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

Legend:
Removed from v.87  
changed lines
  Added in v.346

webmaster@exim.org
ViewVC Help
Powered by ViewVC 1.1.12