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

Diff of /code/trunk/ChangeLog

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

revision 23 by nigel, Sat Feb 24 21:38:41 2007 UTC revision 158 by ph10, Tue May 1 10:30:14 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    
14    Version 7.1 24-Apr-07
15    ---------------------
16    
17     1. Applied Bob Rossi and Daniel G's patches to convert the build system to one
18        that is more "standard", making use of automake and other Autotools. There
19        is some re-arrangement of the files and adjustment of comments consequent
20        on this.
21    
22     2. Part of the patch fixed a problem with the pcregrep tests. The test of -r
23        for recursive directory scanning broke on some systems because the files
24        are not scanned in any specific order and on different systems the order
25        was different. A call to "sort" has been inserted into RunGrepTest for the
26        approprate test as a short-term fix. In the longer term there may be an
27        alternative.
28    
29     3. I had an email from Eric Raymond about problems translating some of PCRE's
30        man pages to HTML (despite the fact that I distribute HTML pages, some
31        people do their own conversions for various reasons). The problems
32        concerned the use of low-level troff macros .br and .in. I have therefore
33        removed all such uses from the man pages (some were redundant, some could
34        be replaced by .nf/.fi pairs). The 132html script that I use to generate
35        HTML has been updated to handle .nf/.fi and to complain if it encounters
36        .br or .in.
37    
38     4. Updated comments in configure.ac that get placed in config.h.in and also
39        arranged for config.h to be included in the distribution, with the name
40        config.h.generic, for the benefit of those who have to compile without
41        Autotools (compare pcre.h, which is now distributed as pcre.h.generic).
42    
43     5. Updated the support (such as it is) for Virtual Pascal, thanks to Stefan
44        Weber: (1) pcre_internal.h was missing some function renames; (2) updated
45        makevp.bat for the current PCRE, using the additional files
46        makevp_c.txt, makevp_l.txt, and pcregexp.pas.
47    
48     6. A Windows user reported a minor discrepancy with test 2, which turned out
49        to be caused by a trailing space on an input line that had got lost in his
50        copy. The trailing space was an accident, so I've just removed it.
51    
52     7. Add -Wl,-R... flags in pcre-config.in for *BSD* systems, as I'm told
53        that is needed.
54    
55     8. Mark ucp_table (in ucptable.h) and ucp_gentype (in pcre_ucp_searchfuncs.c)
56        as "const" (a) because they are and (b) because it helps the PHP
57        maintainers who have recently made a script to detect big data structures
58        in the php code that should be moved to the .rodata section. I remembered
59        to update Builducptable as well, so it won't revert if ucptable.h is ever
60        re-created.
61    
62     9. Added some extra #ifdef SUPPORT_UTF8 conditionals into pcretest.c,
63        pcre_printint.src, pcre_compile.c, pcre_study.c, and pcre_tables.c, in
64        order to be able to cut out the UTF-8 tables in the latter when UTF-8
65        support is not required. This saves 1.5-2K of code, which is important in
66        some applications.
67    
68        Later: more #ifdefs are needed in pcre_ord2utf8.c and pcre_valid_utf8.c
69        so as not to refer to the tables, even though these functions will never be
70        called when UTF-8 support is disabled. Otherwise there are problems with a
71        shared library.
72    
73    10. Fixed two bugs in the emulated memmove() function in pcre_internal.h:
74    
75        (a) It was defining its arguments as char * instead of void *.
76    
77        (b) It was assuming that all moves were upwards in memory; this was true
78            a long time ago when I wrote it, but is no longer the case.
79    
80        The emulated memove() is provided for those environments that have neither
81        memmove() nor bcopy(). I didn't think anyone used it these days, but that
82        is clearly not the case, as these two bugs were recently reported.
83    
84    11. The script PrepareRelease is now distributed: it calls 132html, CleanTxt,
85        and Detrail to create the HTML documentation, the .txt form of the man
86        pages, and it removes trailing spaces from listed files. It also creates
87        pcre.h.generic and config.h.generic from pcre.h and config.h. In the latter
88        case, it wraps all the #defines with #ifndefs. This script should be run
89        before "make dist".
90    
91    12. Fixed two fairly obscure bugs concerned with quantified caseless matching
92        with Unicode property support.
93    
94        (a) For a maximizing quantifier, if the two different cases of the
95            character were of different lengths in their UTF-8 codings (there are
96            some cases like this - I found 11), and the matching function had to
97            back up over a mixture of the two cases, it incorrectly assumed they
98            were both the same length.
99    
100        (b) When PCRE was configured to use the heap rather than the stack for
101            recursion during matching, it was not correctly preserving the data for
102            the other case of a UTF-8 character when checking ahead for a match
103            while processing a minimizing repeat. If the check also involved
104            matching a wide character, but failed, corruption could cause an
105            erroneous result when trying to check for a repeat of the original
106            character.
107    
108    13. Some tidying changes to the testing mechanism:
109    
110        (a) The RunTest script now detects the internal link size and whether there
111            is UTF-8 and UCP support by running ./pcretest -C instead of relying on
112            values substituted by "configure". (The RunGrepTest script already did
113            this for UTF-8.) The configure.ac script no longer substitutes the
114            relevant variables.
115    
116        (b) The debugging options /B and /D in pcretest show the compiled bytecode
117            with length and offset values. This means that the output is different
118            for different internal link sizes. Test 2 is skipped for link sizes
119            other than 2 because of this, bypassing the problem. Unfortunately,
120            there was also a test in test 3 (the locale tests) that used /B and
121            failed for link sizes other than 2. Rather than cut the whole test out,
122            I have added a new /Z option to pcretest that replaces the length and
123            offset values with spaces. This is now used to make test 3 independent
124            of link size. (Test 2 will be tidied up later.)
125    
126    14. If erroroffset was passed as NULL to pcre_compile, it provoked a
127        segmentation fault instead of returning the appropriate error message.
128    
129    15. In multiline mode when the newline sequence was set to "any", the pattern
130        ^$ would give a match between the \r and \n of a subject such as "A\r\nB".
131        This doesn't seem right; it now treats the CRLF combination as the line
132        ending, and so does not match in that case. It's only a pattern such as ^$
133        that would hit this one: something like ^ABC$ would have failed after \r
134        and then tried again after \r\n.
135    
136    16. Changed the comparison command for RunGrepTest from "diff -u" to "diff -ub"
137        in an attempt to make files that differ only in their line terminators
138        compare equal. This works on Linux.
139    
140    17. Under certain error circumstances pcregrep might try to free random memory
141        as it exited. This is now fixed, thanks to valgrind.
142    
143    19. In pcretest, if the pattern /(?m)^$/g<any> was matched against the string
144        "abc\r\n\r\n", it found an unwanted second match after the second \r. This
145        was because its rules for how to advance for /g after matching an empty
146        string at the end of a line did not allow for this case. They now check for
147        it specially.
148    
149    20. pcretest is supposed to handle patterns and data of any length, by
150        extending its buffers when necessary. It was getting this wrong when the
151        buffer for a data line had to be extended.
152    
153    21. Added PCRE_NEWLINE_ANYCRLF which is like ANY, but matches only CR, LF, or
154        CRLF as a newline sequence.
155    
156    22. Code for handling Unicode properties in pcre_dfa_exec() wasn't being cut
157        out by #ifdef SUPPORT_UCP. This did no harm, as it could never be used, but
158        I have nevertheless tidied it up.
159    
160    23. Added some casts to kill warnings from HP-UX ia64 compiler.
161    
162    24. Added a man page for pcre-config.
163    
164    
165    Version 7.0 19-Dec-06
166    ---------------------
167    
168     1. Fixed a signed/unsigned compiler warning in pcre_compile.c, shown up by
169        moving to gcc 4.1.1.
170    
171     2. The -S option for pcretest uses setrlimit(); I had omitted to #include
172        sys/time.h, which is documented as needed for this function. It doesn't
173        seem to matter on Linux, but it showed up on some releases of OS X.
174    
175     3. It seems that there are systems where bytes whose values are greater than
176        127 match isprint() in the "C" locale. The "C" locale should be the
177        default when a C program starts up. In most systems, only ASCII printing
178        characters match isprint(). This difference caused the output from pcretest
179        to vary, making some of the tests fail. I have changed pcretest so that:
180    
181        (a) When it is outputting text in the compiled version of a pattern, bytes
182            other than 32-126 are always shown as hex escapes.
183    
184        (b) When it is outputting text that is a matched part of a subject string,
185            it does the same, unless a different locale has been set for the match
186            (using the /L modifier). In this case, it uses isprint() to decide.
187    
188     4. Fixed a major bug that caused incorrect computation of the amount of memory
189        required for a compiled pattern when options that changed within the
190        pattern affected the logic of the preliminary scan that determines the
191        length. The relevant options are -x, and -i in UTF-8 mode. The result was
192        that the computed length was too small. The symptoms of this bug were
193        either the PCRE error "internal error: code overflow" from pcre_compile(),
194        or a glibc crash with a message such as "pcretest: free(): invalid next
195        size (fast)". Examples of patterns that provoked this bug (shown in
196        pcretest format) are:
197    
198          /(?-x: )/x
199          /(?x)(?-x: \s*#\s*)/
200          /((?i)[\x{c0}])/8
201          /(?i:[\x{c0}])/8
202    
203        HOWEVER: Change 17 below makes this fix obsolete as the memory computation
204        is now done differently.
205    
206     5. Applied patches from Google to: (a) add a QuoteMeta function to the C++
207        wrapper classes; (b) implement a new function in the C++ scanner that is
208        more efficient than the old way of doing things because it avoids levels of
209        recursion in the regex matching; (c) add a paragraph to the documentation
210        for the FullMatch() function.
211    
212     6. The escape sequence \n was being treated as whatever was defined as
213        "newline". Not only was this contrary to the documentation, which states
214        that \n is character 10 (hex 0A), but it also went horribly wrong when
215        "newline" was defined as CRLF. This has been fixed.
216    
217     7. In pcre_dfa_exec.c the value of an unsigned integer (the variable called c)
218        was being set to -1 for the "end of line" case (supposedly a value that no
219        character can have). Though this value is never used (the check for end of
220        line is "zero bytes in current character"), it caused compiler complaints.
221        I've changed it to 0xffffffff.
222    
223     8. In pcre_version.c, the version string was being built by a sequence of
224        C macros that, in the event of PCRE_PRERELEASE being defined as an empty
225        string (as it is for production releases) called a macro with an empty
226        argument. The C standard says the result of this is undefined. The gcc
227        compiler treats it as an empty string (which was what was wanted) but it is
228        reported that Visual C gives an error. The source has been hacked around to
229        avoid this problem.
230    
231     9. On the advice of a Windows user, included <io.h> and <fcntl.h> in Windows
232        builds of pcretest, and changed the call to _setmode() to use _O_BINARY
233        instead of 0x8000. Made all the #ifdefs test both _WIN32 and WIN32 (not all
234        of them did).
235    
236    10. Originally, pcretest opened its input and output without "b"; then I was
237        told that "b" was needed in some environments, so it was added for release
238        5.0 to both the input and output. (It makes no difference on Unix-like
239        systems.) Later I was told that it is wrong for the input on Windows. I've
240        now abstracted the modes into two macros, to make it easier to fiddle with
241        them, and removed "b" from the input mode under Windows.
242    
243    11. Added pkgconfig support for the C++ wrapper library, libpcrecpp.
244    
245    12. Added -help and --help to pcretest as an official way of being reminded
246        of the options.
247    
248    13. Removed some redundant semicolons after macro calls in pcrecpparg.h.in
249        and pcrecpp.cc because they annoy compilers at high warning levels.
250    
251    14. A bit of tidying/refactoring in pcre_exec.c in the main bumpalong loop.
252    
253    15. Fixed an occurrence of == in configure.ac that should have been = (shell
254        scripts are not C programs :-) and which was not noticed because it works
255        on Linux.
256    
257    16. pcretest is supposed to handle any length of pattern and data line (as one
258        line or as a continued sequence of lines) by extending its input buffer if
259        necessary. This feature was broken for very long pattern lines, leading to
260        a string of junk being passed to pcre_compile() if the pattern was longer
261        than about 50K.
262    
263    17. I have done a major re-factoring of the way pcre_compile() computes the
264        amount of memory needed for a compiled pattern. Previously, there was code
265        that made a preliminary scan of the pattern in order to do this. That was
266        OK when PCRE was new, but as the facilities have expanded, it has become
267        harder and harder to keep it in step with the real compile phase, and there
268        have been a number of bugs (see for example, 4 above). I have now found a
269        cunning way of running the real compile function in a "fake" mode that
270        enables it to compute how much memory it would need, while actually only
271        ever using a few hundred bytes of working memory and without too many
272        tests of the mode. This should make future maintenance and development
273        easier. A side effect of this work is that the limit of 200 on the nesting
274        depth of parentheses has been removed (though this was never a serious
275        limitation, I suspect). However, there is a downside: pcre_compile() now
276        runs more slowly than before (30% or more, depending on the pattern). I
277        hope this isn't a big issue. There is no effect on runtime performance.
278    
279    18. Fixed a minor bug in pcretest: if a pattern line was not terminated by a
280        newline (only possible for the last line of a file) and it was a
281        pattern that set a locale (followed by /Lsomething), pcretest crashed.
282    
283    19. Added additional timing features to pcretest. (1) The -tm option now times
284        matching only, not compiling. (2) Both -t and -tm can be followed, as a
285        separate command line item, by a number that specifies the number of
286        repeats to use when timing. The default is 50000; this gives better
287        precision, but takes uncomfortably long for very large patterns.
288    
289    20. Extended pcre_study() to be more clever in cases where a branch of a
290        subpattern has no definite first character. For example, (a*|b*)[cd] would
291        previously give no result from pcre_study(). Now it recognizes that the
292        first character must be a, b, c, or d.
293    
294    21. There was an incorrect error "recursive call could loop indefinitely" if
295        a subpattern (or the entire pattern) that was being tested for matching an
296        empty string contained only one non-empty item after a nested subpattern.
297        For example, the pattern (?>\x{100}*)\d(?R) provoked this error
298        incorrectly, because the \d was being skipped in the check.
299    
300    22. The pcretest program now has a new pattern option /B and a command line
301        option -b, which is equivalent to adding /B to every pattern. This causes
302        it to show the compiled bytecode, without the additional information that
303        -d shows. The effect of -d is now the same as -b with -i (and similarly, /D
304        is the same as /B/I).
305    
306    23. A new optimization is now able automatically to treat some sequences such
307        as a*b as a*+b. More specifically, if something simple (such as a character
308        or a simple class like \d) has an unlimited quantifier, and is followed by
309        something that cannot possibly match the quantified thing, the quantifier
310        is automatically "possessified".
311    
312    24. A recursive reference to a subpattern whose number was greater than 39
313        went wrong under certain circumstances in UTF-8 mode. This bug could also
314        have affected the operation of pcre_study().
315    
316    25. Realized that a little bit of performance could be had by replacing
317        (c & 0xc0) == 0xc0 with c >= 0xc0 when processing UTF-8 characters.
318    
319    26. Timing data from pcretest is now shown to 4 decimal places instead of 3.
320    
321    27. Possessive quantifiers such as a++ were previously implemented by turning
322        them into atomic groups such as ($>a+). Now they have their own opcodes,
323        which improves performance. This includes the automatically created ones
324        from 23 above.
325    
326    28. A pattern such as (?=(\w+))\1: which simulates an atomic group using a
327        lookahead was broken if it was not anchored. PCRE was mistakenly expecting
328        the first matched character to be a colon. This applied both to named and
329        numbered groups.
330    
331    29. The ucpinternal.h header file was missing its idempotency #ifdef.
332    
333    30. I was sent a "project" file called libpcre.a.dev which I understand makes
334        building PCRE on Windows easier, so I have included it in the distribution.
335    
336    31. There is now a check in pcretest against a ridiculously large number being
337        returned by pcre_exec() or pcre_dfa_exec(). If this happens in a /g or /G
338        loop, the loop is abandoned.
339    
340    32. Forward references to subpatterns in conditions such as (?(2)...) where
341        subpattern 2 is defined later cause pcre_compile() to search forwards in
342        the pattern for the relevant set of parentheses. This search went wrong
343        when there were unescaped parentheses in a character class, parentheses
344        escaped with \Q...\E, or parentheses in a #-comment in /x mode.
345    
346    33. "Subroutine" calls and backreferences were previously restricted to
347        referencing subpatterns earlier in the regex. This restriction has now
348        been removed.
349    
350    34. Added a number of extra features that are going to be in Perl 5.10. On the
351        whole, these are just syntactic alternatives for features that PCRE had
352        previously implemented using the Python syntax or my own invention. The
353        other formats are all retained for compatibility.
354    
355        (a) Named groups can now be defined as (?<name>...) or (?'name'...) as well
356            as (?P<name>...). The new forms, as well as being in Perl 5.10, are
357            also .NET compatible.
358    
359        (b) A recursion or subroutine call to a named group can now be defined as
360            (?&name) as well as (?P>name).
361    
362        (c) A backreference to a named group can now be defined as \k<name> or
363            \k'name' as well as (?P=name). The new forms, as well as being in Perl
364            5.10, are also .NET compatible.
365    
366        (d) A conditional reference to a named group can now use the syntax
367            (?(<name>) or (?('name') as well as (?(name).
368    
369        (e) A "conditional group" of the form (?(DEFINE)...) can be used to define
370            groups (named and numbered) that are never evaluated inline, but can be
371            called as "subroutines" from elsewhere. In effect, the DEFINE condition
372            is always false. There may be only one alternative in such a group.
373    
374        (f) A test for recursion can be given as (?(R1).. or (?(R&name)... as well
375            as the simple (?(R). The condition is true only if the most recent
376            recursion is that of the given number or name. It does not search out
377            through the entire recursion stack.
378    
379        (g) The escape \gN or \g{N} has been added, where N is a positive or
380            negative number, specifying an absolute or relative reference.
381    
382    35. Tidied to get rid of some further signed/unsigned compiler warnings and
383        some "unreachable code" warnings.
384    
385    36. Updated the Unicode property tables to Unicode version 5.0.0. Amongst other
386        things, this adds five new scripts.
387    
388    37. Perl ignores orphaned \E escapes completely. PCRE now does the same.
389        There were also incompatibilities regarding the handling of \Q..\E inside
390        character classes, for example with patterns like [\Qa\E-\Qz\E] where the
391        hyphen was adjacent to \Q or \E. I hope I've cleared all this up now.
392    
393    38. Like Perl, PCRE detects when an indefinitely repeated parenthesized group
394        matches an empty string, and forcibly breaks the loop. There were bugs in
395        this code in non-simple cases. For a pattern such as  ^(a()*)*  matched
396        against  aaaa  the result was just "a" rather than "aaaa", for example. Two
397        separate and independent bugs (that affected different cases) have been
398        fixed.
399    
400    39. Refactored the code to abolish the use of different opcodes for small
401        capturing bracket numbers. This is a tidy that I avoided doing when I
402        removed the limit on the number of capturing brackets for 3.5 back in 2001.
403        The new approach is not only tidier, it makes it possible to reduce the
404        memory needed to fix the previous bug (38).
405    
406    40. Implemented PCRE_NEWLINE_ANY to recognize any of the Unicode newline
407        sequences (http://unicode.org/unicode/reports/tr18/) as "newline" when
408        processing dot, circumflex, or dollar metacharacters, or #-comments in /x
409        mode.
410    
411    41. Add \R to match any Unicode newline sequence, as suggested in the Unicode
412        report.
413    
414    42. Applied patch, originally from Ari Pollak, modified by Google, to allow
415        copy construction and assignment in the C++ wrapper.
416    
417    43. Updated pcregrep to support "--newline=any". In the process, I fixed a
418        couple of bugs that could have given wrong results in the "--newline=crlf"
419        case.
420    
421    44. Added a number of casts and did some reorganization of signed/unsigned int
422        variables following suggestions from Dair Grant. Also renamed the variable
423        "this" as "item" because it is a C++ keyword.
424    
425    45. Arranged for dftables to add
426    
427          #include "pcre_internal.h"
428    
429        to pcre_chartables.c because without it, gcc 4.x may remove the array
430        definition from the final binary if PCRE is built into a static library and
431        dead code stripping is activated.
432    
433    46. For an unanchored pattern, if a match attempt fails at the start of a
434        newline sequence, and the newline setting is CRLF or ANY, and the next two
435        characters are CRLF, advance by two characters instead of one.
436    
437    
438    Version 6.7 04-Jul-06
439    ---------------------
440    
441     1. In order to handle tests when input lines are enormously long, pcretest has
442        been re-factored so that it automatically extends its buffers when
443        necessary. The code is crude, but this _is_ just a test program. The
444        default size has been increased from 32K to 50K.
445    
446     2. The code in pcre_study() was using the value of the re argument before
447        testing it for NULL. (Of course, in any sensible call of the function, it
448        won't be NULL.)
449    
450     3. The memmove() emulation function in pcre_internal.h, which is used on
451        systems that lack both memmove() and bcopy() - that is, hardly ever -
452        was missing a "static" storage class specifier.
453    
454     4. When UTF-8 mode was not set, PCRE looped when compiling certain patterns
455        containing an extended class (one that cannot be represented by a bitmap
456        because it contains high-valued characters or Unicode property items, e.g.
457        [\pZ]). Almost always one would set UTF-8 mode when processing such a
458        pattern, but PCRE should not loop if you do not (it no longer does).
459        [Detail: two cases were found: (a) a repeated subpattern containing an
460        extended class; (b) a recursive reference to a subpattern that followed a
461        previous extended class. It wasn't skipping over the extended class
462        correctly when UTF-8 mode was not set.]
463    
464     5. A negated single-character class was not being recognized as fixed-length
465        in lookbehind assertions such as (?<=[^f]), leading to an incorrect
466        compile error "lookbehind assertion is not fixed length".
467    
468     6. The RunPerlTest auxiliary script was showing an unexpected difference
469        between PCRE and Perl for UTF-8 tests. It turns out that it is hard to
470        write a Perl script that can interpret lines of an input file either as
471        byte characters or as UTF-8, which is what "perltest" was being required to
472        do for the non-UTF-8 and UTF-8 tests, respectively. Essentially what you
473        can't do is switch easily at run time between having the "use utf8;" pragma
474        or not. In the end, I fudged it by using the RunPerlTest script to insert
475        "use utf8;" explicitly for the UTF-8 tests.
476    
477     7. In multiline (/m) mode, PCRE was matching ^ after a terminating newline at
478        the end of the subject string, contrary to the documentation and to what
479        Perl does. This was true of both matching functions. Now it matches only at
480        the start of the subject and immediately after *internal* newlines.
481    
482     8. A call of pcre_fullinfo() from pcretest to get the option bits was passing
483        a pointer to an int instead of a pointer to an unsigned long int. This
484        caused problems on 64-bit systems.
485    
486     9. Applied a patch from the folks at Google to pcrecpp.cc, to fix "another
487        instance of the 'standard' template library not being so standard".
488    
489    10. There was no check on the number of named subpatterns nor the maximum
490        length of a subpattern name. The product of these values is used to compute
491        the size of the memory block for a compiled pattern. By supplying a very
492        long subpattern name and a large number of named subpatterns, the size
493        computation could be caused to overflow. This is now prevented by limiting
494        the length of names to 32 characters, and the number of named subpatterns
495        to 10,000.
496    
497    11. Subpatterns that are repeated with specific counts have to be replicated in
498        the compiled pattern. The size of memory for this was computed from the
499        length of the subpattern and the repeat count. The latter is limited to
500        65535, but there was no limit on the former, meaning that integer overflow
501        could in principle occur. The compiled length of a repeated subpattern is
502        now limited to 30,000 bytes in order to prevent this.
503    
504    12. Added the optional facility to have named substrings with the same name.
505    
506    13. Added the ability to use a named substring as a condition, using the
507        Python syntax: (?(name)yes|no). This overloads (?(R)... and names that
508        are numbers (not recommended). Forward references are permitted.
509    
510    14. Added forward references in named backreferences (if you see what I mean).
511    
512    15. In UTF-8 mode, with the PCRE_DOTALL option set, a quantified dot in the
513        pattern could run off the end of the subject. For example, the pattern
514        "(?s)(.{1,5})"8 did this with the subject "ab".
515    
516    16. If PCRE_DOTALL or PCRE_MULTILINE were set, pcre_dfa_exec() behaved as if
517        PCRE_CASELESS was set when matching characters that were quantified with ?
518        or *.
519    
520    17. A character class other than a single negated character that had a minimum
521        but no maximum quantifier - for example [ab]{6,} - was not handled
522        correctly by pce_dfa_exec(). It would match only one character.
523    
524    18. A valid (though odd) pattern that looked like a POSIX character
525        class but used an invalid character after [ (for example [[,abc,]]) caused
526        pcre_compile() to give the error "Failed: internal error: code overflow" or
527        in some cases to crash with a glibc free() error. This could even happen if
528        the pattern terminated after [[ but there just happened to be a sequence of
529        letters, a binary zero, and a closing ] in the memory that followed.
530    
531    19. Perl's treatment of octal escapes in the range \400 to \777 has changed
532        over the years. Originally (before any Unicode support), just the bottom 8
533        bits were taken. Thus, for example, \500 really meant \100. Nowadays the
534        output from "man perlunicode" includes this:
535    
536          The regular expression compiler produces polymorphic opcodes.  That
537          is, the pattern adapts to the data and automatically switches to
538          the Unicode character scheme when presented with Unicode data--or
539          instead uses a traditional byte scheme when presented with byte
540          data.
541    
542        Sadly, a wide octal escape does not cause a switch, and in a string with
543        no other multibyte characters, these octal escapes are treated as before.
544        Thus, in Perl, the pattern  /\500/ actually matches \100 but the pattern
545        /\500|\x{1ff}/ matches \500 or \777 because the whole thing is treated as a
546        Unicode string.
547    
548        I have not perpetrated such confusion in PCRE. Up till now, it took just
549        the bottom 8 bits, as in old Perl. I have now made octal escapes with
550        values greater than \377 illegal in non-UTF-8 mode. In UTF-8 mode they
551        translate to the appropriate multibyte character.
552    
553    29. Applied some refactoring to reduce the number of warnings from Microsoft
554        and Borland compilers. This has included removing the fudge introduced
555        seven years ago for the OS/2 compiler (see 2.02/2 below) because it caused
556        a warning about an unused variable.
557    
558    21. PCRE has not included VT (character 0x0b) in the set of whitespace
559        characters since release 4.0, because Perl (from release 5.004) does not.
560        [Or at least, is documented not to: some releases seem to be in conflict
561        with the documentation.] However, when a pattern was studied with
562        pcre_study() and all its branches started with \s, PCRE still included VT
563        as a possible starting character. Of course, this did no harm; it just
564        caused an unnecessary match attempt.
565    
566    22. Removed a now-redundant internal flag bit that recorded the fact that case
567        dependency changed within the pattern. This was once needed for "required
568        byte" processing, but is no longer used. This recovers a now-scarce options
569        bit. Also moved the least significant internal flag bit to the most-
570        significant bit of the word, which was not previously used (hangover from
571        the days when it was an int rather than a uint) to free up another bit for
572        the future.
573    
574    23. Added support for CRLF line endings as well as CR and LF. As well as the
575        default being selectable at build time, it can now be changed at runtime
576        via the PCRE_NEWLINE_xxx flags. There are now options for pcregrep to
577        specify that it is scanning data with non-default line endings.
578    
579    24. Changed the definition of CXXLINK to make it agree with the definition of
580        LINK in the Makefile, by replacing LDFLAGS to CXXFLAGS.
581    
582    25. Applied Ian Taylor's patches to avoid using another stack frame for tail
583        recursions. This makes a big different to stack usage for some patterns.
584    
585    26. If a subpattern containing a named recursion or subroutine reference such
586        as (?P>B) was quantified, for example (xxx(?P>B)){3}, the calculation of
587        the space required for the compiled pattern went wrong and gave too small a
588        value. Depending on the environment, this could lead to "Failed: internal
589        error: code overflow at offset 49" or "glibc detected double free or
590        corruption" errors.
591    
592    27. Applied patches from Google (a) to support the new newline modes and (b) to
593        advance over multibyte UTF-8 characters in GlobalReplace.
594    
595    28. Change free() to pcre_free() in pcredemo.c. Apparently this makes a
596        difference for some implementation of PCRE in some Windows version.
597    
598    29. Added some extra testing facilities to pcretest:
599    
600        \q<number>   in a data line sets the "match limit" value
601        \Q<number>   in a data line sets the "match recursion limt" value
602        -S <number>  sets the stack size, where <number> is in megabytes
603    
604        The -S option isn't available for Windows.
605    
606    
607    Version 6.6 06-Feb-06
608    ---------------------
609    
610     1. Change 16(a) for 6.5 broke things, because PCRE_DATA_SCOPE was not defined
611        in pcreposix.h. I have copied the definition from pcre.h.
612    
613     2. Change 25 for 6.5 broke compilation in a build directory out-of-tree
614        because pcre.h is no longer a built file.
615    
616     3. Added Jeff Friedl's additional debugging patches to pcregrep. These are
617        not normally included in the compiled code.
618    
619    
620    Version 6.5 01-Feb-06
621    ---------------------
622    
623     1. When using the partial match feature with pcre_dfa_exec(), it was not
624        anchoring the second and subsequent partial matches at the new starting
625        point. This could lead to incorrect results. For example, with the pattern
626        /1234/, partially matching against "123" and then "a4" gave a match.
627    
628     2. Changes to pcregrep:
629    
630        (a) All non-match returns from pcre_exec() were being treated as failures
631            to match the line. Now, unless the error is PCRE_ERROR_NOMATCH, an
632            error message is output. Some extra information is given for the
633            PCRE_ERROR_MATCHLIMIT and PCRE_ERROR_RECURSIONLIMIT errors, which are
634            probably the only errors that are likely to be caused by users (by
635            specifying a regex that has nested indefinite repeats, for instance).
636            If there are more than 20 of these errors, pcregrep is abandoned.
637    
638        (b) A binary zero was treated as data while matching, but terminated the
639            output line if it was written out. This has been fixed: binary zeroes
640            are now no different to any other data bytes.
641    
642        (c) Whichever of the LC_ALL or LC_CTYPE environment variables is set is
643            used to set a locale for matching. The --locale=xxxx long option has
644            been added (no short equivalent) to specify a locale explicitly on the
645            pcregrep command, overriding the environment variables.
646    
647        (d) When -B was used with -n, some line numbers in the output were one less
648            than they should have been.
649    
650        (e) Added the -o (--only-matching) option.
651    
652        (f) If -A or -C was used with -c (count only), some lines of context were
653            accidentally printed for the final match.
654    
655        (g) Added the -H (--with-filename) option.
656    
657        (h) The combination of options -rh failed to suppress file names for files
658            that were found from directory arguments.
659    
660        (i) Added the -D (--devices) and -d (--directories) options.
661    
662        (j) Added the -F (--fixed-strings) option.
663    
664        (k) Allow "-" to be used as a file name for -f as well as for a data file.
665    
666        (l) Added the --colo(u)r option.
667    
668        (m) Added Jeffrey Friedl's -S testing option, but within #ifdefs so that it
669            is not present by default.
670    
671     3. A nasty bug was discovered in the handling of recursive patterns, that is,
672        items such as (?R) or (?1), when the recursion could match a number of
673        alternatives. If it matched one of the alternatives, but subsequently,
674        outside the recursion, there was a failure, the code tried to back up into
675        the recursion. However, because of the way PCRE is implemented, this is not
676        possible, and the result was an incorrect result from the match.
677    
678        In order to prevent this happening, the specification of recursion has
679        been changed so that all such subpatterns are automatically treated as
680        atomic groups. Thus, for example, (?R) is treated as if it were (?>(?R)).
681    
682     4. I had overlooked the fact that, in some locales, there are characters for
683        which isalpha() is true but neither isupper() nor islower() are true. In
684        the fr_FR locale, for instance, the \xAA and \xBA characters (ordmasculine
685        and ordfeminine) are like this. This affected the treatment of \w and \W
686        when they appeared in character classes, but not when they appeared outside
687        a character class. The bit map for "word" characters is now created
688        separately from the results of isalnum() instead of just taking it from the
689        upper, lower, and digit maps. (Plus the underscore character, of course.)
690    
691     5. The above bug also affected the handling of POSIX character classes such as
692        [[:alpha:]] and [[:alnum:]]. These do not have their own bit maps in PCRE's
693        permanent tables. Instead, the bit maps for such a class were previously
694        created as the appropriate unions of the upper, lower, and digit bitmaps.
695        Now they are created by subtraction from the [[:word:]] class, which has
696        its own bitmap.
697    
698     6. The [[:blank:]] character class matches horizontal, but not vertical space.
699        It is created by subtracting the vertical space characters (\x09, \x0a,
700        \x0b, \x0c) from the [[:space:]] bitmap. Previously, however, the
701        subtraction was done in the overall bitmap for a character class, meaning
702        that a class such as [\x0c[:blank:]] was incorrect because \x0c would not
703        be recognized. This bug has been fixed.
704    
705     7. Patches from the folks at Google:
706    
707          (a) pcrecpp.cc: "to handle a corner case that may or may not happen in
708          real life, but is still worth protecting against".
709    
710          (b) pcrecpp.cc: "corrects a bug when negative radixes are used with
711          regular expressions".
712    
713          (c) pcre_scanner.cc: avoid use of std::count() because not all systems
714          have it.
715    
716          (d) Split off pcrecpparg.h from pcrecpp.h and had the former built by
717          "configure" and the latter not, in order to fix a problem somebody had
718          with compiling the Arg class on HP-UX.
719    
720          (e) Improve the error-handling of the C++ wrapper a little bit.
721    
722          (f) New tests for checking recursion limiting.
723    
724     8. The pcre_memmove() function, which is used only if the environment does not
725        have a standard memmove() function (and is therefore rarely compiled),
726        contained two bugs: (a) use of int instead of size_t, and (b) it was not
727        returning a result (though PCRE never actually uses the result).
728    
729     9. In the POSIX regexec() interface, if nmatch is specified as a ridiculously
730        large number - greater than INT_MAX/(3*sizeof(int)) - REG_ESPACE is
731        returned instead of calling malloc() with an overflowing number that would
732        most likely cause subsequent chaos.
733    
734    10. The debugging option of pcretest was not showing the NO_AUTO_CAPTURE flag.
735    
736    11. The POSIX flag REG_NOSUB is now supported. When a pattern that was compiled
737        with this option is matched, the nmatch and pmatch options of regexec() are
738        ignored.
739    
740    12. Added REG_UTF8 to the POSIX interface. This is not defined by POSIX, but is
741        provided in case anyone wants to the the POSIX interface with UTF-8
742        strings.
743    
744    13. Added CXXLDFLAGS to the Makefile parameters to provide settings only on the
745        C++ linking (needed for some HP-UX environments).
746    
747    14. Avoid compiler warnings in get_ucpname() when compiled without UCP support
748        (unused parameter) and in the pcre_printint() function (omitted "default"
749        switch label when the default is to do nothing).
750    
751    15. Added some code to make it possible, when PCRE is compiled as a C++
752        library, to replace subject pointers for pcre_exec() with a smart pointer
753        class, thus making it possible to process discontinuous strings.
754    
755    16. The two macros PCRE_EXPORT and PCRE_DATA_SCOPE are confusing, and perform
756        much the same function. They were added by different people who were trying
757        to make PCRE easy to compile on non-Unix systems. It has been suggested
758        that PCRE_EXPORT be abolished now that there is more automatic apparatus
759        for compiling on Windows systems. I have therefore replaced it with
760        PCRE_DATA_SCOPE. This is set automatically for Windows; if not set it
761        defaults to "extern" for C or "extern C" for C++, which works fine on
762        Unix-like systems. It is now possible to override the value of PCRE_DATA_
763        SCOPE with something explicit in config.h. In addition:
764    
765        (a) pcreposix.h still had just "extern" instead of either of these macros;
766            I have replaced it with PCRE_DATA_SCOPE.
767    
768        (b) Functions such as _pcre_xclass(), which are internal to the library,
769            but external in the C sense, all had PCRE_EXPORT in their definitions.
770            This is apparently wrong for the Windows case, so I have removed it.
771            (It makes no difference on Unix-like systems.)
772    
773    17. Added a new limit, MATCH_LIMIT_RECURSION, which limits the depth of nesting
774        of recursive calls to match(). This is different to MATCH_LIMIT because
775        that limits the total number of calls to match(), not all of which increase
776        the depth of recursion. Limiting the recursion depth limits the amount of
777        stack (or heap if NO_RECURSE is set) that is used. The default can be set
778        when PCRE is compiled, and changed at run time. A patch from Google adds
779        this functionality to the C++ interface.
780    
781    18. Changes to the handling of Unicode character properties:
782    
783        (a) Updated the table to Unicode 4.1.0.
784    
785        (b) Recognize characters that are not in the table as "Cn" (undefined).
786    
787        (c) I revised the way the table is implemented to a much improved format
788            which includes recognition of ranges. It now supports the ranges that
789            are defined in UnicodeData.txt, and it also amalgamates other
790            characters into ranges. This has reduced the number of entries in the
791            table from around 16,000 to around 3,000, thus reducing its size
792            considerably. I realized I did not need to use a tree structure after
793            all - a binary chop search is just as efficient. Having reduced the
794            number of entries, I extended their size from 6 bytes to 8 bytes to
795            allow for more data.
796    
797        (d) Added support for Unicode script names via properties such as \p{Han}.
798    
799    19. In UTF-8 mode, a backslash followed by a non-Ascii character was not
800        matching that character.
801    
802    20. When matching a repeated Unicode property with a minimum greater than zero,
803        (for example \pL{2,}), PCRE could look past the end of the subject if it
804        reached it while seeking the minimum number of characters. This could
805        happen only if some of the characters were more than one byte long, because
806        there is a check for at least the minimum number of bytes.
807    
808    21. Refactored the implementation of \p and \P so as to be more general, to
809        allow for more different types of property in future. This has changed the
810        compiled form incompatibly. Anybody with saved compiled patterns that use
811        \p or \P will have to recompile them.
812    
813    22. Added "Any" and "L&" to the supported property types.
814    
815    23. Recognize \x{...} as a code point specifier, even when not in UTF-8 mode,
816        but give a compile time error if the value is greater than 0xff.
817    
818    24. The man pages for pcrepartial, pcreprecompile, and pcre_compile2 were
819        accidentally not being installed or uninstalled.
820    
821    25. The pcre.h file was built from pcre.h.in, but the only changes that were
822        made were to insert the current release number. This seemed silly, because
823        it made things harder for people building PCRE on systems that don't run
824        "configure". I have turned pcre.h into a distributed file, no longer built
825        by "configure", with the version identification directly included. There is
826        no longer a pcre.h.in file.
827    
828        However, this change necessitated a change to the pcre-config script as
829        well. It is built from pcre-config.in, and one of the substitutions was the
830        release number. I have updated configure.ac so that ./configure now finds
831        the release number by grepping pcre.h.
832    
833    26. Added the ability to run the tests under valgrind.
834    
835    
836    Version 6.4 05-Sep-05
837    ---------------------
838    
839     1. Change 6.0/10/(l) to pcregrep introduced a bug that caused separator lines
840        "--" to be printed when multiple files were scanned, even when none of the
841        -A, -B, or -C options were used. This is not compatible with Gnu grep, so I
842        consider it to be a bug, and have restored the previous behaviour.
843    
844     2. A couple of code tidies to get rid of compiler warnings.
845    
846     3. The pcretest program used to cheat by referring to symbols in the library
847        whose names begin with _pcre_. These are internal symbols that are not
848        really supposed to be visible externally, and in some environments it is
849        possible to suppress them. The cheating is now confined to including
850        certain files from the library's source, which is a bit cleaner.
851    
852     4. Renamed pcre.in as pcre.h.in to go with pcrecpp.h.in; it also makes the
853        file's purpose clearer.
854    
855     5. Reorganized pcre_ucp_findchar().
856    
857    
858    Version 6.3 15-Aug-05
859    ---------------------
860    
861     1. The file libpcre.pc.in did not have general read permission in the tarball.
862    
863     2. There were some problems when building without C++ support:
864    
865        (a) If C++ support was not built, "make install" and "make test" still
866            tried to test it.
867    
868        (b) There were problems when the value of CXX was explicitly set. Some
869            changes have been made to try to fix these, and ...
870    
871        (c) --disable-cpp can now be used to explicitly disable C++ support.
872    
873        (d) The use of @CPP_OBJ@ directly caused a blank line preceded by a
874            backslash in a target when C++ was disabled. This confuses some
875            versions of "make", apparently. Using an intermediate variable solves
876            this. (Same for CPP_LOBJ.)
877    
878     3. $(LINK_FOR_BUILD) now includes $(CFLAGS_FOR_BUILD) and $(LINK)
879        (non-Windows) now includes $(CFLAGS) because these flags are sometimes
880        necessary on certain architectures.
881    
882     4. Added a setting of -export-symbols-regex to the link command to remove
883        those symbols that are exported in the C sense, but actually are local
884        within the library, and not documented. Their names all begin with
885        "_pcre_". This is not a perfect job, because (a) we have to except some
886        symbols that pcretest ("illegally") uses, and (b) the facility isn't always
887        available (and never for static libraries). I have made a note to try to
888        find a way round (a) in the future.
889    
890    
891    Version 6.2 01-Aug-05
892    ---------------------
893    
894     1. There was no test for integer overflow of quantifier values. A construction
895        such as {1111111111111111} would give undefined results. What is worse, if
896        a minimum quantifier for a parenthesized subpattern overflowed and became
897        negative, the calculation of the memory size went wrong. This could have
898        led to memory overwriting.
899    
900     2. Building PCRE using VPATH was broken. Hopefully it is now fixed.
901    
902     3. Added "b" to the 2nd argument of fopen() in dftables.c, for non-Unix-like
903        operating environments where this matters.
904    
905     4. Applied Giuseppe Maxia's patch to add additional features for controlling
906        PCRE options from within the C++ wrapper.
907    
908     5. Named capturing subpatterns were not being correctly counted when a pattern
909        was compiled. This caused two problems: (a) If there were more than 100
910        such subpatterns, the calculation of the memory needed for the whole
911        compiled pattern went wrong, leading to an overflow error. (b) Numerical
912        back references of the form \12, where the number was greater than 9, were
913        not recognized as back references, even though there were sufficient
914        previous subpatterns.
915    
916     6. Two minor patches to pcrecpp.cc in order to allow it to compile on older
917        versions of gcc, e.g. 2.95.4.
918    
919    
920    Version 6.1 21-Jun-05
921    ---------------------
922    
923     1. There was one reference to the variable "posix" in pcretest.c that was not
924        surrounded by "#if !defined NOPOSIX".
925    
926     2. Make it possible to compile pcretest without DFA support, UTF8 support, or
927        the cross-check on the old pcre_info() function, for the benefit of the
928        cut-down version of PCRE that is currently imported into Exim.
929    
930     3. A (silly) pattern starting with (?i)(?-i) caused an internal space
931        allocation error. I've done the easy fix, which wastes 2 bytes for sensible
932        patterns that start (?i) but I don't think that matters. The use of (?i) is
933        just an example; this all applies to the other options as well.
934    
935     4. Since libtool seems to echo the compile commands it is issuing, the output
936        from "make" can be reduced a bit by putting "@" in front of each libtool
937        compile command.
938    
939     5. Patch from the folks at Google for configure.in to be a bit more thorough
940        in checking for a suitable C++ installation before trying to compile the
941        C++ stuff. This should fix a reported problem when a compiler was present,
942        but no suitable headers.
943    
944     6. The man pages all had just "PCRE" as their title. I have changed them to
945        be the relevant file name. I have also arranged that these names are
946        retained in the file doc/pcre.txt, which is a concatenation in text format
947        of all the man pages except the little individual ones for each function.
948    
949     7. The NON-UNIX-USE file had not been updated for the different set of source
950        files that come with release 6. I also added a few comments about the C++
951        wrapper.
952    
953    
954    Version 6.0 07-Jun-05
955    ---------------------
956    
957     1. Some minor internal re-organization to help with my DFA experiments.
958    
959     2. Some missing #ifdef SUPPORT_UCP conditionals in pcretest and printint that
960        didn't matter for the library itself when fully configured, but did matter
961        when compiling without UCP support, or within Exim, where the ucp files are
962        not imported.
963    
964     3. Refactoring of the library code to split up the various functions into
965        different source modules. The addition of the new DFA matching code (see
966        below) to a single monolithic source would have made it really too
967        unwieldy, quite apart from causing all the code to be include in a
968        statically linked application, when only some functions are used. This is
969        relevant even without the DFA addition now that patterns can be compiled in
970        one application and matched in another.
971    
972        The downside of splitting up is that there have to be some external
973        functions and data tables that are used internally in different modules of
974        the library but which are not part of the API. These have all had their
975        names changed to start with "_pcre_" so that they are unlikely to clash
976        with other external names.
977    
978     4. Added an alternate matching function, pcre_dfa_exec(), which matches using
979        a different (DFA) algorithm. Although it is slower than the original
980        function, it does have some advantages for certain types of matching
981        problem.
982    
983     5. Upgrades to pcretest in order to test the features of pcre_dfa_exec(),
984        including restarting after a partial match.
985    
986     6. A patch for pcregrep that defines INVALID_FILE_ATTRIBUTES if it is not
987        defined when compiling for Windows was sent to me. I have put it into the
988        code, though I have no means of testing or verifying it.
989    
990     7. Added the pcre_refcount() auxiliary function.
991    
992     8. Added the PCRE_FIRSTLINE option. This constrains an unanchored pattern to
993        match before or at the first newline in the subject string. In pcretest,
994        the /f option on a pattern can be used to set this.
995    
996     9. A repeated \w when used in UTF-8 mode with characters greater than 256
997        would behave wrongly. This has been present in PCRE since release 4.0.
998    
999    10. A number of changes to the pcregrep command:
1000    
1001        (a) Refactored how -x works; insert ^(...)$ instead of setting
1002            PCRE_ANCHORED and checking the length, in preparation for adding
1003            something similar for -w.
1004    
1005        (b) Added the -w (match as a word) option.
1006    
1007        (c) Refactored the way lines are read and buffered so as to have more
1008            than one at a time available.
1009    
1010        (d) Implemented a pcregrep test script.
1011    
1012        (e) Added the -M (multiline match) option. This allows patterns to match
1013            over several lines of the subject. The buffering ensures that at least
1014            8K, or the rest of the document (whichever is the shorter) is available
1015            for matching (and similarly the previous 8K for lookbehind assertions).
1016    
1017        (f) Changed the --help output so that it now says
1018    
1019              -w, --word-regex(p)
1020    
1021            instead of two lines, one with "regex" and the other with "regexp"
1022            because that confused at least one person since the short forms are the
1023            same. (This required a bit of code, as the output is generated
1024            automatically from a table. It wasn't just a text change.)
1025    
1026        (g) -- can be used to terminate pcregrep options if the next thing isn't an
1027            option but starts with a hyphen. Could be a pattern or a path name
1028            starting with a hyphen, for instance.
1029    
1030        (h) "-" can be given as a file name to represent stdin.
1031    
1032        (i) When file names are being printed, "(standard input)" is used for
1033            the standard input, for compatibility with GNU grep. Previously
1034            "<stdin>" was used.
1035    
1036        (j) The option --label=xxx can be used to supply a name to be used for
1037            stdin when file names are being printed. There is no short form.
1038    
1039        (k) Re-factored the options decoding logic because we are going to add
1040            two more options that take data. Such options can now be given in four
1041            different ways, e.g. "-fname", "-f name", "--file=name", "--file name".
1042    
1043        (l) Added the -A, -B, and -C options for requesting that lines of context
1044            around matches be printed.
1045    
1046        (m) Added the -L option to print the names of files that do not contain
1047            any matching lines, that is, the complement of -l.
1048    
1049        (n) The return code is 2 if any file cannot be opened, but pcregrep does
1050            continue to scan other files.
1051    
1052        (o) The -s option was incorrectly implemented. For compatibility with other
1053            greps, it now suppresses the error message for a non-existent or non-
1054            accessible file (but not the return code). There is a new option called
1055            -q that suppresses the output of matching lines, which was what -s was
1056            previously doing.
1057    
1058        (p) Added --include and --exclude options to specify files for inclusion
1059            and exclusion when recursing.
1060    
1061    11. The Makefile was not using the Autoconf-supported LDFLAGS macro properly.
1062        Hopefully, it now does.
1063    
1064    12. Missing cast in pcre_study().
1065    
1066    13. Added an "uninstall" target to the makefile.
1067    
1068    14. Replaced "extern" in the function prototypes in Makefile.in with
1069        "PCRE_DATA_SCOPE", which defaults to 'extern' or 'extern "C"' in the Unix
1070        world, but is set differently for Windows.
1071    
1072    15. Added a second compiling function called pcre_compile2(). The only
1073        difference is that it has an extra argument, which is a pointer to an
1074        integer error code. When there is a compile-time failure, this is set
1075        non-zero, in addition to the error test pointer being set to point to an
1076        error message. The new argument may be NULL if no error number is required
1077        (but then you may as well call pcre_compile(), which is now just a
1078        wrapper). This facility is provided because some applications need a
1079        numeric error indication, but it has also enabled me to tidy up the way
1080        compile-time errors are handled in the POSIX wrapper.
1081    
1082    16. Added VPATH=.libs to the makefile; this should help when building with one
1083        prefix path and installing with another. (Or so I'm told by someone who
1084        knows more about this stuff than I do.)
1085    
1086    17. Added a new option, REG_DOTALL, to the POSIX function regcomp(). This
1087        passes PCRE_DOTALL to the pcre_compile() function, making the "." character
1088        match everything, including newlines. This is not POSIX-compatible, but
1089        somebody wanted the feature. From pcretest it can be activated by using
1090        both the P and the s flags.
1091    
1092    18. AC_PROG_LIBTOOL appeared twice in Makefile.in. Removed one.
1093    
1094    19. libpcre.pc was being incorrectly installed as executable.
1095    
1096    20. A couple of places in pcretest check for end-of-line by looking for '\n';
1097        it now also looks for '\r' so that it will work unmodified on Windows.
1098    
1099    21. Added Google's contributed C++ wrapper to the distribution.
1100    
1101    22. Added some untidy missing memory free() calls in pcretest, to keep
1102        Electric Fence happy when testing.
1103    
1104    
1105    
1106    Version 5.0 13-Sep-04
1107    ---------------------
1108    
1109     1. Internal change: literal characters are no longer packed up into items
1110        containing multiple characters in a single byte-string. Each character
1111        is now matched using a separate opcode. However, there may be more than one
1112        byte in the character in UTF-8 mode.
1113    
1114     2. The pcre_callout_block structure has two new fields: pattern_position and
1115        next_item_length. These contain the offset in the pattern to the next match
1116        item, and its length, respectively.
1117    
1118     3. The PCRE_AUTO_CALLOUT option for pcre_compile() requests the automatic
1119        insertion of callouts before each pattern item. Added the /C option to
1120        pcretest to make use of this.
1121    
1122     4. On the advice of a Windows user, the lines
1123    
1124          #if defined(_WIN32) || defined(WIN32)
1125          _setmode( _fileno( stdout ), 0x8000 );
1126          #endif  /* defined(_WIN32) || defined(WIN32) */
1127    
1128        have been added to the source of pcretest. This apparently does useful
1129        magic in relation to line terminators.
1130    
1131     5. Changed "r" and "w" in the calls to fopen() in pcretest to "rb" and "wb"
1132        for the benefit of those environments where the "b" makes a difference.
1133    
1134     6. The icc compiler has the same options as gcc, but "configure" doesn't seem
1135        to know about it. I have put a hack into configure.in that adds in code
1136        to set GCC=yes if CC=icc. This seems to end up at a point in the
1137        generated configure script that is early enough to affect the setting of
1138        compiler options, which is what is needed, but I have no means of testing
1139        whether it really works. (The user who reported this had patched the
1140        generated configure script, which of course I cannot do.)
1141    
1142        LATER: After change 22 below (new libtool files), the configure script
1143        seems to know about icc (and also ecc). Therefore, I have commented out
1144        this hack in configure.in.
1145    
1146     7. Added support for pkg-config (2 patches were sent in).
1147    
1148     8. Negated POSIX character classes that used a combination of internal tables
1149        were completely broken. These were [[:^alpha:]], [[:^alnum:]], and
1150        [[:^ascii]]. Typically, they would match almost any characters. The other
1151        POSIX classes were not broken in this way.
1152    
1153     9. Matching the pattern "\b.*?" against "ab cd", starting at offset 1, failed
1154        to find the match, as PCRE was deluded into thinking that the match had to
1155        start at the start point or following a newline. The same bug applied to
1156        patterns with negative forward assertions or any backward assertions
1157        preceding ".*" at the start, unless the pattern required a fixed first
1158        character. This was a failing pattern: "(?!.bcd).*". The bug is now fixed.
1159    
1160    10. In UTF-8 mode, when moving forwards in the subject after a failed match
1161        starting at the last subject character, bytes beyond the end of the subject
1162        string were read.
1163    
1164    11. Renamed the variable "class" as "classbits" to make life easier for C++
1165        users. (Previously there was a macro definition, but it apparently wasn't
1166        enough.)
1167    
1168    12. Added the new field "tables" to the extra data so that tables can be passed
1169        in at exec time, or the internal tables can be re-selected. This allows
1170        a compiled regex to be saved and re-used at a later time by a different
1171        program that might have everything at different addresses.
1172    
1173    13. Modified the pcre-config script so that, when run on Solaris, it shows a
1174        -R library as well as a -L library.
1175    
1176    14. The debugging options of pcretest (-d on the command line or D on a
1177        pattern) showed incorrect output for anything following an extended class
1178        that contained multibyte characters and which was followed by a quantifier.
1179    
1180    15. Added optional support for general category Unicode character properties
1181        via the \p, \P, and \X escapes. Unicode property support implies UTF-8
1182        support. It adds about 90K to the size of the library. The meanings of the
1183        inbuilt class escapes such as \d and \s have NOT been changed.
1184    
1185    16. Updated pcredemo.c to include calls to free() to release the memory for the
1186        compiled pattern.
1187    
1188    17. The generated file chartables.c was being created in the source directory
1189        instead of in the building directory. This caused the build to fail if the
1190        source directory was different from the building directory, and was
1191        read-only.
1192    
1193    18. Added some sample Win commands from Mark Tetrode into the NON-UNIX-USE
1194        file. No doubt somebody will tell me if they don't make sense... Also added
1195        Dan Mooney's comments about building on OpenVMS.
1196    
1197    19. Added support for partial matching via the PCRE_PARTIAL option for
1198        pcre_exec() and the \P data escape in pcretest.
1199    
1200    20. Extended pcretest with 3 new pattern features:
1201    
1202        (i)   A pattern option of the form ">rest-of-line" causes pcretest to
1203              write the compiled pattern to the file whose name is "rest-of-line".
1204              This is a straight binary dump of the data, with the saved pointer to
1205              the character tables forced to be NULL. The study data, if any, is
1206              written too. After writing, pcretest reads a new pattern.
1207    
1208        (ii)  If, instead of a pattern, "<rest-of-line" is given, pcretest reads a
1209              compiled pattern from the given file. There must not be any
1210              occurrences of "<" in the file name (pretty unlikely); if there are,
1211              pcretest will instead treat the initial "<" as a pattern delimiter.
1212              After reading in the pattern, pcretest goes on to read data lines as
1213              usual.
1214    
1215        (iii) The F pattern option causes pcretest to flip the bytes in the 32-bit
1216              and 16-bit fields in a compiled pattern, to simulate a pattern that
1217              was compiled on a host of opposite endianness.
1218    
1219    21. The pcre-exec() function can now cope with patterns that were compiled on
1220        hosts of opposite endianness, with this restriction:
1221    
1222          As for any compiled expression that is saved and used later, the tables
1223          pointer field cannot be preserved; the extra_data field in the arguments
1224          to pcre_exec() should be used to pass in a tables address if a value
1225          other than the default internal tables were used at compile time.
1226    
1227    22. Calling pcre_exec() with a negative value of the "ovecsize" parameter is
1228        now diagnosed as an error. Previously, most of the time, a negative number
1229        would have been treated as zero, but if in addition "ovector" was passed as
1230        NULL, a crash could occur.
1231    
1232    23. Updated the files ltmain.sh, config.sub, config.guess, and aclocal.m4 with
1233        new versions from the libtool 1.5 distribution (the last one is a copy of
1234        a file called libtool.m4). This seems to have fixed the need to patch
1235        "configure" to support Darwin 1.3 (which I used to do). However, I still
1236        had to patch ltmain.sh to ensure that ${SED} is set (it isn't on my
1237        workstation).
1238    
1239    24. Changed the PCRE licence to be the more standard "BSD" licence.
1240    
1241    
1242    Version 4.5 01-Dec-03
1243    ---------------------
1244    
1245     1. There has been some re-arrangement of the code for the match() function so
1246        that it can be compiled in a version that does not call itself recursively.
1247        Instead, it keeps those local variables that need separate instances for
1248        each "recursion" in a frame on the heap, and gets/frees frames whenever it
1249        needs to "recurse". Keeping track of where control must go is done by means
1250        of setjmp/longjmp. The whole thing is implemented by a set of macros that
1251        hide most of the details from the main code, and operates only if
1252        NO_RECURSE is defined while compiling pcre.c. If PCRE is built using the
1253        "configure" mechanism, "--disable-stack-for-recursion" turns on this way of
1254        operating.
1255    
1256        To make it easier for callers to provide specially tailored get/free
1257        functions for this usage, two new functions, pcre_stack_malloc, and
1258        pcre_stack_free, are used. They are always called in strict stacking order,
1259        and the size of block requested is always the same.
1260    
1261        The PCRE_CONFIG_STACKRECURSE info parameter can be used to find out whether
1262        PCRE has been compiled to use the stack or the heap for recursion. The
1263        -C option of pcretest uses this to show which version is compiled.
1264    
1265        A new data escape \S, is added to pcretest; it causes the amounts of store
1266        obtained and freed by both kinds of malloc/free at match time to be added
1267        to the output.
1268    
1269     2. Changed the locale test to use "fr_FR" instead of "fr" because that's
1270        what's available on my current Linux desktop machine.
1271    
1272     3. When matching a UTF-8 string, the test for a valid string at the start has
1273        been extended. If start_offset is not zero, PCRE now checks that it points
1274        to a byte that is the start of a UTF-8 character. If not, it returns
1275        PCRE_ERROR_BADUTF8_OFFSET (-11). Note: the whole string is still checked;
1276        this is necessary because there may be backward assertions in the pattern.
1277        When matching the same subject several times, it may save resources to use
1278        PCRE_NO_UTF8_CHECK on all but the first call if the string is long.
1279    
1280     4. The code for checking the validity of UTF-8 strings has been tightened so
1281        that it rejects (a) strings containing 0xfe or 0xff bytes and (b) strings
1282        containing "overlong sequences".
1283    
1284     5. Fixed a bug (appearing twice) that I could not find any way of exploiting!
1285        I had written "if ((digitab[*p++] && chtab_digit) == 0)" where the "&&"
1286        should have been "&", but it just so happened that all the cases this let
1287        through by mistake were picked up later in the function.
1288    
1289     6. I had used a variable called "isblank" - this is a C99 function, causing
1290        some compilers to warn. To avoid this, I renamed it (as "blankclass").
1291    
1292     7. Cosmetic: (a) only output another newline at the end of pcretest if it is
1293        prompting; (b) run "./pcretest /dev/null" at the start of the test script
1294        so the version is shown; (c) stop "make test" echoing "./RunTest".
1295    
1296     8. Added patches from David Burgess to enable PCRE to run on EBCDIC systems.
1297    
1298     9. The prototype for memmove() for systems that don't have it was using
1299        size_t, but the inclusion of the header that defines size_t was later. I've
1300        moved the #includes for the C headers earlier to avoid this.
1301    
1302    10. Added some adjustments to the code to make it easier to compiler on certain
1303        special systems:
1304    
1305          (a) Some "const" qualifiers were missing.
1306          (b) Added the macro EXPORT before all exported functions; by default this
1307              is defined to be empty.
1308          (c) Changed the dftables auxiliary program (that builds chartables.c) so
1309              that it reads its output file name as an argument instead of writing
1310              to the standard output and assuming this can be redirected.
1311    
1312    11. In UTF-8 mode, if a recursive reference (e.g. (?1)) followed a character
1313        class containing characters with values greater than 255, PCRE compilation
1314        went into a loop.
1315    
1316    12. A recursive reference to a subpattern that was within another subpattern
1317        that had a minimum quantifier of zero caused PCRE to crash. For example,
1318        (x(y(?2))z)? provoked this bug with a subject that got as far as the
1319        recursion. If the recursively-called subpattern itself had a zero repeat,
1320        that was OK.
1321    
1322    13. In pcretest, the buffer for reading a data line was set at 30K, but the
1323        buffer into which it was copied (for escape processing) was still set at
1324        1024, so long lines caused crashes.
1325    
1326    14. A pattern such as /[ab]{1,3}+/ failed to compile, giving the error
1327        "internal error: code overflow...". This applied to any character class
1328        that was followed by a possessive quantifier.
1329    
1330    15. Modified the Makefile to add libpcre.la as a prerequisite for
1331        libpcreposix.la because I was told this is needed for a parallel build to
1332        work.
1333    
1334    16. If a pattern that contained .* following optional items at the start was
1335        studied, the wrong optimizing data was generated, leading to matching
1336        errors. For example, studying /[ab]*.*c/ concluded, erroneously, that any
1337        matching string must start with a or b or c. The correct conclusion for
1338        this pattern is that a match can start with any character.
1339    
1340    
1341    Version 4.4 13-Aug-03
1342    ---------------------
1343    
1344     1. In UTF-8 mode, a character class containing characters with values between
1345        127 and 255 was not handled correctly if the compiled pattern was studied.
1346        In fixing this, I have also improved the studying algorithm for such
1347        classes (slightly).
1348    
1349     2. Three internal functions had redundant arguments passed to them. Removal
1350        might give a very teeny performance improvement.
1351    
1352     3. Documentation bug: the value of the capture_top field in a callout is *one
1353        more than* the number of the hightest numbered captured substring.
1354    
1355     4. The Makefile linked pcretest and pcregrep with -lpcre, which could result
1356        in incorrectly linking with a previously installed version. They now link
1357        explicitly with libpcre.la.
1358    
1359     5. configure.in no longer needs to recognize Cygwin specially.
1360    
1361     6. A problem in pcre.in for Windows platforms is fixed.
1362    
1363     7. If a pattern was successfully studied, and the -d (or /D) flag was given to
1364        pcretest, it used to include the size of the study block as part of its
1365        output. Unfortunately, the structure contains a field that has a different
1366        size on different hardware architectures. This meant that the tests that
1367        showed this size failed. As the block is currently always of a fixed size,
1368        this information isn't actually particularly useful in pcretest output, so
1369        I have just removed it.
1370    
1371     8. Three pre-processor statements accidentally did not start in column 1.
1372        Sadly, there are *still* compilers around that complain, even though
1373        standard C has not required this for well over a decade. Sigh.
1374    
1375     9. In pcretest, the code for checking callouts passed small integers in the
1376        callout_data field, which is a void * field. However, some picky compilers
1377        complained about the casts involved for this on 64-bit systems. Now
1378        pcretest passes the address of the small integer instead, which should get
1379        rid of the warnings.
1380    
1381    10. By default, when in UTF-8 mode, PCRE now checks for valid UTF-8 strings at
1382        both compile and run time, and gives an error if an invalid UTF-8 sequence
1383        is found. There is a option for disabling this check in cases where the
1384        string is known to be correct and/or the maximum performance is wanted.
1385    
1386    11. In response to a bug report, I changed one line in Makefile.in from
1387    
1388            -Wl,--out-implib,.libs/lib@WIN_PREFIX@pcreposix.dll.a \
1389        to
1390            -Wl,--out-implib,.libs/@WIN_PREFIX@libpcreposix.dll.a \
1391    
1392        to look similar to other lines, but I have no way of telling whether this
1393        is the right thing to do, as I do not use Windows. No doubt I'll get told
1394        if it's wrong...
1395    
1396    
1397    Version 4.3 21-May-03
1398    ---------------------
1399    
1400    1. Two instances of @WIN_PREFIX@ omitted from the Windows targets in the
1401       Makefile.
1402    
1403    2. Some refactoring to improve the quality of the code:
1404    
1405       (i)   The utf8_table... variables are now declared "const".
1406    
1407       (ii)  The code for \cx, which used the "case flipping" table to upper case
1408             lower case letters, now just substracts 32. This is ASCII-specific,
1409             but the whole concept of \cx is ASCII-specific, so it seems
1410             reasonable.
1411    
1412       (iii) PCRE was using its character types table to recognize decimal and
1413             hexadecimal digits in the pattern. This is silly, because it handles
1414             only 0-9, a-f, and A-F, but the character types table is locale-
1415             specific, which means strange things might happen. A private
1416             table is now used for this - though it costs 256 bytes, a table is
1417             much faster than multiple explicit tests. Of course, the standard
1418             character types table is still used for matching digits in subject
1419             strings against \d.
1420    
1421       (iv)  Strictly, the identifier ESC_t is reserved by POSIX (all identifiers
1422             ending in _t are). So I've renamed it as ESC_tee.
1423    
1424    3. The first argument for regexec() in the POSIX wrapper should have been
1425       defined as "const".
1426    
1427    4. Changed pcretest to use malloc() for its buffers so that they can be
1428       Electric Fenced for debugging.
1429    
1430    5. There were several places in the code where, in UTF-8 mode, PCRE would try
1431       to read one or more bytes before the start of the subject string. Often this
1432       had no effect on PCRE's behaviour, but in some circumstances it could
1433       provoke a segmentation fault.
1434    
1435    6. A lookbehind at the start of a pattern in UTF-8 mode could also cause PCRE
1436       to try to read one or more bytes before the start of the subject string.
1437    
1438    7. A lookbehind in a pattern matched in non-UTF-8 mode on a PCRE compiled with
1439       UTF-8 support could misbehave in various ways if the subject string
1440       contained bytes with the 0x80 bit set and the 0x40 bit unset in a lookbehind
1441       area. (PCRE was not checking for the UTF-8 mode flag, and trying to move
1442       back over UTF-8 characters.)
1443    
1444    
1445    Version 4.2 14-Apr-03
1446    ---------------------
1447    
1448    1. Typo "#if SUPPORT_UTF8" instead of "#ifdef SUPPORT_UTF8" fixed.
1449    
1450    2. Changes to the building process, supplied by Ronald Landheer-Cieslak
1451         [ON_WINDOWS]: new variable, "#" on non-Windows platforms
1452         [NOT_ON_WINDOWS]: new variable, "#" on Windows platforms
1453         [WIN_PREFIX]: new variable, "cyg" for Cygwin
1454         * Makefile.in: use autoconf substitution for OBJEXT, EXEEXT, BUILD_OBJEXT
1455           and BUILD_EXEEXT
1456         Note: automatic setting of the BUILD variables is not yet working
1457         set CPPFLAGS and BUILD_CPPFLAGS (but don't use yet) - should be used at
1458           compile-time but not at link-time
1459         [LINK]: use for linking executables only
1460         make different versions for Windows and non-Windows
1461         [LINKLIB]: new variable, copy of UNIX-style LINK, used for linking
1462           libraries
1463         [LINK_FOR_BUILD]: new variable
1464         [OBJEXT]: use throughout
1465         [EXEEXT]: use throughout
1466         <winshared>: new target
1467         <wininstall>: new target
1468         <dftables.o>: use native compiler
1469         <dftables>: use native linker
1470         <install>: handle Windows platform correctly
1471         <clean>: ditto
1472         <check>: ditto
1473         copy DLL to top builddir before testing
1474    
1475       As part of these changes, -no-undefined was removed again. This was reported
1476       to give trouble on HP-UX 11.0, so getting rid of it seems like a good idea
1477       in any case.
1478    
1479    3. Some tidies to get rid of compiler warnings:
1480    
1481       . In the match_data structure, match_limit was an unsigned long int, whereas
1482         match_call_count was an int. I've made them both unsigned long ints.
1483    
1484       . In pcretest the fact that a const uschar * doesn't automatically cast to
1485         a void * provoked a warning.
1486    
1487       . Turning on some more compiler warnings threw up some "shadow" variables
1488         and a few more missing casts.
1489    
1490    4. If PCRE was complied with UTF-8 support, but called without the PCRE_UTF8
1491       option, a class that contained a single character with a value between 128
1492       and 255 (e.g. /[\xFF]/) caused PCRE to crash.
1493    
1494    5. If PCRE was compiled with UTF-8 support, but called without the PCRE_UTF8
1495       option, a class that contained several characters, but with at least one
1496       whose value was between 128 and 255 caused PCRE to crash.
1497    
1498    
1499    Version 4.1 12-Mar-03
1500    ---------------------
1501    
1502    1. Compiling with gcc -pedantic found a couple of places where casts were
1503    needed, and a string in dftables.c that was longer than standard compilers are
1504    required to support.
1505    
1506    2. Compiling with Sun's compiler found a few more places where the code could
1507    be tidied up in order to avoid warnings.
1508    
1509    3. The variables for cross-compiling were called HOST_CC and HOST_CFLAGS; the
1510    first of these names is deprecated in the latest Autoconf in favour of the name
1511    CC_FOR_BUILD, because "host" is typically used to mean the system on which the
1512    compiled code will be run. I can't find a reference for HOST_CFLAGS, but by
1513    analogy I have changed it to CFLAGS_FOR_BUILD.
1514    
1515    4. Added -no-undefined to the linking command in the Makefile, because this is
1516    apparently helpful for Windows. To make it work, also added "-L. -lpcre" to the
1517    linking step for the pcreposix library.
1518    
1519    5. PCRE was failing to diagnose the case of two named groups with the same
1520    name.
1521    
1522    6. A problem with one of PCRE's optimizations was discovered. PCRE remembers a
1523    literal character that is needed in the subject for a match, and scans along to
1524    ensure that it is present before embarking on the full matching process. This
1525    saves time in cases of nested unlimited repeats that are never going to match.
1526    Problem: the scan can take a lot of time if the subject is very long (e.g.
1527    megabytes), thus penalizing straightforward matches. It is now done only if the
1528    amount of subject to be scanned is less than 1000 bytes.
1529    
1530    7. A lesser problem with the same optimization is that it was recording the
1531    first character of an anchored pattern as "needed", thus provoking a search
1532    right along the subject, even when the first match of the pattern was going to
1533    fail. The "needed" character is now not set for anchored patterns, unless it
1534    follows something in the pattern that is of non-fixed length. Thus, it still
1535    fulfils its original purpose of finding quick non-matches in cases of nested
1536    unlimited repeats, but isn't used for simple anchored patterns such as /^abc/.
1537    
1538    
1539    Version 4.0 17-Feb-03
1540    ---------------------
1541    
1542    1. If a comment in an extended regex that started immediately after a meta-item
1543    extended to the end of string, PCRE compiled incorrect data. This could lead to
1544    all kinds of weird effects. Example: /#/ was bad; /()#/ was bad; /a#/ was not.
1545    
1546    2. Moved to autoconf 2.53 and libtool 1.4.2.
1547    
1548    3. Perl 5.8 no longer needs "use utf8" for doing UTF-8 things. Consequently,
1549    the special perltest8 script is no longer needed - all the tests can be run
1550    from a single perltest script.
1551    
1552    4. From 5.004, Perl has not included the VT character (0x0b) in the set defined
1553    by \s. It has now been removed in PCRE. This means it isn't recognized as
1554    whitespace in /x regexes too, which is the same as Perl. Note that the POSIX
1555    class [:space:] *does* include VT, thereby creating a mess.
1556    
1557    5. Added the class [:blank:] (a GNU extension from Perl 5.8) to match only
1558    space and tab.
1559    
1560    6. Perl 5.005 was a long time ago. It's time to amalgamate the tests that use
1561    its new features into the main test script, reducing the number of scripts.
1562    
1563    7. Perl 5.8 has changed the meaning of patterns like /a(?i)b/. Earlier versions
1564    were backward compatible, and made the (?i) apply to the whole pattern, as if
1565    /i were given. Now it behaves more logically, and applies the option setting
1566    only to what follows. PCRE has been changed to follow suit. However, if it
1567    finds options settings right at the start of the pattern, it extracts them into
1568    the global options, as before. Thus, they show up in the info data.
1569    
1570    8. Added support for the \Q...\E escape sequence. Characters in between are
1571    treated as literals. This is slightly different from Perl in that $ and @ are
1572    also handled as literals inside the quotes. In Perl, they will cause variable
1573    interpolation. Note the following examples:
1574    
1575        Pattern            PCRE matches      Perl matches
1576    
1577        \Qabc$xyz\E        abc$xyz           abc followed by the contents of $xyz
1578        \Qabc\$xyz\E       abc\$xyz          abc\$xyz
1579        \Qabc\E\$\Qxyz\E   abc$xyz           abc$xyz
1580    
1581    For compatibility with Perl, \Q...\E sequences are recognized inside character
1582    classes as well as outside them.
1583    
1584    9. Re-organized 3 code statements in pcretest to avoid "overflow in
1585    floating-point constant arithmetic" warnings from a Microsoft compiler. Added a
1586    (size_t) cast to one statement in pcretest and one in pcreposix to avoid
1587    signed/unsigned warnings.
1588    
1589    10. SunOS4 doesn't have strtoul(). This was used only for unpicking the -o
1590    option for pcretest, so I've replaced it by a simple function that does just
1591    that job.
1592    
1593    11. pcregrep was ending with code 0 instead of 2 for the commands "pcregrep" or
1594    "pcregrep -".
1595    
1596    12. Added "possessive quantifiers" ?+, *+, ++, and {,}+ which come from Sun's
1597    Java package. This provides some syntactic sugar for simple cases of what my
1598    documentation calls "once-only subpatterns". A pattern such as x*+ is the same
1599    as (?>x*). In other words, if what is inside (?>...) is just a single repeated
1600    item, you can use this simplified notation. Note that only makes sense with
1601    greedy quantifiers. Consequently, the use of the possessive quantifier forces
1602    greediness, whatever the setting of the PCRE_UNGREEDY option.
1603    
1604    13. A change of greediness default within a pattern was not taking effect at
1605    the current level for patterns like /(b+(?U)a+)/. It did apply to parenthesized
1606    subpatterns that followed. Patterns like /b+(?U)a+/ worked because the option
1607    was abstracted outside.
1608    
1609    14. PCRE now supports the \G assertion. It is true when the current matching
1610    position is at the start point of the match. This differs from \A when the
1611    starting offset is non-zero. Used with the /g option of pcretest (or similar
1612    code), it works in the same way as it does for Perl's /g option. If all
1613    alternatives of a regex begin with \G, the expression is anchored to the start
1614    match position, and the "anchored" flag is set in the compiled expression.
1615    
1616    15. Some bugs concerning the handling of certain option changes within patterns
1617    have been fixed. These applied to options other than (?ims). For example,
1618    "a(?x: b c )d" did not match "XabcdY" but did match "Xa b c dY". It should have
1619    been the other way round. Some of this was related to change 7 above.
1620    
1621    16. PCRE now gives errors for /[.x.]/ and /[=x=]/ as unsupported POSIX
1622    features, as Perl does. Previously, PCRE gave the warnings only for /[[.x.]]/
1623    and /[[=x=]]/. PCRE now also gives an error for /[:name:]/ because it supports
1624    POSIX classes only within a class (e.g. /[[:alpha:]]/).
1625    
1626    17. Added support for Perl's \C escape. This matches one byte, even in UTF8
1627    mode. Unlike ".", it always matches newline, whatever the setting of
1628    PCRE_DOTALL. However, PCRE does not permit \C to appear in lookbehind
1629    assertions. Perl allows it, but it doesn't (in general) work because it can't
1630    calculate the length of the lookbehind. At least, that's the case for Perl
1631    5.8.0 - I've been told they are going to document that it doesn't work in
1632    future.
1633    
1634    18. Added an error diagnosis for escapes that PCRE does not support: these are
1635    \L, \l, \N, \P, \p, \U, \u, and \X.
1636    
1637    19. Although correctly diagnosing a missing ']' in a character class, PCRE was
1638    reading past the end of the pattern in cases such as /[abcd/.
1639    
1640    20. PCRE was getting more memory than necessary for patterns with classes that
1641    contained both POSIX named classes and other characters, e.g. /[[:space:]abc/.
1642    
1643    21. Added some code, conditional on #ifdef VPCOMPAT, to make life easier for
1644    compiling PCRE for use with Virtual Pascal.
1645    
1646    22. Small fix to the Makefile to make it work properly if the build is done
1647    outside the source tree.
1648    
1649    23. Added a new extension: a condition to go with recursion. If a conditional
1650    subpattern starts with (?(R) the "true" branch is used if recursion has
1651    happened, whereas the "false" branch is used only at the top level.
1652    
1653    24. When there was a very long string of literal characters (over 255 bytes
1654    without UTF support, over 250 bytes with UTF support), the computation of how
1655    much memory was required could be incorrect, leading to segfaults or other
1656    strange effects.
1657    
1658    25. PCRE was incorrectly assuming anchoring (either to start of subject or to
1659    start of line for a non-DOTALL pattern) when a pattern started with (.*) and
1660    there was a subsequent back reference to those brackets. This meant that, for
1661    example, /(.*)\d+\1/ failed to match "abc123bc". Unfortunately, it isn't
1662    possible to check for precisely this case. All we can do is abandon the
1663    optimization if .* occurs inside capturing brackets when there are any back
1664    references whatsoever. (See below for a better fix that came later.)
1665    
1666    26. The handling of the optimization for finding the first character of a
1667    non-anchored pattern, and for finding a character that is required later in the
1668    match were failing in some cases. This didn't break the matching; it just
1669    failed to optimize when it could. The way this is done has been re-implemented.
1670    
1671    27. Fixed typo in error message for invalid (?R item (it said "(?p").
1672    
1673    28. Added a new feature that provides some of the functionality that Perl
1674    provides with (?{...}). The facility is termed a "callout". The way it is done
1675    in PCRE is for the caller to provide an optional function, by setting
1676    pcre_callout to its entry point. Like pcre_malloc and pcre_free, this is a
1677    global variable. By default it is unset, which disables all calling out. To get
1678    the function called, the regex must include (?C) at appropriate points. This
1679    is, in fact, equivalent to (?C0), and any number <= 255 may be given with (?C).
1680    This provides a means of identifying different callout points. When PCRE
1681    reaches such a point in the regex, if pcre_callout has been set, the external
1682    function is called. It is provided with data in a structure called
1683    pcre_callout_block, which is defined in pcre.h. If the function returns 0,
1684    matching continues; if it returns a non-zero value, the match at the current
1685    point fails. However, backtracking will occur if possible. [This was changed
1686    later and other features added - see item 49 below.]
1687    
1688    29. pcretest is upgraded to test the callout functionality. It provides a
1689    callout function that displays information. By default, it shows the start of
1690    the match and the current position in the text. There are some new data escapes
1691    to vary what happens:
1692    
1693        \C+         in addition, show current contents of captured substrings
1694        \C-         do not supply a callout function
1695        \C!n        return 1 when callout number n is reached
1696        \C!n!m      return 1 when callout number n is reached for the mth time
1697    
1698    30. If pcregrep was called with the -l option and just a single file name, it
1699    output "<stdin>" if a match was found, instead of the file name.
1700    
1701    31. Improve the efficiency of the POSIX API to PCRE. If the number of capturing
1702    slots is less than POSIX_MALLOC_THRESHOLD, use a block on the stack to pass to
1703    pcre_exec(). This saves a malloc/free per call. The default value of
1704    POSIX_MALLOC_THRESHOLD is 10; it can be changed by --with-posix-malloc-threshold
1705    when configuring.
1706    
1707    32. The default maximum size of a compiled pattern is 64K. There have been a
1708    few cases of people hitting this limit. The code now uses macros to handle the
1709    storing of links as offsets within the compiled pattern. It defaults to 2-byte
1710    links, but this can be changed to 3 or 4 bytes by --with-link-size when
1711    configuring. Tests 2 and 5 work only with 2-byte links because they output
1712    debugging information about compiled patterns.
1713    
1714    33. Internal code re-arrangements:
1715    
1716    (a) Moved the debugging function for printing out a compiled regex into
1717        its own source file (printint.c) and used #include to pull it into
1718        pcretest.c and, when DEBUG is defined, into pcre.c, instead of having two
1719        separate copies.
1720    
1721    (b) Defined the list of op-code names for debugging as a macro in
1722        internal.h so that it is next to the definition of the opcodes.
1723    
1724    (c) Defined a table of op-code lengths for simpler skipping along compiled
1725        code. This is again a macro in internal.h so that it is next to the
1726        definition of the opcodes.
1727    
1728    34. Added support for recursive calls to individual subpatterns, along the
1729    lines of Robin Houston's patch (but implemented somewhat differently).
1730    
1731    35. Further mods to the Makefile to help Win32. Also, added code to pcregrep to
1732    allow it to read and process whole directories in Win32. This code was
1733    contributed by Lionel Fourquaux; it has not been tested by me.
1734    
1735    36. Added support for named subpatterns. The Python syntax (?P<name>...) is
1736    used to name a group. Names consist of alphanumerics and underscores, and must
1737    be unique. Back references use the syntax (?P=name) and recursive calls use
1738    (?P>name) which is a PCRE extension to the Python extension. Groups still have
1739    numbers. The function pcre_fullinfo() can be used after compilation to extract
1740    a name/number map. There are three relevant calls:
1741    
1742      PCRE_INFO_NAMEENTRYSIZE        yields the size of each entry in the map
1743      PCRE_INFO_NAMECOUNT            yields the number of entries
1744      PCRE_INFO_NAMETABLE            yields a pointer to the map.
1745    
1746    The map is a vector of fixed-size entries. The size of each entry depends on
1747    the length of the longest name used. The first two bytes of each entry are the
1748    group number, most significant byte first. There follows the corresponding
1749    name, zero terminated. The names are in alphabetical order.
1750    
1751    37. Make the maximum literal string in the compiled code 250 for the non-UTF-8
1752    case instead of 255. Making it the same both with and without UTF-8 support
1753    means that the same test output works with both.
1754    
1755    38. There was a case of malloc(0) in the POSIX testing code in pcretest. Avoid
1756    calling malloc() with a zero argument.
1757    
1758    39. Change 25 above had to resort to a heavy-handed test for the .* anchoring
1759    optimization. I've improved things by keeping a bitmap of backreferences with
1760    numbers 1-31 so that if .* occurs inside capturing brackets that are not in
1761    fact referenced, the optimization can be applied. It is unlikely that a
1762    relevant occurrence of .* (i.e. one which might indicate anchoring or forcing
1763    the match to follow \n) will appear inside brackets with a number greater than
1764    31, but if it does, any back reference > 31 suppresses the optimization.
1765    
1766    40. Added a new compile-time option PCRE_NO_AUTO_CAPTURE. This has the effect
1767    of disabling numbered capturing parentheses. Any opening parenthesis that is
1768    not followed by ? behaves as if it were followed by ?: but named parentheses
1769    can still be used for capturing (and they will acquire numbers in the usual
1770    way).
1771    
1772    41. Redesigned the return codes from the match() function into yes/no/error so
1773    that errors can be passed back from deep inside the nested calls. A malloc
1774    failure while inside a recursive subpattern call now causes the
1775    PCRE_ERROR_NOMEMORY return instead of quietly going wrong.
1776    
1777    42. It is now possible to set a limit on the number of times the match()
1778    function is called in a call to pcre_exec(). This facility makes it possible to
1779    limit the amount of recursion and backtracking, though not in a directly
1780    obvious way, because the match() function is used in a number of different
1781    circumstances. The count starts from zero for each position in the subject
1782    string (for non-anchored patterns). The default limit is, for compatibility, a
1783    large number, namely 10 000 000. You can change this in two ways:
1784    
1785    (a) When configuring PCRE before making, you can use --with-match-limit=n
1786        to set a default value for the compiled library.
1787    
1788    (b) For each call to pcre_exec(), you can pass a pcre_extra block in which
1789        a different value is set. See 45 below.
1790    
1791    If the limit is exceeded, pcre_exec() returns PCRE_ERROR_MATCHLIMIT.
1792    
1793    43. Added a new function pcre_config(int, void *) to enable run-time extraction
1794    of things that can be changed at compile time. The first argument specifies
1795    what is wanted and the second points to where the information is to be placed.
1796    The current list of available information is:
1797    
1798      PCRE_CONFIG_UTF8
1799    
1800    The output is an integer that is set to one if UTF-8 support is available;
1801    otherwise it is set to zero.
1802    
1803      PCRE_CONFIG_NEWLINE
1804    
1805    The output is an integer that it set to the value of the code that is used for
1806    newline. It is either LF (10) or CR (13).
1807    
1808      PCRE_CONFIG_LINK_SIZE
1809    
1810    The output is an integer that contains the number of bytes used for internal
1811    linkage in compiled expressions. The value is 2, 3, or 4. See item 32 above.
1812    
1813      PCRE_CONFIG_POSIX_MALLOC_THRESHOLD
1814    
1815    The output is an integer that contains the threshold above which the POSIX
1816    interface uses malloc() for output vectors. See item 31 above.
1817    
1818      PCRE_CONFIG_MATCH_LIMIT
1819    
1820    The output is an unsigned integer that contains the default limit of the number
1821    of match() calls in a pcre_exec() execution. See 42 above.
1822    
1823    44. pcretest has been upgraded by the addition of the -C option. This causes it
1824    to extract all the available output from the new pcre_config() function, and to
1825    output it. The program then exits immediately.
1826    
1827    45. A need has arisen to pass over additional data with calls to pcre_exec() in
1828    order to support additional features. One way would have been to define
1829    pcre_exec2() (for example) with extra arguments, but this would not have been
1830    extensible, and would also have required all calls to the original function to
1831    be mapped to the new one. Instead, I have chosen to extend the mechanism that
1832    is used for passing in "extra" data from pcre_study().
1833    
1834    The pcre_extra structure is now exposed and defined in pcre.h. It currently
1835    contains the following fields:
1836    
1837      flags         a bitmap indicating which of the following fields are set
1838      study_data    opaque data from pcre_study()
1839      match_limit   a way of specifying a limit on match() calls for a specific
1840                      call to pcre_exec()
1841      callout_data  data for callouts (see 49 below)
1842    
1843    The flag bits are also defined in pcre.h, and are
1844    
1845      PCRE_EXTRA_STUDY_DATA
1846      PCRE_EXTRA_MATCH_LIMIT
1847      PCRE_EXTRA_CALLOUT_DATA
1848    
1849    The pcre_study() function now returns one of these new pcre_extra blocks, with
1850    the actual study data pointed to by the study_data field, and the
1851    PCRE_EXTRA_STUDY_DATA flag set. This can be passed directly to pcre_exec() as
1852    before. That is, this change is entirely upwards-compatible and requires no
1853    change to existing code.
1854    
1855    If you want to pass in additional data to pcre_exec(), you can either place it
1856    in a pcre_extra block provided by pcre_study(), or create your own pcre_extra
1857    block.
1858    
1859    46. pcretest has been extended to test the PCRE_EXTRA_MATCH_LIMIT feature. If a
1860    data string contains the escape sequence \M, pcretest calls pcre_exec() several
1861    times with different match limits, until it finds the minimum value needed for
1862    pcre_exec() to complete. The value is then output. This can be instructive; for
1863    most simple matches the number is quite small, but for pathological cases it
1864    gets very large very quickly.
1865    
1866    47. There's a new option for pcre_fullinfo() called PCRE_INFO_STUDYSIZE. It
1867    returns the size of the data block pointed to by the study_data field in a
1868    pcre_extra block, that is, the value that was passed as the argument to
1869    pcre_malloc() when PCRE was getting memory in which to place the information
1870    created by pcre_study(). The fourth argument should point to a size_t variable.
1871    pcretest has been extended so that this information is shown after a successful
1872    pcre_study() call when information about the compiled regex is being displayed.
1873    
1874    48. Cosmetic change to Makefile: there's no need to have / after $(DESTDIR)
1875    because what follows is always an absolute path. (Later: it turns out that this
1876    is more than cosmetic for MinGW, because it doesn't like empty path
1877    components.)
1878    
1879    49. Some changes have been made to the callout feature (see 28 above):
1880    
1881    (i)  A callout function now has three choices for what it returns:
1882    
1883           0  =>  success, carry on matching
1884         > 0  =>  failure at this point, but backtrack if possible
1885         < 0  =>  serious error, return this value from pcre_exec()
1886    
1887         Negative values should normally be chosen from the set of PCRE_ERROR_xxx
1888         values. In particular, returning PCRE_ERROR_NOMATCH forces a standard
1889         "match failed" error. The error number PCRE_ERROR_CALLOUT is reserved for
1890         use by callout functions. It will never be used by PCRE itself.
1891    
1892    (ii) The pcre_extra structure (see 45 above) has a void * field called
1893         callout_data, with corresponding flag bit PCRE_EXTRA_CALLOUT_DATA. The
1894         pcre_callout_block structure has a field of the same name. The contents of
1895         the field passed in the pcre_extra structure are passed to the callout
1896         function in the corresponding field in the callout block. This makes it
1897         easier to use the same callout-containing regex from multiple threads. For
1898         testing, the pcretest program has a new data escape
1899    
1900           \C*n        pass the number n (may be negative) as callout_data
1901    
1902         If the callout function in pcretest receives a non-zero value as
1903         callout_data, it returns that value.
1904    
1905    50. Makefile wasn't handling CFLAGS properly when compiling dftables. Also,
1906    there were some redundant $(CFLAGS) in commands that are now specified as
1907    $(LINK), which already includes $(CFLAGS).
1908    
1909    51. Extensions to UTF-8 support are listed below. These all apply when (a) PCRE
1910    has been compiled with UTF-8 support *and* pcre_compile() has been compiled
1911    with the PCRE_UTF8 flag. Patterns that are compiled without that flag assume
1912    one-byte characters throughout. Note that case-insensitive matching applies
1913    only to characters whose values are less than 256. PCRE doesn't support the
1914    notion of cases for higher-valued characters.
1915    
1916    (i)   A character class whose characters are all within 0-255 is handled as
1917          a bit map, and the map is inverted for negative classes. Previously, a
1918          character > 255 always failed to match such a class; however it should
1919          match if the class was a negative one (e.g. [^ab]). This has been fixed.
1920    
1921    (ii)  A negated character class with a single character < 255 is coded as
1922          "not this character" (OP_NOT). This wasn't working properly when the test
1923          character was multibyte, either singly or repeated.
1924    
1925    (iii) Repeats of multibyte characters are now handled correctly in UTF-8
1926          mode, for example: \x{100}{2,3}.
1927    
1928    (iv)  The character escapes \b, \B, \d, \D, \s, \S, \w, and \W (either
1929          singly or repeated) now correctly test multibyte characters. However,
1930          PCRE doesn't recognize any characters with values greater than 255 as
1931          digits, spaces, or word characters. Such characters always match \D, \S,
1932          and \W, and never match \d, \s, or \w.
1933    
1934    (v)   Classes may now contain characters and character ranges with values
1935          greater than 255. For example: [ab\x{100}-\x{400}].
1936    
1937    (vi)  pcregrep now has a --utf-8 option (synonym -u) which makes it call
1938          PCRE in UTF-8 mode.
1939    
1940    52. The info request value PCRE_INFO_FIRSTCHAR has been renamed
1941    PCRE_INFO_FIRSTBYTE because it is a byte value. However, the old name is
1942    retained for backwards compatibility. (Note that LASTLITERAL is also a byte
1943    value.)
1944    
1945    53. The single man page has become too large. I have therefore split it up into
1946    a number of separate man pages. These also give rise to individual HTML pages;
1947    these are now put in a separate directory, and there is an index.html page that
1948    lists them all. Some hyperlinking between the pages has been installed.
1949    
1950    54. Added convenience functions for handling named capturing parentheses.
1951    
1952    55. Unknown escapes inside character classes (e.g. [\M]) and escapes that
1953    aren't interpreted therein (e.g. [\C]) are literals in Perl. This is now also
1954    true in PCRE, except when the PCRE_EXTENDED option is set, in which case they
1955    are faulted.
1956    
1957    56. Introduced HOST_CC and HOST_CFLAGS which can be set in the environment when
1958    calling configure. These values are used when compiling the dftables.c program
1959    which is run to generate the source of the default character tables. They
1960    default to the values of CC and CFLAGS. If you are cross-compiling PCRE,
1961    you will need to set these values.
1962    
1963    57. Updated the building process for Windows DLL, as provided by Fred Cox.
1964    
1965    
1966    Version 3.9 02-Jan-02
1967    ---------------------
1968    
1969    1. A bit of extraneous text had somehow crept into the pcregrep documentation.
1970    
1971    2. If --disable-static was given, the building process failed when trying to
1972    build pcretest and pcregrep. (For some reason it was using libtool to compile
1973    them, which is not right, as they aren't part of the library.)
1974    
1975    
1976    Version 3.8 18-Dec-01
1977    ---------------------
1978    
1979    1. The experimental UTF-8 code was completely screwed up. It was packing the
1980    bytes in the wrong order. How dumb can you get?
1981    
1982    
1983    Version 3.7 29-Oct-01
1984    ---------------------
1985    
1986    1. In updating pcretest to check change 1 of version 3.6, I screwed up.
1987    This caused pcretest, when used on the test data, to segfault. Unfortunately,
1988    this didn't happen under Solaris 8, where I normally test things.
1989    
1990    2. The Makefile had to be changed to make it work on BSD systems, where 'make'
1991    doesn't seem to recognize that ./xxx and xxx are the same file. (This entry
1992    isn't in ChangeLog distributed with 3.7 because I forgot when I hastily made
1993    this fix an hour or so after the initial 3.7 release.)
1994    
1995    
1996    Version 3.6 23-Oct-01
1997    ---------------------
1998    
1999    1. Crashed with /(sens|respons)e and \1ibility/ and "sense and sensibility" if
2000    offsets passed as NULL with zero offset count.
2001    
2002    2. The config.guess and config.sub files had not been updated when I moved to
2003    the latest autoconf.
2004    
2005    
2006    Version 3.5 15-Aug-01
2007    ---------------------
2008    
2009    1. Added some missing #if !defined NOPOSIX conditionals in pcretest.c that
2010    had been forgotten.
2011    
2012    2. By using declared but undefined structures, we can avoid using "void"
2013    definitions in pcre.h while keeping the internal definitions of the structures
2014    private.
2015    
2016    3. The distribution is now built using autoconf 2.50 and libtool 1.4. From a
2017    user point of view, this means that both static and shared libraries are built
2018    by default, but this can be individually controlled. More of the work of
2019    handling this static/shared cases is now inside libtool instead of PCRE's make
2020    file.
2021    
2022    4. The pcretest utility is now installed along with pcregrep because it is
2023    useful for users (to test regexs) and by doing this, it automatically gets
2024    relinked by libtool. The documentation has been turned into a man page, so
2025    there are now .1, .txt, and .html versions in /doc.
2026    
2027    5. Upgrades to pcregrep:
2028       (i)   Added long-form option names like gnu grep.
2029       (ii)  Added --help to list all options with an explanatory phrase.
2030       (iii) Added -r, --recursive to recurse into sub-directories.
2031       (iv)  Added -f, --file to read patterns from a file.
2032    
2033    6. pcre_exec() was referring to its "code" argument before testing that
2034    argument for NULL (and giving an error if it was NULL).
2035    
2036    7. Upgraded Makefile.in to allow for compiling in a different directory from
2037    the source directory.
2038    
2039    8. Tiny buglet in pcretest: when pcre_fullinfo() was called to retrieve the
2040    options bits, the pointer it was passed was to an int instead of to an unsigned
2041    long int. This mattered only on 64-bit systems.
2042    
2043    9. Fixed typo (3.4/1) in pcre.h again. Sigh. I had changed pcre.h (which is
2044    generated) instead of pcre.in, which it its source. Also made the same change
2045    in several of the .c files.
2046    
2047    10. A new release of gcc defines printf() as a macro, which broke pcretest
2048    because it had an ifdef in the middle of a string argument for printf(). Fixed
2049    by using separate calls to printf().
2050    
2051    11. Added --enable-newline-is-cr and --enable-newline-is-lf to the configure
2052    script, to force use of CR or LF instead of \n in the source. On non-Unix
2053    systems, the value can be set in config.h.
2054    
2055    12. The limit of 200 on non-capturing parentheses is a _nesting_ limit, not an
2056    absolute limit. Changed the text of the error message to make this clear, and
2057    likewise updated the man page.
2058    
2059    13. The limit of 99 on the number of capturing subpatterns has been removed.
2060    The new limit is 65535, which I hope will not be a "real" limit.
2061    
2062    
2063    Version 3.4 22-Aug-00
2064    ---------------------
2065    
2066    1. Fixed typo in pcre.h: unsigned const char * changed to const unsigned char *.
2067    
2068    2. Diagnose condition (?(0) as an error instead of crashing on matching.
2069    
2070    
2071    Version 3.3 01-Aug-00
2072    ---------------------
2073    
2074    1. If an octal character was given, but the value was greater than \377, it
2075    was not getting masked to the least significant bits, as documented. This could
2076    lead to crashes in some systems.
2077    
2078    2. Perl 5.6 (if not earlier versions) accepts classes like [a-\d] and treats
2079    the hyphen as a literal. PCRE used to give an error; it now behaves like Perl.
2080    
2081    3. Added the functions pcre_free_substring() and pcre_free_substring_list().
2082    These just pass their arguments on to (pcre_free)(), but they are provided
2083    because some uses of PCRE bind it to non-C systems that can call its functions,
2084    but cannot call free() or pcre_free() directly.
2085    
2086    4. Add "make test" as a synonym for "make check". Corrected some comments in
2087    the Makefile.
2088    
2089    5. Add $(DESTDIR)/ in front of all the paths in the "install" target in the
2090    Makefile.
2091    
2092    6. Changed the name of pgrep to pcregrep, because Solaris has introduced a
2093    command called pgrep for grepping around the active processes.
2094    
2095    7. Added the beginnings of support for UTF-8 character strings.
2096    
2097    8. Arranged for the Makefile to pass over the settings of CC, CFLAGS, and
2098    RANLIB to ./ltconfig so that they are used by libtool. I think these are all
2099    the relevant ones. (AR is not passed because ./ltconfig does its own figuring
2100    out for the ar command.)
2101    
2102    
2103    Version 3.2 12-May-00
2104    ---------------------
2105    
2106    This is purely a bug fixing release.
2107    
2108    1. If the pattern /((Z)+|A)*/ was matched agained ZABCDEFG it matched Z instead
2109    of ZA. This was just one example of several cases that could provoke this bug,
2110    which was introduced by change 9 of version 2.00. The code for breaking
2111    infinite loops after an iteration that matches an empty string was't working
2112    correctly.
2113    
2114    2. The pcretest program was not imitating Perl correctly for the pattern /a*/g
2115    when matched against abbab (for example). After matching an empty string, it
2116    wasn't forcing anchoring when setting PCRE_NOTEMPTY for the next attempt; this
2117    caused it to match further down the string than it should.
2118    
2119    3. The code contained an inclusion of sys/types.h. It isn't clear why this
2120    was there because it doesn't seem to be needed, and it causes trouble on some
2121    systems, as it is not a Standard C header. It has been removed.
2122    
2123    4. Made 4 silly changes to the source to avoid stupid compiler warnings that
2124    were reported on the Macintosh. The changes were from
2125    
2126      while ((c = *(++ptr)) != 0 && c != '\n');
2127    to
2128      while ((c = *(++ptr)) != 0 && c != '\n') ;
2129    
2130    Totally extraordinary, but if that's what it takes...
2131    
2132    5. PCRE is being used in one environment where neither memmove() nor bcopy() is
2133    available. Added HAVE_BCOPY and an autoconf test for it; if neither
2134    HAVE_MEMMOVE nor HAVE_BCOPY is set, use a built-in emulation function which
2135    assumes the way PCRE uses memmove() (always moving upwards).
2136    
2137    6. PCRE is being used in one environment where strchr() is not available. There
2138    was only one use in pcre.c, and writing it out to avoid strchr() probably gives
2139    faster code anyway.
2140    
2141    
2142    Version 3.1 09-Feb-00
2143    ---------------------
2144    
2145    The only change in this release is the fixing of some bugs in Makefile.in for
2146    the "install" target:
2147    
2148    (1) It was failing to install pcreposix.h.
2149    
2150    (2) It was overwriting the pcre.3 man page with the pcreposix.3 man page.
2151    
2152    
2153    Version 3.0 01-Feb-00
2154    ---------------------
2155    
2156    1. Add support for the /+ modifier to perltest (to output $` like it does in
2157    pcretest).
2158    
2159    2. Add support for the /g modifier to perltest.
2160    
2161    3. Fix pcretest so that it behaves even more like Perl for /g when the pattern
2162    matches null strings.
2163    
2164    4. Fix perltest so that it doesn't do unwanted things when fed an empty
2165    pattern. Perl treats empty patterns specially - it reuses the most recent
2166    pattern, which is not what we want. Replace // by /(?#)/ in order to avoid this
2167    effect.
2168    
2169    5. The POSIX interface was broken in that it was just handing over the POSIX
2170    captured string vector to pcre_exec(), but (since release 2.00) PCRE has
2171    required a bigger vector, with some working space on the end. This means that
2172    the POSIX wrapper now has to get and free some memory, and copy the results.
2173    
2174    6. Added some simple autoconf support, placing the test data and the
2175    documentation in separate directories, re-organizing some of the
2176    information files, and making it build pcre-config (a GNU standard). Also added
2177    libtool support for building PCRE as a shared library, which is now the
2178    default.
2179    
2180    7. Got rid of the leading zero in the definition of PCRE_MINOR because 08 and
2181    09 are not valid octal constants. Single digits will be used for minor values
2182    less than 10.
2183    
2184    8. Defined REG_EXTENDED and REG_NOSUB as zero in the POSIX header, so that
2185    existing programs that set these in the POSIX interface can use PCRE without
2186    modification.
2187    
2188    9. Added a new function, pcre_fullinfo() with an extensible interface. It can
2189    return all that pcre_info() returns, plus additional data. The pcre_info()
2190    function is retained for compatibility, but is considered to be obsolete.
2191    
2192    10. Added experimental recursion feature (?R) to handle one common case that
2193    Perl 5.6 will be able to do with (?p{...}).
2194    
2195    11. Added support for POSIX character classes like [:alpha:], which Perl is
2196    adopting.
2197    
2198    
2199    Version 2.08 31-Aug-99
2200    ----------------------
2201    
2202    1. When startoffset was not zero and the pattern began with ".*", PCRE was not
2203    trying to match at the startoffset position, but instead was moving forward to
2204    the next newline as if a previous match had failed.
2205    
2206    2. pcretest was not making use of PCRE_NOTEMPTY when repeating for /g and /G,
2207    and could get into a loop if a null string was matched other than at the start
2208    of the subject.
2209    
2210    3. Added definitions of PCRE_MAJOR and PCRE_MINOR to pcre.h so the version can
2211    be distinguished at compile time, and for completeness also added PCRE_DATE.
2212    
2213    5. Added Paul Sokolovsky's minor changes to make it easy to compile a Win32 DLL
2214    in GnuWin32 environments.
2215    
2216    
2217    Version 2.07 29-Jul-99
2218    ----------------------
2219    
2220    1. The documentation is now supplied in plain text form and HTML as well as in
2221    the form of man page sources.
2222    
2223    2. C++ compilers don't like assigning (void *) values to other pointer types.
2224    In particular this affects malloc(). Although there is no problem in Standard
2225    C, I've put in casts to keep C++ compilers happy.
2226    
2227    3. Typo on pcretest.c; a cast of (unsigned char *) in the POSIX regexec() call
2228    should be (const char *).
2229    
2230    4. If NOPOSIX is defined, pcretest.c compiles without POSIX support. This may
2231    be useful for non-Unix systems who don't want to bother with the POSIX stuff.
2232    However, I haven't made this a standard facility. The documentation doesn't
2233    mention it, and the Makefile doesn't support it.
2234    
2235    5. The Makefile now contains an "install" target, with editable destinations at
2236    the top of the file. The pcretest program is not installed.
2237    
2238    6. pgrep -V now gives the PCRE version number and date.
2239    
2240    7. Fixed bug: a zero repetition after a literal string (e.g. /abcde{0}/) was
2241    causing the entire string to be ignored, instead of just the last character.
2242    
2243    8. If a pattern like /"([^\\"]+|\\.)*"/ is applied in the normal way to a
2244    non-matching string, it can take a very, very long time, even for strings of
2245    quite modest length, because of the nested recursion. PCRE now does better in
2246    some of these cases. It does this by remembering the last required literal
2247    character in the pattern, and pre-searching the subject to ensure it is present
2248    before running the real match. In other words, it applies a heuristic to detect
2249    some types of certain failure quickly, and in the above example, if presented
2250    with a string that has no trailing " it gives "no match" very quickly.
2251    
2252    9. A new runtime option PCRE_NOTEMPTY causes null string matches to be ignored;
2253    other alternatives are tried instead.
2254    
2255    
2256    Version 2.06 09-Jun-99
2257    ----------------------
2258    
2259    1. Change pcretest's output for amount of store used to show just the code
2260    space, because the remainder (the data block) varies in size between 32-bit and
2261    64-bit systems.
2262    
2263    2. Added an extra argument to pcre_exec() to supply an offset in the subject to
2264    start matching at. This allows lookbehinds to work when searching for multiple
2265    occurrences in a string.
2266    
2267    3. Added additional options to pcretest for testing multiple occurrences:
2268    
2269       /+   outputs the rest of the string that follows a match
2270       /g   loops for multiple occurrences, using the new startoffset argument
2271       /G   loops for multiple occurrences by passing an incremented pointer
2272    
2273    4. PCRE wasn't doing the "first character" optimization for patterns starting
2274    with \b or \B, though it was doing it for other lookbehind assertions. That is,
2275    it wasn't noticing that a match for a pattern such as /\bxyz/ has to start with
2276    the letter 'x'. On long subject strings, this gives a significant speed-up.
2277    
2278    
2279    Version 2.05 21-Apr-99
2280    ----------------------
2281    
2282    1. Changed the type of magic_number from int to long int so that it works
2283    properly on 16-bit systems.
2284    
2285    2. Fixed a bug which caused patterns starting with .* not to work correctly
2286    when the subject string contained newline characters. PCRE was assuming
2287    anchoring for such patterns in all cases, which is not correct because .* will
2288    not pass a newline unless PCRE_DOTALL is set. It now assumes anchoring only if
2289    DOTALL is set at top level; otherwise it knows that patterns starting with .*
2290    must be retried after every newline in the subject.
2291    
2292    
2293    Version 2.04 18-Feb-99
2294    ----------------------
2295    
2296    1. For parenthesized subpatterns with repeats whose minimum was zero, the
2297    computation of the store needed to hold the pattern was incorrect (too large).
2298    If such patterns were nested a few deep, this could multiply and become a real
2299    problem.
2300    
2301    2. Added /M option to pcretest to show the memory requirement of a specific
2302    pattern. Made -m a synonym of -s (which does this globally) for compatibility.
2303    
2304    3. Subpatterns of the form (regex){n,m} (i.e. limited maximum) were being
2305    compiled in such a way that the backtracking after subsequent failure was
2306    pessimal. Something like (a){0,3} was compiled as (a)?(a)?(a)? instead of
2307    ((a)((a)(a)?)?)? with disastrous performance if the maximum was of any size.
2308    
2309    
2310    Version 2.03 02-Feb-99
2311    ----------------------
2312    
2313    1. Fixed typo and small mistake in man page.
2314    
2315    2. Added 4th condition (GPL supersedes if conflict) and created separate
2316    LICENCE file containing the conditions.
2317    
2318    3. Updated pcretest so that patterns such as /abc\/def/ work like they do in
2319    Perl, that is the internal \ allows the delimiter to be included in the
2320    pattern. Locked out the use of \ as a delimiter. If \ immediately follows
2321    the final delimiter, add \ to the end of the pattern (to test the error).
2322    
2323    4. Added the convenience functions for extracting substrings after a successful
2324    match. Updated pcretest to make it able to test these functions.
2325    
2326    
2327    Version 2.02 14-Jan-99
2328    ----------------------
2329    
2330    1. Initialized the working variables associated with each extraction so that
2331    their saving and restoring doesn't refer to uninitialized store.
2332    
2333    2. Put dummy code into study.c in order to trick the optimizer of the IBM C
2334    compiler for OS/2 into generating correct code. Apparently IBM isn't going to
2335    fix the problem.
2336    
2337    3. Pcretest: the timing code wasn't using LOOPREPEAT for timing execution
2338    calls, and wasn't printing the correct value for compiling calls. Increased the
2339    default value of LOOPREPEAT, and the number of significant figures in the
2340    times.
2341    
2342    4. Changed "/bin/rm" in the Makefile to "-rm" so it works on Windows NT.
2343    
2344    5. Renamed "deftables" as "dftables" to get it down to 8 characters, to avoid
2345    a building problem on Windows NT with a FAT file system.
2346    
2347    
2348    Version 2.01 21-Oct-98
2349    ----------------------
2350    
2351    1. Changed the API for pcre_compile() to allow for the provision of a pointer
2352    to character tables built by pcre_maketables() in the current locale. If NULL
2353    is passed, the default tables are used.
2354    
2355    
2356  Version 2.00 24-Sep-98  Version 2.00 24-Sep-98
2357  ----------------------  ----------------------

Legend:
Removed from v.23  
changed lines
  Added in v.158

webmaster@exim.org
ViewVC Help
Powered by ViewVC 1.1.12