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

Diff of /code/trunk/ChangeLog

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

revision 59 by nigel, Sat Feb 24 21:39:54 2007 UTC revision 164 by ph10, Fri May 4 15:11:14 2007 UTC
# Line 1  Line 1 
1  ChangeLog for PCRE  ChangeLog for PCRE
2  ------------------  ------------------
3    
4    Version 7.2 01-May-07
5    ---------------------
6    
7     1. If the fr_FR locale cannot be found for test 3, try the "french" locale,
8        which is apparently normally available under Windows.
9    
10     2. Re-jig the pcregrep tests with different newline settings in an attempt
11        to make them independent of the local environment's newline setting.
12    
13     3. Add code to configure.ac to remove -g from the CFLAGS default settings.
14    
15     4. Some of the "internals" tests were previously cut out when the link size
16        was not 2, because the output contained actual offsets. The recent new
17        "Z" feature of pcretest means that these can be cut out, making the tests
18        usable with all link sizes.
19    
20     5. Implemented Stan Switzer's goto replacement for longjmp() when not using
21        stack recursion. This gives a massive performance boost under BSD, but just
22        a small improvement under Linux. However, it saves one field in the frame
23        in all cases.
24    
25    
26    Version 7.1 24-Apr-07
27    ---------------------
28    
29     1. Applied Bob Rossi and Daniel G's patches to convert the build system to one
30        that is more "standard", making use of automake and other Autotools. There
31        is some re-arrangement of the files and adjustment of comments consequent
32        on this.
33    
34     2. Part of the patch fixed a problem with the pcregrep tests. The test of -r
35        for recursive directory scanning broke on some systems because the files
36        are not scanned in any specific order and on different systems the order
37        was different. A call to "sort" has been inserted into RunGrepTest for the
38        approprate test as a short-term fix. In the longer term there may be an
39        alternative.
40    
41     3. I had an email from Eric Raymond about problems translating some of PCRE's
42        man pages to HTML (despite the fact that I distribute HTML pages, some
43        people do their own conversions for various reasons). The problems
44        concerned the use of low-level troff macros .br and .in. I have therefore
45        removed all such uses from the man pages (some were redundant, some could
46        be replaced by .nf/.fi pairs). The 132html script that I use to generate
47        HTML has been updated to handle .nf/.fi and to complain if it encounters
48        .br or .in.
49    
50     4. Updated comments in configure.ac that get placed in config.h.in and also
51        arranged for config.h to be included in the distribution, with the name
52        config.h.generic, for the benefit of those who have to compile without
53        Autotools (compare pcre.h, which is now distributed as pcre.h.generic).
54    
55     5. Updated the support (such as it is) for Virtual Pascal, thanks to Stefan
56        Weber: (1) pcre_internal.h was missing some function renames; (2) updated
57        makevp.bat for the current PCRE, using the additional files
58        makevp_c.txt, makevp_l.txt, and pcregexp.pas.
59    
60     6. A Windows user reported a minor discrepancy with test 2, which turned out
61        to be caused by a trailing space on an input line that had got lost in his
62        copy. The trailing space was an accident, so I've just removed it.
63    
64     7. Add -Wl,-R... flags in pcre-config.in for *BSD* systems, as I'm told
65        that is needed.
66    
67     8. Mark ucp_table (in ucptable.h) and ucp_gentype (in pcre_ucp_searchfuncs.c)
68        as "const" (a) because they are and (b) because it helps the PHP
69        maintainers who have recently made a script to detect big data structures
70        in the php code that should be moved to the .rodata section. I remembered
71        to update Builducptable as well, so it won't revert if ucptable.h is ever
72        re-created.
73    
74     9. Added some extra #ifdef SUPPORT_UTF8 conditionals into pcretest.c,
75        pcre_printint.src, pcre_compile.c, pcre_study.c, and pcre_tables.c, in
76        order to be able to cut out the UTF-8 tables in the latter when UTF-8
77        support is not required. This saves 1.5-2K of code, which is important in
78        some applications.
79    
80        Later: more #ifdefs are needed in pcre_ord2utf8.c and pcre_valid_utf8.c
81        so as not to refer to the tables, even though these functions will never be
82        called when UTF-8 support is disabled. Otherwise there are problems with a
83        shared library.
84    
85    10. Fixed two bugs in the emulated memmove() function in pcre_internal.h:
86    
87        (a) It was defining its arguments as char * instead of void *.
88    
89        (b) It was assuming that all moves were upwards in memory; this was true
90            a long time ago when I wrote it, but is no longer the case.
91    
92        The emulated memove() is provided for those environments that have neither
93        memmove() nor bcopy(). I didn't think anyone used it these days, but that
94        is clearly not the case, as these two bugs were recently reported.
95    
96    11. The script PrepareRelease is now distributed: it calls 132html, CleanTxt,
97        and Detrail to create the HTML documentation, the .txt form of the man
98        pages, and it removes trailing spaces from listed files. It also creates
99        pcre.h.generic and config.h.generic from pcre.h and config.h. In the latter
100        case, it wraps all the #defines with #ifndefs. This script should be run
101        before "make dist".
102    
103    12. Fixed two fairly obscure bugs concerned with quantified caseless matching
104        with Unicode property support.
105    
106        (a) For a maximizing quantifier, if the two different cases of the
107            character were of different lengths in their UTF-8 codings (there are
108            some cases like this - I found 11), and the matching function had to
109            back up over a mixture of the two cases, it incorrectly assumed they
110            were both the same length.
111    
112        (b) When PCRE was configured to use the heap rather than the stack for
113            recursion during matching, it was not correctly preserving the data for
114            the other case of a UTF-8 character when checking ahead for a match
115            while processing a minimizing repeat. If the check also involved
116            matching a wide character, but failed, corruption could cause an
117            erroneous result when trying to check for a repeat of the original
118            character.
119    
120    13. Some tidying changes to the testing mechanism:
121    
122        (a) The RunTest script now detects the internal link size and whether there
123            is UTF-8 and UCP support by running ./pcretest -C instead of relying on
124            values substituted by "configure". (The RunGrepTest script already did
125            this for UTF-8.) The configure.ac script no longer substitutes the
126            relevant variables.
127    
128        (b) The debugging options /B and /D in pcretest show the compiled bytecode
129            with length and offset values. This means that the output is different
130            for different internal link sizes. Test 2 is skipped for link sizes
131            other than 2 because of this, bypassing the problem. Unfortunately,
132            there was also a test in test 3 (the locale tests) that used /B and
133            failed for link sizes other than 2. Rather than cut the whole test out,
134            I have added a new /Z option to pcretest that replaces the length and
135            offset values with spaces. This is now used to make test 3 independent
136            of link size. (Test 2 will be tidied up later.)
137    
138    14. If erroroffset was passed as NULL to pcre_compile, it provoked a
139        segmentation fault instead of returning the appropriate error message.
140    
141    15. In multiline mode when the newline sequence was set to "any", the pattern
142        ^$ would give a match between the \r and \n of a subject such as "A\r\nB".
143        This doesn't seem right; it now treats the CRLF combination as the line
144        ending, and so does not match in that case. It's only a pattern such as ^$
145        that would hit this one: something like ^ABC$ would have failed after \r
146        and then tried again after \r\n.
147    
148    16. Changed the comparison command for RunGrepTest from "diff -u" to "diff -ub"
149        in an attempt to make files that differ only in their line terminators
150        compare equal. This works on Linux.
151    
152    17. Under certain error circumstances pcregrep might try to free random memory
153        as it exited. This is now fixed, thanks to valgrind.
154    
155    19. In pcretest, if the pattern /(?m)^$/g<any> was matched against the string
156        "abc\r\n\r\n", it found an unwanted second match after the second \r. This
157        was because its rules for how to advance for /g after matching an empty
158        string at the end of a line did not allow for this case. They now check for
159        it specially.
160    
161    20. pcretest is supposed to handle patterns and data of any length, by
162        extending its buffers when necessary. It was getting this wrong when the
163        buffer for a data line had to be extended.
164    
165    21. Added PCRE_NEWLINE_ANYCRLF which is like ANY, but matches only CR, LF, or
166        CRLF as a newline sequence.
167    
168    22. Code for handling Unicode properties in pcre_dfa_exec() wasn't being cut
169        out by #ifdef SUPPORT_UCP. This did no harm, as it could never be used, but
170        I have nevertheless tidied it up.
171    
172    23. Added some casts to kill warnings from HP-UX ia64 compiler.
173    
174    24. Added a man page for pcre-config.
175    
176    
177    Version 7.0 19-Dec-06
178    ---------------------
179    
180     1. Fixed a signed/unsigned compiler warning in pcre_compile.c, shown up by
181        moving to gcc 4.1.1.
182    
183     2. The -S option for pcretest uses setrlimit(); I had omitted to #include
184        sys/time.h, which is documented as needed for this function. It doesn't
185        seem to matter on Linux, but it showed up on some releases of OS X.
186    
187     3. It seems that there are systems where bytes whose values are greater than
188        127 match isprint() in the "C" locale. The "C" locale should be the
189        default when a C program starts up. In most systems, only ASCII printing
190        characters match isprint(). This difference caused the output from pcretest
191        to vary, making some of the tests fail. I have changed pcretest so that:
192    
193        (a) When it is outputting text in the compiled version of a pattern, bytes
194            other than 32-126 are always shown as hex escapes.
195    
196        (b) When it is outputting text that is a matched part of a subject string,
197            it does the same, unless a different locale has been set for the match
198            (using the /L modifier). In this case, it uses isprint() to decide.
199    
200     4. Fixed a major bug that caused incorrect computation of the amount of memory
201        required for a compiled pattern when options that changed within the
202        pattern affected the logic of the preliminary scan that determines the
203        length. The relevant options are -x, and -i in UTF-8 mode. The result was
204        that the computed length was too small. The symptoms of this bug were
205        either the PCRE error "internal error: code overflow" from pcre_compile(),
206        or a glibc crash with a message such as "pcretest: free(): invalid next
207        size (fast)". Examples of patterns that provoked this bug (shown in
208        pcretest format) are:
209    
210          /(?-x: )/x
211          /(?x)(?-x: \s*#\s*)/
212          /((?i)[\x{c0}])/8
213          /(?i:[\x{c0}])/8
214    
215        HOWEVER: Change 17 below makes this fix obsolete as the memory computation
216        is now done differently.
217    
218     5. Applied patches from Google to: (a) add a QuoteMeta function to the C++
219        wrapper classes; (b) implement a new function in the C++ scanner that is
220        more efficient than the old way of doing things because it avoids levels of
221        recursion in the regex matching; (c) add a paragraph to the documentation
222        for the FullMatch() function.
223    
224     6. The escape sequence \n was being treated as whatever was defined as
225        "newline". Not only was this contrary to the documentation, which states
226        that \n is character 10 (hex 0A), but it also went horribly wrong when
227        "newline" was defined as CRLF. This has been fixed.
228    
229     7. In pcre_dfa_exec.c the value of an unsigned integer (the variable called c)
230        was being set to -1 for the "end of line" case (supposedly a value that no
231        character can have). Though this value is never used (the check for end of
232        line is "zero bytes in current character"), it caused compiler complaints.
233        I've changed it to 0xffffffff.
234    
235     8. In pcre_version.c, the version string was being built by a sequence of
236        C macros that, in the event of PCRE_PRERELEASE being defined as an empty
237        string (as it is for production releases) called a macro with an empty
238        argument. The C standard says the result of this is undefined. The gcc
239        compiler treats it as an empty string (which was what was wanted) but it is
240        reported that Visual C gives an error. The source has been hacked around to
241        avoid this problem.
242    
243     9. On the advice of a Windows user, included <io.h> and <fcntl.h> in Windows
244        builds of pcretest, and changed the call to _setmode() to use _O_BINARY
245        instead of 0x8000. Made all the #ifdefs test both _WIN32 and WIN32 (not all
246        of them did).
247    
248    10. Originally, pcretest opened its input and output without "b"; then I was
249        told that "b" was needed in some environments, so it was added for release
250        5.0 to both the input and output. (It makes no difference on Unix-like
251        systems.) Later I was told that it is wrong for the input on Windows. I've
252        now abstracted the modes into two macros, to make it easier to fiddle with
253        them, and removed "b" from the input mode under Windows.
254    
255    11. Added pkgconfig support for the C++ wrapper library, libpcrecpp.
256    
257    12. Added -help and --help to pcretest as an official way of being reminded
258        of the options.
259    
260    13. Removed some redundant semicolons after macro calls in pcrecpparg.h.in
261        and pcrecpp.cc because they annoy compilers at high warning levels.
262    
263    14. A bit of tidying/refactoring in pcre_exec.c in the main bumpalong loop.
264    
265    15. Fixed an occurrence of == in configure.ac that should have been = (shell
266        scripts are not C programs :-) and which was not noticed because it works
267        on Linux.
268    
269    16. pcretest is supposed to handle any length of pattern and data line (as one
270        line or as a continued sequence of lines) by extending its input buffer if
271        necessary. This feature was broken for very long pattern lines, leading to
272        a string of junk being passed to pcre_compile() if the pattern was longer
273        than about 50K.
274    
275    17. I have done a major re-factoring of the way pcre_compile() computes the
276        amount of memory needed for a compiled pattern. Previously, there was code
277        that made a preliminary scan of the pattern in order to do this. That was
278        OK when PCRE was new, but as the facilities have expanded, it has become
279        harder and harder to keep it in step with the real compile phase, and there
280        have been a number of bugs (see for example, 4 above). I have now found a
281        cunning way of running the real compile function in a "fake" mode that
282        enables it to compute how much memory it would need, while actually only
283        ever using a few hundred bytes of working memory and without too many
284        tests of the mode. This should make future maintenance and development
285        easier. A side effect of this work is that the limit of 200 on the nesting
286        depth of parentheses has been removed (though this was never a serious
287        limitation, I suspect). However, there is a downside: pcre_compile() now
288        runs more slowly than before (30% or more, depending on the pattern). I
289        hope this isn't a big issue. There is no effect on runtime performance.
290    
291    18. Fixed a minor bug in pcretest: if a pattern line was not terminated by a
292        newline (only possible for the last line of a file) and it was a
293        pattern that set a locale (followed by /Lsomething), pcretest crashed.
294    
295    19. Added additional timing features to pcretest. (1) The -tm option now times
296        matching only, not compiling. (2) Both -t and -tm can be followed, as a
297        separate command line item, by a number that specifies the number of
298        repeats to use when timing. The default is 50000; this gives better
299        precision, but takes uncomfortably long for very large patterns.
300    
301    20. Extended pcre_study() to be more clever in cases where a branch of a
302        subpattern has no definite first character. For example, (a*|b*)[cd] would
303        previously give no result from pcre_study(). Now it recognizes that the
304        first character must be a, b, c, or d.
305    
306    21. There was an incorrect error "recursive call could loop indefinitely" if
307        a subpattern (or the entire pattern) that was being tested for matching an
308        empty string contained only one non-empty item after a nested subpattern.
309        For example, the pattern (?>\x{100}*)\d(?R) provoked this error
310        incorrectly, because the \d was being skipped in the check.
311    
312    22. The pcretest program now has a new pattern option /B and a command line
313        option -b, which is equivalent to adding /B to every pattern. This causes
314        it to show the compiled bytecode, without the additional information that
315        -d shows. The effect of -d is now the same as -b with -i (and similarly, /D
316        is the same as /B/I).
317    
318    23. A new optimization is now able automatically to treat some sequences such
319        as a*b as a*+b. More specifically, if something simple (such as a character
320        or a simple class like \d) has an unlimited quantifier, and is followed by
321        something that cannot possibly match the quantified thing, the quantifier
322        is automatically "possessified".
323    
324    24. A recursive reference to a subpattern whose number was greater than 39
325        went wrong under certain circumstances in UTF-8 mode. This bug could also
326        have affected the operation of pcre_study().
327    
328    25. Realized that a little bit of performance could be had by replacing
329        (c & 0xc0) == 0xc0 with c >= 0xc0 when processing UTF-8 characters.
330    
331    26. Timing data from pcretest is now shown to 4 decimal places instead of 3.
332    
333    27. Possessive quantifiers such as a++ were previously implemented by turning
334        them into atomic groups such as ($>a+). Now they have their own opcodes,
335        which improves performance. This includes the automatically created ones
336        from 23 above.
337    
338    28. A pattern such as (?=(\w+))\1: which simulates an atomic group using a
339        lookahead was broken if it was not anchored. PCRE was mistakenly expecting
340        the first matched character to be a colon. This applied both to named and
341        numbered groups.
342    
343    29. The ucpinternal.h header file was missing its idempotency #ifdef.
344    
345    30. I was sent a "project" file called libpcre.a.dev which I understand makes
346        building PCRE on Windows easier, so I have included it in the distribution.
347    
348    31. There is now a check in pcretest against a ridiculously large number being
349        returned by pcre_exec() or pcre_dfa_exec(). If this happens in a /g or /G
350        loop, the loop is abandoned.
351    
352    32. Forward references to subpatterns in conditions such as (?(2)...) where
353        subpattern 2 is defined later cause pcre_compile() to search forwards in
354        the pattern for the relevant set of parentheses. This search went wrong
355        when there were unescaped parentheses in a character class, parentheses
356        escaped with \Q...\E, or parentheses in a #-comment in /x mode.
357    
358    33. "Subroutine" calls and backreferences were previously restricted to
359        referencing subpatterns earlier in the regex. This restriction has now
360        been removed.
361    
362    34. Added a number of extra features that are going to be in Perl 5.10. On the
363        whole, these are just syntactic alternatives for features that PCRE had
364        previously implemented using the Python syntax or my own invention. The
365        other formats are all retained for compatibility.
366    
367        (a) Named groups can now be defined as (?<name>...) or (?'name'...) as well
368            as (?P<name>...). The new forms, as well as being in Perl 5.10, are
369            also .NET compatible.
370    
371        (b) A recursion or subroutine call to a named group can now be defined as
372            (?&name) as well as (?P>name).
373    
374        (c) A backreference to a named group can now be defined as \k<name> or
375            \k'name' as well as (?P=name). The new forms, as well as being in Perl
376            5.10, are also .NET compatible.
377    
378        (d) A conditional reference to a named group can now use the syntax
379            (?(<name>) or (?('name') as well as (?(name).
380    
381        (e) A "conditional group" of the form (?(DEFINE)...) can be used to define
382            groups (named and numbered) that are never evaluated inline, but can be
383            called as "subroutines" from elsewhere. In effect, the DEFINE condition
384            is always false. There may be only one alternative in such a group.
385    
386        (f) A test for recursion can be given as (?(R1).. or (?(R&name)... as well
387            as the simple (?(R). The condition is true only if the most recent
388            recursion is that of the given number or name. It does not search out
389            through the entire recursion stack.
390    
391        (g) The escape \gN or \g{N} has been added, where N is a positive or
392            negative number, specifying an absolute or relative reference.
393    
394    35. Tidied to get rid of some further signed/unsigned compiler warnings and
395        some "unreachable code" warnings.
396    
397    36. Updated the Unicode property tables to Unicode version 5.0.0. Amongst other
398        things, this adds five new scripts.
399    
400    37. Perl ignores orphaned \E escapes completely. PCRE now does the same.
401        There were also incompatibilities regarding the handling of \Q..\E inside
402        character classes, for example with patterns like [\Qa\E-\Qz\E] where the
403        hyphen was adjacent to \Q or \E. I hope I've cleared all this up now.
404    
405    38. Like Perl, PCRE detects when an indefinitely repeated parenthesized group
406        matches an empty string, and forcibly breaks the loop. There were bugs in
407        this code in non-simple cases. For a pattern such as  ^(a()*)*  matched
408        against  aaaa  the result was just "a" rather than "aaaa", for example. Two
409        separate and independent bugs (that affected different cases) have been
410        fixed.
411    
412    39. Refactored the code to abolish the use of different opcodes for small
413        capturing bracket numbers. This is a tidy that I avoided doing when I
414        removed the limit on the number of capturing brackets for 3.5 back in 2001.
415        The new approach is not only tidier, it makes it possible to reduce the
416        memory needed to fix the previous bug (38).
417    
418    40. Implemented PCRE_NEWLINE_ANY to recognize any of the Unicode newline
419        sequences (http://unicode.org/unicode/reports/tr18/) as "newline" when
420        processing dot, circumflex, or dollar metacharacters, or #-comments in /x
421        mode.
422    
423    41. Add \R to match any Unicode newline sequence, as suggested in the Unicode
424        report.
425    
426    42. Applied patch, originally from Ari Pollak, modified by Google, to allow
427        copy construction and assignment in the C++ wrapper.
428    
429    43. Updated pcregrep to support "--newline=any". In the process, I fixed a
430        couple of bugs that could have given wrong results in the "--newline=crlf"
431        case.
432    
433    44. Added a number of casts and did some reorganization of signed/unsigned int
434        variables following suggestions from Dair Grant. Also renamed the variable
435        "this" as "item" because it is a C++ keyword.
436    
437    45. Arranged for dftables to add
438    
439          #include "pcre_internal.h"
440    
441        to pcre_chartables.c because without it, gcc 4.x may remove the array
442        definition from the final binary if PCRE is built into a static library and
443        dead code stripping is activated.
444    
445    46. For an unanchored pattern, if a match attempt fails at the start of a
446        newline sequence, and the newline setting is CRLF or ANY, and the next two
447        characters are CRLF, advance by two characters instead of one.
448    
449    
450    Version 6.7 04-Jul-06
451    ---------------------
452    
453     1. In order to handle tests when input lines are enormously long, pcretest has
454        been re-factored so that it automatically extends its buffers when
455        necessary. The code is crude, but this _is_ just a test program. The
456        default size has been increased from 32K to 50K.
457    
458     2. The code in pcre_study() was using the value of the re argument before
459        testing it for NULL. (Of course, in any sensible call of the function, it
460        won't be NULL.)
461    
462     3. The memmove() emulation function in pcre_internal.h, which is used on
463        systems that lack both memmove() and bcopy() - that is, hardly ever -
464        was missing a "static" storage class specifier.
465    
466     4. When UTF-8 mode was not set, PCRE looped when compiling certain patterns
467        containing an extended class (one that cannot be represented by a bitmap
468        because it contains high-valued characters or Unicode property items, e.g.
469        [\pZ]). Almost always one would set UTF-8 mode when processing such a
470        pattern, but PCRE should not loop if you do not (it no longer does).
471        [Detail: two cases were found: (a) a repeated subpattern containing an
472        extended class; (b) a recursive reference to a subpattern that followed a
473        previous extended class. It wasn't skipping over the extended class
474        correctly when UTF-8 mode was not set.]
475    
476     5. A negated single-character class was not being recognized as fixed-length
477        in lookbehind assertions such as (?<=[^f]), leading to an incorrect
478        compile error "lookbehind assertion is not fixed length".
479    
480     6. The RunPerlTest auxiliary script was showing an unexpected difference
481        between PCRE and Perl for UTF-8 tests. It turns out that it is hard to
482        write a Perl script that can interpret lines of an input file either as
483        byte characters or as UTF-8, which is what "perltest" was being required to
484        do for the non-UTF-8 and UTF-8 tests, respectively. Essentially what you
485        can't do is switch easily at run time between having the "use utf8;" pragma
486        or not. In the end, I fudged it by using the RunPerlTest script to insert
487        "use utf8;" explicitly for the UTF-8 tests.
488    
489     7. In multiline (/m) mode, PCRE was matching ^ after a terminating newline at
490        the end of the subject string, contrary to the documentation and to what
491        Perl does. This was true of both matching functions. Now it matches only at
492        the start of the subject and immediately after *internal* newlines.
493    
494     8. A call of pcre_fullinfo() from pcretest to get the option bits was passing
495        a pointer to an int instead of a pointer to an unsigned long int. This
496        caused problems on 64-bit systems.
497    
498     9. Applied a patch from the folks at Google to pcrecpp.cc, to fix "another
499        instance of the 'standard' template library not being so standard".
500    
501    10. There was no check on the number of named subpatterns nor the maximum
502        length of a subpattern name. The product of these values is used to compute
503        the size of the memory block for a compiled pattern. By supplying a very
504        long subpattern name and a large number of named subpatterns, the size
505        computation could be caused to overflow. This is now prevented by limiting
506        the length of names to 32 characters, and the number of named subpatterns
507        to 10,000.
508    
509    11. Subpatterns that are repeated with specific counts have to be replicated in
510        the compiled pattern. The size of memory for this was computed from the
511        length of the subpattern and the repeat count. The latter is limited to
512        65535, but there was no limit on the former, meaning that integer overflow
513        could in principle occur. The compiled length of a repeated subpattern is
514        now limited to 30,000 bytes in order to prevent this.
515    
516    12. Added the optional facility to have named substrings with the same name.
517    
518    13. Added the ability to use a named substring as a condition, using the
519        Python syntax: (?(name)yes|no). This overloads (?(R)... and names that
520        are numbers (not recommended). Forward references are permitted.
521    
522    14. Added forward references in named backreferences (if you see what I mean).
523    
524    15. In UTF-8 mode, with the PCRE_DOTALL option set, a quantified dot in the
525        pattern could run off the end of the subject. For example, the pattern
526        "(?s)(.{1,5})"8 did this with the subject "ab".
527    
528    16. If PCRE_DOTALL or PCRE_MULTILINE were set, pcre_dfa_exec() behaved as if
529        PCRE_CASELESS was set when matching characters that were quantified with ?
530        or *.
531    
532    17. A character class other than a single negated character that had a minimum
533        but no maximum quantifier - for example [ab]{6,} - was not handled
534        correctly by pce_dfa_exec(). It would match only one character.
535    
536    18. A valid (though odd) pattern that looked like a POSIX character
537        class but used an invalid character after [ (for example [[,abc,]]) caused
538        pcre_compile() to give the error "Failed: internal error: code overflow" or
539        in some cases to crash with a glibc free() error. This could even happen if
540        the pattern terminated after [[ but there just happened to be a sequence of
541        letters, a binary zero, and a closing ] in the memory that followed.
542    
543    19. Perl's treatment of octal escapes in the range \400 to \777 has changed
544        over the years. Originally (before any Unicode support), just the bottom 8
545        bits were taken. Thus, for example, \500 really meant \100. Nowadays the
546        output from "man perlunicode" includes this:
547    
548          The regular expression compiler produces polymorphic opcodes.  That
549          is, the pattern adapts to the data and automatically switches to
550          the Unicode character scheme when presented with Unicode data--or
551          instead uses a traditional byte scheme when presented with byte
552          data.
553    
554        Sadly, a wide octal escape does not cause a switch, and in a string with
555        no other multibyte characters, these octal escapes are treated as before.
556        Thus, in Perl, the pattern  /\500/ actually matches \100 but the pattern
557        /\500|\x{1ff}/ matches \500 or \777 because the whole thing is treated as a
558        Unicode string.
559    
560        I have not perpetrated such confusion in PCRE. Up till now, it took just
561        the bottom 8 bits, as in old Perl. I have now made octal escapes with
562        values greater than \377 illegal in non-UTF-8 mode. In UTF-8 mode they
563        translate to the appropriate multibyte character.
564    
565    29. Applied some refactoring to reduce the number of warnings from Microsoft
566        and Borland compilers. This has included removing the fudge introduced
567        seven years ago for the OS/2 compiler (see 2.02/2 below) because it caused
568        a warning about an unused variable.
569    
570    21. PCRE has not included VT (character 0x0b) in the set of whitespace
571        characters since release 4.0, because Perl (from release 5.004) does not.
572        [Or at least, is documented not to: some releases seem to be in conflict
573        with the documentation.] However, when a pattern was studied with
574        pcre_study() and all its branches started with \s, PCRE still included VT
575        as a possible starting character. Of course, this did no harm; it just
576        caused an unnecessary match attempt.
577    
578    22. Removed a now-redundant internal flag bit that recorded the fact that case
579        dependency changed within the pattern. This was once needed for "required
580        byte" processing, but is no longer used. This recovers a now-scarce options
581        bit. Also moved the least significant internal flag bit to the most-
582        significant bit of the word, which was not previously used (hangover from
583        the days when it was an int rather than a uint) to free up another bit for
584        the future.
585    
586    23. Added support for CRLF line endings as well as CR and LF. As well as the
587        default being selectable at build time, it can now be changed at runtime
588        via the PCRE_NEWLINE_xxx flags. There are now options for pcregrep to
589        specify that it is scanning data with non-default line endings.
590    
591    24. Changed the definition of CXXLINK to make it agree with the definition of
592        LINK in the Makefile, by replacing LDFLAGS to CXXFLAGS.
593    
594    25. Applied Ian Taylor's patches to avoid using another stack frame for tail
595        recursions. This makes a big different to stack usage for some patterns.
596    
597    26. If a subpattern containing a named recursion or subroutine reference such
598        as (?P>B) was quantified, for example (xxx(?P>B)){3}, the calculation of
599        the space required for the compiled pattern went wrong and gave too small a
600        value. Depending on the environment, this could lead to "Failed: internal
601        error: code overflow at offset 49" or "glibc detected double free or
602        corruption" errors.
603    
604    27. Applied patches from Google (a) to support the new newline modes and (b) to
605        advance over multibyte UTF-8 characters in GlobalReplace.
606    
607    28. Change free() to pcre_free() in pcredemo.c. Apparently this makes a
608        difference for some implementation of PCRE in some Windows version.
609    
610    29. Added some extra testing facilities to pcretest:
611    
612        \q<number>   in a data line sets the "match limit" value
613        \Q<number>   in a data line sets the "match recursion limt" value
614        -S <number>  sets the stack size, where <number> is in megabytes
615    
616        The -S option isn't available for Windows.
617    
618    
619    Version 6.6 06-Feb-06
620    ---------------------
621    
622     1. Change 16(a) for 6.5 broke things, because PCRE_DATA_SCOPE was not defined
623        in pcreposix.h. I have copied the definition from pcre.h.
624    
625     2. Change 25 for 6.5 broke compilation in a build directory out-of-tree
626        because pcre.h is no longer a built file.
627    
628     3. Added Jeff Friedl's additional debugging patches to pcregrep. These are
629        not normally included in the compiled code.
630    
631    
632    Version 6.5 01-Feb-06
633    ---------------------
634    
635     1. When using the partial match feature with pcre_dfa_exec(), it was not
636        anchoring the second and subsequent partial matches at the new starting
637        point. This could lead to incorrect results. For example, with the pattern
638        /1234/, partially matching against "123" and then "a4" gave a match.
639    
640     2. Changes to pcregrep:
641    
642        (a) All non-match returns from pcre_exec() were being treated as failures
643            to match the line. Now, unless the error is PCRE_ERROR_NOMATCH, an
644            error message is output. Some extra information is given for the
645            PCRE_ERROR_MATCHLIMIT and PCRE_ERROR_RECURSIONLIMIT errors, which are
646            probably the only errors that are likely to be caused by users (by
647            specifying a regex that has nested indefinite repeats, for instance).
648            If there are more than 20 of these errors, pcregrep is abandoned.
649    
650        (b) A binary zero was treated as data while matching, but terminated the
651            output line if it was written out. This has been fixed: binary zeroes
652            are now no different to any other data bytes.
653    
654        (c) Whichever of the LC_ALL or LC_CTYPE environment variables is set is
655            used to set a locale for matching. The --locale=xxxx long option has
656            been added (no short equivalent) to specify a locale explicitly on the
657            pcregrep command, overriding the environment variables.
658    
659        (d) When -B was used with -n, some line numbers in the output were one less
660            than they should have been.
661    
662        (e) Added the -o (--only-matching) option.
663    
664        (f) If -A or -C was used with -c (count only), some lines of context were
665            accidentally printed for the final match.
666    
667        (g) Added the -H (--with-filename) option.
668    
669        (h) The combination of options -rh failed to suppress file names for files
670            that were found from directory arguments.
671    
672        (i) Added the -D (--devices) and -d (--directories) options.
673    
674        (j) Added the -F (--fixed-strings) option.
675    
676        (k) Allow "-" to be used as a file name for -f as well as for a data file.
677    
678        (l) Added the --colo(u)r option.
679    
680        (m) Added Jeffrey Friedl's -S testing option, but within #ifdefs so that it
681            is not present by default.
682    
683     3. A nasty bug was discovered in the handling of recursive patterns, that is,
684        items such as (?R) or (?1), when the recursion could match a number of
685        alternatives. If it matched one of the alternatives, but subsequently,
686        outside the recursion, there was a failure, the code tried to back up into
687        the recursion. However, because of the way PCRE is implemented, this is not
688        possible, and the result was an incorrect result from the match.
689    
690        In order to prevent this happening, the specification of recursion has
691        been changed so that all such subpatterns are automatically treated as
692        atomic groups. Thus, for example, (?R) is treated as if it were (?>(?R)).
693    
694     4. I had overlooked the fact that, in some locales, there are characters for
695        which isalpha() is true but neither isupper() nor islower() are true. In
696        the fr_FR locale, for instance, the \xAA and \xBA characters (ordmasculine
697        and ordfeminine) are like this. This affected the treatment of \w and \W
698        when they appeared in character classes, but not when they appeared outside
699        a character class. The bit map for "word" characters is now created
700        separately from the results of isalnum() instead of just taking it from the
701        upper, lower, and digit maps. (Plus the underscore character, of course.)
702    
703     5. The above bug also affected the handling of POSIX character classes such as
704        [[:alpha:]] and [[:alnum:]]. These do not have their own bit maps in PCRE's
705        permanent tables. Instead, the bit maps for such a class were previously
706        created as the appropriate unions of the upper, lower, and digit bitmaps.
707        Now they are created by subtraction from the [[:word:]] class, which has
708        its own bitmap.
709    
710     6. The [[:blank:]] character class matches horizontal, but not vertical space.
711        It is created by subtracting the vertical space characters (\x09, \x0a,
712        \x0b, \x0c) from the [[:space:]] bitmap. Previously, however, the
713        subtraction was done in the overall bitmap for a character class, meaning
714        that a class such as [\x0c[:blank:]] was incorrect because \x0c would not
715        be recognized. This bug has been fixed.
716    
717     7. Patches from the folks at Google:
718    
719          (a) pcrecpp.cc: "to handle a corner case that may or may not happen in
720          real life, but is still worth protecting against".
721    
722          (b) pcrecpp.cc: "corrects a bug when negative radixes are used with
723          regular expressions".
724    
725          (c) pcre_scanner.cc: avoid use of std::count() because not all systems
726          have it.
727    
728          (d) Split off pcrecpparg.h from pcrecpp.h and had the former built by
729          "configure" and the latter not, in order to fix a problem somebody had
730          with compiling the Arg class on HP-UX.
731    
732          (e) Improve the error-handling of the C++ wrapper a little bit.
733    
734          (f) New tests for checking recursion limiting.
735    
736     8. The pcre_memmove() function, which is used only if the environment does not
737        have a standard memmove() function (and is therefore rarely compiled),
738        contained two bugs: (a) use of int instead of size_t, and (b) it was not
739        returning a result (though PCRE never actually uses the result).
740    
741     9. In the POSIX regexec() interface, if nmatch is specified as a ridiculously
742        large number - greater than INT_MAX/(3*sizeof(int)) - REG_ESPACE is
743        returned instead of calling malloc() with an overflowing number that would
744        most likely cause subsequent chaos.
745    
746    10. The debugging option of pcretest was not showing the NO_AUTO_CAPTURE flag.
747    
748    11. The POSIX flag REG_NOSUB is now supported. When a pattern that was compiled
749        with this option is matched, the nmatch and pmatch options of regexec() are
750        ignored.
751    
752    12. Added REG_UTF8 to the POSIX interface. This is not defined by POSIX, but is
753        provided in case anyone wants to the the POSIX interface with UTF-8
754        strings.
755    
756    13. Added CXXLDFLAGS to the Makefile parameters to provide settings only on the
757        C++ linking (needed for some HP-UX environments).
758    
759    14. Avoid compiler warnings in get_ucpname() when compiled without UCP support
760        (unused parameter) and in the pcre_printint() function (omitted "default"
761        switch label when the default is to do nothing).
762    
763    15. Added some code to make it possible, when PCRE is compiled as a C++
764        library, to replace subject pointers for pcre_exec() with a smart pointer
765        class, thus making it possible to process discontinuous strings.
766    
767    16. The two macros PCRE_EXPORT and PCRE_DATA_SCOPE are confusing, and perform
768        much the same function. They were added by different people who were trying
769        to make PCRE easy to compile on non-Unix systems. It has been suggested
770        that PCRE_EXPORT be abolished now that there is more automatic apparatus
771        for compiling on Windows systems. I have therefore replaced it with
772        PCRE_DATA_SCOPE. This is set automatically for Windows; if not set it
773        defaults to "extern" for C or "extern C" for C++, which works fine on
774        Unix-like systems. It is now possible to override the value of PCRE_DATA_
775        SCOPE with something explicit in config.h. In addition:
776    
777        (a) pcreposix.h still had just "extern" instead of either of these macros;
778            I have replaced it with PCRE_DATA_SCOPE.
779    
780        (b) Functions such as _pcre_xclass(), which are internal to the library,
781            but external in the C sense, all had PCRE_EXPORT in their definitions.
782            This is apparently wrong for the Windows case, so I have removed it.
783            (It makes no difference on Unix-like systems.)
784    
785    17. Added a new limit, MATCH_LIMIT_RECURSION, which limits the depth of nesting
786        of recursive calls to match(). This is different to MATCH_LIMIT because
787        that limits the total number of calls to match(), not all of which increase
788        the depth of recursion. Limiting the recursion depth limits the amount of
789        stack (or heap if NO_RECURSE is set) that is used. The default can be set
790        when PCRE is compiled, and changed at run time. A patch from Google adds
791        this functionality to the C++ interface.
792    
793    18. Changes to the handling of Unicode character properties:
794    
795        (a) Updated the table to Unicode 4.1.0.
796    
797        (b) Recognize characters that are not in the table as "Cn" (undefined).
798    
799        (c) I revised the way the table is implemented to a much improved format
800            which includes recognition of ranges. It now supports the ranges that
801            are defined in UnicodeData.txt, and it also amalgamates other
802            characters into ranges. This has reduced the number of entries in the
803            table from around 16,000 to around 3,000, thus reducing its size
804            considerably. I realized I did not need to use a tree structure after
805            all - a binary chop search is just as efficient. Having reduced the
806            number of entries, I extended their size from 6 bytes to 8 bytes to
807            allow for more data.
808    
809        (d) Added support for Unicode script names via properties such as \p{Han}.
810    
811    19. In UTF-8 mode, a backslash followed by a non-Ascii character was not
812        matching that character.
813    
814    20. When matching a repeated Unicode property with a minimum greater than zero,
815        (for example \pL{2,}), PCRE could look past the end of the subject if it
816        reached it while seeking the minimum number of characters. This could
817        happen only if some of the characters were more than one byte long, because
818        there is a check for at least the minimum number of bytes.
819    
820    21. Refactored the implementation of \p and \P so as to be more general, to
821        allow for more different types of property in future. This has changed the
822        compiled form incompatibly. Anybody with saved compiled patterns that use
823        \p or \P will have to recompile them.
824    
825    22. Added "Any" and "L&" to the supported property types.
826    
827    23. Recognize \x{...} as a code point specifier, even when not in UTF-8 mode,
828        but give a compile time error if the value is greater than 0xff.
829    
830    24. The man pages for pcrepartial, pcreprecompile, and pcre_compile2 were
831        accidentally not being installed or uninstalled.
832    
833    25. The pcre.h file was built from pcre.h.in, but the only changes that were
834        made were to insert the current release number. This seemed silly, because
835        it made things harder for people building PCRE on systems that don't run
836        "configure". I have turned pcre.h into a distributed file, no longer built
837        by "configure", with the version identification directly included. There is
838        no longer a pcre.h.in file.
839    
840        However, this change necessitated a change to the pcre-config script as
841        well. It is built from pcre-config.in, and one of the substitutions was the
842        release number. I have updated configure.ac so that ./configure now finds
843        the release number by grepping pcre.h.
844    
845    26. Added the ability to run the tests under valgrind.
846    
847    
848    Version 6.4 05-Sep-05
849    ---------------------
850    
851     1. Change 6.0/10/(l) to pcregrep introduced a bug that caused separator lines
852        "--" to be printed when multiple files were scanned, even when none of the
853        -A, -B, or -C options were used. This is not compatible with Gnu grep, so I
854        consider it to be a bug, and have restored the previous behaviour.
855    
856     2. A couple of code tidies to get rid of compiler warnings.
857    
858     3. The pcretest program used to cheat by referring to symbols in the library
859        whose names begin with _pcre_. These are internal symbols that are not
860        really supposed to be visible externally, and in some environments it is
861        possible to suppress them. The cheating is now confined to including
862        certain files from the library's source, which is a bit cleaner.
863    
864     4. Renamed pcre.in as pcre.h.in to go with pcrecpp.h.in; it also makes the
865        file's purpose clearer.
866    
867     5. Reorganized pcre_ucp_findchar().
868    
869    
870    Version 6.3 15-Aug-05
871    ---------------------
872    
873     1. The file libpcre.pc.in did not have general read permission in the tarball.
874    
875     2. There were some problems when building without C++ support:
876    
877        (a) If C++ support was not built, "make install" and "make test" still
878            tried to test it.
879    
880        (b) There were problems when the value of CXX was explicitly set. Some
881            changes have been made to try to fix these, and ...
882    
883        (c) --disable-cpp can now be used to explicitly disable C++ support.
884    
885        (d) The use of @CPP_OBJ@ directly caused a blank line preceded by a
886            backslash in a target when C++ was disabled. This confuses some
887            versions of "make", apparently. Using an intermediate variable solves
888            this. (Same for CPP_LOBJ.)
889    
890     3. $(LINK_FOR_BUILD) now includes $(CFLAGS_FOR_BUILD) and $(LINK)
891        (non-Windows) now includes $(CFLAGS) because these flags are sometimes
892        necessary on certain architectures.
893    
894     4. Added a setting of -export-symbols-regex to the link command to remove
895        those symbols that are exported in the C sense, but actually are local
896        within the library, and not documented. Their names all begin with
897        "_pcre_". This is not a perfect job, because (a) we have to except some
898        symbols that pcretest ("illegally") uses, and (b) the facility isn't always
899        available (and never for static libraries). I have made a note to try to
900        find a way round (a) in the future.
901    
902    
903    Version 6.2 01-Aug-05
904    ---------------------
905    
906     1. There was no test for integer overflow of quantifier values. A construction
907        such as {1111111111111111} would give undefined results. What is worse, if
908        a minimum quantifier for a parenthesized subpattern overflowed and became
909        negative, the calculation of the memory size went wrong. This could have
910        led to memory overwriting.
911    
912     2. Building PCRE using VPATH was broken. Hopefully it is now fixed.
913    
914     3. Added "b" to the 2nd argument of fopen() in dftables.c, for non-Unix-like
915        operating environments where this matters.
916    
917     4. Applied Giuseppe Maxia's patch to add additional features for controlling
918        PCRE options from within the C++ wrapper.
919    
920     5. Named capturing subpatterns were not being correctly counted when a pattern
921        was compiled. This caused two problems: (a) If there were more than 100
922        such subpatterns, the calculation of the memory needed for the whole
923        compiled pattern went wrong, leading to an overflow error. (b) Numerical
924        back references of the form \12, where the number was greater than 9, were
925        not recognized as back references, even though there were sufficient
926        previous subpatterns.
927    
928     6. Two minor patches to pcrecpp.cc in order to allow it to compile on older
929        versions of gcc, e.g. 2.95.4.
930    
931    
932    Version 6.1 21-Jun-05
933    ---------------------
934    
935     1. There was one reference to the variable "posix" in pcretest.c that was not
936        surrounded by "#if !defined NOPOSIX".
937    
938     2. Make it possible to compile pcretest without DFA support, UTF8 support, or
939        the cross-check on the old pcre_info() function, for the benefit of the
940        cut-down version of PCRE that is currently imported into Exim.
941    
942     3. A (silly) pattern starting with (?i)(?-i) caused an internal space
943        allocation error. I've done the easy fix, which wastes 2 bytes for sensible
944        patterns that start (?i) but I don't think that matters. The use of (?i) is
945        just an example; this all applies to the other options as well.
946    
947     4. Since libtool seems to echo the compile commands it is issuing, the output
948        from "make" can be reduced a bit by putting "@" in front of each libtool
949        compile command.
950    
951     5. Patch from the folks at Google for configure.in to be a bit more thorough
952        in checking for a suitable C++ installation before trying to compile the
953        C++ stuff. This should fix a reported problem when a compiler was present,
954        but no suitable headers.
955    
956     6. The man pages all had just "PCRE" as their title. I have changed them to
957        be the relevant file name. I have also arranged that these names are
958        retained in the file doc/pcre.txt, which is a concatenation in text format
959        of all the man pages except the little individual ones for each function.
960    
961     7. The NON-UNIX-USE file had not been updated for the different set of source
962        files that come with release 6. I also added a few comments about the C++
963        wrapper.
964    
965    
966    Version 6.0 07-Jun-05
967    ---------------------
968    
969     1. Some minor internal re-organization to help with my DFA experiments.
970    
971     2. Some missing #ifdef SUPPORT_UCP conditionals in pcretest and printint that
972        didn't matter for the library itself when fully configured, but did matter
973        when compiling without UCP support, or within Exim, where the ucp files are
974        not imported.
975    
976     3. Refactoring of the library code to split up the various functions into
977        different source modules. The addition of the new DFA matching code (see
978        below) to a single monolithic source would have made it really too
979        unwieldy, quite apart from causing all the code to be include in a
980        statically linked application, when only some functions are used. This is
981        relevant even without the DFA addition now that patterns can be compiled in
982        one application and matched in another.
983    
984        The downside of splitting up is that there have to be some external
985        functions and data tables that are used internally in different modules of
986        the library but which are not part of the API. These have all had their
987        names changed to start with "_pcre_" so that they are unlikely to clash
988        with other external names.
989    
990     4. Added an alternate matching function, pcre_dfa_exec(), which matches using
991        a different (DFA) algorithm. Although it is slower than the original
992        function, it does have some advantages for certain types of matching
993        problem.
994    
995     5. Upgrades to pcretest in order to test the features of pcre_dfa_exec(),
996        including restarting after a partial match.
997    
998     6. A patch for pcregrep that defines INVALID_FILE_ATTRIBUTES if it is not
999        defined when compiling for Windows was sent to me. I have put it into the
1000        code, though I have no means of testing or verifying it.
1001    
1002     7. Added the pcre_refcount() auxiliary function.
1003    
1004     8. Added the PCRE_FIRSTLINE option. This constrains an unanchored pattern to
1005        match before or at the first newline in the subject string. In pcretest,
1006        the /f option on a pattern can be used to set this.
1007    
1008     9. A repeated \w when used in UTF-8 mode with characters greater than 256
1009        would behave wrongly. This has been present in PCRE since release 4.0.
1010    
1011    10. A number of changes to the pcregrep command:
1012    
1013        (a) Refactored how -x works; insert ^(...)$ instead of setting
1014            PCRE_ANCHORED and checking the length, in preparation for adding
1015            something similar for -w.
1016    
1017        (b) Added the -w (match as a word) option.
1018    
1019        (c) Refactored the way lines are read and buffered so as to have more
1020            than one at a time available.
1021    
1022        (d) Implemented a pcregrep test script.
1023    
1024        (e) Added the -M (multiline match) option. This allows patterns to match
1025            over several lines of the subject. The buffering ensures that at least
1026            8K, or the rest of the document (whichever is the shorter) is available
1027            for matching (and similarly the previous 8K for lookbehind assertions).
1028    
1029        (f) Changed the --help output so that it now says
1030    
1031              -w, --word-regex(p)
1032    
1033            instead of two lines, one with "regex" and the other with "regexp"
1034            because that confused at least one person since the short forms are the
1035            same. (This required a bit of code, as the output is generated
1036            automatically from a table. It wasn't just a text change.)
1037    
1038        (g) -- can be used to terminate pcregrep options if the next thing isn't an
1039            option but starts with a hyphen. Could be a pattern or a path name
1040            starting with a hyphen, for instance.
1041    
1042        (h) "-" can be given as a file name to represent stdin.
1043    
1044        (i) When file names are being printed, "(standard input)" is used for
1045            the standard input, for compatibility with GNU grep. Previously
1046            "<stdin>" was used.
1047    
1048        (j) The option --label=xxx can be used to supply a name to be used for
1049            stdin when file names are being printed. There is no short form.
1050    
1051        (k) Re-factored the options decoding logic because we are going to add
1052            two more options that take data. Such options can now be given in four
1053            different ways, e.g. "-fname", "-f name", "--file=name", "--file name".
1054    
1055        (l) Added the -A, -B, and -C options for requesting that lines of context
1056            around matches be printed.
1057    
1058        (m) Added the -L option to print the names of files that do not contain
1059            any matching lines, that is, the complement of -l.
1060    
1061        (n) The return code is 2 if any file cannot be opened, but pcregrep does
1062            continue to scan other files.
1063    
1064        (o) The -s option was incorrectly implemented. For compatibility with other
1065            greps, it now suppresses the error message for a non-existent or non-
1066            accessible file (but not the return code). There is a new option called
1067            -q that suppresses the output of matching lines, which was what -s was
1068            previously doing.
1069    
1070        (p) Added --include and --exclude options to specify files for inclusion
1071            and exclusion when recursing.
1072    
1073    11. The Makefile was not using the Autoconf-supported LDFLAGS macro properly.
1074        Hopefully, it now does.
1075    
1076    12. Missing cast in pcre_study().
1077    
1078    13. Added an "uninstall" target to the makefile.
1079    
1080    14. Replaced "extern" in the function prototypes in Makefile.in with
1081        "PCRE_DATA_SCOPE", which defaults to 'extern' or 'extern "C"' in the Unix
1082        world, but is set differently for Windows.
1083    
1084    15. Added a second compiling function called pcre_compile2(). The only
1085        difference is that it has an extra argument, which is a pointer to an
1086        integer error code. When there is a compile-time failure, this is set
1087        non-zero, in addition to the error test pointer being set to point to an
1088        error message. The new argument may be NULL if no error number is required
1089        (but then you may as well call pcre_compile(), which is now just a
1090        wrapper). This facility is provided because some applications need a
1091        numeric error indication, but it has also enabled me to tidy up the way
1092        compile-time errors are handled in the POSIX wrapper.
1093    
1094    16. Added VPATH=.libs to the makefile; this should help when building with one
1095        prefix path and installing with another. (Or so I'm told by someone who
1096        knows more about this stuff than I do.)
1097    
1098    17. Added a new option, REG_DOTALL, to the POSIX function regcomp(). This
1099        passes PCRE_DOTALL to the pcre_compile() function, making the "." character
1100        match everything, including newlines. This is not POSIX-compatible, but
1101        somebody wanted the feature. From pcretest it can be activated by using
1102        both the P and the s flags.
1103    
1104    18. AC_PROG_LIBTOOL appeared twice in Makefile.in. Removed one.
1105    
1106    19. libpcre.pc was being incorrectly installed as executable.
1107    
1108    20. A couple of places in pcretest check for end-of-line by looking for '\n';
1109        it now also looks for '\r' so that it will work unmodified on Windows.
1110    
1111    21. Added Google's contributed C++ wrapper to the distribution.
1112    
1113    22. Added some untidy missing memory free() calls in pcretest, to keep
1114        Electric Fence happy when testing.
1115    
1116    
1117    
1118    Version 5.0 13-Sep-04
1119    ---------------------
1120    
1121     1. Internal change: literal characters are no longer packed up into items
1122        containing multiple characters in a single byte-string. Each character
1123        is now matched using a separate opcode. However, there may be more than one
1124        byte in the character in UTF-8 mode.
1125    
1126     2. The pcre_callout_block structure has two new fields: pattern_position and
1127        next_item_length. These contain the offset in the pattern to the next match
1128        item, and its length, respectively.
1129    
1130     3. The PCRE_AUTO_CALLOUT option for pcre_compile() requests the automatic
1131        insertion of callouts before each pattern item. Added the /C option to
1132        pcretest to make use of this.
1133    
1134     4. On the advice of a Windows user, the lines
1135    
1136          #if defined(_WIN32) || defined(WIN32)
1137          _setmode( _fileno( stdout ), 0x8000 );
1138          #endif  /* defined(_WIN32) || defined(WIN32) */
1139    
1140        have been added to the source of pcretest. This apparently does useful
1141        magic in relation to line terminators.
1142    
1143     5. Changed "r" and "w" in the calls to fopen() in pcretest to "rb" and "wb"
1144        for the benefit of those environments where the "b" makes a difference.
1145    
1146     6. The icc compiler has the same options as gcc, but "configure" doesn't seem
1147        to know about it. I have put a hack into configure.in that adds in code
1148        to set GCC=yes if CC=icc. This seems to end up at a point in the
1149        generated configure script that is early enough to affect the setting of
1150        compiler options, which is what is needed, but I have no means of testing
1151        whether it really works. (The user who reported this had patched the
1152        generated configure script, which of course I cannot do.)
1153    
1154        LATER: After change 22 below (new libtool files), the configure script
1155        seems to know about icc (and also ecc). Therefore, I have commented out
1156        this hack in configure.in.
1157    
1158     7. Added support for pkg-config (2 patches were sent in).
1159    
1160     8. Negated POSIX character classes that used a combination of internal tables
1161        were completely broken. These were [[:^alpha:]], [[:^alnum:]], and
1162        [[:^ascii]]. Typically, they would match almost any characters. The other
1163        POSIX classes were not broken in this way.
1164    
1165     9. Matching the pattern "\b.*?" against "ab cd", starting at offset 1, failed
1166        to find the match, as PCRE was deluded into thinking that the match had to
1167        start at the start point or following a newline. The same bug applied to
1168        patterns with negative forward assertions or any backward assertions
1169        preceding ".*" at the start, unless the pattern required a fixed first
1170        character. This was a failing pattern: "(?!.bcd).*". The bug is now fixed.
1171    
1172    10. In UTF-8 mode, when moving forwards in the subject after a failed match
1173        starting at the last subject character, bytes beyond the end of the subject
1174        string were read.
1175    
1176    11. Renamed the variable "class" as "classbits" to make life easier for C++
1177        users. (Previously there was a macro definition, but it apparently wasn't
1178        enough.)
1179    
1180    12. Added the new field "tables" to the extra data so that tables can be passed
1181        in at exec time, or the internal tables can be re-selected. This allows
1182        a compiled regex to be saved and re-used at a later time by a different
1183        program that might have everything at different addresses.
1184    
1185    13. Modified the pcre-config script so that, when run on Solaris, it shows a
1186        -R library as well as a -L library.
1187    
1188    14. The debugging options of pcretest (-d on the command line or D on a
1189        pattern) showed incorrect output for anything following an extended class
1190        that contained multibyte characters and which was followed by a quantifier.
1191    
1192    15. Added optional support for general category Unicode character properties
1193        via the \p, \P, and \X escapes. Unicode property support implies UTF-8
1194        support. It adds about 90K to the size of the library. The meanings of the
1195        inbuilt class escapes such as \d and \s have NOT been changed.
1196    
1197    16. Updated pcredemo.c to include calls to free() to release the memory for the
1198        compiled pattern.
1199    
1200    17. The generated file chartables.c was being created in the source directory
1201        instead of in the building directory. This caused the build to fail if the
1202        source directory was different from the building directory, and was
1203        read-only.
1204    
1205    18. Added some sample Win commands from Mark Tetrode into the NON-UNIX-USE
1206        file. No doubt somebody will tell me if they don't make sense... Also added
1207        Dan Mooney's comments about building on OpenVMS.
1208    
1209    19. Added support for partial matching via the PCRE_PARTIAL option for
1210        pcre_exec() and the \P data escape in pcretest.
1211    
1212    20. Extended pcretest with 3 new pattern features:
1213    
1214        (i)   A pattern option of the form ">rest-of-line" causes pcretest to
1215              write the compiled pattern to the file whose name is "rest-of-line".
1216              This is a straight binary dump of the data, with the saved pointer to
1217              the character tables forced to be NULL. The study data, if any, is
1218              written too. After writing, pcretest reads a new pattern.
1219    
1220        (ii)  If, instead of a pattern, "<rest-of-line" is given, pcretest reads a
1221              compiled pattern from the given file. There must not be any
1222              occurrences of "<" in the file name (pretty unlikely); if there are,
1223              pcretest will instead treat the initial "<" as a pattern delimiter.
1224              After reading in the pattern, pcretest goes on to read data lines as
1225              usual.
1226    
1227        (iii) The F pattern option causes pcretest to flip the bytes in the 32-bit
1228              and 16-bit fields in a compiled pattern, to simulate a pattern that
1229              was compiled on a host of opposite endianness.
1230    
1231    21. The pcre-exec() function can now cope with patterns that were compiled on
1232        hosts of opposite endianness, with this restriction:
1233    
1234          As for any compiled expression that is saved and used later, the tables
1235          pointer field cannot be preserved; the extra_data field in the arguments
1236          to pcre_exec() should be used to pass in a tables address if a value
1237          other than the default internal tables were used at compile time.
1238    
1239    22. Calling pcre_exec() with a negative value of the "ovecsize" parameter is
1240        now diagnosed as an error. Previously, most of the time, a negative number
1241        would have been treated as zero, but if in addition "ovector" was passed as
1242        NULL, a crash could occur.
1243    
1244    23. Updated the files ltmain.sh, config.sub, config.guess, and aclocal.m4 with
1245        new versions from the libtool 1.5 distribution (the last one is a copy of
1246        a file called libtool.m4). This seems to have fixed the need to patch
1247        "configure" to support Darwin 1.3 (which I used to do). However, I still
1248        had to patch ltmain.sh to ensure that ${SED} is set (it isn't on my
1249        workstation).
1250    
1251    24. Changed the PCRE licence to be the more standard "BSD" licence.
1252    
1253    
1254    Version 4.5 01-Dec-03
1255    ---------------------
1256    
1257     1. There has been some re-arrangement of the code for the match() function so
1258        that it can be compiled in a version that does not call itself recursively.
1259        Instead, it keeps those local variables that need separate instances for
1260        each "recursion" in a frame on the heap, and gets/frees frames whenever it
1261        needs to "recurse". Keeping track of where control must go is done by means
1262        of setjmp/longjmp. The whole thing is implemented by a set of macros that
1263        hide most of the details from the main code, and operates only if
1264        NO_RECURSE is defined while compiling pcre.c. If PCRE is built using the
1265        "configure" mechanism, "--disable-stack-for-recursion" turns on this way of
1266        operating.
1267    
1268        To make it easier for callers to provide specially tailored get/free
1269        functions for this usage, two new functions, pcre_stack_malloc, and
1270        pcre_stack_free, are used. They are always called in strict stacking order,
1271        and the size of block requested is always the same.
1272    
1273        The PCRE_CONFIG_STACKRECURSE info parameter can be used to find out whether
1274        PCRE has been compiled to use the stack or the heap for recursion. The
1275        -C option of pcretest uses this to show which version is compiled.
1276    
1277        A new data escape \S, is added to pcretest; it causes the amounts of store
1278        obtained and freed by both kinds of malloc/free at match time to be added
1279        to the output.
1280    
1281     2. Changed the locale test to use "fr_FR" instead of "fr" because that's
1282        what's available on my current Linux desktop machine.
1283    
1284     3. When matching a UTF-8 string, the test for a valid string at the start has
1285        been extended. If start_offset is not zero, PCRE now checks that it points
1286        to a byte that is the start of a UTF-8 character. If not, it returns
1287        PCRE_ERROR_BADUTF8_OFFSET (-11). Note: the whole string is still checked;
1288        this is necessary because there may be backward assertions in the pattern.
1289        When matching the same subject several times, it may save resources to use
1290        PCRE_NO_UTF8_CHECK on all but the first call if the string is long.
1291    
1292     4. The code for checking the validity of UTF-8 strings has been tightened so
1293        that it rejects (a) strings containing 0xfe or 0xff bytes and (b) strings
1294        containing "overlong sequences".
1295    
1296     5. Fixed a bug (appearing twice) that I could not find any way of exploiting!
1297        I had written "if ((digitab[*p++] && chtab_digit) == 0)" where the "&&"
1298        should have been "&", but it just so happened that all the cases this let
1299        through by mistake were picked up later in the function.
1300    
1301     6. I had used a variable called "isblank" - this is a C99 function, causing
1302        some compilers to warn. To avoid this, I renamed it (as "blankclass").
1303    
1304     7. Cosmetic: (a) only output another newline at the end of pcretest if it is
1305        prompting; (b) run "./pcretest /dev/null" at the start of the test script
1306        so the version is shown; (c) stop "make test" echoing "./RunTest".
1307    
1308     8. Added patches from David Burgess to enable PCRE to run on EBCDIC systems.
1309    
1310     9. The prototype for memmove() for systems that don't have it was using
1311        size_t, but the inclusion of the header that defines size_t was later. I've
1312        moved the #includes for the C headers earlier to avoid this.
1313    
1314    10. Added some adjustments to the code to make it easier to compiler on certain
1315        special systems:
1316    
1317          (a) Some "const" qualifiers were missing.
1318          (b) Added the macro EXPORT before all exported functions; by default this
1319              is defined to be empty.
1320          (c) Changed the dftables auxiliary program (that builds chartables.c) so
1321              that it reads its output file name as an argument instead of writing
1322              to the standard output and assuming this can be redirected.
1323    
1324    11. In UTF-8 mode, if a recursive reference (e.g. (?1)) followed a character
1325        class containing characters with values greater than 255, PCRE compilation
1326        went into a loop.
1327    
1328    12. A recursive reference to a subpattern that was within another subpattern
1329        that had a minimum quantifier of zero caused PCRE to crash. For example,
1330        (x(y(?2))z)? provoked this bug with a subject that got as far as the
1331        recursion. If the recursively-called subpattern itself had a zero repeat,
1332        that was OK.
1333    
1334    13. In pcretest, the buffer for reading a data line was set at 30K, but the
1335        buffer into which it was copied (for escape processing) was still set at
1336        1024, so long lines caused crashes.
1337    
1338    14. A pattern such as /[ab]{1,3}+/ failed to compile, giving the error
1339        "internal error: code overflow...". This applied to any character class
1340        that was followed by a possessive quantifier.
1341    
1342    15. Modified the Makefile to add libpcre.la as a prerequisite for
1343        libpcreposix.la because I was told this is needed for a parallel build to
1344        work.
1345    
1346    16. If a pattern that contained .* following optional items at the start was
1347        studied, the wrong optimizing data was generated, leading to matching
1348        errors. For example, studying /[ab]*.*c/ concluded, erroneously, that any
1349        matching string must start with a or b or c. The correct conclusion for
1350        this pattern is that a match can start with any character.
1351    
1352    
1353    Version 4.4 13-Aug-03
1354    ---------------------
1355    
1356     1. In UTF-8 mode, a character class containing characters with values between
1357        127 and 255 was not handled correctly if the compiled pattern was studied.
1358        In fixing this, I have also improved the studying algorithm for such
1359        classes (slightly).
1360    
1361     2. Three internal functions had redundant arguments passed to them. Removal
1362        might give a very teeny performance improvement.
1363    
1364     3. Documentation bug: the value of the capture_top field in a callout is *one
1365        more than* the number of the hightest numbered captured substring.
1366    
1367     4. The Makefile linked pcretest and pcregrep with -lpcre, which could result
1368        in incorrectly linking with a previously installed version. They now link
1369        explicitly with libpcre.la.
1370    
1371     5. configure.in no longer needs to recognize Cygwin specially.
1372    
1373     6. A problem in pcre.in for Windows platforms is fixed.
1374    
1375     7. If a pattern was successfully studied, and the -d (or /D) flag was given to
1376        pcretest, it used to include the size of the study block as part of its
1377        output. Unfortunately, the structure contains a field that has a different
1378        size on different hardware architectures. This meant that the tests that
1379        showed this size failed. As the block is currently always of a fixed size,
1380        this information isn't actually particularly useful in pcretest output, so
1381        I have just removed it.
1382    
1383     8. Three pre-processor statements accidentally did not start in column 1.
1384        Sadly, there are *still* compilers around that complain, even though
1385        standard C has not required this for well over a decade. Sigh.
1386    
1387     9. In pcretest, the code for checking callouts passed small integers in the
1388        callout_data field, which is a void * field. However, some picky compilers
1389        complained about the casts involved for this on 64-bit systems. Now
1390        pcretest passes the address of the small integer instead, which should get
1391        rid of the warnings.
1392    
1393    10. By default, when in UTF-8 mode, PCRE now checks for valid UTF-8 strings at
1394        both compile and run time, and gives an error if an invalid UTF-8 sequence
1395        is found. There is a option for disabling this check in cases where the
1396        string is known to be correct and/or the maximum performance is wanted.
1397    
1398    11. In response to a bug report, I changed one line in Makefile.in from
1399    
1400            -Wl,--out-implib,.libs/lib@WIN_PREFIX@pcreposix.dll.a \
1401        to
1402            -Wl,--out-implib,.libs/@WIN_PREFIX@libpcreposix.dll.a \
1403    
1404        to look similar to other lines, but I have no way of telling whether this
1405        is the right thing to do, as I do not use Windows. No doubt I'll get told
1406        if it's wrong...
1407    
1408    
1409    Version 4.3 21-May-03
1410    ---------------------
1411    
1412    1. Two instances of @WIN_PREFIX@ omitted from the Windows targets in the
1413       Makefile.
1414    
1415    2. Some refactoring to improve the quality of the code:
1416    
1417       (i)   The utf8_table... variables are now declared "const".
1418    
1419       (ii)  The code for \cx, which used the "case flipping" table to upper case
1420             lower case letters, now just substracts 32. This is ASCII-specific,
1421             but the whole concept of \cx is ASCII-specific, so it seems
1422             reasonable.
1423    
1424       (iii) PCRE was using its character types table to recognize decimal and
1425             hexadecimal digits in the pattern. This is silly, because it handles
1426             only 0-9, a-f, and A-F, but the character types table is locale-
1427             specific, which means strange things might happen. A private
1428             table is now used for this - though it costs 256 bytes, a table is
1429             much faster than multiple explicit tests. Of course, the standard
1430             character types table is still used for matching digits in subject
1431             strings against \d.
1432    
1433       (iv)  Strictly, the identifier ESC_t is reserved by POSIX (all identifiers
1434             ending in _t are). So I've renamed it as ESC_tee.
1435    
1436    3. The first argument for regexec() in the POSIX wrapper should have been
1437       defined as "const".
1438    
1439    4. Changed pcretest to use malloc() for its buffers so that they can be
1440       Electric Fenced for debugging.
1441    
1442    5. There were several places in the code where, in UTF-8 mode, PCRE would try
1443       to read one or more bytes before the start of the subject string. Often this
1444       had no effect on PCRE's behaviour, but in some circumstances it could
1445       provoke a segmentation fault.
1446    
1447    6. A lookbehind at the start of a pattern in UTF-8 mode could also cause PCRE
1448       to try to read one or more bytes before the start of the subject string.
1449    
1450    7. A lookbehind in a pattern matched in non-UTF-8 mode on a PCRE compiled with
1451       UTF-8 support could misbehave in various ways if the subject string
1452       contained bytes with the 0x80 bit set and the 0x40 bit unset in a lookbehind
1453       area. (PCRE was not checking for the UTF-8 mode flag, and trying to move
1454       back over UTF-8 characters.)
1455    
1456    
1457    Version 4.2 14-Apr-03
1458    ---------------------
1459    
1460    1. Typo "#if SUPPORT_UTF8" instead of "#ifdef SUPPORT_UTF8" fixed.
1461    
1462    2. Changes to the building process, supplied by Ronald Landheer-Cieslak
1463         [ON_WINDOWS]: new variable, "#" on non-Windows platforms
1464         [NOT_ON_WINDOWS]: new variable, "#" on Windows platforms
1465         [WIN_PREFIX]: new variable, "cyg" for Cygwin
1466         * Makefile.in: use autoconf substitution for OBJEXT, EXEEXT, BUILD_OBJEXT
1467           and BUILD_EXEEXT
1468         Note: automatic setting of the BUILD variables is not yet working
1469         set CPPFLAGS and BUILD_CPPFLAGS (but don't use yet) - should be used at
1470           compile-time but not at link-time
1471         [LINK]: use for linking executables only
1472         make different versions for Windows and non-Windows
1473         [LINKLIB]: new variable, copy of UNIX-style LINK, used for linking
1474           libraries
1475         [LINK_FOR_BUILD]: new variable
1476         [OBJEXT]: use throughout
1477         [EXEEXT]: use throughout
1478         <winshared>: new target
1479         <wininstall>: new target
1480         <dftables.o>: use native compiler
1481         <dftables>: use native linker
1482         <install>: handle Windows platform correctly
1483         <clean>: ditto
1484         <check>: ditto
1485         copy DLL to top builddir before testing
1486    
1487       As part of these changes, -no-undefined was removed again. This was reported
1488       to give trouble on HP-UX 11.0, so getting rid of it seems like a good idea
1489       in any case.
1490    
1491    3. Some tidies to get rid of compiler warnings:
1492    
1493       . In the match_data structure, match_limit was an unsigned long int, whereas
1494         match_call_count was an int. I've made them both unsigned long ints.
1495    
1496       . In pcretest the fact that a const uschar * doesn't automatically cast to
1497         a void * provoked a warning.
1498    
1499       . Turning on some more compiler warnings threw up some "shadow" variables
1500         and a few more missing casts.
1501    
1502    4. If PCRE was complied with UTF-8 support, but called without the PCRE_UTF8
1503       option, a class that contained a single character with a value between 128
1504       and 255 (e.g. /[\xFF]/) caused PCRE to crash.
1505    
1506    5. If PCRE was compiled with UTF-8 support, but called without the PCRE_UTF8
1507       option, a class that contained several characters, but with at least one
1508       whose value was between 128 and 255 caused PCRE to crash.
1509    
1510    
1511    Version 4.1 12-Mar-03
1512    ---------------------
1513    
1514    1. Compiling with gcc -pedantic found a couple of places where casts were
1515    needed, and a string in dftables.c that was longer than standard compilers are
1516    required to support.
1517    
1518    2. Compiling with Sun's compiler found a few more places where the code could
1519    be tidied up in order to avoid warnings.
1520    
1521    3. The variables for cross-compiling were called HOST_CC and HOST_CFLAGS; the
1522    first of these names is deprecated in the latest Autoconf in favour of the name
1523    CC_FOR_BUILD, because "host" is typically used to mean the system on which the
1524    compiled code will be run. I can't find a reference for HOST_CFLAGS, but by
1525    analogy I have changed it to CFLAGS_FOR_BUILD.
1526    
1527    4. Added -no-undefined to the linking command in the Makefile, because this is
1528    apparently helpful for Windows. To make it work, also added "-L. -lpcre" to the
1529    linking step for the pcreposix library.
1530    
1531    5. PCRE was failing to diagnose the case of two named groups with the same
1532    name.
1533    
1534    6. A problem with one of PCRE's optimizations was discovered. PCRE remembers a
1535    literal character that is needed in the subject for a match, and scans along to
1536    ensure that it is present before embarking on the full matching process. This
1537    saves time in cases of nested unlimited repeats that are never going to match.
1538    Problem: the scan can take a lot of time if the subject is very long (e.g.
1539    megabytes), thus penalizing straightforward matches. It is now done only if the
1540    amount of subject to be scanned is less than 1000 bytes.
1541    
1542    7. A lesser problem with the same optimization is that it was recording the
1543    first character of an anchored pattern as "needed", thus provoking a search
1544    right along the subject, even when the first match of the pattern was going to
1545    fail. The "needed" character is now not set for anchored patterns, unless it
1546    follows something in the pattern that is of non-fixed length. Thus, it still
1547    fulfils its original purpose of finding quick non-matches in cases of nested
1548    unlimited repeats, but isn't used for simple anchored patterns such as /^abc/.
1549    
1550    
1551    Version 4.0 17-Feb-03
1552    ---------------------
1553    
1554    1. If a comment in an extended regex that started immediately after a meta-item
1555    extended to the end of string, PCRE compiled incorrect data. This could lead to
1556    all kinds of weird effects. Example: /#/ was bad; /()#/ was bad; /a#/ was not.
1557    
1558    2. Moved to autoconf 2.53 and libtool 1.4.2.
1559    
1560    3. Perl 5.8 no longer needs "use utf8" for doing UTF-8 things. Consequently,
1561    the special perltest8 script is no longer needed - all the tests can be run
1562    from a single perltest script.
1563    
1564    4. From 5.004, Perl has not included the VT character (0x0b) in the set defined
1565    by \s. It has now been removed in PCRE. This means it isn't recognized as
1566    whitespace in /x regexes too, which is the same as Perl. Note that the POSIX
1567    class [:space:] *does* include VT, thereby creating a mess.
1568    
1569    5. Added the class [:blank:] (a GNU extension from Perl 5.8) to match only
1570    space and tab.
1571    
1572    6. Perl 5.005 was a long time ago. It's time to amalgamate the tests that use
1573    its new features into the main test script, reducing the number of scripts.
1574    
1575    7. Perl 5.8 has changed the meaning of patterns like /a(?i)b/. Earlier versions
1576    were backward compatible, and made the (?i) apply to the whole pattern, as if
1577    /i were given. Now it behaves more logically, and applies the option setting
1578    only to what follows. PCRE has been changed to follow suit. However, if it
1579    finds options settings right at the start of the pattern, it extracts them into
1580    the global options, as before. Thus, they show up in the info data.
1581    
1582    8. Added support for the \Q...\E escape sequence. Characters in between are
1583    treated as literals. This is slightly different from Perl in that $ and @ are
1584    also handled as literals inside the quotes. In Perl, they will cause variable
1585    interpolation. Note the following examples:
1586    
1587        Pattern            PCRE matches      Perl matches
1588    
1589        \Qabc$xyz\E        abc$xyz           abc followed by the contents of $xyz
1590        \Qabc\$xyz\E       abc\$xyz          abc\$xyz
1591        \Qabc\E\$\Qxyz\E   abc$xyz           abc$xyz
1592    
1593    For compatibility with Perl, \Q...\E sequences are recognized inside character
1594    classes as well as outside them.
1595    
1596    9. Re-organized 3 code statements in pcretest to avoid "overflow in
1597    floating-point constant arithmetic" warnings from a Microsoft compiler. Added a
1598    (size_t) cast to one statement in pcretest and one in pcreposix to avoid
1599    signed/unsigned warnings.
1600    
1601    10. SunOS4 doesn't have strtoul(). This was used only for unpicking the -o
1602    option for pcretest, so I've replaced it by a simple function that does just
1603    that job.
1604    
1605    11. pcregrep was ending with code 0 instead of 2 for the commands "pcregrep" or
1606    "pcregrep -".
1607    
1608    12. Added "possessive quantifiers" ?+, *+, ++, and {,}+ which come from Sun's
1609    Java package. This provides some syntactic sugar for simple cases of what my
1610    documentation calls "once-only subpatterns". A pattern such as x*+ is the same
1611    as (?>x*). In other words, if what is inside (?>...) is just a single repeated
1612    item, you can use this simplified notation. Note that only makes sense with
1613    greedy quantifiers. Consequently, the use of the possessive quantifier forces
1614    greediness, whatever the setting of the PCRE_UNGREEDY option.
1615    
1616    13. A change of greediness default within a pattern was not taking effect at
1617    the current level for patterns like /(b+(?U)a+)/. It did apply to parenthesized
1618    subpatterns that followed. Patterns like /b+(?U)a+/ worked because the option
1619    was abstracted outside.
1620    
1621    14. PCRE now supports the \G assertion. It is true when the current matching
1622    position is at the start point of the match. This differs from \A when the
1623    starting offset is non-zero. Used with the /g option of pcretest (or similar
1624    code), it works in the same way as it does for Perl's /g option. If all
1625    alternatives of a regex begin with \G, the expression is anchored to the start
1626    match position, and the "anchored" flag is set in the compiled expression.
1627    
1628    15. Some bugs concerning the handling of certain option changes within patterns
1629    have been fixed. These applied to options other than (?ims). For example,
1630    "a(?x: b c )d" did not match "XabcdY" but did match "Xa b c dY". It should have
1631    been the other way round. Some of this was related to change 7 above.
1632    
1633    16. PCRE now gives errors for /[.x.]/ and /[=x=]/ as unsupported POSIX
1634    features, as Perl does. Previously, PCRE gave the warnings only for /[[.x.]]/
1635    and /[[=x=]]/. PCRE now also gives an error for /[:name:]/ because it supports
1636    POSIX classes only within a class (e.g. /[[:alpha:]]/).
1637    
1638    17. Added support for Perl's \C escape. This matches one byte, even in UTF8
1639    mode. Unlike ".", it always matches newline, whatever the setting of
1640    PCRE_DOTALL. However, PCRE does not permit \C to appear in lookbehind
1641    assertions. Perl allows it, but it doesn't (in general) work because it can't
1642    calculate the length of the lookbehind. At least, that's the case for Perl
1643    5.8.0 - I've been told they are going to document that it doesn't work in
1644    future.
1645    
1646    18. Added an error diagnosis for escapes that PCRE does not support: these are
1647    \L, \l, \N, \P, \p, \U, \u, and \X.
1648    
1649    19. Although correctly diagnosing a missing ']' in a character class, PCRE was
1650    reading past the end of the pattern in cases such as /[abcd/.
1651    
1652    20. PCRE was getting more memory than necessary for patterns with classes that
1653    contained both POSIX named classes and other characters, e.g. /[[:space:]abc/.
1654    
1655    21. Added some code, conditional on #ifdef VPCOMPAT, to make life easier for
1656    compiling PCRE for use with Virtual Pascal.
1657    
1658    22. Small fix to the Makefile to make it work properly if the build is done
1659    outside the source tree.
1660    
1661    23. Added a new extension: a condition to go with recursion. If a conditional
1662    subpattern starts with (?(R) the "true" branch is used if recursion has
1663    happened, whereas the "false" branch is used only at the top level.
1664    
1665    24. When there was a very long string of literal characters (over 255 bytes
1666    without UTF support, over 250 bytes with UTF support), the computation of how
1667    much memory was required could be incorrect, leading to segfaults or other
1668    strange effects.
1669    
1670    25. PCRE was incorrectly assuming anchoring (either to start of subject or to
1671    start of line for a non-DOTALL pattern) when a pattern started with (.*) and
1672    there was a subsequent back reference to those brackets. This meant that, for
1673    example, /(.*)\d+\1/ failed to match "abc123bc". Unfortunately, it isn't
1674    possible to check for precisely this case. All we can do is abandon the
1675    optimization if .* occurs inside capturing brackets when there are any back
1676    references whatsoever. (See below for a better fix that came later.)
1677    
1678    26. The handling of the optimization for finding the first character of a
1679    non-anchored pattern, and for finding a character that is required later in the
1680    match were failing in some cases. This didn't break the matching; it just
1681    failed to optimize when it could. The way this is done has been re-implemented.
1682    
1683    27. Fixed typo in error message for invalid (?R item (it said "(?p").
1684    
1685    28. Added a new feature that provides some of the functionality that Perl
1686    provides with (?{...}). The facility is termed a "callout". The way it is done
1687    in PCRE is for the caller to provide an optional function, by setting
1688    pcre_callout to its entry point. Like pcre_malloc and pcre_free, this is a
1689    global variable. By default it is unset, which disables all calling out. To get
1690    the function called, the regex must include (?C) at appropriate points. This
1691    is, in fact, equivalent to (?C0), and any number <= 255 may be given with (?C).
1692    This provides a means of identifying different callout points. When PCRE
1693    reaches such a point in the regex, if pcre_callout has been set, the external
1694    function is called. It is provided with data in a structure called
1695    pcre_callout_block, which is defined in pcre.h. If the function returns 0,
1696    matching continues; if it returns a non-zero value, the match at the current
1697    point fails. However, backtracking will occur if possible. [This was changed
1698    later and other features added - see item 49 below.]
1699    
1700    29. pcretest is upgraded to test the callout functionality. It provides a
1701    callout function that displays information. By default, it shows the start of
1702    the match and the current position in the text. There are some new data escapes
1703    to vary what happens:
1704    
1705        \C+         in addition, show current contents of captured substrings
1706        \C-         do not supply a callout function
1707        \C!n        return 1 when callout number n is reached
1708        \C!n!m      return 1 when callout number n is reached for the mth time
1709    
1710    30. If pcregrep was called with the -l option and just a single file name, it
1711    output "<stdin>" if a match was found, instead of the file name.
1712    
1713    31. Improve the efficiency of the POSIX API to PCRE. If the number of capturing
1714    slots is less than POSIX_MALLOC_THRESHOLD, use a block on the stack to pass to
1715    pcre_exec(). This saves a malloc/free per call. The default value of
1716    POSIX_MALLOC_THRESHOLD is 10; it can be changed by --with-posix-malloc-threshold
1717    when configuring.
1718    
1719    32. The default maximum size of a compiled pattern is 64K. There have been a
1720    few cases of people hitting this limit. The code now uses macros to handle the
1721    storing of links as offsets within the compiled pattern. It defaults to 2-byte
1722    links, but this can be changed to 3 or 4 bytes by --with-link-size when
1723    configuring. Tests 2 and 5 work only with 2-byte links because they output
1724    debugging information about compiled patterns.
1725    
1726    33. Internal code re-arrangements:
1727    
1728    (a) Moved the debugging function for printing out a compiled regex into
1729        its own source file (printint.c) and used #include to pull it into
1730        pcretest.c and, when DEBUG is defined, into pcre.c, instead of having two
1731        separate copies.
1732    
1733    (b) Defined the list of op-code names for debugging as a macro in
1734        internal.h so that it is next to the definition of the opcodes.
1735    
1736    (c) Defined a table of op-code lengths for simpler skipping along compiled
1737        code. This is again a macro in internal.h so that it is next to the
1738        definition of the opcodes.
1739    
1740    34. Added support for recursive calls to individual subpatterns, along the
1741    lines of Robin Houston's patch (but implemented somewhat differently).
1742    
1743    35. Further mods to the Makefile to help Win32. Also, added code to pcregrep to
1744    allow it to read and process whole directories in Win32. This code was
1745    contributed by Lionel Fourquaux; it has not been tested by me.
1746    
1747    36. Added support for named subpatterns. The Python syntax (?P<name>...) is
1748    used to name a group. Names consist of alphanumerics and underscores, and must
1749    be unique. Back references use the syntax (?P=name) and recursive calls use
1750    (?P>name) which is a PCRE extension to the Python extension. Groups still have
1751    numbers. The function pcre_fullinfo() can be used after compilation to extract
1752    a name/number map. There are three relevant calls:
1753    
1754      PCRE_INFO_NAMEENTRYSIZE        yields the size of each entry in the map
1755      PCRE_INFO_NAMECOUNT            yields the number of entries
1756      PCRE_INFO_NAMETABLE            yields a pointer to the map.
1757    
1758    The map is a vector of fixed-size entries. The size of each entry depends on
1759    the length of the longest name used. The first two bytes of each entry are the
1760    group number, most significant byte first. There follows the corresponding
1761    name, zero terminated. The names are in alphabetical order.
1762    
1763    37. Make the maximum literal string in the compiled code 250 for the non-UTF-8
1764    case instead of 255. Making it the same both with and without UTF-8 support
1765    means that the same test output works with both.
1766    
1767    38. There was a case of malloc(0) in the POSIX testing code in pcretest. Avoid
1768    calling malloc() with a zero argument.
1769    
1770    39. Change 25 above had to resort to a heavy-handed test for the .* anchoring
1771    optimization. I've improved things by keeping a bitmap of backreferences with
1772    numbers 1-31 so that if .* occurs inside capturing brackets that are not in
1773    fact referenced, the optimization can be applied. It is unlikely that a
1774    relevant occurrence of .* (i.e. one which might indicate anchoring or forcing
1775    the match to follow \n) will appear inside brackets with a number greater than
1776    31, but if it does, any back reference > 31 suppresses the optimization.
1777    
1778    40. Added a new compile-time option PCRE_NO_AUTO_CAPTURE. This has the effect
1779    of disabling numbered capturing parentheses. Any opening parenthesis that is
1780    not followed by ? behaves as if it were followed by ?: but named parentheses
1781    can still be used for capturing (and they will acquire numbers in the usual
1782    way).
1783    
1784    41. Redesigned the return codes from the match() function into yes/no/error so
1785    that errors can be passed back from deep inside the nested calls. A malloc
1786    failure while inside a recursive subpattern call now causes the
1787    PCRE_ERROR_NOMEMORY return instead of quietly going wrong.
1788    
1789    42. It is now possible to set a limit on the number of times the match()
1790    function is called in a call to pcre_exec(). This facility makes it possible to
1791    limit the amount of recursion and backtracking, though not in a directly
1792    obvious way, because the match() function is used in a number of different
1793    circumstances. The count starts from zero for each position in the subject
1794    string (for non-anchored patterns). The default limit is, for compatibility, a
1795    large number, namely 10 000 000. You can change this in two ways:
1796    
1797    (a) When configuring PCRE before making, you can use --with-match-limit=n
1798        to set a default value for the compiled library.
1799    
1800    (b) For each call to pcre_exec(), you can pass a pcre_extra block in which
1801        a different value is set. See 45 below.
1802    
1803    If the limit is exceeded, pcre_exec() returns PCRE_ERROR_MATCHLIMIT.
1804    
1805    43. Added a new function pcre_config(int, void *) to enable run-time extraction
1806    of things that can be changed at compile time. The first argument specifies
1807    what is wanted and the second points to where the information is to be placed.
1808    The current list of available information is:
1809    
1810      PCRE_CONFIG_UTF8
1811    
1812    The output is an integer that is set to one if UTF-8 support is available;
1813    otherwise it is set to zero.
1814    
1815      PCRE_CONFIG_NEWLINE
1816    
1817    The output is an integer that it set to the value of the code that is used for
1818    newline. It is either LF (10) or CR (13).
1819    
1820      PCRE_CONFIG_LINK_SIZE
1821    
1822    The output is an integer that contains the number of bytes used for internal
1823    linkage in compiled expressions. The value is 2, 3, or 4. See item 32 above.
1824    
1825      PCRE_CONFIG_POSIX_MALLOC_THRESHOLD
1826    
1827    The output is an integer that contains the threshold above which the POSIX
1828    interface uses malloc() for output vectors. See item 31 above.
1829    
1830      PCRE_CONFIG_MATCH_LIMIT
1831    
1832    The output is an unsigned integer that contains the default limit of the number
1833    of match() calls in a pcre_exec() execution. See 42 above.
1834    
1835    44. pcretest has been upgraded by the addition of the -C option. This causes it
1836    to extract all the available output from the new pcre_config() function, and to
1837    output it. The program then exits immediately.
1838    
1839    45. A need has arisen to pass over additional data with calls to pcre_exec() in
1840    order to support additional features. One way would have been to define
1841    pcre_exec2() (for example) with extra arguments, but this would not have been
1842    extensible, and would also have required all calls to the original function to
1843    be mapped to the new one. Instead, I have chosen to extend the mechanism that
1844    is used for passing in "extra" data from pcre_study().
1845    
1846    The pcre_extra structure is now exposed and defined in pcre.h. It currently
1847    contains the following fields:
1848    
1849      flags         a bitmap indicating which of the following fields are set
1850      study_data    opaque data from pcre_study()
1851      match_limit   a way of specifying a limit on match() calls for a specific
1852                      call to pcre_exec()
1853      callout_data  data for callouts (see 49 below)
1854    
1855    The flag bits are also defined in pcre.h, and are
1856    
1857      PCRE_EXTRA_STUDY_DATA
1858      PCRE_EXTRA_MATCH_LIMIT
1859      PCRE_EXTRA_CALLOUT_DATA
1860    
1861    The pcre_study() function now returns one of these new pcre_extra blocks, with
1862    the actual study data pointed to by the study_data field, and the
1863    PCRE_EXTRA_STUDY_DATA flag set. This can be passed directly to pcre_exec() as
1864    before. That is, this change is entirely upwards-compatible and requires no
1865    change to existing code.
1866    
1867    If you want to pass in additional data to pcre_exec(), you can either place it
1868    in a pcre_extra block provided by pcre_study(), or create your own pcre_extra
1869    block.
1870    
1871    46. pcretest has been extended to test the PCRE_EXTRA_MATCH_LIMIT feature. If a
1872    data string contains the escape sequence \M, pcretest calls pcre_exec() several
1873    times with different match limits, until it finds the minimum value needed for
1874    pcre_exec() to complete. The value is then output. This can be instructive; for
1875    most simple matches the number is quite small, but for pathological cases it
1876    gets very large very quickly.
1877    
1878    47. There's a new option for pcre_fullinfo() called PCRE_INFO_STUDYSIZE. It
1879    returns the size of the data block pointed to by the study_data field in a
1880    pcre_extra block, that is, the value that was passed as the argument to
1881    pcre_malloc() when PCRE was getting memory in which to place the information
1882    created by pcre_study(). The fourth argument should point to a size_t variable.
1883    pcretest has been extended so that this information is shown after a successful
1884    pcre_study() call when information about the compiled regex is being displayed.
1885    
1886    48. Cosmetic change to Makefile: there's no need to have / after $(DESTDIR)
1887    because what follows is always an absolute path. (Later: it turns out that this
1888    is more than cosmetic for MinGW, because it doesn't like empty path
1889    components.)
1890    
1891    49. Some changes have been made to the callout feature (see 28 above):
1892    
1893    (i)  A callout function now has three choices for what it returns:
1894    
1895           0  =>  success, carry on matching
1896         > 0  =>  failure at this point, but backtrack if possible
1897         < 0  =>  serious error, return this value from pcre_exec()
1898    
1899         Negative values should normally be chosen from the set of PCRE_ERROR_xxx
1900         values. In particular, returning PCRE_ERROR_NOMATCH forces a standard
1901         "match failed" error. The error number PCRE_ERROR_CALLOUT is reserved for
1902         use by callout functions. It will never be used by PCRE itself.
1903    
1904    (ii) The pcre_extra structure (see 45 above) has a void * field called
1905         callout_data, with corresponding flag bit PCRE_EXTRA_CALLOUT_DATA. The
1906         pcre_callout_block structure has a field of the same name. The contents of
1907         the field passed in the pcre_extra structure are passed to the callout
1908         function in the corresponding field in the callout block. This makes it
1909         easier to use the same callout-containing regex from multiple threads. For
1910         testing, the pcretest program has a new data escape
1911    
1912           \C*n        pass the number n (may be negative) as callout_data
1913    
1914         If the callout function in pcretest receives a non-zero value as
1915         callout_data, it returns that value.
1916    
1917    50. Makefile wasn't handling CFLAGS properly when compiling dftables. Also,
1918    there were some redundant $(CFLAGS) in commands that are now specified as
1919    $(LINK), which already includes $(CFLAGS).
1920    
1921    51. Extensions to UTF-8 support are listed below. These all apply when (a) PCRE
1922    has been compiled with UTF-8 support *and* pcre_compile() has been compiled
1923    with the PCRE_UTF8 flag. Patterns that are compiled without that flag assume
1924    one-byte characters throughout. Note that case-insensitive matching applies
1925    only to characters whose values are less than 256. PCRE doesn't support the
1926    notion of cases for higher-valued characters.
1927    
1928    (i)   A character class whose characters are all within 0-255 is handled as
1929          a bit map, and the map is inverted for negative classes. Previously, a
1930          character > 255 always failed to match such a class; however it should
1931          match if the class was a negative one (e.g. [^ab]). This has been fixed.
1932    
1933    (ii)  A negated character class with a single character < 255 is coded as
1934          "not this character" (OP_NOT). This wasn't working properly when the test
1935          character was multibyte, either singly or repeated.
1936    
1937    (iii) Repeats of multibyte characters are now handled correctly in UTF-8
1938          mode, for example: \x{100}{2,3}.
1939    
1940    (iv)  The character escapes \b, \B, \d, \D, \s, \S, \w, and \W (either
1941          singly or repeated) now correctly test multibyte characters. However,
1942          PCRE doesn't recognize any characters with values greater than 255 as
1943          digits, spaces, or word characters. Such characters always match \D, \S,
1944          and \W, and never match \d, \s, or \w.
1945    
1946    (v)   Classes may now contain characters and character ranges with values
1947          greater than 255. For example: [ab\x{100}-\x{400}].
1948    
1949    (vi)  pcregrep now has a --utf-8 option (synonym -u) which makes it call
1950          PCRE in UTF-8 mode.
1951    
1952    52. The info request value PCRE_INFO_FIRSTCHAR has been renamed
1953    PCRE_INFO_FIRSTBYTE because it is a byte value. However, the old name is
1954    retained for backwards compatibility. (Note that LASTLITERAL is also a byte
1955    value.)
1956    
1957    53. The single man page has become too large. I have therefore split it up into
1958    a number of separate man pages. These also give rise to individual HTML pages;
1959    these are now put in a separate directory, and there is an index.html page that
1960    lists them all. Some hyperlinking between the pages has been installed.
1961    
1962    54. Added convenience functions for handling named capturing parentheses.
1963    
1964    55. Unknown escapes inside character classes (e.g. [\M]) and escapes that
1965    aren't interpreted therein (e.g. [\C]) are literals in Perl. This is now also
1966    true in PCRE, except when the PCRE_EXTENDED option is set, in which case they
1967    are faulted.
1968    
1969    56. Introduced HOST_CC and HOST_CFLAGS which can be set in the environment when
1970    calling configure. These values are used when compiling the dftables.c program
1971    which is run to generate the source of the default character tables. They
1972    default to the values of CC and CFLAGS. If you are cross-compiling PCRE,
1973    you will need to set these values.
1974    
1975    57. Updated the building process for Windows DLL, as provided by Fred Cox.
1976    
1977    
1978    Version 3.9 02-Jan-02
1979    ---------------------
1980    
1981    1. A bit of extraneous text had somehow crept into the pcregrep documentation.
1982    
1983    2. If --disable-static was given, the building process failed when trying to
1984    build pcretest and pcregrep. (For some reason it was using libtool to compile
1985    them, which is not right, as they aren't part of the library.)
1986    
1987    
1988  Version 3.8 18-Dec-01  Version 3.8 18-Dec-01
1989  ---------------------  ---------------------
1990    

Legend:
Removed from v.59  
changed lines
  Added in v.164

webmaster@exim.org
ViewVC Help
Powered by ViewVC 1.1.12