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

Legend:
Removed from v.87  
changed lines
  Added in v.351

  ViewVC Help
Powered by ViewVC 1.1.5