1293        dependency changed within the pattern. This was once needed for "required
1294        byte" processing, but is no longer used. This recovers a now-scarce options
1295        bit. Also moved the least significant internal flag bit to the most-
1296        significant bit of the word, which was not previously used (hangover from
1297        the days when it was an int rather than a uint) to free up another bit for
1298        the future.
1299    
1300    23. Added support for CRLF line endings as well as CR and LF. As well as the
1301        default being selectable at build time, it can now be changed at runtime
1302        via the PCRE_NEWLINE_xxx flags. There are now options for pcregrep to
1303        specify that it is scanning data with non-default line endings.
1304    
1305    24. Changed the definition of CXXLINK to make it agree with the definition of
1306        LINK in the Makefile, by replacing LDFLAGS to CXXFLAGS.
1307    
1308    25. Applied Ian Taylor's patches to avoid using another stack frame for tail
1309        recursions. This makes a big different to stack usage for some patterns.
1310    
1311    26. If a subpattern containing a named recursion or subroutine reference such
1312        as (?P>B) was quantified, for example (xxx(?P>B)){3}, the calculation of
1313        the space required for the compiled pattern went wrong and gave too small a
1314        value. Depending on the environment, this could lead to "Failed: internal
1315        error: code overflow at offset 49" or "glibc detected double free or
1316        corruption" errors.
1317    
1318    27. Applied patches from Google (a) to support the new newline modes and (b) to
1319        advance over multibyte UTF-8 characters in GlobalReplace.
1320    
1321    28. Change free() to pcre_free() in pcredemo.c. Apparently this makes a
1322        difference for some implementation of PCRE in some Windows version.
1323    
1324    29. Added some extra testing facilities to pcretest:
1325    
1326        \q<number>   in a data line sets the "match limit" value
1327        \Q<number>   in a data line sets the "match recursion limt" value
1328        -S <number>  sets the stack size, where <number> is in megabytes
1329    
1330        The -S option isn't available for Windows.
1331    
1332    
1333    Version 6.6 06-Feb-06
1334    ---------------------
1335    
1336     1. Change 16(a) for 6.5 broke things, because PCRE_DATA_SCOPE was not defined
1337        in pcreposix.h. I have copied the definition from pcre.h.
1338    
1339     2. Change 25 for 6.5 broke compilation in a build directory out-of-tree
1340        because pcre.h is no longer a built file.
1341    
1342     3. Added Jeff Friedl's additional debugging patches to pcregrep. These are
1343        not normally included in the compiled code.
1344    
1345    
1346  Version 6.5 01-Feb-06  Version 6.5 01-Feb-06
1347  ---------------------  ---------------------
1348    

Legend:
Removed from v.87  
changed lines
  Added in v.386

webmaster@exim.org
ViewVC Help
Powered by ViewVC 1.1.12