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

Diff of /code/trunk/ChangeLog

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

revision 33 by nigel, Sat Feb 24 21:39:01 2007 UTC revision 104 by ph10, Tue Mar 6 16:25:15 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. There
9        is some re-arrangement of the files and adjustment of comments consequent
10        on this.
11    
12     2. Part of the patch fixed a problem with the pcregrep tests. The test of -r
13        for recursive directory scanning broke on some systems because the files
14        are not scanned in any specific order and on different systems the order
15        was different. A call to "sort" has been inserted into RunGrepTest for the
16        approprate test as a short-term fix. In the longer term there may be an
17        alternative.
18    
19     3. I had an email from Eric Raymond about problems translating some of PCRE's
20        man pages to HTML (despite the fact that I distribute HTML pages, some
21        people do their own conversions for various reasons). The problems
22        concerned the use of low-level troff macros .br and .in. I have therefore
23        removed all such uses from the man pages (some were redundant, some could
24        be replaced by .nf/.fi pairs). The maintain/132html script that I use to
25        generate HTML has been updated to handle .nf/.fi and to complain if it
26        encounters .br or .in.
27    
28     4. Updated comments in configure.ac that get placed in config.h.in and also
29        arranged for config.h to be included in the distribution, for the benefit
30        of those who have to compile without Autotools (compare pcre.h).
31    
32     5. Updated the support (such as it is) for Virtual Pascal, thanks to Stefan
33        Weber: (1) pcre_internal.h was missing some function renames; (2) updated
34        makevp.bat for the current PCRE, using the additional files !compile.txt,
35        !linklib.txt, and pcregexp.pas.
36    
37     6. A Windows user reported a minor discrepancy with test 2, which turned out
38        to be caused by a trailing space on an input line that had got lost in his
39        copy. The trailing space was an accident, so I've just removed it.
40    
41     7. Add -Wl,-R... flags in pcre-config.in for *BSD* systems, as I'm told
42        that is needed.
43    
44    
45    Version 7.0 19-Dec-06
46    ---------------------
47    
48     1. Fixed a signed/unsigned compiler warning in pcre_compile.c, shown up by
49        moving to gcc 4.1.1.
50    
51     2. The -S option for pcretest uses setrlimit(); I had omitted to #include
52        sys/time.h, which is documented as needed for this function. It doesn't
53        seem to matter on Linux, but it showed up on some releases of OS X.
54    
55     3. It seems that there are systems where bytes whose values are greater than
56        127 match isprint() in the "C" locale. The "C" locale should be the
57        default when a C program starts up. In most systems, only ASCII printing
58        characters match isprint(). This difference caused the output from pcretest
59        to vary, making some of the tests fail. I have changed pcretest so that:
60    
61        (a) When it is outputting text in the compiled version of a pattern, bytes
62            other than 32-126 are always shown as hex escapes.
63    
64        (b) When it is outputting text that is a matched part of a subject string,
65            it does the same, unless a different locale has been set for the match
66            (using the /L modifier). In this case, it uses isprint() to decide.
67    
68     4. Fixed a major bug that caused incorrect computation of the amount of memory
69        required for a compiled pattern when options that changed within the
70        pattern affected the logic of the preliminary scan that determines the
71        length. The relevant options are -x, and -i in UTF-8 mode. The result was
72        that the computed length was too small. The symptoms of this bug were
73        either the PCRE error "internal error: code overflow" from pcre_compile(),
74        or a glibc crash with a message such as "pcretest: free(): invalid next
75        size (fast)". Examples of patterns that provoked this bug (shown in
76        pcretest format) are:
77    
78          /(?-x: )/x
79          /(?x)(?-x: \s*#\s*)/
80          /((?i)[\x{c0}])/8
81          /(?i:[\x{c0}])/8
82    
83        HOWEVER: Change 17 below makes this fix obsolete as the memory computation
84        is now done differently.
85    
86     5. Applied patches from Google to: (a) add a QuoteMeta function to the C++
87        wrapper classes; (b) implement a new function in the C++ scanner that is
88        more efficient than the old way of doing things because it avoids levels of
89        recursion in the regex matching; (c) add a paragraph to the documentation
90        for the FullMatch() function.
91    
92     6. The escape sequence \n was being treated as whatever was defined as
93        "newline". Not only was this contrary to the documentation, which states
94        that \n is character 10 (hex 0A), but it also went horribly wrong when
95        "newline" was defined as CRLF. This has been fixed.
96    
97     7. In pcre_dfa_exec.c the value of an unsigned integer (the variable called c)
98        was being set to -1 for the "end of line" case (supposedly a value that no
99        character can have). Though this value is never used (the check for end of
100        line is "zero bytes in current character"), it caused compiler complaints.
101        I've changed it to 0xffffffff.
102    
103     8. In pcre_version.c, the version string was being built by a sequence of
104        C macros that, in the event of PCRE_PRERELEASE being defined as an empty
105        string (as it is for production releases) called a macro with an empty
106        argument. The C standard says the result of this is undefined. The gcc
107        compiler treats it as an empty string (which was what was wanted) but it is
108        reported that Visual C gives an error. The source has been hacked around to
109        avoid this problem.
110    
111     9. On the advice of a Windows user, included <io.h> and <fcntl.h> in Windows
112        builds of pcretest, and changed the call to _setmode() to use _O_BINARY
113        instead of 0x8000. Made all the #ifdefs test both _WIN32 and WIN32 (not all
114        of them did).
115    
116    10. Originally, pcretest opened its input and output without "b"; then I was
117        told that "b" was needed in some environments, so it was added for release
118        5.0 to both the input and output. (It makes no difference on Unix-like
119        systems.) Later I was told that it is wrong for the input on Windows. I've
120        now abstracted the modes into two macros, to make it easier to fiddle with
121        them, and removed "b" from the input mode under Windows.
122    
123    11. Added pkgconfig support for the C++ wrapper library, libpcrecpp.
124    
125    12. Added -help and --help to pcretest as an official way of being reminded
126        of the options.
127    
128    13. Removed some redundant semicolons after macro calls in pcrecpparg.h.in
129        and pcrecpp.cc because they annoy compilers at high warning levels.
130    
131    14. A bit of tidying/refactoring in pcre_exec.c in the main bumpalong loop.
132    
133    15. Fixed an occurrence of == in configure.ac that should have been = (shell
134        scripts are not C programs :-) and which was not noticed because it works
135        on Linux.
136    
137    16. pcretest is supposed to handle any length of pattern and data line (as one
138        line or as a continued sequence of lines) by extending its input buffer if
139        necessary. This feature was broken for very long pattern lines, leading to
140        a string of junk being passed to pcre_compile() if the pattern was longer
141        than about 50K.
142    
143    17. I have done a major re-factoring of the way pcre_compile() computes the
144        amount of memory needed for a compiled pattern. Previously, there was code
145        that made a preliminary scan of the pattern in order to do this. That was
146        OK when PCRE was new, but as the facilities have expanded, it has become
147        harder and harder to keep it in step with the real compile phase, and there
148        have been a number of bugs (see for example, 4 above). I have now found a
149        cunning way of running the real compile function in a "fake" mode that
150        enables it to compute how much memory it would need, while actually only
151        ever using a few hundred bytes of working memory and without too many
152        tests of the mode. This should make future maintenance and development
153        easier. A side effect of this work is that the limit of 200 on the nesting
154        depth of parentheses has been removed (though this was never a serious
155        limitation, I suspect). However, there is a downside: pcre_compile() now
156        runs more slowly than before (30% or more, depending on the pattern). I
157        hope this isn't a big issue. There is no effect on runtime performance.
158    
159    18. Fixed a minor bug in pcretest: if a pattern line was not terminated by a
160        newline (only possible for the last line of a file) and it was a
161        pattern that set a locale (followed by /Lsomething), pcretest crashed.
162    
163    19. Added additional timing features to pcretest. (1) The -tm option now times
164        matching only, not compiling. (2) Both -t and -tm can be followed, as a
165        separate command line item, by a number that specifies the number of
166        repeats to use when timing. The default is 50000; this gives better
167        precision, but takes uncomfortably long for very large patterns.
168    
169    20. Extended pcre_study() to be more clever in cases where a branch of a
170        subpattern has no definite first character. For example, (a*|b*)[cd] would
171        previously give no result from pcre_study(). Now it recognizes that the
172        first character must be a, b, c, or d.
173    
174    21. There was an incorrect error "recursive call could loop indefinitely" if
175        a subpattern (or the entire pattern) that was being tested for matching an
176        empty string contained only one non-empty item after a nested subpattern.
177        For example, the pattern (?>\x{100}*)\d(?R) provoked this error
178        incorrectly, because the \d was being skipped in the check.
179    
180    22. The pcretest program now has a new pattern option /B and a command line
181        option -b, which is equivalent to adding /B to every pattern. This causes
182        it to show the compiled bytecode, without the additional information that
183        -d shows. The effect of -d is now the same as -b with -i (and similarly, /D
184        is the same as /B/I).
185    
186    23. A new optimization is now able automatically to treat some sequences such
187        as a*b as a*+b. More specifically, if something simple (such as a character
188        or a simple class like \d) has an unlimited quantifier, and is followed by
189        something that cannot possibly match the quantified thing, the quantifier
190        is automatically "possessified".
191    
192    24. A recursive reference to a subpattern whose number was greater than 39
193        went wrong under certain circumstances in UTF-8 mode. This bug could also
194        have affected the operation of pcre_study().
195    
196    25. Realized that a little bit of performance could be had by replacing
197        (c & 0xc0) == 0xc0 with c >= 0xc0 when processing UTF-8 characters.
198    
199    26. Timing data from pcretest is now shown to 4 decimal places instead of 3.
200    
201    27. Possessive quantifiers such as a++ were previously implemented by turning
202        them into atomic groups such as ($>a+). Now they have their own opcodes,
203        which improves performance. This includes the automatically created ones
204        from 23 above.
205    
206    28. A pattern such as (?=(\w+))\1: which simulates an atomic group using a
207        lookahead was broken if it was not anchored. PCRE was mistakenly expecting
208        the first matched character to be a colon. This applied both to named and
209        numbered groups.
210    
211    29. The ucpinternal.h header file was missing its idempotency #ifdef.
212    
213    30. I was sent a "project" file called libpcre.a.dev which I understand makes
214        building PCRE on Windows easier, so I have included it in the distribution.
215    
216    31. There is now a check in pcretest against a ridiculously large number being
217        returned by pcre_exec() or pcre_dfa_exec(). If this happens in a /g or /G
218        loop, the loop is abandoned.
219    
220    32. Forward references to subpatterns in conditions such as (?(2)...) where
221        subpattern 2 is defined later cause pcre_compile() to search forwards in
222        the pattern for the relevant set of parentheses. This search went wrong
223        when there were unescaped parentheses in a character class, parentheses
224        escaped with \Q...\E, or parentheses in a #-comment in /x mode.
225    
226    33. "Subroutine" calls and backreferences were previously restricted to
227        referencing subpatterns earlier in the regex. This restriction has now
228        been removed.
229    
230    34. Added a number of extra features that are going to be in Perl 5.10. On the
231        whole, these are just syntactic alternatives for features that PCRE had
232        previously implemented using the Python syntax or my own invention. The
233        other formats are all retained for compatibility.
234    
235        (a) Named groups can now be defined as (?<name>...) or (?'name'...) as well
236            as (?P<name>...). The new forms, as well as being in Perl 5.10, are
237            also .NET compatible.
238    
239        (b) A recursion or subroutine call to a named group can now be defined as
240            (?&name) as well as (?P>name).
241    
242        (c) A backreference to a named group can now be defined as \k<name> or
243            \k'name' as well as (?P=name). The new forms, as well as being in Perl
244            5.10, are also .NET compatible.
245    
246        (d) A conditional reference to a named group can now use the syntax
247            (?(<name>) or (?('name') as well as (?(name).
248    
249        (e) A "conditional group" of the form (?(DEFINE)...) can be used to define
250            groups (named and numbered) that are never evaluated inline, but can be
251            called as "subroutines" from elsewhere. In effect, the DEFINE condition
252            is always false. There may be only one alternative in such a group.
253    
254        (f) A test for recursion can be given as (?(R1).. or (?(R&name)... as well
255            as the simple (?(R). The condition is true only if the most recent
256            recursion is that of the given number or name. It does not search out
257            through the entire recursion stack.
258    
259        (g) The escape \gN or \g{N} has been added, where N is a positive or
260            negative number, specifying an absolute or relative reference.
261    
262    35. Tidied to get rid of some further signed/unsigned compiler warnings and
263        some "unreachable code" warnings.
264    
265    36. Updated the Unicode property tables to Unicode version 5.0.0. Amongst other
266        things, this adds five new scripts.
267    
268    37. Perl ignores orphaned \E escapes completely. PCRE now does the same.
269        There were also incompatibilities regarding the handling of \Q..\E inside
270        character classes, for example with patterns like [\Qa\E-\Qz\E] where the
271        hyphen was adjacent to \Q or \E. I hope I've cleared all this up now.
272    
273    38. Like Perl, PCRE detects when an indefinitely repeated parenthesized group
274        matches an empty string, and forcibly breaks the loop. There were bugs in
275        this code in non-simple cases. For a pattern such as  ^(a()*)*  matched
276        against  aaaa  the result was just "a" rather than "aaaa", for example. Two
277        separate and independent bugs (that affected different cases) have been
278        fixed.
279    
280    39. Refactored the code to abolish the use of different opcodes for small
281        capturing bracket numbers. This is a tidy that I avoided doing when I
282        removed the limit on the number of capturing brackets for 3.5 back in 2001.
283        The new approach is not only tidier, it makes it possible to reduce the
284        memory needed to fix the previous bug (38).
285    
286    40. Implemented PCRE_NEWLINE_ANY to recognize any of the Unicode newline
287        sequences (http://unicode.org/unicode/reports/tr18/) as "newline" when
288        processing dot, circumflex, or dollar metacharacters, or #-comments in /x
289        mode.
290    
291    41. Add \R to match any Unicode newline sequence, as suggested in the Unicode
292        report.
293    
294    42. Applied patch, originally from Ari Pollak, modified by Google, to allow
295        copy construction and assignment in the C++ wrapper.
296    
297    43. Updated pcregrep to support "--newline=any". In the process, I fixed a
298        couple of bugs that could have given wrong results in the "--newline=crlf"
299        case.
300    
301    44. Added a number of casts and did some reorganization of signed/unsigned int
302        variables following suggestions from Dair Grant. Also renamed the variable
303        "this" as "item" because it is a C++ keyword.
304    
305    45. Arranged for dftables to add
306    
307          #include "pcre_internal.h"
308    
309        to pcre_chartables.c because without it, gcc 4.x may remove the array
310        definition from the final binary if PCRE is built into a static library and
311        dead code stripping is activated.
312    
313    46. For an unanchored pattern, if a match attempt fails at the start of a
314        newline sequence, and the newline setting is CRLF or ANY, and the next two
315        characters are CRLF, advance by two characters instead of one.
316    
317    
318    Version 6.7 04-Jul-06
319    ---------------------
320    
321     1. In order to handle tests when input lines are enormously long, pcretest has
322        been re-factored so that it automatically extends its buffers when
323        necessary. The code is crude, but this _is_ just a test program. The
324        default size has been increased from 32K to 50K.
325    
326     2. The code in pcre_study() was using the value of the re argument before
327        testing it for NULL. (Of course, in any sensible call of the function, it
328        won't be NULL.)
329    
330     3. The memmove() emulation function in pcre_internal.h, which is used on
331        systems that lack both memmove() and bcopy() - that is, hardly ever -
332        was missing a "static" storage class specifier.
333    
334     4. When UTF-8 mode was not set, PCRE looped when compiling certain patterns
335        containing an extended class (one that cannot be represented by a bitmap
336        because it contains high-valued characters or Unicode property items, e.g.
337        [\pZ]). Almost always one would set UTF-8 mode when processing such a
338        pattern, but PCRE should not loop if you do not (it no longer does).
339        [Detail: two cases were found: (a) a repeated subpattern containing an
340        extended class; (b) a recursive reference to a subpattern that followed a
341        previous extended class. It wasn't skipping over the extended class
342        correctly when UTF-8 mode was not set.]
343    
344     5. A negated single-character class was not being recognized as fixed-length
345        in lookbehind assertions such as (?<=[^f]), leading to an incorrect
346        compile error "lookbehind assertion is not fixed length".
347    
348     6. The RunPerlTest auxiliary script was showing an unexpected difference
349        between PCRE and Perl for UTF-8 tests. It turns out that it is hard to
350        write a Perl script that can interpret lines of an input file either as
351        byte characters or as UTF-8, which is what "perltest" was being required to
352        do for the non-UTF-8 and UTF-8 tests, respectively. Essentially what you
353        can't do is switch easily at run time between having the "use utf8;" pragma
354        or not. In the end, I fudged it by using the RunPerlTest script to insert
355        "use utf8;" explicitly for the UTF-8 tests.
356    
357     7. In multiline (/m) mode, PCRE was matching ^ after a terminating newline at
358        the end of the subject string, contrary to the documentation and to what
359        Perl does. This was true of both matching functions. Now it matches only at
360        the start of the subject and immediately after *internal* newlines.
361    
362     8. A call of pcre_fullinfo() from pcretest to get the option bits was passing
363        a pointer to an int instead of a pointer to an unsigned long int. This
364        caused problems on 64-bit systems.
365    
366     9. Applied a patch from the folks at Google to pcrecpp.cc, to fix "another
367        instance of the 'standard' template library not being so standard".
368    
369    10. There was no check on the number of named subpatterns nor the maximum
370        length of a subpattern name. The product of these values is used to compute
371        the size of the memory block for a compiled pattern. By supplying a very
372        long subpattern name and a large number of named subpatterns, the size
373        computation could be caused to overflow. This is now prevented by limiting
374        the length of names to 32 characters, and the number of named subpatterns
375        to 10,000.
376    
377    11. Subpatterns that are repeated with specific counts have to be replicated in
378        the compiled pattern. The size of memory for this was computed from the
379        length of the subpattern and the repeat count. The latter is limited to
380        65535, but there was no limit on the former, meaning that integer overflow
381        could in principle occur. The compiled length of a repeated subpattern is
382        now limited to 30,000 bytes in order to prevent this.
383    
384    12. Added the optional facility to have named substrings with the same name.
385    
386    13. Added the ability to use a named substring as a condition, using the
387        Python syntax: (?(name)yes|no). This overloads (?(R)... and names that
388        are numbers (not recommended). Forward references are permitted.
389    
390    14. Added forward references in named backreferences (if you see what I mean).
391    
392    15. In UTF-8 mode, with the PCRE_DOTALL option set, a quantified dot in the
393        pattern could run off the end of the subject. For example, the pattern
394        "(?s)(.{1,5})"8 did this with the subject "ab".
395    
396    16. If PCRE_DOTALL or PCRE_MULTILINE were set, pcre_dfa_exec() behaved as if
397        PCRE_CASELESS was set when matching characters that were quantified with ?
398        or *.
399    
400    17. A character class other than a single negated character that had a minimum
401        but no maximum quantifier - for example [ab]{6,} - was not handled
402        correctly by pce_dfa_exec(). It would match only one character.
403    
404    18. A valid (though odd) pattern that looked like a POSIX character
405        class but used an invalid character after [ (for example [[,abc,]]) caused
406        pcre_compile() to give the error "Failed: internal error: code overflow" or
407        in some cases to crash with a glibc free() error. This could even happen if
408        the pattern terminated after [[ but there just happened to be a sequence of
409        letters, a binary zero, and a closing ] in the memory that followed.
410    
411    19. Perl's treatment of octal escapes in the range \400 to \777 has changed
412        over the years. Originally (before any Unicode support), just the bottom 8
413        bits were taken. Thus, for example, \500 really meant \100. Nowadays the
414        output from "man perlunicode" includes this:
415    
416          The regular expression compiler produces polymorphic opcodes.  That
417          is, the pattern adapts to the data and automatically switches to
418          the Unicode character scheme when presented with Unicode data--or
419          instead uses a traditional byte scheme when presented with byte
420          data.
421    
422        Sadly, a wide octal escape does not cause a switch, and in a string with
423        no other multibyte characters, these octal escapes are treated as before.
424        Thus, in Perl, the pattern  /\500/ actually matches \100 but the pattern
425        /\500|\x{1ff}/ matches \500 or \777 because the whole thing is treated as a
426        Unicode string.
427    
428        I have not perpetrated such confusion in PCRE. Up till now, it took just
429        the bottom 8 bits, as in old Perl. I have now made octal escapes with
430        values greater than \377 illegal in non-UTF-8 mode. In UTF-8 mode they
431        translate to the appropriate multibyte character.
432    
433    29. Applied some refactoring to reduce the number of warnings from Microsoft
434        and Borland compilers. This has included removing the fudge introduced
435        seven years ago for the OS/2 compiler (see 2.02/2 below) because it caused
436        a warning about an unused variable.
437    
438    21. PCRE has not included VT (character 0x0b) in the set of whitespace
439        characters since release 4.0, because Perl (from release 5.004) does not.
440        [Or at least, is documented not to: some releases seem to be in conflict
441        with the documentation.] However, when a pattern was studied with
442        pcre_study() and all its branches started with \s, PCRE still included VT
443        as a possible starting character. Of course, this did no harm; it just
444        caused an unnecessary match attempt.
445    
446    22. Removed a now-redundant internal flag bit that recorded the fact that case
447        dependency changed within the pattern. This was once needed for "required
448        byte" processing, but is no longer used. This recovers a now-scarce options
449        bit. Also moved the least significant internal flag bit to the most-
450        significant bit of the word, which was not previously used (hangover from
451        the days when it was an int rather than a uint) to free up another bit for
452        the future.
453    
454    23. Added support for CRLF line endings as well as CR and LF. As well as the
455        default being selectable at build time, it can now be changed at runtime
456        via the PCRE_NEWLINE_xxx flags. There are now options for pcregrep to
457        specify that it is scanning data with non-default line endings.
458    
459    24. Changed the definition of CXXLINK to make it agree with the definition of
460        LINK in the Makefile, by replacing LDFLAGS to CXXFLAGS.
461    
462    25. Applied Ian Taylor's patches to avoid using another stack frame for tail
463        recursions. This makes a big different to stack usage for some patterns.
464    
465    26. If a subpattern containing a named recursion or subroutine reference such
466        as (?P>B) was quantified, for example (xxx(?P>B)){3}, the calculation of
467        the space required for the compiled pattern went wrong and gave too small a
468        value. Depending on the environment, this could lead to "Failed: internal
469        error: code overflow at offset 49" or "glibc detected double free or
470        corruption" errors.
471    
472    27. Applied patches from Google (a) to support the new newline modes and (b) to
473        advance over multibyte UTF-8 characters in GlobalReplace.
474    
475    28. Change free() to pcre_free() in pcredemo.c. Apparently this makes a
476        difference for some implementation of PCRE in some Windows version.
477    
478    29. Added some extra testing facilities to pcretest:
479    
480        \q<number>   in a data line sets the "match limit" value
481        \Q<number>   in a data line sets the "match recursion limt" value
482        -S <number>  sets the stack size, where <number> is in megabytes
483    
484        The -S option isn't available for Windows.
485    
486    
487    Version 6.6 06-Feb-06
488    ---------------------
489    
490     1. Change 16(a) for 6.5 broke things, because PCRE_DATA_SCOPE was not defined
491        in pcreposix.h. I have copied the definition from pcre.h.
492    
493     2. Change 25 for 6.5 broke compilation in a build directory out-of-tree
494        because pcre.h is no longer a built file.
495    
496     3. Added Jeff Friedl's additional debugging patches to pcregrep. These are
497        not normally included in the compiled code.
498    
499    
500    Version 6.5 01-Feb-06
501    ---------------------
502    
503     1. When using the partial match feature with pcre_dfa_exec(), it was not
504        anchoring the second and subsequent partial matches at the new starting
505        point. This could lead to incorrect results. For example, with the pattern
506        /1234/, partially matching against "123" and then "a4" gave a match.
507    
508     2. Changes to pcregrep:
509    
510        (a) All non-match returns from pcre_exec() were being treated as failures
511            to match the line. Now, unless the error is PCRE_ERROR_NOMATCH, an
512            error message is output. Some extra information is given for the
513            PCRE_ERROR_MATCHLIMIT and PCRE_ERROR_RECURSIONLIMIT errors, which are
514            probably the only errors that are likely to be caused by users (by
515            specifying a regex that has nested indefinite repeats, for instance).
516            If there are more than 20 of these errors, pcregrep is abandoned.
517    
518        (b) A binary zero was treated as data while matching, but terminated the
519            output line if it was written out. This has been fixed: binary zeroes
520            are now no different to any other data bytes.
521    
522        (c) Whichever of the LC_ALL or LC_CTYPE environment variables is set is
523            used to set a locale for matching. The --locale=xxxx long option has
524            been added (no short equivalent) to specify a locale explicitly on the
525            pcregrep command, overriding the environment variables.
526    
527        (d) When -B was used with -n, some line numbers in the output were one less
528            than they should have been.
529    
530        (e) Added the -o (--only-matching) option.
531    
532        (f) If -A or -C was used with -c (count only), some lines of context were
533            accidentally printed for the final match.
534    
535        (g) Added the -H (--with-filename) option.
536    
537        (h) The combination of options -rh failed to suppress file names for files
538            that were found from directory arguments.
539    
540        (i) Added the -D (--devices) and -d (--directories) options.
541    
542        (j) Added the -F (--fixed-strings) option.
543    
544        (k) Allow "-" to be used as a file name for -f as well as for a data file.
545    
546        (l) Added the --colo(u)r option.
547    
548        (m) Added Jeffrey Friedl's -S testing option, but within #ifdefs so that it
549            is not present by default.
550    
551     3. A nasty bug was discovered in the handling of recursive patterns, that is,
552        items such as (?R) or (?1), when the recursion could match a number of
553        alternatives. If it matched one of the alternatives, but subsequently,
554        outside the recursion, there was a failure, the code tried to back up into
555        the recursion. However, because of the way PCRE is implemented, this is not
556        possible, and the result was an incorrect result from the match.
557    
558        In order to prevent this happening, the specification of recursion has
559        been changed so that all such subpatterns are automatically treated as
560        atomic groups. Thus, for example, (?R) is treated as if it were (?>(?R)).
561    
562     4. I had overlooked the fact that, in some locales, there are characters for
563        which isalpha() is true but neither isupper() nor islower() are true. In
564        the fr_FR locale, for instance, the \xAA and \xBA characters (ordmasculine
565        and ordfeminine) are like this. This affected the treatment of \w and \W
566        when they appeared in character classes, but not when they appeared outside
567        a character class. The bit map for "word" characters is now created
568        separately from the results of isalnum() instead of just taking it from the
569        upper, lower, and digit maps. (Plus the underscore character, of course.)
570    
571     5. The above bug also affected the handling of POSIX character classes such as
572        [[:alpha:]] and [[:alnum:]]. These do not have their own bit maps in PCRE's
573        permanent tables. Instead, the bit maps for such a class were previously
574        created as the appropriate unions of the upper, lower, and digit bitmaps.
575        Now they are created by subtraction from the [[:word:]] class, which has
576        its own bitmap.
577    
578     6. The [[:blank:]] character class matches horizontal, but not vertical space.
579        It is created by subtracting the vertical space characters (\x09, \x0a,
580        \x0b, \x0c) from the [[:space:]] bitmap. Previously, however, the
581        subtraction was done in the overall bitmap for a character class, meaning
582        that a class such as [\x0c[:blank:]] was incorrect because \x0c would not
583        be recognized. This bug has been fixed.
584    
585     7. Patches from the folks at Google:
586    
587          (a) pcrecpp.cc: "to handle a corner case that may or may not happen in
588          real life, but is still worth protecting against".
589    
590          (b) pcrecpp.cc: "corrects a bug when negative radixes are used with
591          regular expressions".
592    
593          (c) pcre_scanner.cc: avoid use of std::count() because not all systems
594          have it.
595    
596          (d) Split off pcrecpparg.h from pcrecpp.h and had the former built by
597          "configure" and the latter not, in order to fix a problem somebody had
598          with compiling the Arg class on HP-UX.
599    
600          (e) Improve the error-handling of the C++ wrapper a little bit.
601    
602          (f) New tests for checking recursion limiting.
603    
604     8. The pcre_memmove() function, which is used only if the environment does not
605        have a standard memmove() function (and is therefore rarely compiled),
606        contained two bugs: (a) use of int instead of size_t, and (b) it was not
607        returning a result (though PCRE never actually uses the result).
608    
609     9. In the POSIX regexec() interface, if nmatch is specified as a ridiculously
610        large number - greater than INT_MAX/(3*sizeof(int)) - REG_ESPACE is
611        returned instead of calling malloc() with an overflowing number that would
612        most likely cause subsequent chaos.
613    
614    10. The debugging option of pcretest was not showing the NO_AUTO_CAPTURE flag.
615    
616    11. The POSIX flag REG_NOSUB is now supported. When a pattern that was compiled
617        with this option is matched, the nmatch and pmatch options of regexec() are
618        ignored.
619    
620    12. Added REG_UTF8 to the POSIX interface. This is not defined by POSIX, but is
621        provided in case anyone wants to the the POSIX interface with UTF-8
622        strings.
623    
624    13. Added CXXLDFLAGS to the Makefile parameters to provide settings only on the
625        C++ linking (needed for some HP-UX environments).
626    
627    14. Avoid compiler warnings in get_ucpname() when compiled without UCP support
628        (unused parameter) and in the pcre_printint() function (omitted "default"
629        switch label when the default is to do nothing).
630    
631    15. Added some code to make it possible, when PCRE is compiled as a C++
632        library, to replace subject pointers for pcre_exec() with a smart pointer
633        class, thus making it possible to process discontinuous strings.
634    
635    16. The two macros PCRE_EXPORT and PCRE_DATA_SCOPE are confusing, and perform
636        much the same function. They were added by different people who were trying
637        to make PCRE easy to compile on non-Unix systems. It has been suggested
638        that PCRE_EXPORT be abolished now that there is more automatic apparatus
639        for compiling on Windows systems. I have therefore replaced it with
640        PCRE_DATA_SCOPE. This is set automatically for Windows; if not set it
641        defaults to "extern" for C or "extern C" for C++, which works fine on
642        Unix-like systems. It is now possible to override the value of PCRE_DATA_
643        SCOPE with something explicit in config.h. In addition:
644    
645        (a) pcreposix.h still had just "extern" instead of either of these macros;
646            I have replaced it with PCRE_DATA_SCOPE.
647    
648        (b) Functions such as _pcre_xclass(), which are internal to the library,
649            but external in the C sense, all had PCRE_EXPORT in their definitions.
650            This is apparently wrong for the Windows case, so I have removed it.
651            (It makes no difference on Unix-like systems.)
652    
653    17. Added a new limit, MATCH_LIMIT_RECURSION, which limits the depth of nesting
654        of recursive calls to match(). This is different to MATCH_LIMIT because
655        that limits the total number of calls to match(), not all of which increase
656        the depth of recursion. Limiting the recursion depth limits the amount of
657        stack (or heap if NO_RECURSE is set) that is used. The default can be set
658        when PCRE is compiled, and changed at run time. A patch from Google adds
659        this functionality to the C++ interface.
660    
661    18. Changes to the handling of Unicode character properties:
662    
663        (a) Updated the table to Unicode 4.1.0.
664    
665        (b) Recognize characters that are not in the table as "Cn" (undefined).
666    
667        (c) I revised the way the table is implemented to a much improved format
668            which includes recognition of ranges. It now supports the ranges that
669            are defined in UnicodeData.txt, and it also amalgamates other
670            characters into ranges. This has reduced the number of entries in the
671            table from around 16,000 to around 3,000, thus reducing its size
672            considerably. I realized I did not need to use a tree structure after
673            all - a binary chop search is just as efficient. Having reduced the
674            number of entries, I extended their size from 6 bytes to 8 bytes to
675            allow for more data.
676    
677        (d) Added support for Unicode script names via properties such as \p{Han}.
678    
679    19. In UTF-8 mode, a backslash followed by a non-Ascii character was not
680        matching that character.
681    
682    20. When matching a repeated Unicode property with a minimum greater than zero,
683        (for example \pL{2,}), PCRE could look past the end of the subject if it
684        reached it while seeking the minimum number of characters. This could
685        happen only if some of the characters were more than one byte long, because
686        there is a check for at least the minimum number of bytes.
687    
688    21. Refactored the implementation of \p and \P so as to be more general, to
689        allow for more different types of property in future. This has changed the
690        compiled form incompatibly. Anybody with saved compiled patterns that use
691        \p or \P will have to recompile them.
692    
693    22. Added "Any" and "L&" to the supported property types.
694    
695    23. Recognize \x{...} as a code point specifier, even when not in UTF-8 mode,
696        but give a compile time error if the value is greater than 0xff.
697    
698    24. The man pages for pcrepartial, pcreprecompile, and pcre_compile2 were
699        accidentally not being installed or uninstalled.
700    
701    25. The pcre.h file was built from pcre.h.in, but the only changes that were
702        made were to insert the current release number. This seemed silly, because
703        it made things harder for people building PCRE on systems that don't run
704        "configure". I have turned pcre.h into a distributed file, no longer built
705        by "configure", with the version identification directly included. There is
706        no longer a pcre.h.in file.
707    
708        However, this change necessitated a change to the pcre-config script as
709        well. It is built from pcre-config.in, and one of the substitutions was the
710        release number. I have updated configure.ac so that ./configure now finds
711        the release number by grepping pcre.h.
712    
713    26. Added the ability to run the tests under valgrind.
714    
715    
716    Version 6.4 05-Sep-05
717    ---------------------
718    
719     1. Change 6.0/10/(l) to pcregrep introduced a bug that caused separator lines
720        "--" to be printed when multiple files were scanned, even when none of the
721        -A, -B, or -C options were used. This is not compatible with Gnu grep, so I
722        consider it to be a bug, and have restored the previous behaviour.
723    
724     2. A couple of code tidies to get rid of compiler warnings.
725    
726     3. The pcretest program used to cheat by referring to symbols in the library
727        whose names begin with _pcre_. These are internal symbols that are not
728        really supposed to be visible externally, and in some environments it is
729        possible to suppress them. The cheating is now confined to including
730        certain files from the library's source, which is a bit cleaner.
731    
732     4. Renamed pcre.in as pcre.h.in to go with pcrecpp.h.in; it also makes the
733        file's purpose clearer.
734    
735     5. Reorganized pcre_ucp_findchar().
736    
737    
738    Version 6.3 15-Aug-05
739    ---------------------
740    
741     1. The file libpcre.pc.in did not have general read permission in the tarball.
742    
743     2. There were some problems when building without C++ support:
744    
745        (a) If C++ support was not built, "make install" and "make test" still
746            tried to test it.
747    
748        (b) There were problems when the value of CXX was explicitly set. Some
749            changes have been made to try to fix these, and ...
750    
751        (c) --disable-cpp can now be used to explicitly disable C++ support.
752    
753        (d) The use of @CPP_OBJ@ directly caused a blank line preceded by a
754            backslash in a target when C++ was disabled. This confuses some
755            versions of "make", apparently. Using an intermediate variable solves
756            this. (Same for CPP_LOBJ.)
757    
758     3. $(LINK_FOR_BUILD) now includes $(CFLAGS_FOR_BUILD) and $(LINK)
759        (non-Windows) now includes $(CFLAGS) because these flags are sometimes
760        necessary on certain architectures.
761    
762     4. Added a setting of -export-symbols-regex to the link command to remove
763        those symbols that are exported in the C sense, but actually are local
764        within the library, and not documented. Their names all begin with
765        "_pcre_". This is not a perfect job, because (a) we have to except some
766        symbols that pcretest ("illegally") uses, and (b) the facility isn't always
767        available (and never for static libraries). I have made a note to try to
768        find a way round (a) in the future.
769    
770    
771    Version 6.2 01-Aug-05
772    ---------------------
773    
774     1. There was no test for integer overflow of quantifier values. A construction
775        such as {1111111111111111} would give undefined results. What is worse, if
776        a minimum quantifier for a parenthesized subpattern overflowed and became
777        negative, the calculation of the memory size went wrong. This could have
778        led to memory overwriting.
779    
780     2. Building PCRE using VPATH was broken. Hopefully it is now fixed.
781    
782     3. Added "b" to the 2nd argument of fopen() in dftables.c, for non-Unix-like
783        operating environments where this matters.
784    
785     4. Applied Giuseppe Maxia's patch to add additional features for controlling
786        PCRE options from within the C++ wrapper.
787    
788     5. Named capturing subpatterns were not being correctly counted when a pattern
789        was compiled. This caused two problems: (a) If there were more than 100
790        such subpatterns, the calculation of the memory needed for the whole
791        compiled pattern went wrong, leading to an overflow error. (b) Numerical
792        back references of the form \12, where the number was greater than 9, were
793        not recognized as back references, even though there were sufficient
794        previous subpatterns.
795    
796     6. Two minor patches to pcrecpp.cc in order to allow it to compile on older
797        versions of gcc, e.g. 2.95.4.
798    
799    
800    Version 6.1 21-Jun-05
801    ---------------------
802    
803     1. There was one reference to the variable "posix" in pcretest.c that was not
804        surrounded by "#if !defined NOPOSIX".
805    
806     2. Make it possible to compile pcretest without DFA support, UTF8 support, or
807        the cross-check on the old pcre_info() function, for the benefit of the
808        cut-down version of PCRE that is currently imported into Exim.
809    
810     3. A (silly) pattern starting with (?i)(?-i) caused an internal space
811        allocation error. I've done the easy fix, which wastes 2 bytes for sensible
812        patterns that start (?i) but I don't think that matters. The use of (?i) is
813        just an example; this all applies to the other options as well.
814    
815     4. Since libtool seems to echo the compile commands it is issuing, the output
816        from "make" can be reduced a bit by putting "@" in front of each libtool
817        compile command.
818    
819     5. Patch from the folks at Google for configure.in to be a bit more thorough
820        in checking for a suitable C++ installation before trying to compile the
821        C++ stuff. This should fix a reported problem when a compiler was present,
822        but no suitable headers.
823    
824     6. The man pages all had just "PCRE" as their title. I have changed them to
825        be the relevant file name. I have also arranged that these names are
826        retained in the file doc/pcre.txt, which is a concatenation in text format
827        of all the man pages except the little individual ones for each function.
828    
829     7. The NON-UNIX-USE file had not been updated for the different set of source
830        files that come with release 6. I also added a few comments about the C++
831        wrapper.
832    
833    
834    Version 6.0 07-Jun-05
835    ---------------------
836    
837     1. Some minor internal re-organization to help with my DFA experiments.
838    
839     2. Some missing #ifdef SUPPORT_UCP conditionals in pcretest and printint that
840        didn't matter for the library itself when fully configured, but did matter
841        when compiling without UCP support, or within Exim, where the ucp files are
842        not imported.
843    
844     3. Refactoring of the library code to split up the various functions into
845        different source modules. The addition of the new DFA matching code (see
846        below) to a single monolithic source would have made it really too
847        unwieldy, quite apart from causing all the code to be include in a
848        statically linked application, when only some functions are used. This is
849        relevant even without the DFA addition now that patterns can be compiled in
850        one application and matched in another.
851    
852        The downside of splitting up is that there have to be some external
853        functions and data tables that are used internally in different modules of
854        the library but which are not part of the API. These have all had their
855        names changed to start with "_pcre_" so that they are unlikely to clash
856        with other external names.
857    
858     4. Added an alternate matching function, pcre_dfa_exec(), which matches using
859        a different (DFA) algorithm. Although it is slower than the original
860        function, it does have some advantages for certain types of matching
861        problem.
862    
863     5. Upgrades to pcretest in order to test the features of pcre_dfa_exec(),
864        including restarting after a partial match.
865    
866     6. A patch for pcregrep that defines INVALID_FILE_ATTRIBUTES if it is not
867        defined when compiling for Windows was sent to me. I have put it into the
868        code, though I have no means of testing or verifying it.
869    
870     7. Added the pcre_refcount() auxiliary function.
871    
872     8. Added the PCRE_FIRSTLINE option. This constrains an unanchored pattern to
873        match before or at the first newline in the subject string. In pcretest,
874        the /f option on a pattern can be used to set this.
875    
876     9. A repeated \w when used in UTF-8 mode with characters greater than 256
877        would behave wrongly. This has been present in PCRE since release 4.0.
878    
879    10. A number of changes to the pcregrep command:
880    
881        (a) Refactored how -x works; insert ^(...)$ instead of setting
882            PCRE_ANCHORED and checking the length, in preparation for adding
883            something similar for -w.
884    
885        (b) Added the -w (match as a word) option.
886    
887        (c) Refactored the way lines are read and buffered so as to have more
888            than one at a time available.
889    
890        (d) Implemented a pcregrep test script.
891    
892        (e) Added the -M (multiline match) option. This allows patterns to match
893            over several lines of the subject. The buffering ensures that at least
894            8K, or the rest of the document (whichever is the shorter) is available
895            for matching (and similarly the previous 8K for lookbehind assertions).
896    
897        (f) Changed the --help output so that it now says
898    
899              -w, --word-regex(p)
900    
901            instead of two lines, one with "regex" and the other with "regexp"
902            because that confused at least one person since the short forms are the
903            same. (This required a bit of code, as the output is generated
904            automatically from a table. It wasn't just a text change.)
905    
906        (g) -- can be used to terminate pcregrep options if the next thing isn't an
907            option but starts with a hyphen. Could be a pattern or a path name
908            starting with a hyphen, for instance.
909    
910        (h) "-" can be given as a file name to represent stdin.
911    
912        (i) When file names are being printed, "(standard input)" is used for
913            the standard input, for compatibility with GNU grep. Previously
914            "<stdin>" was used.
915    
916        (j) The option --label=xxx can be used to supply a name to be used for
917            stdin when file names are being printed. There is no short form.
918    
919        (k) Re-factored the options decoding logic because we are going to add
920            two more options that take data. Such options can now be given in four
921            different ways, e.g. "-fname", "-f name", "--file=name", "--file name".
922    
923        (l) Added the -A, -B, and -C options for requesting that lines of context
924            around matches be printed.
925    
926        (m) Added the -L option to print the names of files that do not contain
927            any matching lines, that is, the complement of -l.
928    
929        (n) The return code is 2 if any file cannot be opened, but pcregrep does
930            continue to scan other files.
931    
932        (o) The -s option was incorrectly implemented. For compatibility with other
933            greps, it now suppresses the error message for a non-existent or non-
934            accessible file (but not the return code). There is a new option called
935            -q that suppresses the output of matching lines, which was what -s was
936            previously doing.
937    
938        (p) Added --include and --exclude options to specify files for inclusion
939            and exclusion when recursing.
940    
941    11. The Makefile was not using the Autoconf-supported LDFLAGS macro properly.
942        Hopefully, it now does.
943    
944    12. Missing cast in pcre_study().
945    
946    13. Added an "uninstall" target to the makefile.
947    
948    14. Replaced "extern" in the function prototypes in Makefile.in with
949        "PCRE_DATA_SCOPE", which defaults to 'extern' or 'extern "C"' in the Unix
950        world, but is set differently for Windows.
951    
952    15. Added a second compiling function called pcre_compile2(). The only
953        difference is that it has an extra argument, which is a pointer to an
954        integer error code. When there is a compile-time failure, this is set
955        non-zero, in addition to the error test pointer being set to point to an
956        error message. The new argument may be NULL if no error number is required
957        (but then you may as well call pcre_compile(), which is now just a
958        wrapper). This facility is provided because some applications need a
959        numeric error indication, but it has also enabled me to tidy up the way
960        compile-time errors are handled in the POSIX wrapper.
961    
962    16. Added VPATH=.libs to the makefile; this should help when building with one
963        prefix path and installing with another. (Or so I'm told by someone who
964        knows more about this stuff than I do.)
965    
966    17. Added a new option, REG_DOTALL, to the POSIX function regcomp(). This
967        passes PCRE_DOTALL to the pcre_compile() function, making the "." character
968        match everything, including newlines. This is not POSIX-compatible, but
969        somebody wanted the feature. From pcretest it can be activated by using
970        both the P and the s flags.
971    
972    18. AC_PROG_LIBTOOL appeared twice in Makefile.in. Removed one.
973    
974    19. libpcre.pc was being incorrectly installed as executable.
975    
976    20. A couple of places in pcretest check for end-of-line by looking for '\n';
977        it now also looks for '\r' so that it will work unmodified on Windows.
978    
979    21. Added Google's contributed C++ wrapper to the distribution.
980    
981    22. Added some untidy missing memory free() calls in pcretest, to keep
982        Electric Fence happy when testing.
983    
984    
985    
986    Version 5.0 13-Sep-04
987    ---------------------
988    
989     1. Internal change: literal characters are no longer packed up into items
990        containing multiple characters in a single byte-string. Each character
991        is now matched using a separate opcode. However, there may be more than one
992        byte in the character in UTF-8 mode.
993    
994     2. The pcre_callout_block structure has two new fields: pattern_position and
995        next_item_length. These contain the offset in the pattern to the next match
996        item, and its length, respectively.
997    
998     3. The PCRE_AUTO_CALLOUT option for pcre_compile() requests the automatic
999        insertion of callouts before each pattern item. Added the /C option to
1000        pcretest to make use of this.
1001    
1002     4. On the advice of a Windows user, the lines
1003    
1004          #if defined(_WIN32) || defined(WIN32)
1005          _setmode( _fileno( stdout ), 0x8000 );
1006          #endif  /* defined(_WIN32) || defined(WIN32) */
1007    
1008        have been added to the source of pcretest. This apparently does useful
1009        magic in relation to line terminators.
1010    
1011     5. Changed "r" and "w" in the calls to fopen() in pcretest to "rb" and "wb"
1012        for the benefit of those environments where the "b" makes a difference.
1013    
1014     6. The icc compiler has the same options as gcc, but "configure" doesn't seem
1015        to know about it. I have put a hack into configure.in that adds in code
1016        to set GCC=yes if CC=icc. This seems to end up at a point in the
1017        generated configure script that is early enough to affect the setting of
1018        compiler options, which is what is needed, but I have no means of testing
1019        whether it really works. (The user who reported this had patched the
1020        generated configure script, which of course I cannot do.)
1021    
1022        LATER: After change 22 below (new libtool files), the configure script
1023        seems to know about icc (and also ecc). Therefore, I have commented out
1024        this hack in configure.in.
1025    
1026     7. Added support for pkg-config (2 patches were sent in).
1027    
1028     8. Negated POSIX character classes that used a combination of internal tables
1029        were completely broken. These were [[:^alpha:]], [[:^alnum:]], and
1030        [[:^ascii]]. Typically, they would match almost any characters. The other
1031        POSIX classes were not broken in this way.
1032    
1033     9. Matching the pattern "\b.*?" against "ab cd", starting at offset 1, failed
1034        to find the match, as PCRE was deluded into thinking that the match had to
1035        start at the start point or following a newline. The same bug applied to
1036        patterns with negative forward assertions or any backward assertions
1037        preceding ".*" at the start, unless the pattern required a fixed first
1038        character. This was a failing pattern: "(?!.bcd).*". The bug is now fixed.
1039    
1040    10. In UTF-8 mode, when moving forwards in the subject after a failed match
1041        starting at the last subject character, bytes beyond the end of the subject
1042        string were read.
1043    
1044    11. Renamed the variable "class" as "classbits" to make life easier for C++
1045        users. (Previously there was a macro definition, but it apparently wasn't
1046        enough.)
1047    
1048    12. Added the new field "tables" to the extra data so that tables can be passed
1049        in at exec time, or the internal tables can be re-selected. This allows
1050        a compiled regex to be saved and re-used at a later time by a different
1051        program that might have everything at different addresses.
1052    
1053    13. Modified the pcre-config script so that, when run on Solaris, it shows a
1054        -R library as well as a -L library.
1055    
1056    14. The debugging options of pcretest (-d on the command line or D on a
1057        pattern) showed incorrect output for anything following an extended class
1058        that contained multibyte characters and which was followed by a quantifier.
1059    
1060    15. Added optional support for general category Unicode character properties
1061        via the \p, \P, and \X escapes. Unicode property support implies UTF-8
1062        support. It adds about 90K to the size of the library. The meanings of the
1063        inbuilt class escapes such as \d and \s have NOT been changed.
1064    
1065    16. Updated pcredemo.c to include calls to free() to release the memory for the
1066        compiled pattern.
1067    
1068    17. The generated file chartables.c was being created in the source directory
1069        instead of in the building directory. This caused the build to fail if the
1070        source directory was different from the building directory, and was
1071        read-only.
1072    
1073    18. Added some sample Win commands from Mark Tetrode into the NON-UNIX-USE
1074        file. No doubt somebody will tell me if they don't make sense... Also added
1075        Dan Mooney's comments about building on OpenVMS.
1076    
1077    19. Added support for partial matching via the PCRE_PARTIAL option for
1078        pcre_exec() and the \P data escape in pcretest.
1079    
1080    20. Extended pcretest with 3 new pattern features:
1081    
1082        (i)   A pattern option of the form ">rest-of-line" causes pcretest to
1083              write the compiled pattern to the file whose name is "rest-of-line".
1084              This is a straight binary dump of the data, with the saved pointer to
1085              the character tables forced to be NULL. The study data, if any, is
1086              written too. After writing, pcretest reads a new pattern.
1087    
1088        (ii)  If, instead of a pattern, "<rest-of-line" is given, pcretest reads a
1089              compiled pattern from the given file. There must not be any
1090              occurrences of "<" in the file name (pretty unlikely); if there are,
1091              pcretest will instead treat the initial "<" as a pattern delimiter.
1092              After reading in the pattern, pcretest goes on to read data lines as
1093              usual.
1094    
1095        (iii) The F pattern option causes pcretest to flip the bytes in the 32-bit
1096              and 16-bit fields in a compiled pattern, to simulate a pattern that
1097              was compiled on a host of opposite endianness.
1098    
1099    21. The pcre-exec() function can now cope with patterns that were compiled on
1100        hosts of opposite endianness, with this restriction:
1101    
1102          As for any compiled expression that is saved and used later, the tables
1103          pointer field cannot be preserved; the extra_data field in the arguments
1104          to pcre_exec() should be used to pass in a tables address if a value
1105          other than the default internal tables were used at compile time.
1106    
1107    22. Calling pcre_exec() with a negative value of the "ovecsize" parameter is
1108        now diagnosed as an error. Previously, most of the time, a negative number
1109        would have been treated as zero, but if in addition "ovector" was passed as
1110        NULL, a crash could occur.
1111    
1112    23. Updated the files ltmain.sh, config.sub, config.guess, and aclocal.m4 with
1113        new versions from the libtool 1.5 distribution (the last one is a copy of
1114        a file called libtool.m4). This seems to have fixed the need to patch
1115        "configure" to support Darwin 1.3 (which I used to do). However, I still
1116        had to patch ltmain.sh to ensure that ${SED} is set (it isn't on my
1117        workstation).
1118    
1119    24. Changed the PCRE licence to be the more standard "BSD" licence.
1120    
1121    
1122    Version 4.5 01-Dec-03
1123    ---------------------
1124    
1125     1. There has been some re-arrangement of the code for the match() function so
1126        that it can be compiled in a version that does not call itself recursively.
1127        Instead, it keeps those local variables that need separate instances for
1128        each "recursion" in a frame on the heap, and gets/frees frames whenever it
1129        needs to "recurse". Keeping track of where control must go is done by means
1130        of setjmp/longjmp. The whole thing is implemented by a set of macros that
1131        hide most of the details from the main code, and operates only if
1132        NO_RECURSE is defined while compiling pcre.c. If PCRE is built using the
1133        "configure" mechanism, "--disable-stack-for-recursion" turns on this way of
1134        operating.
1135    
1136        To make it easier for callers to provide specially tailored get/free
1137        functions for this usage, two new functions, pcre_stack_malloc, and
1138        pcre_stack_free, are used. They are always called in strict stacking order,
1139        and the size of block requested is always the same.
1140    
1141        The PCRE_CONFIG_STACKRECURSE info parameter can be used to find out whether
1142        PCRE has been compiled to use the stack or the heap for recursion. The
1143        -C option of pcretest uses this to show which version is compiled.
1144    
1145        A new data escape \S, is added to pcretest; it causes the amounts of store
1146        obtained and freed by both kinds of malloc/free at match time to be added
1147        to the output.
1148    
1149     2. Changed the locale test to use "fr_FR" instead of "fr" because that's
1150        what's available on my current Linux desktop machine.
1151    
1152     3. When matching a UTF-8 string, the test for a valid string at the start has
1153        been extended. If start_offset is not zero, PCRE now checks that it points
1154        to a byte that is the start of a UTF-8 character. If not, it returns
1155        PCRE_ERROR_BADUTF8_OFFSET (-11). Note: the whole string is still checked;
1156        this is necessary because there may be backward assertions in the pattern.
1157        When matching the same subject several times, it may save resources to use
1158        PCRE_NO_UTF8_CHECK on all but the first call if the string is long.
1159    
1160     4. The code for checking the validity of UTF-8 strings has been tightened so
1161        that it rejects (a) strings containing 0xfe or 0xff bytes and (b) strings
1162        containing "overlong sequences".
1163    
1164     5. Fixed a bug (appearing twice) that I could not find any way of exploiting!
1165        I had written "if ((digitab[*p++] && chtab_digit) == 0)" where the "&&"
1166        should have been "&", but it just so happened that all the cases this let
1167        through by mistake were picked up later in the function.
1168    
1169     6. I had used a variable called "isblank" - this is a C99 function, causing
1170        some compilers to warn. To avoid this, I renamed it (as "blankclass").
1171    
1172     7. Cosmetic: (a) only output another newline at the end of pcretest if it is
1173        prompting; (b) run "./pcretest /dev/null" at the start of the test script
1174        so the version is shown; (c) stop "make test" echoing "./RunTest".
1175    
1176     8. Added patches from David Burgess to enable PCRE to run on EBCDIC systems.
1177    
1178     9. The prototype for memmove() for systems that don't have it was using
1179        size_t, but the inclusion of the header that defines size_t was later. I've
1180        moved the #includes for the C headers earlier to avoid this.
1181    
1182    10. Added some adjustments to the code to make it easier to compiler on certain
1183        special systems:
1184    
1185          (a) Some "const" qualifiers were missing.
1186          (b) Added the macro EXPORT before all exported functions; by default this
1187              is defined to be empty.
1188          (c) Changed the dftables auxiliary program (that builds chartables.c) so
1189              that it reads its output file name as an argument instead of writing
1190              to the standard output and assuming this can be redirected.
1191    
1192    11. In UTF-8 mode, if a recursive reference (e.g. (?1)) followed a character
1193        class containing characters with values greater than 255, PCRE compilation
1194        went into a loop.
1195    
1196    12. A recursive reference to a subpattern that was within another subpattern
1197        that had a minimum quantifier of zero caused PCRE to crash. For example,
1198        (x(y(?2))z)? provoked this bug with a subject that got as far as the
1199        recursion. If the recursively-called subpattern itself had a zero repeat,
1200        that was OK.
1201    
1202    13. In pcretest, the buffer for reading a data line was set at 30K, but the
1203        buffer into which it was copied (for escape processing) was still set at
1204        1024, so long lines caused crashes.
1205    
1206    14. A pattern such as /[ab]{1,3}+/ failed to compile, giving the error
1207        "internal error: code overflow...". This applied to any character class
1208        that was followed by a possessive quantifier.
1209    
1210    15. Modified the Makefile to add libpcre.la as a prerequisite for
1211        libpcreposix.la because I was told this is needed for a parallel build to
1212        work.
1213    
1214    16. If a pattern that contained .* following optional items at the start was
1215        studied, the wrong optimizing data was generated, leading to matching
1216        errors. For example, studying /[ab]*.*c/ concluded, erroneously, that any
1217        matching string must start with a or b or c. The correct conclusion for
1218        this pattern is that a match can start with any character.
1219    
1220    
1221    Version 4.4 13-Aug-03
1222    ---------------------
1223    
1224     1. In UTF-8 mode, a character class containing characters with values between
1225        127 and 255 was not handled correctly if the compiled pattern was studied.
1226        In fixing this, I have also improved the studying algorithm for such
1227        classes (slightly).
1228    
1229     2. Three internal functions had redundant arguments passed to them. Removal
1230        might give a very teeny performance improvement.
1231    
1232     3. Documentation bug: the value of the capture_top field in a callout is *one
1233        more than* the number of the hightest numbered captured substring.
1234    
1235     4. The Makefile linked pcretest and pcregrep with -lpcre, which could result
1236        in incorrectly linking with a previously installed version. They now link
1237        explicitly with libpcre.la.
1238    
1239     5. configure.in no longer needs to recognize Cygwin specially.
1240    
1241     6. A problem in pcre.in for Windows platforms is fixed.
1242    
1243     7. If a pattern was successfully studied, and the -d (or /D) flag was given to
1244        pcretest, it used to include the size of the study block as part of its
1245        output. Unfortunately, the structure contains a field that has a different
1246        size on different hardware architectures. This meant that the tests that
1247        showed this size failed. As the block is currently always of a fixed size,
1248        this information isn't actually particularly useful in pcretest output, so
1249        I have just removed it.
1250    
1251     8. Three pre-processor statements accidentally did not start in column 1.
1252        Sadly, there are *still* compilers around that complain, even though
1253        standard C has not required this for well over a decade. Sigh.
1254    
1255     9. In pcretest, the code for checking callouts passed small integers in the
1256        callout_data field, which is a void * field. However, some picky compilers
1257        complained about the casts involved for this on 64-bit systems. Now
1258        pcretest passes the address of the small integer instead, which should get
1259        rid of the warnings.
1260    
1261    10. By default, when in UTF-8 mode, PCRE now checks for valid UTF-8 strings at
1262        both compile and run time, and gives an error if an invalid UTF-8 sequence
1263        is found. There is a option for disabling this check in cases where the
1264        string is known to be correct and/or the maximum performance is wanted.
1265    
1266    11. In response to a bug report, I changed one line in Makefile.in from
1267    
1268            -Wl,--out-implib,.libs/lib@WIN_PREFIX@pcreposix.dll.a \
1269        to
1270            -Wl,--out-implib,.libs/@WIN_PREFIX@libpcreposix.dll.a \
1271    
1272        to look similar to other lines, but I have no way of telling whether this
1273        is the right thing to do, as I do not use Windows. No doubt I'll get told
1274        if it's wrong...
1275    
1276    
1277    Version 4.3 21-May-03
1278    ---------------------
1279    
1280    1. Two instances of @WIN_PREFIX@ omitted from the Windows targets in the
1281       Makefile.
1282    
1283    2. Some refactoring to improve the quality of the code:
1284    
1285       (i)   The utf8_table... variables are now declared "const".
1286    
1287       (ii)  The code for \cx, which used the "case flipping" table to upper case
1288             lower case letters, now just substracts 32. This is ASCII-specific,
1289             but the whole concept of \cx is ASCII-specific, so it seems
1290             reasonable.
1291    
1292       (iii) PCRE was using its character types table to recognize decimal and
1293             hexadecimal digits in the pattern. This is silly, because it handles
1294             only 0-9, a-f, and A-F, but the character types table is locale-
1295             specific, which means strange things might happen. A private
1296             table is now used for this - though it costs 256 bytes, a table is
1297             much faster than multiple explicit tests. Of course, the standard
1298             character types table is still used for matching digits in subject
1299             strings against \d.
1300    
1301       (iv)  Strictly, the identifier ESC_t is reserved by POSIX (all identifiers
1302             ending in _t are). So I've renamed it as ESC_tee.
1303    
1304    3. The first argument for regexec() in the POSIX wrapper should have been
1305       defined as "const".
1306    
1307    4. Changed pcretest to use malloc() for its buffers so that they can be
1308       Electric Fenced for debugging.
1309    
1310    5. There were several places in the code where, in UTF-8 mode, PCRE would try
1311       to read one or more bytes before the start of the subject string. Often this
1312       had no effect on PCRE's behaviour, but in some circumstances it could
1313       provoke a segmentation fault.
1314    
1315    6. A lookbehind at the start of a pattern in UTF-8 mode could also cause PCRE
1316       to try to read one or more bytes before the start of the subject string.
1317    
1318    7. A lookbehind in a pattern matched in non-UTF-8 mode on a PCRE compiled with
1319       UTF-8 support could misbehave in various ways if the subject string
1320       contained bytes with the 0x80 bit set and the 0x40 bit unset in a lookbehind
1321       area. (PCRE was not checking for the UTF-8 mode flag, and trying to move
1322       back over UTF-8 characters.)
1323    
1324    
1325    Version 4.2 14-Apr-03
1326    ---------------------
1327    
1328    1. Typo "#if SUPPORT_UTF8" instead of "#ifdef SUPPORT_UTF8" fixed.
1329    
1330    2. Changes to the building process, supplied by Ronald Landheer-Cieslak
1331         [ON_WINDOWS]: new variable, "#" on non-Windows platforms
1332         [NOT_ON_WINDOWS]: new variable, "#" on Windows platforms
1333         [WIN_PREFIX]: new variable, "cyg" for Cygwin
1334         * Makefile.in: use autoconf substitution for OBJEXT, EXEEXT, BUILD_OBJEXT
1335           and BUILD_EXEEXT
1336         Note: automatic setting of the BUILD variables is not yet working
1337         set CPPFLAGS and BUILD_CPPFLAGS (but don't use yet) - should be used at
1338           compile-time but not at link-time
1339         [LINK]: use for linking executables only
1340         make different versions for Windows and non-Windows
1341         [LINKLIB]: new variable, copy of UNIX-style LINK, used for linking
1342           libraries
1343         [LINK_FOR_BUILD]: new variable
1344         [OBJEXT]: use throughout
1345         [EXEEXT]: use throughout
1346         <winshared>: new target
1347         <wininstall>: new target
1348         <dftables.o>: use native compiler
1349         <dftables>: use native linker
1350         <install>: handle Windows platform correctly
1351         <clean>: ditto
1352         <check>: ditto
1353         copy DLL to top builddir before testing
1354    
1355       As part of these changes, -no-undefined was removed again. This was reported
1356       to give trouble on HP-UX 11.0, so getting rid of it seems like a good idea
1357       in any case.
1358    
1359    3. Some tidies to get rid of compiler warnings:
1360    
1361       . In the match_data structure, match_limit was an unsigned long int, whereas
1362         match_call_count was an int. I've made them both unsigned long ints.
1363    
1364       . In pcretest the fact that a const uschar * doesn't automatically cast to
1365         a void * provoked a warning.
1366    
1367       . Turning on some more compiler warnings threw up some "shadow" variables
1368         and a few more missing casts.
1369    
1370    4. If PCRE was complied with UTF-8 support, but called without the PCRE_UTF8
1371       option, a class that contained a single character with a value between 128
1372       and 255 (e.g. /[\xFF]/) caused PCRE to crash.
1373    
1374    5. If PCRE was compiled with UTF-8 support, but called without the PCRE_UTF8
1375       option, a class that contained several characters, but with at least one
1376       whose value was between 128 and 255 caused PCRE to crash.
1377    
1378    
1379    Version 4.1 12-Mar-03
1380    ---------------------
1381    
1382    1. Compiling with gcc -pedantic found a couple of places where casts were
1383    needed, and a string in dftables.c that was longer than standard compilers are
1384    required to support.
1385    
1386    2. Compiling with Sun's compiler found a few more places where the code could
1387    be tidied up in order to avoid warnings.
1388    
1389    3. The variables for cross-compiling were called HOST_CC and HOST_CFLAGS; the
1390    first of these names is deprecated in the latest Autoconf in favour of the name
1391    CC_FOR_BUILD, because "host" is typically used to mean the system on which the
1392    compiled code will be run. I can't find a reference for HOST_CFLAGS, but by
1393    analogy I have changed it to CFLAGS_FOR_BUILD.
1394    
1395    4. Added -no-undefined to the linking command in the Makefile, because this is
1396    apparently helpful for Windows. To make it work, also added "-L. -lpcre" to the
1397    linking step for the pcreposix library.
1398    
1399    5. PCRE was failing to diagnose the case of two named groups with the same
1400    name.
1401    
1402    6. A problem with one of PCRE's optimizations was discovered. PCRE remembers a
1403    literal character that is needed in the subject for a match, and scans along to
1404    ensure that it is present before embarking on the full matching process. This
1405    saves time in cases of nested unlimited repeats that are never going to match.
1406    Problem: the scan can take a lot of time if the subject is very long (e.g.
1407    megabytes), thus penalizing straightforward matches. It is now done only if the
1408    amount of subject to be scanned is less than 1000 bytes.
1409    
1410    7. A lesser problem with the same optimization is that it was recording the
1411    first character of an anchored pattern as "needed", thus provoking a search
1412    right along the subject, even when the first match of the pattern was going to
1413    fail. The "needed" character is now not set for anchored patterns, unless it
1414    follows something in the pattern that is of non-fixed length. Thus, it still
1415    fulfils its original purpose of finding quick non-matches in cases of nested
1416    unlimited repeats, but isn't used for simple anchored patterns such as /^abc/.
1417    
1418    
1419    Version 4.0 17-Feb-03
1420    ---------------------
1421    
1422    1. If a comment in an extended regex that started immediately after a meta-item
1423    extended to the end of string, PCRE compiled incorrect data. This could lead to
1424    all kinds of weird effects. Example: /#/ was bad; /()#/ was bad; /a#/ was not.
1425    
1426    2. Moved to autoconf 2.53 and libtool 1.4.2.
1427    
1428    3. Perl 5.8 no longer needs "use utf8" for doing UTF-8 things. Consequently,
1429    the special perltest8 script is no longer needed - all the tests can be run
1430    from a single perltest script.
1431    
1432    4. From 5.004, Perl has not included the VT character (0x0b) in the set defined
1433    by \s. It has now been removed in PCRE. This means it isn't recognized as
1434    whitespace in /x regexes too, which is the same as Perl. Note that the POSIX
1435    class [:space:] *does* include VT, thereby creating a mess.
1436    
1437    5. Added the class [:blank:] (a GNU extension from Perl 5.8) to match only
1438    space and tab.
1439    
1440    6. Perl 5.005 was a long time ago. It's time to amalgamate the tests that use
1441    its new features into the main test script, reducing the number of scripts.
1442    
1443    7. Perl 5.8 has changed the meaning of patterns like /a(?i)b/. Earlier versions
1444    were backward compatible, and made the (?i) apply to the whole pattern, as if
1445    /i were given. Now it behaves more logically, and applies the option setting
1446    only to what follows. PCRE has been changed to follow suit. However, if it
1447    finds options settings right at the start of the pattern, it extracts them into
1448    the global options, as before. Thus, they show up in the info data.
1449    
1450    8. Added support for the \Q...\E escape sequence. Characters in between are
1451    treated as literals. This is slightly different from Perl in that $ and @ are
1452    also handled as literals inside the quotes. In Perl, they will cause variable
1453    interpolation. Note the following examples:
1454    
1455        Pattern            PCRE matches      Perl matches
1456    
1457        \Qabc$xyz\E        abc$xyz           abc followed by the contents of $xyz
1458        \Qabc\$xyz\E       abc\$xyz          abc\$xyz
1459        \Qabc\E\$\Qxyz\E   abc$xyz           abc$xyz
1460    
1461    For compatibility with Perl, \Q...\E sequences are recognized inside character
1462    classes as well as outside them.
1463    
1464    9. Re-organized 3 code statements in pcretest to avoid "overflow in
1465    floating-point constant arithmetic" warnings from a Microsoft compiler. Added a
1466    (size_t) cast to one statement in pcretest and one in pcreposix to avoid
1467    signed/unsigned warnings.
1468    
1469    10. SunOS4 doesn't have strtoul(). This was used only for unpicking the -o
1470    option for pcretest, so I've replaced it by a simple function that does just
1471    that job.
1472    
1473    11. pcregrep was ending with code 0 instead of 2 for the commands "pcregrep" or
1474    "pcregrep -".
1475    
1476    12. Added "possessive quantifiers" ?+, *+, ++, and {,}+ which come from Sun's
1477    Java package. This provides some syntactic sugar for simple cases of what my
1478    documentation calls "once-only subpatterns". A pattern such as x*+ is the same
1479    as (?>x*). In other words, if what is inside (?>...) is just a single repeated
1480    item, you can use this simplified notation. Note that only makes sense with
1481    greedy quantifiers. Consequently, the use of the possessive quantifier forces
1482    greediness, whatever the setting of the PCRE_UNGREEDY option.
1483    
1484    13. A change of greediness default within a pattern was not taking effect at
1485    the current level for patterns like /(b+(?U)a+)/. It did apply to parenthesized
1486    subpatterns that followed. Patterns like /b+(?U)a+/ worked because the option
1487    was abstracted outside.
1488    
1489    14. PCRE now supports the \G assertion. It is true when the current matching
1490    position is at the start point of the match. This differs from \A when the
1491    starting offset is non-zero. Used with the /g option of pcretest (or similar
1492    code), it works in the same way as it does for Perl's /g option. If all
1493    alternatives of a regex begin with \G, the expression is anchored to the start
1494    match position, and the "anchored" flag is set in the compiled expression.
1495    
1496    15. Some bugs concerning the handling of certain option changes within patterns
1497    have been fixed. These applied to options other than (?ims). For example,
1498    "a(?x: b c )d" did not match "XabcdY" but did match "Xa b c dY". It should have
1499    been the other way round. Some of this was related to change 7 above.
1500    
1501    16. PCRE now gives errors for /[.x.]/ and /[=x=]/ as unsupported POSIX
1502    features, as Perl does. Previously, PCRE gave the warnings only for /[[.x.]]/
1503    and /[[=x=]]/. PCRE now also gives an error for /[:name:]/ because it supports
1504    POSIX classes only within a class (e.g. /[[:alpha:]]/).
1505    
1506    17. Added support for Perl's \C escape. This matches one byte, even in UTF8
1507    mode. Unlike ".", it always matches newline, whatever the setting of
1508    PCRE_DOTALL. However, PCRE does not permit \C to appear in lookbehind
1509    assertions. Perl allows it, but it doesn't (in general) work because it can't
1510    calculate the length of the lookbehind. At least, that's the case for Perl
1511    5.8.0 - I've been told they are going to document that it doesn't work in
1512    future.
1513    
1514    18. Added an error diagnosis for escapes that PCRE does not support: these are
1515    \L, \l, \N, \P, \p, \U, \u, and \X.
1516    
1517    19. Although correctly diagnosing a missing ']' in a character class, PCRE was
1518    reading past the end of the pattern in cases such as /[abcd/.
1519    
1520    20. PCRE was getting more memory than necessary for patterns with classes that
1521    contained both POSIX named classes and other characters, e.g. /[[:space:]abc/.
1522    
1523    21. Added some code, conditional on #ifdef VPCOMPAT, to make life easier for
1524    compiling PCRE for use with Virtual Pascal.
1525    
1526    22. Small fix to the Makefile to make it work properly if the build is done
1527    outside the source tree.
1528    
1529    23. Added a new extension: a condition to go with recursion. If a conditional
1530    subpattern starts with (?(R) the "true" branch is used if recursion has
1531    happened, whereas the "false" branch is used only at the top level.
1532    
1533    24. When there was a very long string of literal characters (over 255 bytes
1534    without UTF support, over 250 bytes with UTF support), the computation of how
1535    much memory was required could be incorrect, leading to segfaults or other
1536    strange effects.
1537    
1538    25. PCRE was incorrectly assuming anchoring (either to start of subject or to
1539    start of line for a non-DOTALL pattern) when a pattern started with (.*) and
1540    there was a subsequent back reference to those brackets. This meant that, for
1541    example, /(.*)\d+\1/ failed to match "abc123bc". Unfortunately, it isn't
1542    possible to check for precisely this case. All we can do is abandon the
1543    optimization if .* occurs inside capturing brackets when there are any back
1544    references whatsoever. (See below for a better fix that came later.)
1545    
1546    26. The handling of the optimization for finding the first character of a
1547    non-anchored pattern, and for finding a character that is required later in the
1548    match were failing in some cases. This didn't break the matching; it just
1549    failed to optimize when it could. The way this is done has been re-implemented.
1550    
1551    27. Fixed typo in error message for invalid (?R item (it said "(?p").
1552    
1553    28. Added a new feature that provides some of the functionality that Perl
1554    provides with (?{...}). The facility is termed a "callout". The way it is done
1555    in PCRE is for the caller to provide an optional function, by setting
1556    pcre_callout to its entry point. Like pcre_malloc and pcre_free, this is a
1557    global variable. By default it is unset, which disables all calling out. To get
1558    the function called, the regex must include (?C) at appropriate points. This
1559    is, in fact, equivalent to (?C0), and any number <= 255 may be given with (?C).
1560    This provides a means of identifying different callout points. When PCRE
1561    reaches such a point in the regex, if pcre_callout has been set, the external
1562    function is called. It is provided with data in a structure called
1563    pcre_callout_block, which is defined in pcre.h. If the function returns 0,
1564    matching continues; if it returns a non-zero value, the match at the current
1565    point fails. However, backtracking will occur if possible. [This was changed
1566    later and other features added - see item 49 below.]
1567    
1568    29. pcretest is upgraded to test the callout functionality. It provides a
1569    callout function that displays information. By default, it shows the start of
1570    the match and the current position in the text. There are some new data escapes
1571    to vary what happens:
1572    
1573        \C+         in addition, show current contents of captured substrings
1574        \C-         do not supply a callout function
1575        \C!n        return 1 when callout number n is reached
1576        \C!n!m      return 1 when callout number n is reached for the mth time
1577    
1578    30. If pcregrep was called with the -l option and just a single file name, it
1579    output "<stdin>" if a match was found, instead of the file name.
1580    
1581    31. Improve the efficiency of the POSIX API to PCRE. If the number of capturing
1582    slots is less than POSIX_MALLOC_THRESHOLD, use a block on the stack to pass to
1583    pcre_exec(). This saves a malloc/free per call. The default value of
1584    POSIX_MALLOC_THRESHOLD is 10; it can be changed by --with-posix-malloc-threshold
1585    when configuring.
1586    
1587    32. The default maximum size of a compiled pattern is 64K. There have been a
1588    few cases of people hitting this limit. The code now uses macros to handle the
1589    storing of links as offsets within the compiled pattern. It defaults to 2-byte
1590    links, but this can be changed to 3 or 4 bytes by --with-link-size when
1591    configuring. Tests 2 and 5 work only with 2-byte links because they output
1592    debugging information about compiled patterns.
1593    
1594    33. Internal code re-arrangements:
1595    
1596    (a) Moved the debugging function for printing out a compiled regex into
1597        its own source file (printint.c) and used #include to pull it into
1598        pcretest.c and, when DEBUG is defined, into pcre.c, instead of having two
1599        separate copies.
1600    
1601    (b) Defined the list of op-code names for debugging as a macro in
1602        internal.h so that it is next to the definition of the opcodes.
1603    
1604    (c) Defined a table of op-code lengths for simpler skipping along compiled
1605        code. This is again a macro in internal.h so that it is next to the
1606        definition of the opcodes.
1607    
1608    34. Added support for recursive calls to individual subpatterns, along the
1609    lines of Robin Houston's patch (but implemented somewhat differently).
1610    
1611    35. Further mods to the Makefile to help Win32. Also, added code to pcregrep to
1612    allow it to read and process whole directories in Win32. This code was
1613    contributed by Lionel Fourquaux; it has not been tested by me.
1614    
1615    36. Added support for named subpatterns. The Python syntax (?P<name>...) is
1616    used to name a group. Names consist of alphanumerics and underscores, and must
1617    be unique. Back references use the syntax (?P=name) and recursive calls use
1618    (?P>name) which is a PCRE extension to the Python extension. Groups still have
1619    numbers. The function pcre_fullinfo() can be used after compilation to extract
1620    a name/number map. There are three relevant calls:
1621    
1622      PCRE_INFO_NAMEENTRYSIZE        yields the size of each entry in the map
1623      PCRE_INFO_NAMECOUNT            yields the number of entries
1624      PCRE_INFO_NAMETABLE            yields a pointer to the map.
1625    
1626    The map is a vector of fixed-size entries. The size of each entry depends on
1627    the length of the longest name used. The first two bytes of each entry are the
1628    group number, most significant byte first. There follows the corresponding
1629    name, zero terminated. The names are in alphabetical order.
1630    
1631    37. Make the maximum literal string in the compiled code 250 for the non-UTF-8
1632    case instead of 255. Making it the same both with and without UTF-8 support
1633    means that the same test output works with both.
1634    
1635    38. There was a case of malloc(0) in the POSIX testing code in pcretest. Avoid
1636    calling malloc() with a zero argument.
1637    
1638    39. Change 25 above had to resort to a heavy-handed test for the .* anchoring
1639    optimization. I've improved things by keeping a bitmap of backreferences with
1640    numbers 1-31 so that if .* occurs inside capturing brackets that are not in
1641    fact referenced, the optimization can be applied. It is unlikely that a
1642    relevant occurrence of .* (i.e. one which might indicate anchoring or forcing
1643    the match to follow \n) will appear inside brackets with a number greater than
1644    31, but if it does, any back reference > 31 suppresses the optimization.
1645    
1646    40. Added a new compile-time option PCRE_NO_AUTO_CAPTURE. This has the effect
1647    of disabling numbered capturing parentheses. Any opening parenthesis that is
1648    not followed by ? behaves as if it were followed by ?: but named parentheses
1649    can still be used for capturing (and they will acquire numbers in the usual
1650    way).
1651    
1652    41. Redesigned the return codes from the match() function into yes/no/error so
1653    that errors can be passed back from deep inside the nested calls. A malloc
1654    failure while inside a recursive subpattern call now causes the
1655    PCRE_ERROR_NOMEMORY return instead of quietly going wrong.
1656    
1657    42. It is now possible to set a limit on the number of times the match()
1658    function is called in a call to pcre_exec(). This facility makes it possible to
1659    limit the amount of recursion and backtracking, though not in a directly
1660    obvious way, because the match() function is used in a number of different
1661    circumstances. The count starts from zero for each position in the subject
1662    string (for non-anchored patterns). The default limit is, for compatibility, a
1663    large number, namely 10 000 000. You can change this in two ways:
1664    
1665    (a) When configuring PCRE before making, you can use --with-match-limit=n
1666        to set a default value for the compiled library.
1667    
1668    (b) For each call to pcre_exec(), you can pass a pcre_extra block in which
1669        a different value is set. See 45 below.
1670    
1671    If the limit is exceeded, pcre_exec() returns PCRE_ERROR_MATCHLIMIT.
1672    
1673    43. Added a new function pcre_config(int, void *) to enable run-time extraction
1674    of things that can be changed at compile time. The first argument specifies
1675    what is wanted and the second points to where the information is to be placed.
1676    The current list of available information is:
1677    
1678      PCRE_CONFIG_UTF8
1679    
1680    The output is an integer that is set to one if UTF-8 support is available;
1681    otherwise it is set to zero.
1682    
1683      PCRE_CONFIG_NEWLINE
1684    
1685    The output is an integer that it set to the value of the code that is used for
1686    newline. It is either LF (10) or CR (13).
1687    
1688      PCRE_CONFIG_LINK_SIZE
1689    
1690    The output is an integer that contains the number of bytes used for internal
1691    linkage in compiled expressions. The value is 2, 3, or 4. See item 32 above.
1692    
1693      PCRE_CONFIG_POSIX_MALLOC_THRESHOLD
1694    
1695    The output is an integer that contains the threshold above which the POSIX
1696    interface uses malloc() for output vectors. See item 31 above.
1697    
1698      PCRE_CONFIG_MATCH_LIMIT
1699    
1700    The output is an unsigned integer that contains the default limit of the number
1701    of match() calls in a pcre_exec() execution. See 42 above.
1702    
1703    44. pcretest has been upgraded by the addition of the -C option. This causes it
1704    to extract all the available output from the new pcre_config() function, and to
1705    output it. The program then exits immediately.
1706    
1707    45. A need has arisen to pass over additional data with calls to pcre_exec() in
1708    order to support additional features. One way would have been to define
1709    pcre_exec2() (for example) with extra arguments, but this would not have been
1710    extensible, and would also have required all calls to the original function to
1711    be mapped to the new one. Instead, I have chosen to extend the mechanism that
1712    is used for passing in "extra" data from pcre_study().
1713    
1714    The pcre_extra structure is now exposed and defined in pcre.h. It currently
1715    contains the following fields:
1716    
1717      flags         a bitmap indicating which of the following fields are set
1718      study_data    opaque data from pcre_study()
1719      match_limit   a way of specifying a limit on match() calls for a specific
1720                      call to pcre_exec()
1721      callout_data  data for callouts (see 49 below)
1722    
1723    The flag bits are also defined in pcre.h, and are
1724    
1725      PCRE_EXTRA_STUDY_DATA
1726      PCRE_EXTRA_MATCH_LIMIT
1727      PCRE_EXTRA_CALLOUT_DATA
1728    
1729    The pcre_study() function now returns one of these new pcre_extra blocks, with
1730    the actual study data pointed to by the study_data field, and the
1731    PCRE_EXTRA_STUDY_DATA flag set. This can be passed directly to pcre_exec() as
1732    before. That is, this change is entirely upwards-compatible and requires no
1733    change to existing code.
1734    
1735    If you want to pass in additional data to pcre_exec(), you can either place it
1736    in a pcre_extra block provided by pcre_study(), or create your own pcre_extra
1737    block.
1738    
1739    46. pcretest has been extended to test the PCRE_EXTRA_MATCH_LIMIT feature. If a
1740    data string contains the escape sequence \M, pcretest calls pcre_exec() several
1741    times with different match limits, until it finds the minimum value needed for
1742    pcre_exec() to complete. The value is then output. This can be instructive; for
1743    most simple matches the number is quite small, but for pathological cases it
1744    gets very large very quickly.
1745    
1746    47. There's a new option for pcre_fullinfo() called PCRE_INFO_STUDYSIZE. It
1747    returns the size of the data block pointed to by the study_data field in a
1748    pcre_extra block, that is, the value that was passed as the argument to
1749    pcre_malloc() when PCRE was getting memory in which to place the information
1750    created by pcre_study(). The fourth argument should point to a size_t variable.
1751    pcretest has been extended so that this information is shown after a successful
1752    pcre_study() call when information about the compiled regex is being displayed.
1753    
1754    48. Cosmetic change to Makefile: there's no need to have / after $(DESTDIR)
1755    because what follows is always an absolute path. (Later: it turns out that this
1756    is more than cosmetic for MinGW, because it doesn't like empty path
1757    components.)
1758    
1759    49. Some changes have been made to the callout feature (see 28 above):
1760    
1761    (i)  A callout function now has three choices for what it returns:
1762    
1763           0  =>  success, carry on matching
1764         > 0  =>  failure at this point, but backtrack if possible
1765         < 0  =>  serious error, return this value from pcre_exec()
1766    
1767         Negative values should normally be chosen from the set of PCRE_ERROR_xxx
1768         values. In particular, returning PCRE_ERROR_NOMATCH forces a standard
1769         "match failed" error. The error number PCRE_ERROR_CALLOUT is reserved for
1770         use by callout functions. It will never be used by PCRE itself.
1771    
1772    (ii) The pcre_extra structure (see 45 above) has a void * field called
1773         callout_data, with corresponding flag bit PCRE_EXTRA_CALLOUT_DATA. The
1774         pcre_callout_block structure has a field of the same name. The contents of
1775         the field passed in the pcre_extra structure are passed to the callout
1776         function in the corresponding field in the callout block. This makes it
1777         easier to use the same callout-containing regex from multiple threads. For
1778         testing, the pcretest program has a new data escape
1779    
1780           \C*n        pass the number n (may be negative) as callout_data
1781    
1782         If the callout function in pcretest receives a non-zero value as
1783         callout_data, it returns that value.
1784    
1785    50. Makefile wasn't handling CFLAGS properly when compiling dftables. Also,
1786    there were some redundant $(CFLAGS) in commands that are now specified as
1787    $(LINK), which already includes $(CFLAGS).
1788    
1789    51. Extensions to UTF-8 support are listed below. These all apply when (a) PCRE
1790    has been compiled with UTF-8 support *and* pcre_compile() has been compiled
1791    with the PCRE_UTF8 flag. Patterns that are compiled without that flag assume
1792    one-byte characters throughout. Note that case-insensitive matching applies
1793    only to characters whose values are less than 256. PCRE doesn't support the
1794    notion of cases for higher-valued characters.
1795    
1796    (i)   A character class whose characters are all within 0-255 is handled as
1797          a bit map, and the map is inverted for negative classes. Previously, a
1798          character > 255 always failed to match such a class; however it should
1799          match if the class was a negative one (e.g. [^ab]). This has been fixed.
1800    
1801    (ii)  A negated character class with a single character < 255 is coded as
1802          "not this character" (OP_NOT). This wasn't working properly when the test
1803          character was multibyte, either singly or repeated.
1804    
1805    (iii) Repeats of multibyte characters are now handled correctly in UTF-8
1806          mode, for example: \x{100}{2,3}.
1807    
1808    (iv)  The character escapes \b, \B, \d, \D, \s, \S, \w, and \W (either
1809          singly or repeated) now correctly test multibyte characters. However,
1810          PCRE doesn't recognize any characters with values greater than 255 as
1811          digits, spaces, or word characters. Such characters always match \D, \S,
1812          and \W, and never match \d, \s, or \w.
1813    
1814    (v)   Classes may now contain characters and character ranges with values
1815          greater than 255. For example: [ab\x{100}-\x{400}].
1816    
1817    (vi)  pcregrep now has a --utf-8 option (synonym -u) which makes it call
1818          PCRE in UTF-8 mode.
1819    
1820    52. The info request value PCRE_INFO_FIRSTCHAR has been renamed
1821    PCRE_INFO_FIRSTBYTE because it is a byte value. However, the old name is
1822    retained for backwards compatibility. (Note that LASTLITERAL is also a byte
1823    value.)
1824    
1825    53. The single man page has become too large. I have therefore split it up into
1826    a number of separate man pages. These also give rise to individual HTML pages;
1827    these are now put in a separate directory, and there is an index.html page that
1828    lists them all. Some hyperlinking between the pages has been installed.
1829    
1830    54. Added convenience functions for handling named capturing parentheses.
1831    
1832    55. Unknown escapes inside character classes (e.g. [\M]) and escapes that
1833    aren't interpreted therein (e.g. [\C]) are literals in Perl. This is now also
1834    true in PCRE, except when the PCRE_EXTENDED option is set, in which case they
1835    are faulted.
1836    
1837    56. Introduced HOST_CC and HOST_CFLAGS which can be set in the environment when
1838    calling configure. These values are used when compiling the dftables.c program
1839    which is run to generate the source of the default character tables. They
1840    default to the values of CC and CFLAGS. If you are cross-compiling PCRE,
1841    you will need to set these values.
1842    
1843    57. Updated the building process for Windows DLL, as provided by Fred Cox.
1844    
1845    
1846    Version 3.9 02-Jan-02
1847    ---------------------
1848    
1849    1. A bit of extraneous text had somehow crept into the pcregrep documentation.
1850    
1851    2. If --disable-static was given, the building process failed when trying to
1852    build pcretest and pcregrep. (For some reason it was using libtool to compile
1853    them, which is not right, as they aren't part of the library.)
1854    
1855    
1856    Version 3.8 18-Dec-01
1857    ---------------------
1858    
1859    1. The experimental UTF-8 code was completely screwed up. It was packing the
1860    bytes in the wrong order. How dumb can you get?
1861    
1862    
1863    Version 3.7 29-Oct-01
1864    ---------------------
1865    
1866    1. In updating pcretest to check change 1 of version 3.6, I screwed up.
1867    This caused pcretest, when used on the test data, to segfault. Unfortunately,
1868    this didn't happen under Solaris 8, where I normally test things.
1869    
1870    2. The Makefile had to be changed to make it work on BSD systems, where 'make'
1871    doesn't seem to recognize that ./xxx and xxx are the same file. (This entry
1872    isn't in ChangeLog distributed with 3.7 because I forgot when I hastily made
1873    this fix an hour or so after the initial 3.7 release.)
1874    
1875    
1876    Version 3.6 23-Oct-01
1877    ---------------------
1878    
1879    1. Crashed with /(sens|respons)e and \1ibility/ and "sense and sensibility" if
1880    offsets passed as NULL with zero offset count.
1881    
1882    2. The config.guess and config.sub files had not been updated when I moved to
1883    the latest autoconf.
1884    
1885    
1886    Version 3.5 15-Aug-01
1887    ---------------------
1888    
1889    1. Added some missing #if !defined NOPOSIX conditionals in pcretest.c that
1890    had been forgotten.
1891    
1892    2. By using declared but undefined structures, we can avoid using "void"
1893    definitions in pcre.h while keeping the internal definitions of the structures
1894    private.
1895    
1896    3. The distribution is now built using autoconf 2.50 and libtool 1.4. From a
1897    user point of view, this means that both static and shared libraries are built
1898    by default, but this can be individually controlled. More of the work of
1899    handling this static/shared cases is now inside libtool instead of PCRE's make
1900    file.
1901    
1902    4. The pcretest utility is now installed along with pcregrep because it is
1903    useful for users (to test regexs) and by doing this, it automatically gets
1904    relinked by libtool. The documentation has been turned into a man page, so
1905    there are now .1, .txt, and .html versions in /doc.
1906    
1907    5. Upgrades to pcregrep:
1908       (i)   Added long-form option names like gnu grep.
1909       (ii)  Added --help to list all options with an explanatory phrase.
1910       (iii) Added -r, --recursive to recurse into sub-directories.
1911       (iv)  Added -f, --file to read patterns from a file.
1912    
1913    6. pcre_exec() was referring to its "code" argument before testing that
1914    argument for NULL (and giving an error if it was NULL).
1915    
1916    7. Upgraded Makefile.in to allow for compiling in a different directory from
1917    the source directory.
1918    
1919    8. Tiny buglet in pcretest: when pcre_fullinfo() was called to retrieve the
1920    options bits, the pointer it was passed was to an int instead of to an unsigned
1921    long int. This mattered only on 64-bit systems.
1922    
1923    9. Fixed typo (3.4/1) in pcre.h again. Sigh. I had changed pcre.h (which is
1924    generated) instead of pcre.in, which it its source. Also made the same change
1925    in several of the .c files.
1926    
1927    10. A new release of gcc defines printf() as a macro, which broke pcretest
1928    because it had an ifdef in the middle of a string argument for printf(). Fixed
1929    by using separate calls to printf().
1930    
1931    11. Added --enable-newline-is-cr and --enable-newline-is-lf to the configure
1932    script, to force use of CR or LF instead of \n in the source. On non-Unix
1933    systems, the value can be set in config.h.
1934    
1935    12. The limit of 200 on non-capturing parentheses is a _nesting_ limit, not an
1936    absolute limit. Changed the text of the error message to make this clear, and
1937    likewise updated the man page.
1938    
1939    13. The limit of 99 on the number of capturing subpatterns has been removed.
1940    The new limit is 65535, which I hope will not be a "real" limit.
1941    
1942    
1943    Version 3.4 22-Aug-00
1944    ---------------------
1945    
1946    1. Fixed typo in pcre.h: unsigned const char * changed to const unsigned char *.
1947    
1948    2. Diagnose condition (?(0) as an error instead of crashing on matching.
1949    
1950    
1951    Version 3.3 01-Aug-00
1952    ---------------------
1953    
1954    1. If an octal character was given, but the value was greater than \377, it
1955    was not getting masked to the least significant bits, as documented. This could
1956    lead to crashes in some systems.
1957    
1958    2. Perl 5.6 (if not earlier versions) accepts classes like [a-\d] and treats
1959    the hyphen as a literal. PCRE used to give an error; it now behaves like Perl.
1960    
1961    3. Added the functions pcre_free_substring() and pcre_free_substring_list().
1962    These just pass their arguments on to (pcre_free)(), but they are provided
1963    because some uses of PCRE bind it to non-C systems that can call its functions,
1964    but cannot call free() or pcre_free() directly.
1965    
1966    4. Add "make test" as a synonym for "make check". Corrected some comments in
1967    the Makefile.
1968    
1969    5. Add $(DESTDIR)/ in front of all the paths in the "install" target in the
1970    Makefile.
1971    
1972    6. Changed the name of pgrep to pcregrep, because Solaris has introduced a
1973    command called pgrep for grepping around the active processes.
1974    
1975    7. Added the beginnings of support for UTF-8 character strings.
1976    
1977    8. Arranged for the Makefile to pass over the settings of CC, CFLAGS, and
1978    RANLIB to ./ltconfig so that they are used by libtool. I think these are all
1979    the relevant ones. (AR is not passed because ./ltconfig does its own figuring
1980    out for the ar command.)
1981    
1982    
1983    Version 3.2 12-May-00
1984    ---------------------
1985    
1986    This is purely a bug fixing release.
1987    
1988    1. If the pattern /((Z)+|A)*/ was matched agained ZABCDEFG it matched Z instead
1989    of ZA. This was just one example of several cases that could provoke this bug,
1990    which was introduced by change 9 of version 2.00. The code for breaking
1991    infinite loops after an iteration that matches an empty string was't working
1992    correctly.
1993    
1994    2. The pcretest program was not imitating Perl correctly for the pattern /a*/g
1995    when matched against abbab (for example). After matching an empty string, it
1996    wasn't forcing anchoring when setting PCRE_NOTEMPTY for the next attempt; this
1997    caused it to match further down the string than it should.
1998    
1999    3. The code contained an inclusion of sys/types.h. It isn't clear why this
2000    was there because it doesn't seem to be needed, and it causes trouble on some
2001    systems, as it is not a Standard C header. It has been removed.
2002    
2003    4. Made 4 silly changes to the source to avoid stupid compiler warnings that
2004    were reported on the Macintosh. The changes were from
2005    
2006      while ((c = *(++ptr)) != 0 && c != '\n');
2007    to
2008      while ((c = *(++ptr)) != 0 && c != '\n') ;
2009    
2010    Totally extraordinary, but if that's what it takes...
2011    
2012    5. PCRE is being used in one environment where neither memmove() nor bcopy() is
2013    available. Added HAVE_BCOPY and an autoconf test for it; if neither
2014    HAVE_MEMMOVE nor HAVE_BCOPY is set, use a built-in emulation function which
2015    assumes the way PCRE uses memmove() (always moving upwards).
2016    
2017    6. PCRE is being used in one environment where strchr() is not available. There
2018    was only one use in pcre.c, and writing it out to avoid strchr() probably gives
2019    faster code anyway.
2020    
2021    
2022    Version 3.1 09-Feb-00
2023    ---------------------
2024    
2025    The only change in this release is the fixing of some bugs in Makefile.in for
2026    the "install" target:
2027    
2028    (1) It was failing to install pcreposix.h.
2029    
2030    (2) It was overwriting the pcre.3 man page with the pcreposix.3 man page.
2031    
2032    
2033    Version 3.0 01-Feb-00
2034    ---------------------
2035    
2036    1. Add support for the /+ modifier to perltest (to output $` like it does in
2037    pcretest).
2038    
2039    2. Add support for the /g modifier to perltest.
2040    
2041    3. Fix pcretest so that it behaves even more like Perl for /g when the pattern
2042    matches null strings.
2043    
2044    4. Fix perltest so that it doesn't do unwanted things when fed an empty
2045    pattern. Perl treats empty patterns specially - it reuses the most recent
2046    pattern, which is not what we want. Replace // by /(?#)/ in order to avoid this
2047    effect.
2048    
2049    5. The POSIX interface was broken in that it was just handing over the POSIX
2050    captured string vector to pcre_exec(), but (since release 2.00) PCRE has
2051    required a bigger vector, with some working space on the end. This means that
2052    the POSIX wrapper now has to get and free some memory, and copy the results.
2053    
2054    6. Added some simple autoconf support, placing the test data and the
2055    documentation in separate directories, re-organizing some of the
2056    information files, and making it build pcre-config (a GNU standard). Also added
2057    libtool support for building PCRE as a shared library, which is now the
2058    default.
2059    
2060    7. Got rid of the leading zero in the definition of PCRE_MINOR because 08 and
2061    09 are not valid octal constants. Single digits will be used for minor values
2062    less than 10.
2063    
2064    8. Defined REG_EXTENDED and REG_NOSUB as zero in the POSIX header, so that
2065    existing programs that set these in the POSIX interface can use PCRE without
2066    modification.
2067    
2068    9. Added a new function, pcre_fullinfo() with an extensible interface. It can
2069    return all that pcre_info() returns, plus additional data. The pcre_info()
2070    function is retained for compatibility, but is considered to be obsolete.
2071    
2072    10. Added experimental recursion feature (?R) to handle one common case that
2073    Perl 5.6 will be able to do with (?p{...}).
2074    
2075    11. Added support for POSIX character classes like [:alpha:], which Perl is
2076    adopting.
2077    
2078    
2079    Version 2.08 31-Aug-99
2080    ----------------------
2081    
2082    1. When startoffset was not zero and the pattern began with ".*", PCRE was not
2083    trying to match at the startoffset position, but instead was moving forward to
2084    the next newline as if a previous match had failed.
2085    
2086    2. pcretest was not making use of PCRE_NOTEMPTY when repeating for /g and /G,
2087    and could get into a loop if a null string was matched other than at the start
2088    of the subject.
2089    
2090    3. Added definitions of PCRE_MAJOR and PCRE_MINOR to pcre.h so the version can
2091    be distinguished at compile time, and for completeness also added PCRE_DATE.
2092    
2093    5. Added Paul Sokolovsky's minor changes to make it easy to compile a Win32 DLL
2094    in GnuWin32 environments.
2095    
2096    
2097    Version 2.07 29-Jul-99
2098    ----------------------
2099    
2100    1. The documentation is now supplied in plain text form and HTML as well as in
2101    the form of man page sources.
2102    
2103    2. C++ compilers don't like assigning (void *) values to other pointer types.
2104    In particular this affects malloc(). Although there is no problem in Standard
2105    C, I've put in casts to keep C++ compilers happy.
2106    
2107    3. Typo on pcretest.c; a cast of (unsigned char *) in the POSIX regexec() call
2108    should be (const char *).
2109    
2110    4. If NOPOSIX is defined, pcretest.c compiles without POSIX support. This may
2111    be useful for non-Unix systems who don't want to bother with the POSIX stuff.
2112    However, I haven't made this a standard facility. The documentation doesn't
2113    mention it, and the Makefile doesn't support it.
2114    
2115    5. The Makefile now contains an "install" target, with editable destinations at
2116    the top of the file. The pcretest program is not installed.
2117    
2118    6. pgrep -V now gives the PCRE version number and date.
2119    
2120    7. Fixed bug: a zero repetition after a literal string (e.g. /abcde{0}/) was
2121    causing the entire string to be ignored, instead of just the last character.
2122    
2123    8. If a pattern like /"([^\\"]+|\\.)*"/ is applied in the normal way to a
2124    non-matching string, it can take a very, very long time, even for strings of
2125    quite modest length, because of the nested recursion. PCRE now does better in
2126    some of these cases. It does this by remembering the last required literal
2127    character in the pattern, and pre-searching the subject to ensure it is present
2128    before running the real match. In other words, it applies a heuristic to detect
2129    some types of certain failure quickly, and in the above example, if presented
2130    with a string that has no trailing " it gives "no match" very quickly.
2131    
2132    9. A new runtime option PCRE_NOTEMPTY causes null string matches to be ignored;
2133    other alternatives are tried instead.
2134    
2135    
2136    Version 2.06 09-Jun-99
2137    ----------------------
2138    
2139    1. Change pcretest's output for amount of store used to show just the code
2140    space, because the remainder (the data block) varies in size between 32-bit and
2141    64-bit systems.
2142    
2143    2. Added an extra argument to pcre_exec() to supply an offset in the subject to
2144    start matching at. This allows lookbehinds to work when searching for multiple
2145    occurrences in a string.
2146    
2147    3. Added additional options to pcretest for testing multiple occurrences:
2148    
2149       /+   outputs the rest of the string that follows a match
2150       /g   loops for multiple occurrences, using the new startoffset argument
2151       /G   loops for multiple occurrences by passing an incremented pointer
2152    
2153    4. PCRE wasn't doing the "first character" optimization for patterns starting
2154    with \b or \B, though it was doing it for other lookbehind assertions. That is,
2155    it wasn't noticing that a match for a pattern such as /\bxyz/ has to start with
2156    the letter 'x'. On long subject strings, this gives a significant speed-up.
2157    
2158    
2159  Version 2.05 21-Apr-99  Version 2.05 21-Apr-99
2160  ----------------------  ----------------------

Legend:
Removed from v.33  
changed lines
  Added in v.104

webmaster@exim.org
ViewVC Help
Powered by ViewVC 1.1.12