/[pcre]/code/branches/pcre16/ChangeLog
ViewVC logotype

Diff of /code/branches/pcre16/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 98 by ph10, Mon Mar 5 15:55:02 2007 UTC
# Line 1  Line 1 
1  ChangeLog for PCRE  ChangeLog for PCRE
2  ------------------  ------------------
3    
4    Version 7.1 05-Mar-07
5    ---------------------
6    
7     1. Applied Bob Rossi and Daniel G's patches to convert the build system to one
8        that is more "standard", making use of automake and other autotools.
9    
10    
11    Version 7.0 19-Dec-06
12    ---------------------
13    
14     1. Fixed a signed/unsigned compiler warning in pcre_compile.c, shown up by
15        moving to gcc 4.1.1.
16    
17     2. The -S option for pcretest uses setrlimit(); I had omitted to #include
18        sys/time.h, which is documented as needed for this function. It doesn't
19        seem to matter on Linux, but it showed up on some releases of OS X.
20    
21     3. It seems that there are systems where bytes whose values are greater than
22        127 match isprint() in the "C" locale. The "C" locale should be the
23        default when a C program starts up. In most systems, only ASCII printing
24        characters match isprint(). This difference caused the output from pcretest
25        to vary, making some of the tests fail. I have changed pcretest so that:
26    
27        (a) When it is outputting text in the compiled version of a pattern, bytes
28            other than 32-126 are always shown as hex escapes.
29    
30        (b) When it is outputting text that is a matched part of a subject string,
31            it does the same, unless a different locale has been set for the match
32            (using the /L modifier). In this case, it uses isprint() to decide.
33    
34     4. Fixed a major bug that caused incorrect computation of the amount of memory
35        required for a compiled pattern when options that changed within the
36        pattern affected the logic of the preliminary scan that determines the
37        length. The relevant options are -x, and -i in UTF-8 mode. The result was
38        that the computed length was too small. The symptoms of this bug were
39        either the PCRE error "internal error: code overflow" from pcre_compile(),
40        or a glibc crash with a message such as "pcretest: free(): invalid next
41        size (fast)". Examples of patterns that provoked this bug (shown in
42        pcretest format) are:
43    
44          /(?-x: )/x
45          /(?x)(?-x: \s*#\s*)/
46          /((?i)[\x{c0}])/8
47          /(?i:[\x{c0}])/8
48    
49        HOWEVER: Change 17 below makes this fix obsolete as the memory computation
50        is now done differently.
51    
52     5. Applied patches from Google to: (a) add a QuoteMeta function to the C++
53        wrapper classes; (b) implement a new function in the C++ scanner that is
54        more efficient than the old way of doing things because it avoids levels of
55        recursion in the regex matching; (c) add a paragraph to the documentation
56        for the FullMatch() function.
57    
58     6. The escape sequence \n was being treated as whatever was defined as
59        "newline". Not only was this contrary to the documentation, which states
60        that \n is character 10 (hex 0A), but it also went horribly wrong when
61        "newline" was defined as CRLF. This has been fixed.
62    
63     7. In pcre_dfa_exec.c the value of an unsigned integer (the variable called c)
64        was being set to -1 for the "end of line" case (supposedly a value that no
65        character can have). Though this value is never used (the check for end of
66        line is "zero bytes in current character"), it caused compiler complaints.
67        I've changed it to 0xffffffff.
68    
69     8. In pcre_version.c, the version string was being built by a sequence of
70        C macros that, in the event of PCRE_PRERELEASE being defined as an empty
71        string (as it is for production releases) called a macro with an empty
72        argument. The C standard says the result of this is undefined. The gcc
73        compiler treats it as an empty string (which was what was wanted) but it is
74        reported that Visual C gives an error. The source has been hacked around to
75        avoid this problem.
76    
77     9. On the advice of a Windows user, included <io.h> and <fcntl.h> in Windows
78        builds of pcretest, and changed the call to _setmode() to use _O_BINARY
79        instead of 0x8000. Made all the #ifdefs test both _WIN32 and WIN32 (not all
80        of them did).
81    
82    10. Originally, pcretest opened its input and output without "b"; then I was
83        told that "b" was needed in some environments, so it was added for release
84        5.0 to both the input and output. (It makes no difference on Unix-like
85        systems.) Later I was told that it is wrong for the input on Windows. I've
86        now abstracted the modes into two macros, to make it easier to fiddle with
87        them, and removed "b" from the input mode under Windows.
88    
89    11. Added pkgconfig support for the C++ wrapper library, libpcrecpp.
90    
91    12. Added -help and --help to pcretest as an official way of being reminded
92        of the options.
93    
94    13. Removed some redundant semicolons after macro calls in pcrecpparg.h.in
95        and pcrecpp.cc because they annoy compilers at high warning levels.
96    
97    14. A bit of tidying/refactoring in pcre_exec.c in the main bumpalong loop.
98    
99    15. Fixed an occurrence of == in configure.ac that should have been = (shell
100        scripts are not C programs :-) and which was not noticed because it works
101        on Linux.
102    
103    16. pcretest is supposed to handle any length of pattern and data line (as one
104        line or as a continued sequence of lines) by extending its input buffer if
105        necessary. This feature was broken for very long pattern lines, leading to
106        a string of junk being passed to pcre_compile() if the pattern was longer
107        than about 50K.
108    
109    17. I have done a major re-factoring of the way pcre_compile() computes the
110        amount of memory needed for a compiled pattern. Previously, there was code
111        that made a preliminary scan of the pattern in order to do this. That was
112        OK when PCRE was new, but as the facilities have expanded, it has become
113        harder and harder to keep it in step with the real compile phase, and there
114        have been a number of bugs (see for example, 4 above). I have now found a
115        cunning way of running the real compile function in a "fake" mode that
116        enables it to compute how much memory it would need, while actually only
117        ever using a few hundred bytes of working memory and without too many
118        tests of the mode. This should make future maintenance and development
119        easier. A side effect of this work is that the limit of 200 on the nesting
120        depth of parentheses has been removed (though this was never a serious
121        limitation, I suspect). However, there is a downside: pcre_compile() now
122        runs more slowly than before (30% or more, depending on the pattern). I
123        hope this isn't a big issue. There is no effect on runtime performance.
124    
125    18. Fixed a minor bug in pcretest: if a pattern line was not terminated by a
126        newline (only possible for the last line of a file) and it was a
127        pattern that set a locale (followed by /Lsomething), pcretest crashed.
128    
129    19. Added additional timing features to pcretest. (1) The -tm option now times
130        matching only, not compiling. (2) Both -t and -tm can be followed, as a
131        separate command line item, by a number that specifies the number of
132        repeats to use when timing. The default is 50000; this gives better
133        precision, but takes uncomfortably long for very large patterns.
134    
135    20. Extended pcre_study() to be more clever in cases where a branch of a
136        subpattern has no definite first character. For example, (a*|b*)[cd] would
137        previously give no result from pcre_study(). Now it recognizes that the
138        first character must be a, b, c, or d.
139    
140    21. There was an incorrect error "recursive call could loop indefinitely" if
141        a subpattern (or the entire pattern) that was being tested for matching an
142        empty string contained only one non-empty item after a nested subpattern.
143        For example, the pattern (?>\x{100}*)\d(?R) provoked this error
144        incorrectly, because the \d was being skipped in the check.
145    
146    22. The pcretest program now has a new pattern option /B and a command line
147        option -b, which is equivalent to adding /B to every pattern. This causes
148        it to show the compiled bytecode, without the additional information that
149        -d shows. The effect of -d is now the same as -b with -i (and similarly, /D
150        is the same as /B/I).
151    
152    23. A new optimization is now able automatically to treat some sequences such
153        as a*b as a*+b. More specifically, if something simple (such as a character
154        or a simple class like \d) has an unlimited quantifier, and is followed by
155        something that cannot possibly match the quantified thing, the quantifier
156        is automatically "possessified".
157    
158    24. A recursive reference to a subpattern whose number was greater than 39
159        went wrong under certain circumstances in UTF-8 mode. This bug could also
160        have affected the operation of pcre_study().
161    
162    25. Realized that a little bit of performance could be had by replacing
163        (c & 0xc0) == 0xc0 with c >= 0xc0 when processing UTF-8 characters.
164    
165    26. Timing data from pcretest is now shown to 4 decimal places instead of 3.
166    
167    27. Possessive quantifiers such as a++ were previously implemented by turning
168        them into atomic groups such as ($>a+). Now they have their own opcodes,
169        which improves performance. This includes the automatically created ones
170        from 23 above.
171    
172    28. A pattern such as (?=(\w+))\1: which simulates an atomic group using a
173        lookahead was broken if it was not anchored. PCRE was mistakenly expecting
174        the first matched character to be a colon. This applied both to named and
175        numbered groups.
176    
177    29. The ucpinternal.h header file was missing its idempotency #ifdef.
178    
179    30. I was sent a "project" file called libpcre.a.dev which I understand makes
180        building PCRE on Windows easier, so I have included it in the distribution.
181    
182    31. There is now a check in pcretest against a ridiculously large number being
183        returned by pcre_exec() or pcre_dfa_exec(). If this happens in a /g or /G
184        loop, the loop is abandoned.
185    
186    32. Forward references to subpatterns in conditions such as (?(2)...) where
187        subpattern 2 is defined later cause pcre_compile() to search forwards in
188        the pattern for the relevant set of parentheses. This search went wrong
189        when there were unescaped parentheses in a character class, parentheses
190        escaped with \Q...\E, or parentheses in a #-comment in /x mode.
191    
192    33. "Subroutine" calls and backreferences were previously restricted to
193        referencing subpatterns earlier in the regex. This restriction has now
194        been removed.
195    
196    34. Added a number of extra features that are going to be in Perl 5.10. On the
197        whole, these are just syntactic alternatives for features that PCRE had
198        previously implemented using the Python syntax or my own invention. The
199        other formats are all retained for compatibility.
200    
201        (a) Named groups can now be defined as (?<name>...) or (?'name'...) as well
202            as (?P<name>...). The new forms, as well as being in Perl 5.10, are
203            also .NET compatible.
204    
205        (b) A recursion or subroutine call to a named group can now be defined as
206            (?&name) as well as (?P>name).
207    
208        (c) A backreference to a named group can now be defined as \k<name> or
209            \k'name' as well as (?P=name). The new forms, as well as being in Perl
210            5.10, are also .NET compatible.
211    
212        (d) A conditional reference to a named group can now use the syntax
213            (?(<name>) or (?('name') as well as (?(name).
214    
215        (e) A "conditional group" of the form (?(DEFINE)...) can be used to define
216            groups (named and numbered) that are never evaluated inline, but can be
217            called as "subroutines" from elsewhere. In effect, the DEFINE condition
218            is always false. There may be only one alternative in such a group.
219    
220        (f) A test for recursion can be given as (?(R1).. or (?(R&name)... as well
221            as the simple (?(R). The condition is true only if the most recent
222            recursion is that of the given number or name. It does not search out
223            through the entire recursion stack.
224    
225        (g) The escape \gN or \g{N} has been added, where N is a positive or
226            negative number, specifying an absolute or relative reference.
227    
228    35. Tidied to get rid of some further signed/unsigned compiler warnings and
229        some "unreachable code" warnings.
230    
231    36. Updated the Unicode property tables to Unicode version 5.0.0. Amongst other
232        things, this adds five new scripts.
233    
234    37. Perl ignores orphaned \E escapes completely. PCRE now does the same.
235        There were also incompatibilities regarding the handling of \Q..\E inside
236        character classes, for example with patterns like [\Qa\E-\Qz\E] where the
237        hyphen was adjacent to \Q or \E. I hope I've cleared all this up now.
238    
239    38. Like Perl, PCRE detects when an indefinitely repeated parenthesized group
240        matches an empty string, and forcibly breaks the loop. There were bugs in
241        this code in non-simple cases. For a pattern such as  ^(a()*)*  matched
242        against  aaaa  the result was just "a" rather than "aaaa", for example. Two
243        separate and independent bugs (that affected different cases) have been
244        fixed.
245    
246    39. Refactored the code to abolish the use of different opcodes for small
247        capturing bracket numbers. This is a tidy that I avoided doing when I
248        removed the limit on the number of capturing brackets for 3.5 back in 2001.
249        The new approach is not only tidier, it makes it possible to reduce the
250        memory needed to fix the previous bug (38).
251    
252    40. Implemented PCRE_NEWLINE_ANY to recognize any of the Unicode newline
253        sequences (http://unicode.org/unicode/reports/tr18/) as "newline" when
254        processing dot, circumflex, or dollar metacharacters, or #-comments in /x
255        mode.
256    
257    41. Add \R to match any Unicode newline sequence, as suggested in the Unicode
258        report.
259    
260    42. Applied patch, originally from Ari Pollak, modified by Google, to allow
261        copy construction and assignment in the C++ wrapper.
262    
263    43. Updated pcregrep to support "--newline=any". In the process, I fixed a
264        couple of bugs that could have given wrong results in the "--newline=crlf"
265        case.
266    
267    44. Added a number of casts and did some reorganization of signed/unsigned int
268        variables following suggestions from Dair Grant. Also renamed the variable
269        "this" as "item" because it is a C++ keyword.
270    
271    45. Arranged for dftables to add
272    
273          #include "pcre_internal.h"
274    
275        to pcre_chartables.c because without it, gcc 4.x may remove the array
276        definition from the final binary if PCRE is built into a static library and
277        dead code stripping is activated.
278    
279    46. For an unanchored pattern, if a match attempt fails at the start of a
280        newline sequence, and the newline setting is CRLF or ANY, and the next two
281        characters are CRLF, advance by two characters instead of one.
282    
283    
284    Version 6.7 04-Jul-06
285    ---------------------
286    
287     1. In order to handle tests when input lines are enormously long, pcretest has
288        been re-factored so that it automatically extends its buffers when
289        necessary. The code is crude, but this _is_ just a test program. The
290        default size has been increased from 32K to 50K.
291    
292     2. The code in pcre_study() was using the value of the re argument before
293        testing it for NULL. (Of course, in any sensible call of the function, it
294        won't be NULL.)
295    
296     3. The memmove() emulation function in pcre_internal.h, which is used on
297        systems that lack both memmove() and bcopy() - that is, hardly ever -
298        was missing a "static" storage class specifier.
299    
300     4. When UTF-8 mode was not set, PCRE looped when compiling certain patterns
301        containing an extended class (one that cannot be represented by a bitmap
302        because it contains high-valued characters or Unicode property items, e.g.
303        [\pZ]). Almost always one would set UTF-8 mode when processing such a
304        pattern, but PCRE should not loop if you do not (it no longer does).
305        [Detail: two cases were found: (a) a repeated subpattern containing an
306        extended class; (b) a recursive reference to a subpattern that followed a
307        previous extended class. It wasn't skipping over the extended class
308        correctly when UTF-8 mode was not set.]
309    
310     5. A negated single-character class was not being recognized as fixed-length
311        in lookbehind assertions such as (?<=[^f]), leading to an incorrect
312        compile error "lookbehind assertion is not fixed length".
313    
314     6. The RunPerlTest auxiliary script was showing an unexpected difference
315        between PCRE and Perl for UTF-8 tests. It turns out that it is hard to
316        write a Perl script that can interpret lines of an input file either as
317        byte characters or as UTF-8, which is what "perltest" was being required to
318        do for the non-UTF-8 and UTF-8 tests, respectively. Essentially what you
319        can't do is switch easily at run time between having the "use utf8;" pragma
320        or not. In the end, I fudged it by using the RunPerlTest script to insert
321        "use utf8;" explicitly for the UTF-8 tests.
322    
323     7. In multiline (/m) mode, PCRE was matching ^ after a terminating newline at
324        the end of the subject string, contrary to the documentation and to what
325        Perl does. This was true of both matching functions. Now it matches only at
326        the start of the subject and immediately after *internal* newlines.
327    
328     8. A call of pcre_fullinfo() from pcretest to get the option bits was passing
329        a pointer to an int instead of a pointer to an unsigned long int. This
330        caused problems on 64-bit systems.
331    
332     9. Applied a patch from the folks at Google to pcrecpp.cc, to fix "another
333        instance of the 'standard' template library not being so standard".
334    
335    10. There was no check on the number of named subpatterns nor the maximum
336        length of a subpattern name. The product of these values is used to compute
337        the size of the memory block for a compiled pattern. By supplying a very
338        long subpattern name and a large number of named subpatterns, the size
339        computation could be caused to overflow. This is now prevented by limiting
340        the length of names to 32 characters, and the number of named subpatterns
341        to 10,000.
342    
343    11. Subpatterns that are repeated with specific counts have to be replicated in
344        the compiled pattern. The size of memory for this was computed from the
345        length of the subpattern and the repeat count. The latter is limited to
346        65535, but there was no limit on the former, meaning that integer overflow
347        could in principle occur. The compiled length of a repeated subpattern is
348        now limited to 30,000 bytes in order to prevent this.
349    
350    12. Added the optional facility to have named substrings with the same name.
351    
352    13. Added the ability to use a named substring as a condition, using the
353        Python syntax: (?(name)yes|no). This overloads (?(R)... and names that
354        are numbers (not recommended). Forward references are permitted.
355    
356    14. Added forward references in named backreferences (if you see what I mean).
357    
358    15. In UTF-8 mode, with the PCRE_DOTALL option set, a quantified dot in the
359        pattern could run off the end of the subject. For example, the pattern
360        "(?s)(.{1,5})"8 did this with the subject "ab".
361    
362    16. If PCRE_DOTALL or PCRE_MULTILINE were set, pcre_dfa_exec() behaved as if
363        PCRE_CASELESS was set when matching characters that were quantified with ?
364        or *.
365    
366    17. A character class other than a single negated character that had a minimum
367        but no maximum quantifier - for example [ab]{6,} - was not handled
368        correctly by pce_dfa_exec(). It would match only one character.
369    
370    18. A valid (though odd) pattern that looked like a POSIX character
371        class but used an invalid character after [ (for example [[,abc,]]) caused
372        pcre_compile() to give the error "Failed: internal error: code overflow" or
373        in some cases to crash with a glibc free() error. This could even happen if
374        the pattern terminated after [[ but there just happened to be a sequence of
375        letters, a binary zero, and a closing ] in the memory that followed.
376    
377    19. Perl's treatment of octal escapes in the range \400 to \777 has changed
378        over the years. Originally (before any Unicode support), just the bottom 8
379        bits were taken. Thus, for example, \500 really meant \100. Nowadays the
380        output from "man perlunicode" includes this:
381    
382          The regular expression compiler produces polymorphic opcodes.  That
383          is, the pattern adapts to the data and automatically switches to
384          the Unicode character scheme when presented with Unicode data--or
385          instead uses a traditional byte scheme when presented with byte
386          data.
387    
388        Sadly, a wide octal escape does not cause a switch, and in a string with
389        no other multibyte characters, these octal escapes are treated as before.
390        Thus, in Perl, the pattern  /\500/ actually matches \100 but the pattern
391        /\500|\x{1ff}/ matches \500 or \777 because the whole thing is treated as a
392        Unicode string.
393    
394        I have not perpetrated such confusion in PCRE. Up till now, it took just
395        the bottom 8 bits, as in old Perl. I have now made octal escapes with
396        values greater than \377 illegal in non-UTF-8 mode. In UTF-8 mode they
397        translate to the appropriate multibyte character.
398    
399    29. Applied some refactoring to reduce the number of warnings from Microsoft
400        and Borland compilers. This has included removing the fudge introduced
401        seven years ago for the OS/2 compiler (see 2.02/2 below) because it caused
402        a warning about an unused variable.
403    
404    21. PCRE has not included VT (character 0x0b) in the set of whitespace
405        characters since release 4.0, because Perl (from release 5.004) does not.
406        [Or at least, is documented not to: some releases seem to be in conflict
407        with the documentation.] However, when a pattern was studied with
408        pcre_study() and all its branches started with \s, PCRE still included VT
409        as a possible starting character. Of course, this did no harm; it just
410        caused an unnecessary match attempt.
411    
412    22. Removed a now-redundant internal flag bit that recorded the fact that case
413        dependency changed within the pattern. This was once needed for "required
414        byte" processing, but is no longer used. This recovers a now-scarce options
415        bit. Also moved the least significant internal flag bit to the most-
416        significant bit of the word, which was not previously used (hangover from
417        the days when it was an int rather than a uint) to free up another bit for
418        the future.
419    
420    23. Added support for CRLF line endings as well as CR and LF. As well as the
421        default being selectable at build time, it can now be changed at runtime
422        via the PCRE_NEWLINE_xxx flags. There are now options for pcregrep to
423        specify that it is scanning data with non-default line endings.
424    
425    24. Changed the definition of CXXLINK to make it agree with the definition of
426        LINK in the Makefile, by replacing LDFLAGS to CXXFLAGS.
427    
428    25. Applied Ian Taylor's patches to avoid using another stack frame for tail
429        recursions. This makes a big different to stack usage for some patterns.
430    
431    26. If a subpattern containing a named recursion or subroutine reference such
432        as (?P>B) was quantified, for example (xxx(?P>B)){3}, the calculation of
433        the space required for the compiled pattern went wrong and gave too small a
434        value. Depending on the environment, this could lead to "Failed: internal
435        error: code overflow at offset 49" or "glibc detected double free or
436        corruption" errors.
437    
438    27. Applied patches from Google (a) to support the new newline modes and (b) to
439        advance over multibyte UTF-8 characters in GlobalReplace.
440    
441    28. Change free() to pcre_free() in pcredemo.c. Apparently this makes a
442        difference for some implementation of PCRE in some Windows version.
443    
444    29. Added some extra testing facilities to pcretest:
445    
446        \q<number>   in a data line sets the "match limit" value
447        \Q<number>   in a data line sets the "match recursion limt" value
448        -S <number>  sets the stack size, where <number> is in megabytes
449    
450        The -S option isn't available for Windows.
451    
452    
453    Version 6.6 06-Feb-06
454    ---------------------
455    
456     1. Change 16(a) for 6.5 broke things, because PCRE_DATA_SCOPE was not defined
457        in pcreposix.h. I have copied the definition from pcre.h.
458    
459     2. Change 25 for 6.5 broke compilation in a build directory out-of-tree
460        because pcre.h is no longer a built file.
461    
462     3. Added Jeff Friedl's additional debugging patches to pcregrep. These are
463        not normally included in the compiled code.
464    
465    
466  Version 6.5 01-Feb-06  Version 6.5 01-Feb-06
467  ---------------------  ---------------------
468    

Legend:
Removed from v.87  
changed lines
  Added in v.98

  ViewVC Help
Powered by ViewVC 1.1.5