/[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 219 by ph10, Thu Aug 16 11:46:40 2007 UTC
# Line 1  Line 1 
1  ChangeLog for PCRE  ChangeLog for PCRE
2  ------------------  ------------------
3    
4    Version 7.3 09-Aug-07
5    ---------------------
6    
7     1. In the rejigging of the build system that eventually resulted in 7.1, the
8        line "#include <pcre.h>" was included in pcre_internal.h. The use of angle
9        brackets there is not right, since it causes compilers to look for an
10        installed pcre.h, not the version that is in the source that is being
11        compiled (which of course may be different). I have changed it back to:
12    
13          #include "pcre.h"
14    
15        I have a vague recollection that the change was concerned with compiling in
16        different directories, but in the new build system, that is taken care of
17        by the VPATH setting the Makefile.
18    
19     2. The pattern .*$ when run in not-DOTALL UTF-8 mode with newline=any failed
20        when the subject happened to end in the byte 0x85 (e.g. if the last
21        character was \x{1ec5}). *Character* 0x85 is one of the "any" newline
22        characters but of course it shouldn't be taken as a newline when it is part
23        of another character. The bug was that, for an unlimited repeat of . in
24        not-DOTALL UTF-8 mode, PCRE was advancing by bytes rather than by
25        characters when looking for a newline.
26    
27     3. A small performance improvement in the DOTALL UTF-8 mode .* case.
28    
29     4. Debugging: adjusted the names of opcodes for different kinds of parentheses
30        in debug output.
31    
32     5. Arrange to use "%I64d" instead of "%lld" and "%I64u" instead of "%llu" for
33        long printing in the pcrecpp unittest when running under MinGW.
34    
35     6. ESC_K was left out of the EBCDIC table.
36    
37     7. Change 7.0/38 introduced a new limit on the number of nested non-capturing
38        parentheses; I made it 1000, which seemed large enough. Unfortunately, the
39        limit also applies to "virtual nesting" when a pattern is recursive, and in
40        this case 1000 isn't so big. I have been able to remove this limit at the
41        expense of backing off one optimization in certain circumstances. Normally,
42        when pcre_exec() would call its internal match() function recursively and
43        immediately return the result unconditionally, it uses a "tail recursion"
44        feature to save stack. However, when a subpattern that can match an empty
45        string has an unlimited repetition quantifier, it no longer makes this
46        optimization. That gives it a stack frame in which to save the data for
47        checking that an empty string has been matched. Previously this was taken
48        from the 1000-entry workspace that had been reserved. So now there is no
49        explicit limit, but more stack is used.
50    
51     8. Applied Daniel's patches to solve problems with the import/export magic
52        syntax that is required for Windows, and which was going wrong for the
53        pcreposix and pcrecpp parts of the library. These were overlooked when this
54        problem was solved for the main library.
55    
56     9. There were some crude static tests to avoid integer overflow when computing
57        the size of patterns that contain repeated groups with explicit upper
58        limits. As the maximum quantifier is 65535, the maximum group length was
59        set at 30,000 so that the product of these two numbers did not overflow a
60        32-bit integer. However, it turns out that people want to use groups that
61        are longer than 30,000 bytes (though not repeat them that many times).
62        Change 7.0/17 (the refactoring of the way the pattern size is computed) has
63        made it possible to implement the integer overflow checks in a much more
64        dynamic way, which I have now done. The artificial limitation on group
65        length has been removed - we now have only the limit on the total length of
66        the compiled pattern, which depends on the LINK_SIZE setting.
67    
68    10. Fixed a bug in the documentation for get/copy named substring when
69        duplicate names are permitted. If none of the named substrings are set, the
70        functions return PCRE_ERROR_NOSUBSTRING (7); the doc said they returned an
71        empty string.
72    
73    11. Because Perl interprets \Q...\E at a high level, and ignores orphan \E
74        instances, patterns such as [\Q\E] or [\E] or even [^\E] cause an error,
75        because the ] is interpreted as the first data character and the
76        terminating ] is not found. PCRE has been made compatible with Perl in this
77        regard. Previously, it interpreted [\Q\E] as an empty class, and [\E] could
78        cause memory overwriting.
79    
80    10. Like Perl, PCRE automatically breaks an unlimited repeat after an empty
81        string has been matched (to stop an infinite loop). It was not recognizing
82        a conditional subpattern that could match an empty string if that
83        subpattern was within another subpattern. For example, it looped when
84        trying to match  (((?(1)X|))*)  but it was OK with  ((?(1)X|)*)  where the
85        condition was not nested. This bug has been fixed.
86    
87    12. A pattern like \X?\d or \P{L}?\d in non-UTF-8 mode could cause a backtrack
88        past the start of the subject in the presence of bytes with the top bit
89        set, for example "\x8aBCD".
90    
91    13. Added Perl 5.10 experimental backtracking controls (*FAIL), (*F), (*PRUNE),
92        (*SKIP), (*THEN), (*COMMIT), and (*ACCEPT).
93    
94    14. Optimized (?!) to (*FAIL).
95    
96    15. Updated the test for a valid UTF-8 string to conform to the later RFC 3629.
97        This restricts code points to be within the range 0 to 0x10FFFF, excluding
98        the "low surrogate" sequence 0xD800 to 0xDFFF. Previously, PCRE allowed the
99        full range 0 to 0x7FFFFFFF, as defined by RFC 2279. Internally, it still
100        does: it's just the validity check that is more restrictive.
101    
102    16. Inserted checks for integer overflows during escape sequence (backslash)
103        processing, and also fixed erroneous offset values for syntax errors during
104        backslash processing.
105    
106    17. Fixed another case of looking too far back in non-UTF-8 mode (cf 12 above)
107        for patterns like [\PPP\x8a]{1,}\x80 with the subject "A\x80".
108    
109    18. An unterminated class in a pattern like (?1)\c[ with a "forward reference"
110        caused an overrun.
111    
112    19. A pattern like (?:[\PPa*]*){8,} which had an "extended class" (one with
113        something other than just ASCII characters) inside a group that had an
114        unlimited repeat caused a loop at compile time (while checking to see
115        whether the group could match an empty string).
116    
117    20. Debugging a pattern containing \p or \P could cause a crash. For example,
118        [\P{Any}] did so. (Error in the code for printing property names.)
119    
120    21. An orphan \E inside a character class could cause a crash.
121    
122    22. A repeated capturing bracket such as (A)? could cause a wild memory
123        reference during compilation.
124    
125    23. There are several functions in pcre_compile() that scan along a compiled
126        expression for various reasons (e.g. to see if it's fixed length for look
127        behind). There were bugs in these functions when a repeated \p or \P was
128        present in the pattern. These operators have additional parameters compared
129        with \d, etc, and these were not being taken into account when moving along
130        the compiled data. Specifically:
131    
132        (a) A item such as \p{Yi}{3} in a lookbehind was not treated as fixed
133            length.
134    
135        (b) An item such as \pL+ within a repeated group could cause crashes or
136            loops.
137    
138        (c) A pattern such as \p{Yi}+(\P{Yi}+)(?1) could give an incorrect
139            "reference to non-existent subpattern" error.
140    
141    24. A repeated \S or \W in UTF-8 mode could give wrong answers when multibyte
142        characters were involved (for example /\S{2}/8g with "A\x{a3}BC").
143    
144    
145    Version 7.2 19-Jun-07
146    ---------------------
147    
148     1. If the fr_FR locale cannot be found for test 3, try the "french" locale,
149        which is apparently normally available under Windows.
150    
151     2. Re-jig the pcregrep tests with different newline settings in an attempt
152        to make them independent of the local environment's newline setting.
153    
154     3. Add code to configure.ac to remove -g from the CFLAGS default settings.
155    
156     4. Some of the "internals" tests were previously cut out when the link size
157        was not 2, because the output contained actual offsets. The recent new
158        "Z" feature of pcretest means that these can be cut out, making the tests
159        usable with all link sizes.
160    
161     5. Implemented Stan Switzer's goto replacement for longjmp() when not using
162        stack recursion. This gives a massive performance boost under BSD, but just
163        a small improvement under Linux. However, it saves one field in the frame
164        in all cases.
165    
166     6. Added more features from the forthcoming Perl 5.10:
167    
168        (a) (?-n) (where n is a string of digits) is a relative subroutine or
169            recursion call. It refers to the nth most recently opened parentheses.
170    
171        (b) (?+n) is also a relative subroutine call; it refers to the nth next
172            to be opened parentheses.
173    
174        (c) Conditions that refer to capturing parentheses can be specified
175            relatively, for example, (?(-2)... or (?(+3)...
176    
177        (d) \K resets the start of the current match so that everything before
178            is not part of it.
179    
180        (e) \k{name} is synonymous with \k<name> and \k'name' (.NET compatible).
181    
182        (f) \g{name} is another synonym - part of Perl 5.10's unification of
183            reference syntax.
184    
185        (g) (?| introduces a group in which the numbering of parentheses in each
186            alternative starts with the same number.
187    
188        (h) \h, \H, \v, and \V match horizontal and vertical whitespace.
189    
190     7. Added two new calls to pcre_fullinfo(): PCRE_INFO_OKPARTIAL and
191        PCRE_INFO_JCHANGED.
192    
193     8. A pattern such as  (.*(.)?)*  caused pcre_exec() to fail by either not
194        terminating or by crashing. Diagnosed by Viktor Griph; it was in the code
195        for detecting groups that can match an empty string.
196    
197     9. A pattern with a very large number of alternatives (more than several
198        hundred) was running out of internal workspace during the pre-compile
199        phase, where pcre_compile() figures out how much memory will be needed. A
200        bit of new cunning has reduced the workspace needed for groups with
201        alternatives. The 1000-alternative test pattern now uses 12 bytes of
202        workspace instead of running out of the 4096 that are available.
203    
204    10. Inserted some missing (unsigned int) casts to get rid of compiler warnings.
205    
206    11. Applied patch from Google to remove an optimization that didn't quite work.
207        The report of the bug said:
208    
209          pcrecpp::RE("a*").FullMatch("aaa") matches, while
210          pcrecpp::RE("a*?").FullMatch("aaa") does not, and
211          pcrecpp::RE("a*?\\z").FullMatch("aaa") does again.
212    
213    12. If \p or \P was used in non-UTF-8 mode on a character greater than 127
214        it matched the wrong number of bytes.
215    
216    
217    Version 7.1 24-Apr-07
218    ---------------------
219    
220     1. Applied Bob Rossi and Daniel G's patches to convert the build system to one
221        that is more "standard", making use of automake and other Autotools. There
222        is some re-arrangement of the files and adjustment of comments consequent
223        on this.
224    
225     2. Part of the patch fixed a problem with the pcregrep tests. The test of -r
226        for recursive directory scanning broke on some systems because the files
227        are not scanned in any specific order and on different systems the order
228        was different. A call to "sort" has been inserted into RunGrepTest for the
229        approprate test as a short-term fix. In the longer term there may be an
230        alternative.
231    
232     3. I had an email from Eric Raymond about problems translating some of PCRE's
233        man pages to HTML (despite the fact that I distribute HTML pages, some
234        people do their own conversions for various reasons). The problems
235        concerned the use of low-level troff macros .br and .in. I have therefore
236        removed all such uses from the man pages (some were redundant, some could
237        be replaced by .nf/.fi pairs). The 132html script that I use to generate
238        HTML has been updated to handle .nf/.fi and to complain if it encounters
239        .br or .in.
240    
241     4. Updated comments in configure.ac that get placed in config.h.in and also
242        arranged for config.h to be included in the distribution, with the name
243        config.h.generic, for the benefit of those who have to compile without
244        Autotools (compare pcre.h, which is now distributed as pcre.h.generic).
245    
246     5. Updated the support (such as it is) for Virtual Pascal, thanks to Stefan
247        Weber: (1) pcre_internal.h was missing some function renames; (2) updated
248        makevp.bat for the current PCRE, using the additional files
249        makevp_c.txt, makevp_l.txt, and pcregexp.pas.
250    
251     6. A Windows user reported a minor discrepancy with test 2, which turned out
252        to be caused by a trailing space on an input line that had got lost in his
253        copy. The trailing space was an accident, so I've just removed it.
254    
255     7. Add -Wl,-R... flags in pcre-config.in for *BSD* systems, as I'm told
256        that is needed.
257    
258     8. Mark ucp_table (in ucptable.h) and ucp_gentype (in pcre_ucp_searchfuncs.c)
259        as "const" (a) because they are and (b) because it helps the PHP
260        maintainers who have recently made a script to detect big data structures
261        in the php code that should be moved to the .rodata section. I remembered
262        to update Builducptable as well, so it won't revert if ucptable.h is ever
263        re-created.
264    
265     9. Added some extra #ifdef SUPPORT_UTF8 conditionals into pcretest.c,
266        pcre_printint.src, pcre_compile.c, pcre_study.c, and pcre_tables.c, in
267        order to be able to cut out the UTF-8 tables in the latter when UTF-8
268        support is not required. This saves 1.5-2K of code, which is important in
269        some applications.
270    
271        Later: more #ifdefs are needed in pcre_ord2utf8.c and pcre_valid_utf8.c
272        so as not to refer to the tables, even though these functions will never be
273        called when UTF-8 support is disabled. Otherwise there are problems with a
274        shared library.
275    
276    10. Fixed two bugs in the emulated memmove() function in pcre_internal.h:
277    
278        (a) It was defining its arguments as char * instead of void *.
279    
280        (b) It was assuming that all moves were upwards in memory; this was true
281            a long time ago when I wrote it, but is no longer the case.
282    
283        The emulated memove() is provided for those environments that have neither
284        memmove() nor bcopy(). I didn't think anyone used it these days, but that
285        is clearly not the case, as these two bugs were recently reported.
286    
287    11. The script PrepareRelease is now distributed: it calls 132html, CleanTxt,
288        and Detrail to create the HTML documentation, the .txt form of the man
289        pages, and it removes trailing spaces from listed files. It also creates
290        pcre.h.generic and config.h.generic from pcre.h and config.h. In the latter
291        case, it wraps all the #defines with #ifndefs. This script should be run
292        before "make dist".
293    
294    12. Fixed two fairly obscure bugs concerned with quantified caseless matching
295        with Unicode property support.
296    
297        (a) For a maximizing quantifier, if the two different cases of the
298            character were of different lengths in their UTF-8 codings (there are
299            some cases like this - I found 11), and the matching function had to
300            back up over a mixture of the two cases, it incorrectly assumed they
301            were both the same length.
302    
303        (b) When PCRE was configured to use the heap rather than the stack for
304            recursion during matching, it was not correctly preserving the data for
305            the other case of a UTF-8 character when checking ahead for a match
306            while processing a minimizing repeat. If the check also involved
307            matching a wide character, but failed, corruption could cause an
308            erroneous result when trying to check for a repeat of the original
309            character.
310    
311    13. Some tidying changes to the testing mechanism:
312    
313        (a) The RunTest script now detects the internal link size and whether there
314            is UTF-8 and UCP support by running ./pcretest -C instead of relying on
315            values substituted by "configure". (The RunGrepTest script already did
316            this for UTF-8.) The configure.ac script no longer substitutes the
317            relevant variables.
318    
319        (b) The debugging options /B and /D in pcretest show the compiled bytecode
320            with length and offset values. This means that the output is different
321            for different internal link sizes. Test 2 is skipped for link sizes
322            other than 2 because of this, bypassing the problem. Unfortunately,
323            there was also a test in test 3 (the locale tests) that used /B and
324            failed for link sizes other than 2. Rather than cut the whole test out,
325            I have added a new /Z option to pcretest that replaces the length and
326            offset values with spaces. This is now used to make test 3 independent
327            of link size. (Test 2 will be tidied up later.)
328    
329    14. If erroroffset was passed as NULL to pcre_compile, it provoked a
330        segmentation fault instead of returning the appropriate error message.
331    
332    15. In multiline mode when the newline sequence was set to "any", the pattern
333        ^$ would give a match between the \r and \n of a subject such as "A\r\nB".
334        This doesn't seem right; it now treats the CRLF combination as the line
335        ending, and so does not match in that case. It's only a pattern such as ^$
336        that would hit this one: something like ^ABC$ would have failed after \r
337        and then tried again after \r\n.
338    
339    16. Changed the comparison command for RunGrepTest from "diff -u" to "diff -ub"
340        in an attempt to make files that differ only in their line terminators
341        compare equal. This works on Linux.
342    
343    17. Under certain error circumstances pcregrep might try to free random memory
344        as it exited. This is now fixed, thanks to valgrind.
345    
346    19. In pcretest, if the pattern /(?m)^$/g<any> was matched against the string
347        "abc\r\n\r\n", it found an unwanted second match after the second \r. This
348        was because its rules for how to advance for /g after matching an empty
349        string at the end of a line did not allow for this case. They now check for
350        it specially.
351    
352    20. pcretest is supposed to handle patterns and data of any length, by
353        extending its buffers when necessary. It was getting this wrong when the
354        buffer for a data line had to be extended.
355    
356    21. Added PCRE_NEWLINE_ANYCRLF which is like ANY, but matches only CR, LF, or
357        CRLF as a newline sequence.
358    
359    22. Code for handling Unicode properties in pcre_dfa_exec() wasn't being cut
360        out by #ifdef SUPPORT_UCP. This did no harm, as it could never be used, but
361        I have nevertheless tidied it up.
362    
363    23. Added some casts to kill warnings from HP-UX ia64 compiler.
364    
365    24. Added a man page for pcre-config.
366    
367    
368    Version 7.0 19-Dec-06
369    ---------------------
370    
371     1. Fixed a signed/unsigned compiler warning in pcre_compile.c, shown up by
372        moving to gcc 4.1.1.
373    
374     2. The -S option for pcretest uses setrlimit(); I had omitted to #include
375        sys/time.h, which is documented as needed for this function. It doesn't
376        seem to matter on Linux, but it showed up on some releases of OS X.
377    
378     3. It seems that there are systems where bytes whose values are greater than
379        127 match isprint() in the "C" locale. The "C" locale should be the
380        default when a C program starts up. In most systems, only ASCII printing
381        characters match isprint(). This difference caused the output from pcretest
382        to vary, making some of the tests fail. I have changed pcretest so that:
383    
384        (a) When it is outputting text in the compiled version of a pattern, bytes
385            other than 32-126 are always shown as hex escapes.
386    
387        (b) When it is outputting text that is a matched part of a subject string,
388            it does the same, unless a different locale has been set for the match
389            (using the /L modifier). In this case, it uses isprint() to decide.
390    
391     4. Fixed a major bug that caused incorrect computation of the amount of memory
392        required for a compiled pattern when options that changed within the
393        pattern affected the logic of the preliminary scan that determines the
394        length. The relevant options are -x, and -i in UTF-8 mode. The result was
395        that the computed length was too small. The symptoms of this bug were
396        either the PCRE error "internal error: code overflow" from pcre_compile(),
397        or a glibc crash with a message such as "pcretest: free(): invalid next
398        size (fast)". Examples of patterns that provoked this bug (shown in
399        pcretest format) are:
400    
401          /(?-x: )/x
402          /(?x)(?-x: \s*#\s*)/
403          /((?i)[\x{c0}])/8
404          /(?i:[\x{c0}])/8
405    
406        HOWEVER: Change 17 below makes this fix obsolete as the memory computation
407        is now done differently.
408    
409     5. Applied patches from Google to: (a) add a QuoteMeta function to the C++
410        wrapper classes; (b) implement a new function in the C++ scanner that is
411        more efficient than the old way of doing things because it avoids levels of
412        recursion in the regex matching; (c) add a paragraph to the documentation
413        for the FullMatch() function.
414    
415     6. The escape sequence \n was being treated as whatever was defined as
416        "newline". Not only was this contrary to the documentation, which states
417        that \n is character 10 (hex 0A), but it also went horribly wrong when
418        "newline" was defined as CRLF. This has been fixed.
419    
420     7. In pcre_dfa_exec.c the value of an unsigned integer (the variable called c)
421        was being set to -1 for the "end of line" case (supposedly a value that no
422        character can have). Though this value is never used (the check for end of
423        line is "zero bytes in current character"), it caused compiler complaints.
424        I've changed it to 0xffffffff.
425    
426     8. In pcre_version.c, the version string was being built by a sequence of
427        C macros that, in the event of PCRE_PRERELEASE being defined as an empty
428        string (as it is for production releases) called a macro with an empty
429        argument. The C standard says the result of this is undefined. The gcc
430        compiler treats it as an empty string (which was what was wanted) but it is
431        reported that Visual C gives an error. The source has been hacked around to
432        avoid this problem.
433    
434     9. On the advice of a Windows user, included <io.h> and <fcntl.h> in Windows
435        builds of pcretest, and changed the call to _setmode() to use _O_BINARY
436        instead of 0x8000. Made all the #ifdefs test both _WIN32 and WIN32 (not all
437        of them did).
438    
439    10. Originally, pcretest opened its input and output without "b"; then I was
440        told that "b" was needed in some environments, so it was added for release
441        5.0 to both the input and output. (It makes no difference on Unix-like
442        systems.) Later I was told that it is wrong for the input on Windows. I've
443        now abstracted the modes into two macros, to make it easier to fiddle with
444        them, and removed "b" from the input mode under Windows.
445    
446    11. Added pkgconfig support for the C++ wrapper library, libpcrecpp.
447    
448    12. Added -help and --help to pcretest as an official way of being reminded
449        of the options.
450    
451    13. Removed some redundant semicolons after macro calls in pcrecpparg.h.in
452        and pcrecpp.cc because they annoy compilers at high warning levels.
453    
454    14. A bit of tidying/refactoring in pcre_exec.c in the main bumpalong loop.
455    
456    15. Fixed an occurrence of == in configure.ac that should have been = (shell
457        scripts are not C programs :-) and which was not noticed because it works
458        on Linux.
459    
460    16. pcretest is supposed to handle any length of pattern and data line (as one
461        line or as a continued sequence of lines) by extending its input buffer if
462        necessary. This feature was broken for very long pattern lines, leading to
463        a string of junk being passed to pcre_compile() if the pattern was longer
464        than about 50K.
465    
466    17. I have done a major re-factoring of the way pcre_compile() computes the
467        amount of memory needed for a compiled pattern. Previously, there was code
468        that made a preliminary scan of the pattern in order to do this. That was
469        OK when PCRE was new, but as the facilities have expanded, it has become
470        harder and harder to keep it in step with the real compile phase, and there
471        have been a number of bugs (see for example, 4 above). I have now found a
472        cunning way of running the real compile function in a "fake" mode that
473        enables it to compute how much memory it would need, while actually only
474        ever using a few hundred bytes of working memory and without too many
475        tests of the mode. This should make future maintenance and development
476        easier. A side effect of this work is that the limit of 200 on the nesting
477        depth of parentheses has been removed (though this was never a serious
478        limitation, I suspect). However, there is a downside: pcre_compile() now
479        runs more slowly than before (30% or more, depending on the pattern). I
480        hope this isn't a big issue. There is no effect on runtime performance.
481    
482    18. Fixed a minor bug in pcretest: if a pattern line was not terminated by a
483        newline (only possible for the last line of a file) and it was a
484        pattern that set a locale (followed by /Lsomething), pcretest crashed.
485    
486    19. Added additional timing features to pcretest. (1) The -tm option now times
487        matching only, not compiling. (2) Both -t and -tm can be followed, as a
488        separate command line item, by a number that specifies the number of
489        repeats to use when timing. The default is 50000; this gives better
490        precision, but takes uncomfortably long for very large patterns.
491    
492    20. Extended pcre_study() to be more clever in cases where a branch of a
493        subpattern has no definite first character. For example, (a*|b*)[cd] would
494        previously give no result from pcre_study(). Now it recognizes that the
495        first character must be a, b, c, or d.
496    
497    21. There was an incorrect error "recursive call could loop indefinitely" if
498        a subpattern (or the entire pattern) that was being tested for matching an
499        empty string contained only one non-empty item after a nested subpattern.
500        For example, the pattern (?>\x{100}*)\d(?R) provoked this error
501        incorrectly, because the \d was being skipped in the check.
502    
503    22. The pcretest program now has a new pattern option /B and a command line
504        option -b, which is equivalent to adding /B to every pattern. This causes
505        it to show the compiled bytecode, without the additional information that
506        -d shows. The effect of -d is now the same as -b with -i (and similarly, /D
507        is the same as /B/I).
508    
509    23. A new optimization is now able automatically to treat some sequences such
510        as a*b as a*+b. More specifically, if something simple (such as a character
511        or a simple class like \d) has an unlimited quantifier, and is followed by
512        something that cannot possibly match the quantified thing, the quantifier
513        is automatically "possessified".
514    
515    24. A recursive reference to a subpattern whose number was greater than 39
516        went wrong under certain circumstances in UTF-8 mode. This bug could also
517        have affected the operation of pcre_study().
518    
519    25. Realized that a little bit of performance could be had by replacing
520        (c & 0xc0) == 0xc0 with c >= 0xc0 when processing UTF-8 characters.
521    
522    26. Timing data from pcretest is now shown to 4 decimal places instead of 3.
523    
524    27. Possessive quantifiers such as a++ were previously implemented by turning
525        them into atomic groups such as ($>a+). Now they have their own opcodes,
526        which improves performance. This includes the automatically created ones
527        from 23 above.
528    
529    28. A pattern such as (?=(\w+))\1: which simulates an atomic group using a
530        lookahead was broken if it was not anchored. PCRE was mistakenly expecting
531        the first matched character to be a colon. This applied both to named and
532        numbered groups.
533    
534    29. The ucpinternal.h header file was missing its idempotency #ifdef.
535    
536    30. I was sent a "project" file called libpcre.a.dev which I understand makes
537        building PCRE on Windows easier, so I have included it in the distribution.
538    
539    31. There is now a check in pcretest against a ridiculously large number being
540        returned by pcre_exec() or pcre_dfa_exec(). If this happens in a /g or /G
541        loop, the loop is abandoned.
542    
543    32. Forward references to subpatterns in conditions such as (?(2)...) where
544        subpattern 2 is defined later cause pcre_compile() to search forwards in
545        the pattern for the relevant set of parentheses. This search went wrong
546        when there were unescaped parentheses in a character class, parentheses
547        escaped with \Q...\E, or parentheses in a #-comment in /x mode.
548    
549    33. "Subroutine" calls and backreferences were previously restricted to
550        referencing subpatterns earlier in the regex. This restriction has now
551        been removed.
552    
553    34. Added a number of extra features that are going to be in Perl 5.10. On the
554        whole, these are just syntactic alternatives for features that PCRE had
555        previously implemented using the Python syntax or my own invention. The
556        other formats are all retained for compatibility.
557    
558        (a) Named groups can now be defined as (?<name>...) or (?'name'...) as well
559            as (?P<name>...). The new forms, as well as being in Perl 5.10, are
560            also .NET compatible.
561    
562        (b) A recursion or subroutine call to a named group can now be defined as
563            (?&name) as well as (?P>name).
564    
565        (c) A backreference to a named group can now be defined as \k<name> or
566            \k'name' as well as (?P=name). The new forms, as well as being in Perl
567            5.10, are also .NET compatible.
568    
569        (d) A conditional reference to a named group can now use the syntax
570            (?(<name>) or (?('name') as well as (?(name).
571    
572        (e) A "conditional group" of the form (?(DEFINE)...) can be used to define
573            groups (named and numbered) that are never evaluated inline, but can be
574            called as "subroutines" from elsewhere. In effect, the DEFINE condition
575            is always false. There may be only one alternative in such a group.
576    
577        (f) A test for recursion can be given as (?(R1).. or (?(R&name)... as well
578            as the simple (?(R). The condition is true only if the most recent
579            recursion is that of the given number or name. It does not search out
580            through the entire recursion stack.
581    
582        (g) The escape \gN or \g{N} has been added, where N is a positive or
583            negative number, specifying an absolute or relative reference.
584    
585    35. Tidied to get rid of some further signed/unsigned compiler warnings and
586        some "unreachable code" warnings.
587    
588    36. Updated the Unicode property tables to Unicode version 5.0.0. Amongst other
589        things, this adds five new scripts.
590    
591    37. Perl ignores orphaned \E escapes completely. PCRE now does the same.
592        There were also incompatibilities regarding the handling of \Q..\E inside
593        character classes, for example with patterns like [\Qa\E-\Qz\E] where the
594        hyphen was adjacent to \Q or \E. I hope I've cleared all this up now.
595    
596    38. Like Perl, PCRE detects when an indefinitely repeated parenthesized group
597        matches an empty string, and forcibly breaks the loop. There were bugs in
598        this code in non-simple cases. For a pattern such as  ^(a()*)*  matched
599        against  aaaa  the result was just "a" rather than "aaaa", for example. Two
600        separate and independent bugs (that affected different cases) have been
601        fixed.
602    
603    39. Refactored the code to abolish the use of different opcodes for small
604        capturing bracket numbers. This is a tidy that I avoided doing when I
605        removed the limit on the number of capturing brackets for 3.5 back in 2001.
606        The new approach is not only tidier, it makes it possible to reduce the
607        memory needed to fix the previous bug (38).
608    
609    40. Implemented PCRE_NEWLINE_ANY to recognize any of the Unicode newline
610        sequences (http://unicode.org/unicode/reports/tr18/) as "newline" when
611        processing dot, circumflex, or dollar metacharacters, or #-comments in /x
612        mode.
613    
614    41. Add \R to match any Unicode newline sequence, as suggested in the Unicode
615        report.
616    
617    42. Applied patch, originally from Ari Pollak, modified by Google, to allow
618        copy construction and assignment in the C++ wrapper.
619    
620    43. Updated pcregrep to support "--newline=any". In the process, I fixed a
621        couple of bugs that could have given wrong results in the "--newline=crlf"
622        case.
623    
624    44. Added a number of casts and did some reorganization of signed/unsigned int
625        variables following suggestions from Dair Grant. Also renamed the variable
626        "this" as "item" because it is a C++ keyword.
627    
628    45. Arranged for dftables to add
629    
630          #include "pcre_internal.h"
631    
632        to pcre_chartables.c because without it, gcc 4.x may remove the array
633        definition from the final binary if PCRE is built into a static library and
634        dead code stripping is activated.
635    
636    46. For an unanchored pattern, if a match attempt fails at the start of a
637        newline sequence, and the newline setting is CRLF or ANY, and the next two
638        characters are CRLF, advance by two characters instead of one.
639    
640    
641    Version 6.7 04-Jul-06
642    ---------------------
643    
644     1. In order to handle tests when input lines are enormously long, pcretest has
645        been re-factored so that it automatically extends its buffers when
646        necessary. The code is crude, but this _is_ just a test program. The
647        default size has been increased from 32K to 50K.
648    
649     2. The code in pcre_study() was using the value of the re argument before
650        testing it for NULL. (Of course, in any sensible call of the function, it
651        won't be NULL.)
652    
653     3. The memmove() emulation function in pcre_internal.h, which is used on
654        systems that lack both memmove() and bcopy() - that is, hardly ever -
655        was missing a "static" storage class specifier.
656    
657     4. When UTF-8 mode was not set, PCRE looped when compiling certain patterns
658        containing an extended class (one that cannot be represented by a bitmap
659        because it contains high-valued characters or Unicode property items, e.g.
660        [\pZ]). Almost always one would set UTF-8 mode when processing such a
661        pattern, but PCRE should not loop if you do not (it no longer does).
662        [Detail: two cases were found: (a) a repeated subpattern containing an
663        extended class; (b) a recursive reference to a subpattern that followed a
664        previous extended class. It wasn't skipping over the extended class
665        correctly when UTF-8 mode was not set.]
666    
667     5. A negated single-character class was not being recognized as fixed-length
668        in lookbehind assertions such as (?<=[^f]), leading to an incorrect
669        compile error "lookbehind assertion is not fixed length".
670    
671     6. The RunPerlTest auxiliary script was showing an unexpected difference
672        between PCRE and Perl for UTF-8 tests. It turns out that it is hard to
673        write a Perl script that can interpret lines of an input file either as
674        byte characters or as UTF-8, which is what "perltest" was being required to
675        do for the non-UTF-8 and UTF-8 tests, respectively. Essentially what you
676        can't do is switch easily at run time between having the "use utf8;" pragma
677        or not. In the end, I fudged it by using the RunPerlTest script to insert
678        "use utf8;" explicitly for the UTF-8 tests.
679    
680     7. In multiline (/m) mode, PCRE was matching ^ after a terminating newline at
681        the end of the subject string, contrary to the documentation and to what
682        Perl does. This was true of both matching functions. Now it matches only at
683        the start of the subject and immediately after *internal* newlines.
684    
685     8. A call of pcre_fullinfo() from pcretest to get the option bits was passing
686        a pointer to an int instead of a pointer to an unsigned long int. This
687        caused problems on 64-bit systems.
688    
689     9. Applied a patch from the folks at Google to pcrecpp.cc, to fix "another
690        instance of the 'standard' template library not being so standard".
691    
692    10. There was no check on the number of named subpatterns nor the maximum
693        length of a subpattern name. The product of these values is used to compute
694        the size of the memory block for a compiled pattern. By supplying a very
695        long subpattern name and a large number of named subpatterns, the size
696        computation could be caused to overflow. This is now prevented by limiting
697        the length of names to 32 characters, and the number of named subpatterns
698        to 10,000.
699    
700    11. Subpatterns that are repeated with specific counts have to be replicated in
701        the compiled pattern. The size of memory for this was computed from the
702        length of the subpattern and the repeat count. The latter is limited to
703        65535, but there was no limit on the former, meaning that integer overflow
704        could in principle occur. The compiled length of a repeated subpattern is
705        now limited to 30,000 bytes in order to prevent this.
706    
707    12. Added the optional facility to have named substrings with the same name.
708    
709    13. Added the ability to use a named substring as a condition, using the
710        Python syntax: (?(name)yes|no). This overloads (?(R)... and names that
711        are numbers (not recommended). Forward references are permitted.
712    
713    14. Added forward references in named backreferences (if you see what I mean).
714    
715    15. In UTF-8 mode, with the PCRE_DOTALL option set, a quantified dot in the
716        pattern could run off the end of the subject. For example, the pattern
717        "(?s)(.{1,5})"8 did this with the subject "ab".
718    
719    16. If PCRE_DOTALL or PCRE_MULTILINE were set, pcre_dfa_exec() behaved as if
720        PCRE_CASELESS was set when matching characters that were quantified with ?
721        or *.
722    
723    17. A character class other than a single negated character that had a minimum
724        but no maximum quantifier - for example [ab]{6,} - was not handled
725        correctly by pce_dfa_exec(). It would match only one character.
726    
727    18. A valid (though odd) pattern that looked like a POSIX character
728        class but used an invalid character after [ (for example [[,abc,]]) caused
729        pcre_compile() to give the error "Failed: internal error: code overflow" or
730        in some cases to crash with a glibc free() error. This could even happen if
731        the pattern terminated after [[ but there just happened to be a sequence of
732        letters, a binary zero, and a closing ] in the memory that followed.
733    
734    19. Perl's treatment of octal escapes in the range \400 to \777 has changed
735        over the years. Originally (before any Unicode support), just the bottom 8
736        bits were taken. Thus, for example, \500 really meant \100. Nowadays the
737        output from "man perlunicode" includes this:
738    
739          The regular expression compiler produces polymorphic opcodes.  That
740          is, the pattern adapts to the data and automatically switches to
741          the Unicode character scheme when presented with Unicode data--or
742          instead uses a traditional byte scheme when presented with byte
743          data.
744    
745        Sadly, a wide octal escape does not cause a switch, and in a string with
746        no other multibyte characters, these octal escapes are treated as before.
747        Thus, in Perl, the pattern  /\500/ actually matches \100 but the pattern
748        /\500|\x{1ff}/ matches \500 or \777 because the whole thing is treated as a
749        Unicode string.
750    
751        I have not perpetrated such confusion in PCRE. Up till now, it took just
752        the bottom 8 bits, as in old Perl. I have now made octal escapes with
753        values greater than \377 illegal in non-UTF-8 mode. In UTF-8 mode they
754        translate to the appropriate multibyte character.
755    
756    29. Applied some refactoring to reduce the number of warnings from Microsoft
757        and Borland compilers. This has included removing the fudge introduced
758        seven years ago for the OS/2 compiler (see 2.02/2 below) because it caused
759        a warning about an unused variable.
760    
761    21. PCRE has not included VT (character 0x0b) in the set of whitespace
762        characters since release 4.0, because Perl (from release 5.004) does not.
763        [Or at least, is documented not to: some releases seem to be in conflict
764        with the documentation.] However, when a pattern was studied with
765        pcre_study() and all its branches started with \s, PCRE still included VT
766        as a possible starting character. Of course, this did no harm; it just
767        caused an unnecessary match attempt.
768    
769    22. Removed a now-redundant internal flag bit that recorded the fact that case
770        dependency changed within the pattern. This was once needed for "required
771        byte" processing, but is no longer used. This recovers a now-scarce options
772        bit. Also moved the least significant internal flag bit to the most-
773        significant bit of the word, which was not previously used (hangover from
774        the days when it was an int rather than a uint) to free up another bit for
775        the future.
776    
777    23. Added support for CRLF line endings as well as CR and LF. As well as the
778        default being selectable at build time, it can now be changed at runtime
779        via the PCRE_NEWLINE_xxx flags. There are now options for pcregrep to
780        specify that it is scanning data with non-default line endings.
781    
782    24. Changed the definition of CXXLINK to make it agree with the definition of
783        LINK in the Makefile, by replacing LDFLAGS to CXXFLAGS.
784    
785    25. Applied Ian Taylor's patches to avoid using another stack frame for tail
786        recursions. This makes a big different to stack usage for some patterns.
787    
788    26. If a subpattern containing a named recursion or subroutine reference such
789        as (?P>B) was quantified, for example (xxx(?P>B)){3}, the calculation of
790        the space required for the compiled pattern went wrong and gave too small a
791        value. Depending on the environment, this could lead to "Failed: internal
792        error: code overflow at offset 49" or "glibc detected double free or
793        corruption" errors.
794    
795    27. Applied patches from Google (a) to support the new newline modes and (b) to
796        advance over multibyte UTF-8 characters in GlobalReplace.
797    
798    28. Change free() to pcre_free() in pcredemo.c. Apparently this makes a
799        difference for some implementation of PCRE in some Windows version.
800    
801    29. Added some extra testing facilities to pcretest:
802    
803        \q<number>   in a data line sets the "match limit" value
804        \Q<number>   in a data line sets the "match recursion limt" value
805        -S <number>  sets the stack size, where <number> is in megabytes
806    
807        The -S option isn't available for Windows.
808    
809    
810    Version 6.6 06-Feb-06
811    ---------------------
812    
813     1. Change 16(a) for 6.5 broke things, because PCRE_DATA_SCOPE was not defined
814        in pcreposix.h. I have copied the definition from pcre.h.
815    
816     2. Change 25 for 6.5 broke compilation in a build directory out-of-tree
817        because pcre.h is no longer a built file.
818    
819     3. Added Jeff Friedl's additional debugging patches to pcregrep. These are
820        not normally included in the compiled code.
821    
822    
823  Version 6.5 01-Feb-06  Version 6.5 01-Feb-06
824  ---------------------  ---------------------
825    

Legend:
Removed from v.87  
changed lines
  Added in v.219

webmaster@exim.org
ViewVC Help
Powered by ViewVC 1.1.12