ViewVC logotype

Contents of /code/trunk/ChangeLog

Parent Directory Parent Directory | Revision Log Revision Log

Revision 115 - (hide annotations) (download)
Fri Mar 9 12:23:37 2007 UTC (8 years, 1 month ago) by ph10
File size: 123620 byte(s)
Fix two obscure, but nasty bugs concerned with caseless matching with Unicode 
property support.

1 nigel 3 ChangeLog for PCRE
2     ------------------
4 ph10 98 Version 7.1 05-Mar-07
5     ---------------------
7 ph10 111 1. Applied Bob Rossi and Daniel G's patches to convert the build system to one
8     that is more "standard", making use of automake and other autotools. There
9 ph10 99 is some re-arrangement of the files and adjustment of comments consequent
10     on this.
11 ph10 111
12     2. Part of the patch fixed a problem with the pcregrep tests. The test of -r
13     for recursive directory scanning broke on some systems because the files
14     are not scanned in any specific order and on different systems the order
15     was different. A call to "sort" has been inserted into RunGrepTest for the
16     approprate test as a short-term fix. In the longer term there may be an
17 ph10 100 alternative.
18 ph10 111
19 ph10 100 3. I had an email from Eric Raymond about problems translating some of PCRE's
20 ph10 111 man pages to HTML (despite the fact that I distribute HTML pages, some
21     people do their own conversions for various reasons). The problems
22     concerned the use of low-level troff macros .br and .in. I have therefore
23     removed all such uses from the man pages (some were redundant, some could
24 ph10 113 be replaced by .nf/.fi pairs). The 132html script that I use to generate
25     HTML has been updated to handle .nf/.fi and to complain if it encounters
26     .br or .in.
27 ph10 111
28 ph10 100 4. Updated comments in configure.ac that get placed in config.h.in and also
29 ph10 111 arranged for config.h to be included in the distribution, with the name
30     config.h.generic, for the benefit of those who have to compile without
31     Autotools (compare pcre.h, which is now distributed as pcre.h.generic).
33     5. Updated the support (such as it is) for Virtual Pascal, thanks to Stefan
34     Weber: (1) pcre_internal.h was missing some function renames; (2) updated
35     makevp.bat for the current PCRE, using the additional files !compile.txt,
36 ph10 101 !linklib.txt, and pcregexp.pas.
37 ph10 111
38     6. A Windows user reported a minor discrepancy with test 2, which turned out
39     to be caused by a trailing space on an input line that had got lost in his
40 ph10 102 copy. The trailing space was an accident, so I've just removed it.
41 ph10 111
42 ph10 104 7. Add -Wl,-R... flags in pcre-config.in for *BSD* systems, as I'm told
43 ph10 111 that is needed.
45 ph10 105 8. Mark ucp_table (in ucptable.h) and ucp_gentype (in pcre_ucp_searchfuncs.c)
46     as "const" (a) because they are and (b) because it helps the PHP
47     maintainers who have recently made a script to detect big data structures
48 ph10 111 in the php code that should be moved to the .rodata section. I remembered
49     to update Builducptable as well, so it won't revert if ucptable.h is ever
50 ph10 105 re-created.
51 ph10 111
52     9. Added some extra #ifdef SUPPORT_UTF8 conditionals into pcretest.c,
53 ph10 107 pcre_printint.src, pcre_compile.c, pcre_study.c, and pcre_tables.c, in
54 ph10 111 order to be able to cut out the UTF-8 tables in the latter when UTF-8
55     support is not required. This saves 1.5-2K of code, which is important in
56 ph10 107 some applications.
57 ph10 111
58 ph10 110 Later: more #ifdefs are needed in pcre_ord2utf8.c and pcre_valid_utf8.c
59 ph10 111 so as not to refer to the tables, even though these functions will never be
60     called when UTF-8 support is disabled. Otherwise there are problems with a
61     shared library.
63     10. The emulated memmove() function in pcre_internal.h (provided for those
64     environments that have neither memmove() nor bcopy()) was defining its
65 ph10 108 arguments as char * instead of void *.
67 ph10 111 11. The script PrepareRelease is now distributed: it calls 132html, CleanTxt,
68     and Detrail to create the HTML documentation, the .txt form of the man
69     pages, and it removes trailing spaces from listed files. It also creates
70     pcre.h.generic and config.h.generic from pcre.h and config.h. In the latter
71     case, it wraps all the #defines with #ifndefs. This script should be run
72     before "make dist".
73 ph10 115
74     12. Fixed two fairly obscure bugs concerned with quantified caseless matching
75     with Unicode property support.
77     (a) For a maximizing quantifier, if the two different cases of the
78     character were of different lengths in their UTF-8 codings (there are
79     some cases like this - I found 11), and the matching function had to
80     back up over a mixture of the two cases, it incorrectly assumed they
81     were both the same length.
83     (b) When PCRE was configured to use the heap rather than the stack for
84     recursion during matching, it was not correctly preserving the data for
85     the other case of a UTF-8 character when checking ahead for a match
86     while processing a minimizing repeat. If the check also involved
87     matching a wide character, but failed, corruption could cause an
88     erroneous result when trying to check for a repeat of the original
89     character.
90 ph10 98
91 ph10 111
92 nigel 93 Version 7.0 19-Dec-06
93     ---------------------
95     1. Fixed a signed/unsigned compiler warning in pcre_compile.c, shown up by
96     moving to gcc 4.1.1.
98     2. The -S option for pcretest uses setrlimit(); I had omitted to #include
99     sys/time.h, which is documented as needed for this function. It doesn't
100     seem to matter on Linux, but it showed up on some releases of OS X.
102     3. It seems that there are systems where bytes whose values are greater than
103     127 match isprint() in the "C" locale. The "C" locale should be the
104     default when a C program starts up. In most systems, only ASCII printing
105     characters match isprint(). This difference caused the output from pcretest
106     to vary, making some of the tests fail. I have changed pcretest so that:
108     (a) When it is outputting text in the compiled version of a pattern, bytes
109     other than 32-126 are always shown as hex escapes.
111     (b) When it is outputting text that is a matched part of a subject string,
112     it does the same, unless a different locale has been set for the match
113     (using the /L modifier). In this case, it uses isprint() to decide.
115     4. Fixed a major bug that caused incorrect computation of the amount of memory
116     required for a compiled pattern when options that changed within the
117     pattern affected the logic of the preliminary scan that determines the
118     length. The relevant options are -x, and -i in UTF-8 mode. The result was
119     that the computed length was too small. The symptoms of this bug were
120     either the PCRE error "internal error: code overflow" from pcre_compile(),
121     or a glibc crash with a message such as "pcretest: free(): invalid next
122     size (fast)". Examples of patterns that provoked this bug (shown in
123     pcretest format) are:
125     /(?-x: )/x
126     /(?x)(?-x: \s*#\s*)/
127     /((?i)[\x{c0}])/8
128     /(?i:[\x{c0}])/8
130     HOWEVER: Change 17 below makes this fix obsolete as the memory computation
131     is now done differently.
133     5. Applied patches from Google to: (a) add a QuoteMeta function to the C++
134     wrapper classes; (b) implement a new function in the C++ scanner that is
135     more efficient than the old way of doing things because it avoids levels of
136     recursion in the regex matching; (c) add a paragraph to the documentation
137     for the FullMatch() function.
139     6. The escape sequence \n was being treated as whatever was defined as
140     "newline". Not only was this contrary to the documentation, which states
141     that \n is character 10 (hex 0A), but it also went horribly wrong when
142     "newline" was defined as CRLF. This has been fixed.
144     7. In pcre_dfa_exec.c the value of an unsigned integer (the variable called c)
145     was being set to -1 for the "end of line" case (supposedly a value that no
146     character can have). Though this value is never used (the check for end of
147     line is "zero bytes in current character"), it caused compiler complaints.
148     I've changed it to 0xffffffff.
150     8. In pcre_version.c, the version string was being built by a sequence of
151     C macros that, in the event of PCRE_PRERELEASE being defined as an empty
152     string (as it is for production releases) called a macro with an empty
153     argument. The C standard says the result of this is undefined. The gcc
154     compiler treats it as an empty string (which was what was wanted) but it is
155     reported that Visual C gives an error. The source has been hacked around to
156     avoid this problem.
158     9. On the advice of a Windows user, included <io.h> and <fcntl.h> in Windows
159     builds of pcretest, and changed the call to _setmode() to use _O_BINARY
160     instead of 0x8000. Made all the #ifdefs test both _WIN32 and WIN32 (not all
161     of them did).
163     10. Originally, pcretest opened its input and output without "b"; then I was
164     told that "b" was needed in some environments, so it was added for release
165     5.0 to both the input and output. (It makes no difference on Unix-like
166     systems.) Later I was told that it is wrong for the input on Windows. I've
167     now abstracted the modes into two macros, to make it easier to fiddle with
168     them, and removed "b" from the input mode under Windows.
170     11. Added pkgconfig support for the C++ wrapper library, libpcrecpp.
172     12. Added -help and --help to pcretest as an official way of being reminded
173     of the options.
175     13. Removed some redundant semicolons after macro calls in pcrecpparg.h.in
176     and pcrecpp.cc because they annoy compilers at high warning levels.
178     14. A bit of tidying/refactoring in pcre_exec.c in the main bumpalong loop.
180     15. Fixed an occurrence of == in configure.ac that should have been = (shell
181     scripts are not C programs :-) and which was not noticed because it works
182     on Linux.
184     16. pcretest is supposed to handle any length of pattern and data line (as one
185     line or as a continued sequence of lines) by extending its input buffer if
186     necessary. This feature was broken for very long pattern lines, leading to
187     a string of junk being passed to pcre_compile() if the pattern was longer
188     than about 50K.
190     17. I have done a major re-factoring of the way pcre_compile() computes the
191     amount of memory needed for a compiled pattern. Previously, there was code
192     that made a preliminary scan of the pattern in order to do this. That was
193     OK when PCRE was new, but as the facilities have expanded, it has become
194     harder and harder to keep it in step with the real compile phase, and there
195     have been a number of bugs (see for example, 4 above). I have now found a
196     cunning way of running the real compile function in a "fake" mode that
197     enables it to compute how much memory it would need, while actually only
198     ever using a few hundred bytes of working memory and without too many
199     tests of the mode. This should make future maintenance and development
200     easier. A side effect of this work is that the limit of 200 on the nesting
201     depth of parentheses has been removed (though this was never a serious
202     limitation, I suspect). However, there is a downside: pcre_compile() now
203     runs more slowly than before (30% or more, depending on the pattern). I
204     hope this isn't a big issue. There is no effect on runtime performance.
206     18. Fixed a minor bug in pcretest: if a pattern line was not terminated by a
207     newline (only possible for the last line of a file) and it was a
208     pattern that set a locale (followed by /Lsomething), pcretest crashed.
210     19. Added additional timing features to pcretest. (1) The -tm option now times
211     matching only, not compiling. (2) Both -t and -tm can be followed, as a
212     separate command line item, by a number that specifies the number of
213     repeats to use when timing. The default is 50000; this gives better
214     precision, but takes uncomfortably long for very large patterns.
216     20. Extended pcre_study() to be more clever in cases where a branch of a
217     subpattern has no definite first character. For example, (a*|b*)[cd] would
218     previously give no result from pcre_study(). Now it recognizes that the
219     first character must be a, b, c, or d.
221     21. There was an incorrect error "recursive call could loop indefinitely" if
222     a subpattern (or the entire pattern) that was being tested for matching an
223     empty string contained only one non-empty item after a nested subpattern.
224     For example, the pattern (?>\x{100}*)\d(?R) provoked this error
225     incorrectly, because the \d was being skipped in the check.
227     22. The pcretest program now has a new pattern option /B and a command line
228     option -b, which is equivalent to adding /B to every pattern. This causes
229     it to show the compiled bytecode, without the additional information that
230     -d shows. The effect of -d is now the same as -b with -i (and similarly, /D
231     is the same as /B/I).
233     23. A new optimization is now able automatically to treat some sequences such
234     as a*b as a*+b. More specifically, if something simple (such as a character
235     or a simple class like \d) has an unlimited quantifier, and is followed by
236     something that cannot possibly match the quantified thing, the quantifier
237     is automatically "possessified".
239     24. A recursive reference to a subpattern whose number was greater than 39
240     went wrong under certain circumstances in UTF-8 mode. This bug could also
241     have affected the operation of pcre_study().
243     25. Realized that a little bit of performance could be had by replacing
244     (c & 0xc0) == 0xc0 with c >= 0xc0 when processing UTF-8 characters.
246     26. Timing data from pcretest is now shown to 4 decimal places instead of 3.
248     27. Possessive quantifiers such as a++ were previously implemented by turning
249     them into atomic groups such as ($>a+). Now they have their own opcodes,
250     which improves performance. This includes the automatically created ones
251     from 23 above.
253     28. A pattern such as (?=(\w+))\1: which simulates an atomic group using a
254     lookahead was broken if it was not anchored. PCRE was mistakenly expecting
255     the first matched character to be a colon. This applied both to named and
256     numbered groups.
258     29. The ucpinternal.h header file was missing its idempotency #ifdef.
260     30. I was sent a "project" file called libpcre.a.dev which I understand makes
261     building PCRE on Windows easier, so I have included it in the distribution.
263     31. There is now a check in pcretest against a ridiculously large number being
264     returned by pcre_exec() or pcre_dfa_exec(). If this happens in a /g or /G
265     loop, the loop is abandoned.
267     32. Forward references to subpatterns in conditions such as (?(2)...) where
268     subpattern 2 is defined later cause pcre_compile() to search forwards in
269     the pattern for the relevant set of parentheses. This search went wrong
270     when there were unescaped parentheses in a character class, parentheses
271     escaped with \Q...\E, or parentheses in a #-comment in /x mode.
273     33. "Subroutine" calls and backreferences were previously restricted to
274     referencing subpatterns earlier in the regex. This restriction has now
275     been removed.
277     34. Added a number of extra features that are going to be in Perl 5.10. On the
278     whole, these are just syntactic alternatives for features that PCRE had
279     previously implemented using the Python syntax or my own invention. The
280     other formats are all retained for compatibility.
282     (a) Named groups can now be defined as (?<name>...) or (?'name'...) as well
283     as (?P<name>...). The new forms, as well as being in Perl 5.10, are
284     also .NET compatible.
286     (b) A recursion or subroutine call to a named group can now be defined as
287     (?&name) as well as (?P>name).
289     (c) A backreference to a named group can now be defined as \k<name> or
290     \k'name' as well as (?P=name). The new forms, as well as being in Perl
291     5.10, are also .NET compatible.
293     (d) A conditional reference to a named group can now use the syntax
294     (?(<name>) or (?('name') as well as (?(name).
296     (e) A "conditional group" of the form (?(DEFINE)...) can be used to define
297     groups (named and numbered) that are never evaluated inline, but can be
298     called as "subroutines" from elsewhere. In effect, the DEFINE condition
299     is always false. There may be only one alternative in such a group.
301     (f) A test for recursion can be given as (?(R1).. or (?(R&name)... as well
302     as the simple (?(R). The condition is true only if the most recent
303     recursion is that of the given number or name. It does not search out
304     through the entire recursion stack.
306     (g) The escape \gN or \g{N} has been added, where N is a positive or
307     negative number, specifying an absolute or relative reference.
309     35. Tidied to get rid of some further signed/unsigned compiler warnings and
310     some "unreachable code" warnings.
312     36. Updated the Unicode property tables to Unicode version 5.0.0. Amongst other
313     things, this adds five new scripts.
315     37. Perl ignores orphaned \E escapes completely. PCRE now does the same.
316     There were also incompatibilities regarding the handling of \Q..\E inside
317     character classes, for example with patterns like [\Qa\E-\Qz\E] where the
318     hyphen was adjacent to \Q or \E. I hope I've cleared all this up now.
320     38. Like Perl, PCRE detects when an indefinitely repeated parenthesized group
321     matches an empty string, and forcibly breaks the loop. There were bugs in
322     this code in non-simple cases. For a pattern such as ^(a()*)* matched
323     against aaaa the result was just "a" rather than "aaaa", for example. Two
324     separate and independent bugs (that affected different cases) have been
325     fixed.
327     39. Refactored the code to abolish the use of different opcodes for small
328     capturing bracket numbers. This is a tidy that I avoided doing when I
329     removed the limit on the number of capturing brackets for 3.5 back in 2001.
330     The new approach is not only tidier, it makes it possible to reduce the
331     memory needed to fix the previous bug (38).
333     40. Implemented PCRE_NEWLINE_ANY to recognize any of the Unicode newline
334     sequences (http://unicode.org/unicode/reports/tr18/) as "newline" when
335     processing dot, circumflex, or dollar metacharacters, or #-comments in /x
336     mode.
338     41. Add \R to match any Unicode newline sequence, as suggested in the Unicode
339     report.
341     42. Applied patch, originally from Ari Pollak, modified by Google, to allow
342     copy construction and assignment in the C++ wrapper.
344     43. Updated pcregrep to support "--newline=any". In the process, I fixed a
345     couple of bugs that could have given wrong results in the "--newline=crlf"
346     case.
348     44. Added a number of casts and did some reorganization of signed/unsigned int
349     variables following suggestions from Dair Grant. Also renamed the variable
350     "this" as "item" because it is a C++ keyword.
352     45. Arranged for dftables to add
354     #include "pcre_internal.h"
356     to pcre_chartables.c because without it, gcc 4.x may remove the array
357     definition from the final binary if PCRE is built into a static library and
358     dead code stripping is activated.
360     46. For an unanchored pattern, if a match attempt fails at the start of a
361     newline sequence, and the newline setting is CRLF or ANY, and the next two
362     characters are CRLF, advance by two characters instead of one.
365 nigel 91 Version 6.7 04-Jul-06
366     ---------------------
368     1. In order to handle tests when input lines are enormously long, pcretest has
369     been re-factored so that it automatically extends its buffers when
370     necessary. The code is crude, but this _is_ just a test program. The
371     default size has been increased from 32K to 50K.
373     2. The code in pcre_study() was using the value of the re argument before
374     testing it for NULL. (Of course, in any sensible call of the function, it
375     won't be NULL.)
377     3. The memmove() emulation function in pcre_internal.h, which is used on
378     systems that lack both memmove() and bcopy() - that is, hardly ever -
379     was missing a "static" storage class specifier.
381     4. When UTF-8 mode was not set, PCRE looped when compiling certain patterns
382     containing an extended class (one that cannot be represented by a bitmap
383     because it contains high-valued characters or Unicode property items, e.g.
384     [\pZ]). Almost always one would set UTF-8 mode when processing such a
385     pattern, but PCRE should not loop if you do not (it no longer does).
386     [Detail: two cases were found: (a) a repeated subpattern containing an
387     extended class; (b) a recursive reference to a subpattern that followed a
388     previous extended class. It wasn't skipping over the extended class
389     correctly when UTF-8 mode was not set.]
391     5. A negated single-character class was not being recognized as fixed-length
392     in lookbehind assertions such as (?<=[^f]), leading to an incorrect
393     compile error "lookbehind assertion is not fixed length".
395     6. The RunPerlTest auxiliary script was showing an unexpected difference
396     between PCRE and Perl for UTF-8 tests. It turns out that it is hard to
397     write a Perl script that can interpret lines of an input file either as
398     byte characters or as UTF-8, which is what "perltest" was being required to
399     do for the non-UTF-8 and UTF-8 tests, respectively. Essentially what you
400     can't do is switch easily at run time between having the "use utf8;" pragma
401     or not. In the end, I fudged it by using the RunPerlTest script to insert
402     "use utf8;" explicitly for the UTF-8 tests.
404     7. In multiline (/m) mode, PCRE was matching ^ after a terminating newline at
405     the end of the subject string, contrary to the documentation and to what
406     Perl does. This was true of both matching functions. Now it matches only at
407     the start of the subject and immediately after *internal* newlines.
409     8. A call of pcre_fullinfo() from pcretest to get the option bits was passing
410     a pointer to an int instead of a pointer to an unsigned long int. This
411     caused problems on 64-bit systems.
413     9. Applied a patch from the folks at Google to pcrecpp.cc, to fix "another
414     instance of the 'standard' template library not being so standard".
416     10. There was no check on the number of named subpatterns nor the maximum
417     length of a subpattern name. The product of these values is used to compute
418     the size of the memory block for a compiled pattern. By supplying a very
419     long subpattern name and a large number of named subpatterns, the size
420     computation could be caused to overflow. This is now prevented by limiting
421     the length of names to 32 characters, and the number of named subpatterns
422     to 10,000.
424     11. Subpatterns that are repeated with specific counts have to be replicated in
425     the compiled pattern. The size of memory for this was computed from the
426     length of the subpattern and the repeat count. The latter is limited to
427     65535, but there was no limit on the former, meaning that integer overflow
428     could in principle occur. The compiled length of a repeated subpattern is
429     now limited to 30,000 bytes in order to prevent this.
431     12. Added the optional facility to have named substrings with the same name.
433     13. Added the ability to use a named substring as a condition, using the
434     Python syntax: (?(name)yes|no). This overloads (?(R)... and names that
435     are numbers (not recommended). Forward references are permitted.
437     14. Added forward references in named backreferences (if you see what I mean).
439     15. In UTF-8 mode, with the PCRE_DOTALL option set, a quantified dot in the
440     pattern could run off the end of the subject. For example, the pattern
441     "(?s)(.{1,5})"8 did this with the subject "ab".
443     16. If PCRE_DOTALL or PCRE_MULTILINE were set, pcre_dfa_exec() behaved as if
444     PCRE_CASELESS was set when matching characters that were quantified with ?
445     or *.
447     17. A character class other than a single negated character that had a minimum
448     but no maximum quantifier - for example [ab]{6,} - was not handled
449     correctly by pce_dfa_exec(). It would match only one character.
451     18. A valid (though odd) pattern that looked like a POSIX character
452     class but used an invalid character after [ (for example [[,abc,]]) caused
453     pcre_compile() to give the error "Failed: internal error: code overflow" or
454     in some cases to crash with a glibc free() error. This could even happen if
455     the pattern terminated after [[ but there just happened to be a sequence of
456     letters, a binary zero, and a closing ] in the memory that followed.
458     19. Perl's treatment of octal escapes in the range \400 to \777 has changed
459     over the years. Originally (before any Unicode support), just the bottom 8
460     bits were taken. Thus, for example, \500 really meant \100. Nowadays the
461     output from "man perlunicode" includes this:
463     The regular expression compiler produces polymorphic opcodes. That
464     is, the pattern adapts to the data and automatically switches to
465     the Unicode character scheme when presented with Unicode data--or
466     instead uses a traditional byte scheme when presented with byte
467     data.
469     Sadly, a wide octal escape does not cause a switch, and in a string with
470     no other multibyte characters, these octal escapes are treated as before.
471     Thus, in Perl, the pattern /\500/ actually matches \100 but the pattern
472     /\500|\x{1ff}/ matches \500 or \777 because the whole thing is treated as a
473     Unicode string.
475     I have not perpetrated such confusion in PCRE. Up till now, it took just
476     the bottom 8 bits, as in old Perl. I have now made octal escapes with
477     values greater than \377 illegal in non-UTF-8 mode. In UTF-8 mode they
478     translate to the appropriate multibyte character.
480     29. Applied some refactoring to reduce the number of warnings from Microsoft
481     and Borland compilers. This has included removing the fudge introduced
482     seven years ago for the OS/2 compiler (see 2.02/2 below) because it caused
483     a warning about an unused variable.
485     21. PCRE has not included VT (character 0x0b) in the set of whitespace
486     characters since release 4.0, because Perl (from release 5.004) does not.
487     [Or at least, is documented not to: some releases seem to be in conflict
488     with the documentation.] However, when a pattern was studied with
489     pcre_study() and all its branches started with \s, PCRE still included VT
490     as a possible starting character. Of course, this did no harm; it just
491     caused an unnecessary match attempt.
493     22. Removed a now-redundant internal flag bit that recorded the fact that case
494     dependency changed within the pattern. This was once needed for "required
495     byte" processing, but is no longer used. This recovers a now-scarce options
496     bit. Also moved the least significant internal flag bit to the most-
497     significant bit of the word, which was not previously used (hangover from
498     the days when it was an int rather than a uint) to free up another bit for
499     the future.
501     23. Added support for CRLF line endings as well as CR and LF. As well as the
502     default being selectable at build time, it can now be changed at runtime
503     via the PCRE_NEWLINE_xxx flags. There are now options for pcregrep to
504     specify that it is scanning data with non-default line endings.
506     24. Changed the definition of CXXLINK to make it agree with the definition of
507     LINK in the Makefile, by replacing LDFLAGS to CXXFLAGS.
509     25. Applied Ian Taylor's patches to avoid using another stack frame for tail
510     recursions. This makes a big different to stack usage for some patterns.
512     26. If a subpattern containing a named recursion or subroutine reference such
513     as (?P>B) was quantified, for example (xxx(?P>B)){3}, the calculation of
514     the space required for the compiled pattern went wrong and gave too small a
515     value. Depending on the environment, this could lead to "Failed: internal
516     error: code overflow at offset 49" or "glibc detected double free or
517     corruption" errors.
519     27. Applied patches from Google (a) to support the new newline modes and (b) to
520     advance over multibyte UTF-8 characters in GlobalReplace.
522     28. Change free() to pcre_free() in pcredemo.c. Apparently this makes a
523     difference for some implementation of PCRE in some Windows version.
525     29. Added some extra testing facilities to pcretest:
527     \q<number> in a data line sets the "match limit" value
528     \Q<number> in a data line sets the "match recursion limt" value
529     -S <number> sets the stack size, where <number> is in megabytes
531     The -S option isn't available for Windows.
534 nigel 89 Version 6.6 06-Feb-06
535     ---------------------
537     1. Change 16(a) for 6.5 broke things, because PCRE_DATA_SCOPE was not defined
538     in pcreposix.h. I have copied the definition from pcre.h.
540     2. Change 25 for 6.5 broke compilation in a build directory out-of-tree
541     because pcre.h is no longer a built file.
543     3. Added Jeff Friedl's additional debugging patches to pcregrep. These are
544     not normally included in the compiled code.
547 nigel 87 Version 6.5 01-Feb-06
548     ---------------------
550     1. When using the partial match feature with pcre_dfa_exec(), it was not
551     anchoring the second and subsequent partial matches at the new starting
552     point. This could lead to incorrect results. For example, with the pattern
553     /1234/, partially matching against "123" and then "a4" gave a match.
555     2. Changes to pcregrep:
557     (a) All non-match returns from pcre_exec() were being treated as failures
558     to match the line. Now, unless the error is PCRE_ERROR_NOMATCH, an
559     error message is output. Some extra information is given for the
561     probably the only errors that are likely to be caused by users (by
562     specifying a regex that has nested indefinite repeats, for instance).
563     If there are more than 20 of these errors, pcregrep is abandoned.
565     (b) A binary zero was treated as data while matching, but terminated the
566     output line if it was written out. This has been fixed: binary zeroes
567     are now no different to any other data bytes.
569     (c) Whichever of the LC_ALL or LC_CTYPE environment variables is set is
570     used to set a locale for matching. The --locale=xxxx long option has
571     been added (no short equivalent) to specify a locale explicitly on the
572     pcregrep command, overriding the environment variables.
574     (d) When -B was used with -n, some line numbers in the output were one less
575     than they should have been.
577     (e) Added the -o (--only-matching) option.
579     (f) If -A or -C was used with -c (count only), some lines of context were
580     accidentally printed for the final match.
582     (g) Added the -H (--with-filename) option.
584     (h) The combination of options -rh failed to suppress file names for files
585     that were found from directory arguments.
587     (i) Added the -D (--devices) and -d (--directories) options.
589     (j) Added the -F (--fixed-strings) option.
591     (k) Allow "-" to be used as a file name for -f as well as for a data file.
593     (l) Added the --colo(u)r option.
595     (m) Added Jeffrey Friedl's -S testing option, but within #ifdefs so that it
596     is not present by default.
598     3. A nasty bug was discovered in the handling of recursive patterns, that is,
599     items such as (?R) or (?1), when the recursion could match a number of
600     alternatives. If it matched one of the alternatives, but subsequently,
601     outside the recursion, there was a failure, the code tried to back up into
602     the recursion. However, because of the way PCRE is implemented, this is not
603     possible, and the result was an incorrect result from the match.
605     In order to prevent this happening, the specification of recursion has
606     been changed so that all such subpatterns are automatically treated as
607     atomic groups. Thus, for example, (?R) is treated as if it were (?>(?R)).
609     4. I had overlooked the fact that, in some locales, there are characters for
610     which isalpha() is true but neither isupper() nor islower() are true. In
611     the fr_FR locale, for instance, the \xAA and \xBA characters (ordmasculine
612     and ordfeminine) are like this. This affected the treatment of \w and \W
613     when they appeared in character classes, but not when they appeared outside
614     a character class. The bit map for "word" characters is now created
615     separately from the results of isalnum() instead of just taking it from the
616     upper, lower, and digit maps. (Plus the underscore character, of course.)
618     5. The above bug also affected the handling of POSIX character classes such as
619     [[:alpha:]] and [[:alnum:]]. These do not have their own bit maps in PCRE's
620     permanent tables. Instead, the bit maps for such a class were previously
621     created as the appropriate unions of the upper, lower, and digit bitmaps.
622     Now they are created by subtraction from the [[:word:]] class, which has
623     its own bitmap.
625     6. The [[:blank:]] character class matches horizontal, but not vertical space.
626     It is created by subtracting the vertical space characters (\x09, \x0a,
627     \x0b, \x0c) from the [[:space:]] bitmap. Previously, however, the
628     subtraction was done in the overall bitmap for a character class, meaning
629     that a class such as [\x0c[:blank:]] was incorrect because \x0c would not
630     be recognized. This bug has been fixed.
632     7. Patches from the folks at Google:
634     (a) pcrecpp.cc: "to handle a corner case that may or may not happen in
635     real life, but is still worth protecting against".
637     (b) pcrecpp.cc: "corrects a bug when negative radixes are used with
638     regular expressions".
640     (c) pcre_scanner.cc: avoid use of std::count() because not all systems
641     have it.
643     (d) Split off pcrecpparg.h from pcrecpp.h and had the former built by
644     "configure" and the latter not, in order to fix a problem somebody had
645     with compiling the Arg class on HP-UX.
647     (e) Improve the error-handling of the C++ wrapper a little bit.
649     (f) New tests for checking recursion limiting.
651     8. The pcre_memmove() function, which is used only if the environment does not
652     have a standard memmove() function (and is therefore rarely compiled),
653     contained two bugs: (a) use of int instead of size_t, and (b) it was not
654     returning a result (though PCRE never actually uses the result).
656     9. In the POSIX regexec() interface, if nmatch is specified as a ridiculously
657     large number - greater than INT_MAX/(3*sizeof(int)) - REG_ESPACE is
658     returned instead of calling malloc() with an overflowing number that would
659     most likely cause subsequent chaos.
661     10. The debugging option of pcretest was not showing the NO_AUTO_CAPTURE flag.
663     11. The POSIX flag REG_NOSUB is now supported. When a pattern that was compiled
664     with this option is matched, the nmatch and pmatch options of regexec() are
665     ignored.
667     12. Added REG_UTF8 to the POSIX interface. This is not defined by POSIX, but is
668     provided in case anyone wants to the the POSIX interface with UTF-8
669     strings.
671     13. Added CXXLDFLAGS to the Makefile parameters to provide settings only on the
672     C++ linking (needed for some HP-UX environments).
674     14. Avoid compiler warnings in get_ucpname() when compiled without UCP support
675     (unused parameter) and in the pcre_printint() function (omitted "default"
676     switch label when the default is to do nothing).
678     15. Added some code to make it possible, when PCRE is compiled as a C++
679     library, to replace subject pointers for pcre_exec() with a smart pointer
680     class, thus making it possible to process discontinuous strings.
682     16. The two macros PCRE_EXPORT and PCRE_DATA_SCOPE are confusing, and perform
683     much the same function. They were added by different people who were trying
684     to make PCRE easy to compile on non-Unix systems. It has been suggested
685     that PCRE_EXPORT be abolished now that there is more automatic apparatus
686     for compiling on Windows systems. I have therefore replaced it with
687     PCRE_DATA_SCOPE. This is set automatically for Windows; if not set it
688     defaults to "extern" for C or "extern C" for C++, which works fine on
689     Unix-like systems. It is now possible to override the value of PCRE_DATA_
690     SCOPE with something explicit in config.h. In addition:
692     (a) pcreposix.h still had just "extern" instead of either of these macros;
693     I have replaced it with PCRE_DATA_SCOPE.
695     (b) Functions such as _pcre_xclass(), which are internal to the library,
696     but external in the C sense, all had PCRE_EXPORT in their definitions.
697     This is apparently wrong for the Windows case, so I have removed it.
698     (It makes no difference on Unix-like systems.)
700     17. Added a new limit, MATCH_LIMIT_RECURSION, which limits the depth of nesting
701     of recursive calls to match(). This is different to MATCH_LIMIT because
702     that limits the total number of calls to match(), not all of which increase
703     the depth of recursion. Limiting the recursion depth limits the amount of
704     stack (or heap if NO_RECURSE is set) that is used. The default can be set
705     when PCRE is compiled, and changed at run time. A patch from Google adds
706     this functionality to the C++ interface.
708     18. Changes to the handling of Unicode character properties:
710     (a) Updated the table to Unicode 4.1.0.
712     (b) Recognize characters that are not in the table as "Cn" (undefined).
714     (c) I revised the way the table is implemented to a much improved format
715     which includes recognition of ranges. It now supports the ranges that
716     are defined in UnicodeData.txt, and it also amalgamates other
717     characters into ranges. This has reduced the number of entries in the
718     table from around 16,000 to around 3,000, thus reducing its size
719     considerably. I realized I did not need to use a tree structure after
720     all - a binary chop search is just as efficient. Having reduced the
721     number of entries, I extended their size from 6 bytes to 8 bytes to
722     allow for more data.
724     (d) Added support for Unicode script names via properties such as \p{Han}.
726     19. In UTF-8 mode, a backslash followed by a non-Ascii character was not
727     matching that character.
729     20. When matching a repeated Unicode property with a minimum greater than zero,
730     (for example \pL{2,}), PCRE could look past the end of the subject if it
731     reached it while seeking the minimum number of characters. This could
732     happen only if some of the characters were more than one byte long, because
733     there is a check for at least the minimum number of bytes.
735     21. Refactored the implementation of \p and \P so as to be more general, to
736     allow for more different types of property in future. This has changed the
737     compiled form incompatibly. Anybody with saved compiled patterns that use
738     \p or \P will have to recompile them.
740     22. Added "Any" and "L&" to the supported property types.
742     23. Recognize \x{...} as a code point specifier, even when not in UTF-8 mode,
743     but give a compile time error if the value is greater than 0xff.
745     24. The man pages for pcrepartial, pcreprecompile, and pcre_compile2 were
746     accidentally not being installed or uninstalled.
748     25. The pcre.h file was built from pcre.h.in, but the only changes that were
749     made were to insert the current release number. This seemed silly, because
750     it made things harder for people building PCRE on systems that don't run
751     "configure". I have turned pcre.h into a distributed file, no longer built
752     by "configure", with the version identification directly included. There is
753     no longer a pcre.h.in file.
755     However, this change necessitated a change to the pcre-config script as
756     well. It is built from pcre-config.in, and one of the substitutions was the
757     release number. I have updated configure.ac so that ./configure now finds
758     the release number by grepping pcre.h.
760     26. Added the ability to run the tests under valgrind.
763 nigel 85 Version 6.4 05-Sep-05
764     ---------------------
766     1. Change 6.0/10/(l) to pcregrep introduced a bug that caused separator lines
767     "--" to be printed when multiple files were scanned, even when none of the
768     -A, -B, or -C options were used. This is not compatible with Gnu grep, so I
769     consider it to be a bug, and have restored the previous behaviour.
771     2. A couple of code tidies to get rid of compiler warnings.
773     3. The pcretest program used to cheat by referring to symbols in the library
774     whose names begin with _pcre_. These are internal symbols that are not
775     really supposed to be visible externally, and in some environments it is
776     possible to suppress them. The cheating is now confined to including
777     certain files from the library's source, which is a bit cleaner.
779     4. Renamed pcre.in as pcre.h.in to go with pcrecpp.h.in; it also makes the
780     file's purpose clearer.
782     5. Reorganized pcre_ucp_findchar().
785 nigel 83 Version 6.3 15-Aug-05
786     ---------------------
788     1. The file libpcre.pc.in did not have general read permission in the tarball.
790     2. There were some problems when building without C++ support:
792     (a) If C++ support was not built, "make install" and "make test" still
793     tried to test it.
795     (b) There were problems when the value of CXX was explicitly set. Some
796     changes have been made to try to fix these, and ...
798     (c) --disable-cpp can now be used to explicitly disable C++ support.
800     (d) The use of @CPP_OBJ@ directly caused a blank line preceded by a
801     backslash in a target when C++ was disabled. This confuses some
802     versions of "make", apparently. Using an intermediate variable solves
803     this. (Same for CPP_LOBJ.)
805     3. $(LINK_FOR_BUILD) now includes $(CFLAGS_FOR_BUILD) and $(LINK)
806     (non-Windows) now includes $(CFLAGS) because these flags are sometimes
807     necessary on certain architectures.
809     4. Added a setting of -export-symbols-regex to the link command to remove
810     those symbols that are exported in the C sense, but actually are local
811     within the library, and not documented. Their names all begin with
812     "_pcre_". This is not a perfect job, because (a) we have to except some
813     symbols that pcretest ("illegally") uses, and (b) the facility isn't always
814     available (and never for static libraries). I have made a note to try to
815     find a way round (a) in the future.
818 nigel 81 Version 6.2 01-Aug-05
819     ---------------------
821     1. There was no test for integer overflow of quantifier values. A construction
822     such as {1111111111111111} would give undefined results. What is worse, if
823     a minimum quantifier for a parenthesized subpattern overflowed and became
824     negative, the calculation of the memory size went wrong. This could have
825     led to memory overwriting.
827     2. Building PCRE using VPATH was broken. Hopefully it is now fixed.
829     3. Added "b" to the 2nd argument of fopen() in dftables.c, for non-Unix-like
830     operating environments where this matters.
832     4. Applied Giuseppe Maxia's patch to add additional features for controlling
833     PCRE options from within the C++ wrapper.
835     5. Named capturing subpatterns were not being correctly counted when a pattern
836     was compiled. This caused two problems: (a) If there were more than 100
837     such subpatterns, the calculation of the memory needed for the whole
838     compiled pattern went wrong, leading to an overflow error. (b) Numerical
839     back references of the form \12, where the number was greater than 9, were
840     not recognized as back references, even though there were sufficient
841     previous subpatterns.
843     6. Two minor patches to pcrecpp.cc in order to allow it to compile on older
844     versions of gcc, e.g. 2.95.4.
847 nigel 79 Version 6.1 21-Jun-05
848     ---------------------
850     1. There was one reference to the variable "posix" in pcretest.c that was not
851     surrounded by "#if !defined NOPOSIX".
853     2. Make it possible to compile pcretest without DFA support, UTF8 support, or
854     the cross-check on the old pcre_info() function, for the benefit of the
855     cut-down version of PCRE that is currently imported into Exim.
857     3. A (silly) pattern starting with (?i)(?-i) caused an internal space
858     allocation error. I've done the easy fix, which wastes 2 bytes for sensible
859     patterns that start (?i) but I don't think that matters. The use of (?i) is
860     just an example; this all applies to the other options as well.
862     4. Since libtool seems to echo the compile commands it is issuing, the output
863     from "make" can be reduced a bit by putting "@" in front of each libtool
864     compile command.
866     5. Patch from the folks at Google for configure.in to be a bit more thorough
867     in checking for a suitable C++ installation before trying to compile the
868     C++ stuff. This should fix a reported problem when a compiler was present,
869     but no suitable headers.
871     6. The man pages all had just "PCRE" as their title. I have changed them to
872     be the relevant file name. I have also arranged that these names are
873     retained in the file doc/pcre.txt, which is a concatenation in text format
874     of all the man pages except the little individual ones for each function.
876     7. The NON-UNIX-USE file had not been updated for the different set of source
877     files that come with release 6. I also added a few comments about the C++
878     wrapper.
881 nigel 77 Version 6.0 07-Jun-05
882     ---------------------
884     1. Some minor internal re-organization to help with my DFA experiments.
886     2. Some missing #ifdef SUPPORT_UCP conditionals in pcretest and printint that
887     didn't matter for the library itself when fully configured, but did matter
888     when compiling without UCP support, or within Exim, where the ucp files are
889     not imported.
891     3. Refactoring of the library code to split up the various functions into
892     different source modules. The addition of the new DFA matching code (see
893     below) to a single monolithic source would have made it really too
894     unwieldy, quite apart from causing all the code to be include in a
895     statically linked application, when only some functions are used. This is
896     relevant even without the DFA addition now that patterns can be compiled in
897     one application and matched in another.
899     The downside of splitting up is that there have to be some external
900     functions and data tables that are used internally in different modules of
901     the library but which are not part of the API. These have all had their
902     names changed to start with "_pcre_" so that they are unlikely to clash
903     with other external names.
905     4. Added an alternate matching function, pcre_dfa_exec(), which matches using
906     a different (DFA) algorithm. Although it is slower than the original
907     function, it does have some advantages for certain types of matching
908     problem.
910     5. Upgrades to pcretest in order to test the features of pcre_dfa_exec(),
911     including restarting after a partial match.
913     6. A patch for pcregrep that defines INVALID_FILE_ATTRIBUTES if it is not
914     defined when compiling for Windows was sent to me. I have put it into the
915     code, though I have no means of testing or verifying it.
917     7. Added the pcre_refcount() auxiliary function.
919     8. Added the PCRE_FIRSTLINE option. This constrains an unanchored pattern to
920     match before or at the first newline in the subject string. In pcretest,
921     the /f option on a pattern can be used to set this.
923     9. A repeated \w when used in UTF-8 mode with characters greater than 256
924     would behave wrongly. This has been present in PCRE since release 4.0.
926     10. A number of changes to the pcregrep command:
928     (a) Refactored how -x works; insert ^(...)$ instead of setting
929     PCRE_ANCHORED and checking the length, in preparation for adding
930     something similar for -w.
932     (b) Added the -w (match as a word) option.
934     (c) Refactored the way lines are read and buffered so as to have more
935     than one at a time available.
937     (d) Implemented a pcregrep test script.
939     (e) Added the -M (multiline match) option. This allows patterns to match
940     over several lines of the subject. The buffering ensures that at least
941     8K, or the rest of the document (whichever is the shorter) is available
942     for matching (and similarly the previous 8K for lookbehind assertions).
944     (f) Changed the --help output so that it now says
946     -w, --word-regex(p)
948     instead of two lines, one with "regex" and the other with "regexp"
949     because that confused at least one person since the short forms are the
950     same. (This required a bit of code, as the output is generated
951     automatically from a table. It wasn't just a text change.)
953     (g) -- can be used to terminate pcregrep options if the next thing isn't an
954     option but starts with a hyphen. Could be a pattern or a path name
955     starting with a hyphen, for instance.
957     (h) "-" can be given as a file name to represent stdin.
959     (i) When file names are being printed, "(standard input)" is used for
960     the standard input, for compatibility with GNU grep. Previously
961     "<stdin>" was used.
963     (j) The option --label=xxx can be used to supply a name to be used for
964     stdin when file names are being printed. There is no short form.
966     (k) Re-factored the options decoding logic because we are going to add
967     two more options that take data. Such options can now be given in four
968     different ways, e.g. "-fname", "-f name", "--file=name", "--file name".
970     (l) Added the -A, -B, and -C options for requesting that lines of context
971     around matches be printed.
973     (m) Added the -L option to print the names of files that do not contain
974     any matching lines, that is, the complement of -l.
976     (n) The return code is 2 if any file cannot be opened, but pcregrep does
977     continue to scan other files.
979     (o) The -s option was incorrectly implemented. For compatibility with other
980     greps, it now suppresses the error message for a non-existent or non-
981     accessible file (but not the return code). There is a new option called
982     -q that suppresses the output of matching lines, which was what -s was
983     previously doing.
985     (p) Added --include and --exclude options to specify files for inclusion
986     and exclusion when recursing.
988     11. The Makefile was not using the Autoconf-supported LDFLAGS macro properly.
989     Hopefully, it now does.
991     12. Missing cast in pcre_study().
993     13. Added an "uninstall" target to the makefile.
995     14. Replaced "extern" in the function prototypes in Makefile.in with
996     "PCRE_DATA_SCOPE", which defaults to 'extern' or 'extern "C"' in the Unix
997     world, but is set differently for Windows.
999     15. Added a second compiling function called pcre_compile2(). The only
1000     difference is that it has an extra argument, which is a pointer to an
1001     integer error code. When there is a compile-time failure, this is set
1002     non-zero, in addition to the error test pointer being set to point to an
1003     error message. The new argument may be NULL if no error number is required
1004     (but then you may as well call pcre_compile(), which is now just a
1005     wrapper). This facility is provided because some applications need a
1006     numeric error indication, but it has also enabled me to tidy up the way
1007     compile-time errors are handled in the POSIX wrapper.
1009     16. Added VPATH=.libs to the makefile; this should help when building with one
1010     prefix path and installing with another. (Or so I'm told by someone who
1011     knows more about this stuff than I do.)
1013     17. Added a new option, REG_DOTALL, to the POSIX function regcomp(). This
1014     passes PCRE_DOTALL to the pcre_compile() function, making the "." character
1015     match everything, including newlines. This is not POSIX-compatible, but
1016     somebody wanted the feature. From pcretest it can be activated by using
1017     both the P and the s flags.
1019     18. AC_PROG_LIBTOOL appeared twice in Makefile.in. Removed one.
1021     19. libpcre.pc was being incorrectly installed as executable.
1023     20. A couple of places in pcretest check for end-of-line by looking for '\n';
1024     it now also looks for '\r' so that it will work unmodified on Windows.
1026     21. Added Google's contributed C++ wrapper to the distribution.
1028     22. Added some untidy missing memory free() calls in pcretest, to keep
1029     Electric Fence happy when testing.
1033 nigel 75 Version 5.0 13-Sep-04
1034     ---------------------
1036     1. Internal change: literal characters are no longer packed up into items
1037     containing multiple characters in a single byte-string. Each character
1038     is now matched using a separate opcode. However, there may be more than one
1039     byte in the character in UTF-8 mode.
1041     2. The pcre_callout_block structure has two new fields: pattern_position and
1042     next_item_length. These contain the offset in the pattern to the next match
1043     item, and its length, respectively.
1045     3. The PCRE_AUTO_CALLOUT option for pcre_compile() requests the automatic
1046     insertion of callouts before each pattern item. Added the /C option to
1047     pcretest to make use of this.
1049     4. On the advice of a Windows user, the lines
1051     #if defined(_WIN32) || defined(WIN32)
1052     _setmode( _fileno( stdout ), 0x8000 );
1053     #endif /* defined(_WIN32) || defined(WIN32) */
1055     have been added to the source of pcretest. This apparently does useful
1056     magic in relation to line terminators.
1058     5. Changed "r" and "w" in the calls to fopen() in pcretest to "rb" and "wb"
1059     for the benefit of those environments where the "b" makes a difference.
1061     6. The icc compiler has the same options as gcc, but "configure" doesn't seem
1062     to know about it. I have put a hack into configure.in that adds in code
1063     to set GCC=yes if CC=icc. This seems to end up at a point in the
1064     generated configure script that is early enough to affect the setting of
1065     compiler options, which is what is needed, but I have no means of testing
1066     whether it really works. (The user who reported this had patched the
1067     generated configure script, which of course I cannot do.)
1069     LATER: After change 22 below (new libtool files), the configure script
1070     seems to know about icc (and also ecc). Therefore, I have commented out
1071     this hack in configure.in.
1073     7. Added support for pkg-config (2 patches were sent in).
1075     8. Negated POSIX character classes that used a combination of internal tables
1076     were completely broken. These were [[:^alpha:]], [[:^alnum:]], and
1077     [[:^ascii]]. Typically, they would match almost any characters. The other
1078     POSIX classes were not broken in this way.
1080     9. Matching the pattern "\b.*?" against "ab cd", starting at offset 1, failed
1081     to find the match, as PCRE was deluded into thinking that the match had to
1082     start at the start point or following a newline. The same bug applied to
1083     patterns with negative forward assertions or any backward assertions
1084     preceding ".*" at the start, unless the pattern required a fixed first
1085     character. This was a failing pattern: "(?!.bcd).*". The bug is now fixed.
1087     10. In UTF-8 mode, when moving forwards in the subject after a failed match
1088     starting at the last subject character, bytes beyond the end of the subject
1089     string were read.
1091     11. Renamed the variable "class" as "classbits" to make life easier for C++
1092     users. (Previously there was a macro definition, but it apparently wasn't
1093     enough.)
1095     12. Added the new field "tables" to the extra data so that tables can be passed
1096     in at exec time, or the internal tables can be re-selected. This allows
1097     a compiled regex to be saved and re-used at a later time by a different
1098     program that might have everything at different addresses.
1100     13. Modified the pcre-config script so that, when run on Solaris, it shows a
1101     -R library as well as a -L library.
1103     14. The debugging options of pcretest (-d on the command line or D on a
1104     pattern) showed incorrect output for anything following an extended class
1105     that contained multibyte characters and which was followed by a quantifier.
1107     15. Added optional support for general category Unicode character properties
1108     via the \p, \P, and \X escapes. Unicode property support implies UTF-8
1109     support. It adds about 90K to the size of the library. The meanings of the
1110     inbuilt class escapes such as \d and \s have NOT been changed.
1112     16. Updated pcredemo.c to include calls to free() to release the memory for the
1113     compiled pattern.
1115     17. The generated file chartables.c was being created in the source directory
1116     instead of in the building directory. This caused the build to fail if the
1117     source directory was different from the building directory, and was
1118     read-only.
1120     18. Added some sample Win commands from Mark Tetrode into the NON-UNIX-USE
1121     file. No doubt somebody will tell me if they don't make sense... Also added
1122     Dan Mooney's comments about building on OpenVMS.
1124     19. Added support for partial matching via the PCRE_PARTIAL option for
1125     pcre_exec() and the \P data escape in pcretest.
1127     20. Extended pcretest with 3 new pattern features:
1129     (i) A pattern option of the form ">rest-of-line" causes pcretest to
1130     write the compiled pattern to the file whose name is "rest-of-line".
1131     This is a straight binary dump of the data, with the saved pointer to
1132     the character tables forced to be NULL. The study data, if any, is
1133     written too. After writing, pcretest reads a new pattern.
1135     (ii) If, instead of a pattern, "<rest-of-line" is given, pcretest reads a
1136     compiled pattern from the given file. There must not be any
1137     occurrences of "<" in the file name (pretty unlikely); if there are,
1138     pcretest will instead treat the initial "<" as a pattern delimiter.
1139     After reading in the pattern, pcretest goes on to read data lines as
1140     usual.
1142     (iii) The F pattern option causes pcretest to flip the bytes in the 32-bit
1143     and 16-bit fields in a compiled pattern, to simulate a pattern that
1144     was compiled on a host of opposite endianness.
1146     21. The pcre-exec() function can now cope with patterns that were compiled on
1147     hosts of opposite endianness, with this restriction:
1149     As for any compiled expression that is saved and used later, the tables
1150     pointer field cannot be preserved; the extra_data field in the arguments
1151     to pcre_exec() should be used to pass in a tables address if a value
1152     other than the default internal tables were used at compile time.
1154     22. Calling pcre_exec() with a negative value of the "ovecsize" parameter is
1155     now diagnosed as an error. Previously, most of the time, a negative number
1156     would have been treated as zero, but if in addition "ovector" was passed as
1157     NULL, a crash could occur.
1159     23. Updated the files ltmain.sh, config.sub, config.guess, and aclocal.m4 with
1160     new versions from the libtool 1.5 distribution (the last one is a copy of
1161     a file called libtool.m4). This seems to have fixed the need to patch
1162     "configure" to support Darwin 1.3 (which I used to do). However, I still
1163     had to patch ltmain.sh to ensure that ${SED} is set (it isn't on my
1164     workstation).
1166     24. Changed the PCRE licence to be the more standard "BSD" licence.
1169 nigel 73 Version 4.5 01-Dec-03
1170     ---------------------
1172     1. There has been some re-arrangement of the code for the match() function so
1173     that it can be compiled in a version that does not call itself recursively.
1174     Instead, it keeps those local variables that need separate instances for
1175     each "recursion" in a frame on the heap, and gets/frees frames whenever it
1176     needs to "recurse". Keeping track of where control must go is done by means
1177     of setjmp/longjmp. The whole thing is implemented by a set of macros that
1178     hide most of the details from the main code, and operates only if
1179     NO_RECURSE is defined while compiling pcre.c. If PCRE is built using the
1180     "configure" mechanism, "--disable-stack-for-recursion" turns on this way of
1181     operating.
1183     To make it easier for callers to provide specially tailored get/free
1184     functions for this usage, two new functions, pcre_stack_malloc, and
1185     pcre_stack_free, are used. They are always called in strict stacking order,
1186     and the size of block requested is always the same.
1188     The PCRE_CONFIG_STACKRECURSE info parameter can be used to find out whether
1189     PCRE has been compiled to use the stack or the heap for recursion. The
1190     -C option of pcretest uses this to show which version is compiled.
1192     A new data escape \S, is added to pcretest; it causes the amounts of store
1193     obtained and freed by both kinds of malloc/free at match time to be added
1194     to the output.
1196     2. Changed the locale test to use "fr_FR" instead of "fr" because that's
1197     what's available on my current Linux desktop machine.
1199     3. When matching a UTF-8 string, the test for a valid string at the start has
1200     been extended. If start_offset is not zero, PCRE now checks that it points
1201     to a byte that is the start of a UTF-8 character. If not, it returns
1202     PCRE_ERROR_BADUTF8_OFFSET (-11). Note: the whole string is still checked;
1203     this is necessary because there may be backward assertions in the pattern.
1204     When matching the same subject several times, it may save resources to use
1205     PCRE_NO_UTF8_CHECK on all but the first call if the string is long.
1207     4. The code for checking the validity of UTF-8 strings has been tightened so
1208     that it rejects (a) strings containing 0xfe or 0xff bytes and (b) strings
1209     containing "overlong sequences".
1211     5. Fixed a bug (appearing twice) that I could not find any way of exploiting!
1212     I had written "if ((digitab[*p++] && chtab_digit) == 0)" where the "&&"
1213     should have been "&", but it just so happened that all the cases this let
1214     through by mistake were picked up later in the function.
1216     6. I had used a variable called "isblank" - this is a C99 function, causing
1217     some compilers to warn. To avoid this, I renamed it (as "blankclass").
1219     7. Cosmetic: (a) only output another newline at the end of pcretest if it is
1220     prompting; (b) run "./pcretest /dev/null" at the start of the test script
1221     so the version is shown; (c) stop "make test" echoing "./RunTest".
1223     8. Added patches from David Burgess to enable PCRE to run on EBCDIC systems.
1225     9. The prototype for memmove() for systems that don't have it was using
1226     size_t, but the inclusion of the header that defines size_t was later. I've
1227     moved the #includes for the C headers earlier to avoid this.
1229     10. Added some adjustments to the code to make it easier to compiler on certain
1230     special systems:
1232     (a) Some "const" qualifiers were missing.
1233     (b) Added the macro EXPORT before all exported functions; by default this
1234     is defined to be empty.
1235     (c) Changed the dftables auxiliary program (that builds chartables.c) so
1236     that it reads its output file name as an argument instead of writing
1237     to the standard output and assuming this can be redirected.
1239     11. In UTF-8 mode, if a recursive reference (e.g. (?1)) followed a character
1240     class containing characters with values greater than 255, PCRE compilation
1241     went into a loop.
1243     12. A recursive reference to a subpattern that was within another subpattern
1244     that had a minimum quantifier of zero caused PCRE to crash. For example,
1245     (x(y(?2))z)? provoked this bug with a subject that got as far as the
1246     recursion. If the recursively-called subpattern itself had a zero repeat,
1247     that was OK.
1249     13. In pcretest, the buffer for reading a data line was set at 30K, but the
1250     buffer into which it was copied (for escape processing) was still set at
1251     1024, so long lines caused crashes.
1253     14. A pattern such as /[ab]{1,3}+/ failed to compile, giving the error
1254     "internal error: code overflow...". This applied to any character class
1255     that was followed by a possessive quantifier.
1257     15. Modified the Makefile to add libpcre.la as a prerequisite for
1258     libpcreposix.la because I was told this is needed for a parallel build to
1259     work.
1261     16. If a pattern that contained .* following optional items at the start was
1262     studied, the wrong optimizing data was generated, leading to matching
1263     errors. For example, studying /[ab]*.*c/ concluded, erroneously, that any
1264     matching string must start with a or b or c. The correct conclusion for
1265     this pattern is that a match can start with any character.
1268 nigel 71 Version 4.4 13-Aug-03
1269     ---------------------
1271     1. In UTF-8 mode, a character class containing characters with values between
1272     127 and 255 was not handled correctly if the compiled pattern was studied.
1273     In fixing this, I have also improved the studying algorithm for such
1274     classes (slightly).
1276     2. Three internal functions had redundant arguments passed to them. Removal
1277     might give a very teeny performance improvement.
1279     3. Documentation bug: the value of the capture_top field in a callout is *one
1280     more than* the number of the hightest numbered captured substring.
1282     4. The Makefile linked pcretest and pcregrep with -lpcre, which could result
1283     in incorrectly linking with a previously installed version. They now link
1284     explicitly with libpcre.la.
1286     5. configure.in no longer needs to recognize Cygwin specially.
1288     6. A problem in pcre.in for Windows platforms is fixed.
1290     7. If a pattern was successfully studied, and the -d (or /D) flag was given to
1291     pcretest, it used to include the size of the study block as part of its
1292     output. Unfortunately, the structure contains a field that has a different
1293     size on different hardware architectures. This meant that the tests that
1294     showed this size failed. As the block is currently always of a fixed size,
1295     this information isn't actually particularly useful in pcretest output, so
1296     I have just removed it.
1298     8. Three pre-processor statements accidentally did not start in column 1.
1299     Sadly, there are *still* compilers around that complain, even though
1300     standard C has not required this for well over a decade. Sigh.
1302     9. In pcretest, the code for checking callouts passed small integers in the
1303     callout_data field, which is a void * field. However, some picky compilers
1304     complained about the casts involved for this on 64-bit systems. Now
1305     pcretest passes the address of the small integer instead, which should get
1306     rid of the warnings.
1308     10. By default, when in UTF-8 mode, PCRE now checks for valid UTF-8 strings at
1309     both compile and run time, and gives an error if an invalid UTF-8 sequence
1310     is found. There is a option for disabling this check in cases where the
1311     string is known to be correct and/or the maximum performance is wanted.
1313     11. In response to a bug report, I changed one line in Makefile.in from
1315     -Wl,--out-implib,.libs/lib@WIN_PREFIX@pcreposix.dll.a \
1316     to
1317     -Wl,--out-implib,.libs/@WIN_PREFIX@libpcreposix.dll.a \
1319     to look similar to other lines, but I have no way of telling whether this
1320     is the right thing to do, as I do not use Windows. No doubt I'll get told
1321     if it's wrong...
1324 nigel 69 Version 4.3 21-May-03
1325     ---------------------
1327     1. Two instances of @WIN_PREFIX@ omitted from the Windows targets in the
1328     Makefile.
1330     2. Some refactoring to improve the quality of the code:
1332     (i) The utf8_table... variables are now declared "const".
1334     (ii) The code for \cx, which used the "case flipping" table to upper case
1335     lower case letters, now just substracts 32. This is ASCII-specific,
1336     but the whole concept of \cx is ASCII-specific, so it seems
1337     reasonable.
1339     (iii) PCRE was using its character types table to recognize decimal and
1340     hexadecimal digits in the pattern. This is silly, because it handles
1341     only 0-9, a-f, and A-F, but the character types table is locale-
1342     specific, which means strange things might happen. A private
1343     table is now used for this - though it costs 256 bytes, a table is
1344     much faster than multiple explicit tests. Of course, the standard
1345     character types table is still used for matching digits in subject
1346     strings against \d.
1348     (iv) Strictly, the identifier ESC_t is reserved by POSIX (all identifiers
1349     ending in _t are). So I've renamed it as ESC_tee.
1351     3. The first argument for regexec() in the POSIX wrapper should have been
1352     defined as "const".
1354     4. Changed pcretest to use malloc() for its buffers so that they can be
1355     Electric Fenced for debugging.
1357     5. There were several places in the code where, in UTF-8 mode, PCRE would try
1358     to read one or more bytes before the start of the subject string. Often this
1359     had no effect on PCRE's behaviour, but in some circumstances it could
1360     provoke a segmentation fault.
1362     6. A lookbehind at the start of a pattern in UTF-8 mode could also cause PCRE
1363     to try to read one or more bytes before the start of the subject string.
1365     7. A lookbehind in a pattern matched in non-UTF-8 mode on a PCRE compiled with
1366     UTF-8 support could misbehave in various ways if the subject string
1367     contained bytes with the 0x80 bit set and the 0x40 bit unset in a lookbehind
1368     area. (PCRE was not checking for the UTF-8 mode flag, and trying to move
1369     back over UTF-8 characters.)
1372 nigel 67 Version 4.2 14-Apr-03
1373     ---------------------
1375     1. Typo "#if SUPPORT_UTF8" instead of "#ifdef SUPPORT_UTF8" fixed.
1377     2. Changes to the building process, supplied by Ronald Landheer-Cieslak
1378     [ON_WINDOWS]: new variable, "#" on non-Windows platforms
1379     [NOT_ON_WINDOWS]: new variable, "#" on Windows platforms
1380     [WIN_PREFIX]: new variable, "cyg" for Cygwin
1381     * Makefile.in: use autoconf substitution for OBJEXT, EXEEXT, BUILD_OBJEXT
1382     and BUILD_EXEEXT
1383     Note: automatic setting of the BUILD variables is not yet working
1384     set CPPFLAGS and BUILD_CPPFLAGS (but don't use yet) - should be used at
1385     compile-time but not at link-time
1386     [LINK]: use for linking executables only
1387     make different versions for Windows and non-Windows
1388     [LINKLIB]: new variable, copy of UNIX-style LINK, used for linking
1389     libraries
1390     [LINK_FOR_BUILD]: new variable
1391     [OBJEXT]: use throughout
1392     [EXEEXT]: use throughout
1393     <winshared>: new target
1394     <wininstall>: new target
1395     <dftables.o>: use native compiler
1396     <dftables>: use native linker
1397     <install>: handle Windows platform correctly
1398     <clean>: ditto
1399     <check>: ditto
1400     copy DLL to top builddir before testing
1402     As part of these changes, -no-undefined was removed again. This was reported
1403     to give trouble on HP-UX 11.0, so getting rid of it seems like a good idea
1404     in any case.
1406     3. Some tidies to get rid of compiler warnings:
1408     . In the match_data structure, match_limit was an unsigned long int, whereas
1409     match_call_count was an int. I've made them both unsigned long ints.
1411     . In pcretest the fact that a const uschar * doesn't automatically cast to
1412     a void * provoked a warning.
1414     . Turning on some more compiler warnings threw up some "shadow" variables
1415     and a few more missing casts.
1417     4. If PCRE was complied with UTF-8 support, but called without the PCRE_UTF8
1418     option, a class that contained a single character with a value between 128
1419     and 255 (e.g. /[\xFF]/) caused PCRE to crash.
1421     5. If PCRE was compiled with UTF-8 support, but called without the PCRE_UTF8
1422     option, a class that contained several characters, but with at least one
1423     whose value was between 128 and 255 caused PCRE to crash.
1426 nigel 65 Version 4.1 12-Mar-03
1427     ---------------------
1428 nigel 63
1429 nigel 65 1. Compiling with gcc -pedantic found a couple of places where casts were
1430     needed, and a string in dftables.c that was longer than standard compilers are
1431     required to support.
1433     2. Compiling with Sun's compiler found a few more places where the code could
1434     be tidied up in order to avoid warnings.
1436     3. The variables for cross-compiling were called HOST_CC and HOST_CFLAGS; the
1437     first of these names is deprecated in the latest Autoconf in favour of the name
1438     CC_FOR_BUILD, because "host" is typically used to mean the system on which the
1439     compiled code will be run. I can't find a reference for HOST_CFLAGS, but by
1440     analogy I have changed it to CFLAGS_FOR_BUILD.
1442     4. Added -no-undefined to the linking command in the Makefile, because this is
1443     apparently helpful for Windows. To make it work, also added "-L. -lpcre" to the
1444     linking step for the pcreposix library.
1446     5. PCRE was failing to diagnose the case of two named groups with the same
1447     name.
1449     6. A problem with one of PCRE's optimizations was discovered. PCRE remembers a
1450     literal character that is needed in the subject for a match, and scans along to
1451     ensure that it is present before embarking on the full matching process. This
1452     saves time in cases of nested unlimited repeats that are never going to match.
1453     Problem: the scan can take a lot of time if the subject is very long (e.g.
1454     megabytes), thus penalizing straightforward matches. It is now done only if the
1455     amount of subject to be scanned is less than 1000 bytes.
1457     7. A lesser problem with the same optimization is that it was recording the
1458     first character of an anchored pattern as "needed", thus provoking a search
1459     right along the subject, even when the first match of the pattern was going to
1460     fail. The "needed" character is now not set for anchored patterns, unless it
1461     follows something in the pattern that is of non-fixed length. Thus, it still
1462     fulfils its original purpose of finding quick non-matches in cases of nested
1463     unlimited repeats, but isn't used for simple anchored patterns such as /^abc/.
1466     Version 4.0 17-Feb-03
1467     ---------------------
1469 nigel 63 1. If a comment in an extended regex that started immediately after a meta-item
1470     extended to the end of string, PCRE compiled incorrect data. This could lead to
1471     all kinds of weird effects. Example: /#/ was bad; /()#/ was bad; /a#/ was not.
1473     2. Moved to autoconf 2.53 and libtool 1.4.2.
1475     3. Perl 5.8 no longer needs "use utf8" for doing UTF-8 things. Consequently,
1476     the special perltest8 script is no longer needed - all the tests can be run
1477     from a single perltest script.
1479     4. From 5.004, Perl has not included the VT character (0x0b) in the set defined
1480     by \s. It has now been removed in PCRE. This means it isn't recognized as
1481     whitespace in /x regexes too, which is the same as Perl. Note that the POSIX
1482     class [:space:] *does* include VT, thereby creating a mess.
1484     5. Added the class [:blank:] (a GNU extension from Perl 5.8) to match only
1485     space and tab.
1487     6. Perl 5.005 was a long time ago. It's time to amalgamate the tests that use
1488     its new features into the main test script, reducing the number of scripts.
1490     7. Perl 5.8 has changed the meaning of patterns like /a(?i)b/. Earlier versions
1491     were backward compatible, and made the (?i) apply to the whole pattern, as if
1492     /i were given. Now it behaves more logically, and applies the option setting
1493     only to what follows. PCRE has been changed to follow suit. However, if it
1494     finds options settings right at the start of the pattern, it extracts them into
1495     the global options, as before. Thus, they show up in the info data.
1497     8. Added support for the \Q...\E escape sequence. Characters in between are
1498     treated as literals. This is slightly different from Perl in that $ and @ are
1499     also handled as literals inside the quotes. In Perl, they will cause variable
1500     interpolation. Note the following examples:
1502     Pattern PCRE matches Perl matches
1504     \Qabc$xyz\E abc$xyz abc followed by the contents of $xyz
1505     \Qabc\$xyz\E abc\$xyz abc\$xyz
1506     \Qabc\E\$\Qxyz\E abc$xyz abc$xyz
1508     For compatibility with Perl, \Q...\E sequences are recognized inside character
1509     classes as well as outside them.
1511     9. Re-organized 3 code statements in pcretest to avoid "overflow in
1512     floating-point constant arithmetic" warnings from a Microsoft compiler. Added a
1513     (size_t) cast to one statement in pcretest and one in pcreposix to avoid
1514     signed/unsigned warnings.
1516     10. SunOS4 doesn't have strtoul(). This was used only for unpicking the -o
1517     option for pcretest, so I've replaced it by a simple function that does just
1518     that job.
1520     11. pcregrep was ending with code 0 instead of 2 for the commands "pcregrep" or
1521     "pcregrep -".
1523     12. Added "possessive quantifiers" ?+, *+, ++, and {,}+ which come from Sun's
1524     Java package. This provides some syntactic sugar for simple cases of what my
1525     documentation calls "once-only subpatterns". A pattern such as x*+ is the same
1526     as (?>x*). In other words, if what is inside (?>...) is just a single repeated
1527     item, you can use this simplified notation. Note that only makes sense with
1528     greedy quantifiers. Consequently, the use of the possessive quantifier forces
1529     greediness, whatever the setting of the PCRE_UNGREEDY option.
1531     13. A change of greediness default within a pattern was not taking effect at
1532     the current level for patterns like /(b+(?U)a+)/. It did apply to parenthesized
1533     subpatterns that followed. Patterns like /b+(?U)a+/ worked because the option
1534     was abstracted outside.
1536     14. PCRE now supports the \G assertion. It is true when the current matching
1537     position is at the start point of the match. This differs from \A when the
1538     starting offset is non-zero. Used with the /g option of pcretest (or similar
1539     code), it works in the same way as it does for Perl's /g option. If all
1540     alternatives of a regex begin with \G, the expression is anchored to the start
1541     match position, and the "anchored" flag is set in the compiled expression.
1543     15. Some bugs concerning the handling of certain option changes within patterns
1544     have been fixed. These applied to options other than (?ims). For example,
1545     "a(?x: b c )d" did not match "XabcdY" but did match "Xa b c dY". It should have
1546     been the other way round. Some of this was related to change 7 above.
1548     16. PCRE now gives errors for /[.x.]/ and /[=x=]/ as unsupported POSIX
1549     features, as Perl does. Previously, PCRE gave the warnings only for /[[.x.]]/
1550     and /[[=x=]]/. PCRE now also gives an error for /[:name:]/ because it supports
1551     POSIX classes only within a class (e.g. /[[:alpha:]]/).
1553     17. Added support for Perl's \C escape. This matches one byte, even in UTF8
1554     mode. Unlike ".", it always matches newline, whatever the setting of
1555     PCRE_DOTALL. However, PCRE does not permit \C to appear in lookbehind
1556     assertions. Perl allows it, but it doesn't (in general) work because it can't
1557     calculate the length of the lookbehind. At least, that's the case for Perl
1558     5.8.0 - I've been told they are going to document that it doesn't work in
1559     future.
1561     18. Added an error diagnosis for escapes that PCRE does not support: these are
1562     \L, \l, \N, \P, \p, \U, \u, and \X.
1564     19. Although correctly diagnosing a missing ']' in a character class, PCRE was
1565     reading past the end of the pattern in cases such as /[abcd/.
1567     20. PCRE was getting more memory than necessary for patterns with classes that
1568     contained both POSIX named classes and other characters, e.g. /[[:space:]abc/.
1570     21. Added some code, conditional on #ifdef VPCOMPAT, to make life easier for
1571     compiling PCRE for use with Virtual Pascal.
1573     22. Small fix to the Makefile to make it work properly if the build is done
1574     outside the source tree.
1576     23. Added a new extension: a condition to go with recursion. If a conditional
1577     subpattern starts with (?(R) the "true" branch is used if recursion has
1578     happened, whereas the "false" branch is used only at the top level.
1580     24. When there was a very long string of literal characters (over 255 bytes
1581     without UTF support, over 250 bytes with UTF support), the computation of how
1582     much memory was required could be incorrect, leading to segfaults or other
1583     strange effects.
1585     25. PCRE was incorrectly assuming anchoring (either to start of subject or to
1586     start of line for a non-DOTALL pattern) when a pattern started with (.*) and
1587     there was a subsequent back reference to those brackets. This meant that, for
1588     example, /(.*)\d+\1/ failed to match "abc123bc". Unfortunately, it isn't
1589     possible to check for precisely this case. All we can do is abandon the
1590     optimization if .* occurs inside capturing brackets when there are any back
1591     references whatsoever. (See below for a better fix that came later.)
1593     26. The handling of the optimization for finding the first character of a
1594     non-anchored pattern, and for finding a character that is required later in the
1595     match were failing in some cases. This didn't break the matching; it just
1596     failed to optimize when it could. The way this is done has been re-implemented.
1598     27. Fixed typo in error message for invalid (?R item (it said "(?p").
1600     28. Added a new feature that provides some of the functionality that Perl
1601     provides with (?{...}). The facility is termed a "callout". The way it is done
1602     in PCRE is for the caller to provide an optional function, by setting
1603     pcre_callout to its entry point. Like pcre_malloc and pcre_free, this is a
1604     global variable. By default it is unset, which disables all calling out. To get
1605     the function called, the regex must include (?C) at appropriate points. This
1606     is, in fact, equivalent to (?C0), and any number <= 255 may be given with (?C).
1607     This provides a means of identifying different callout points. When PCRE
1608     reaches such a point in the regex, if pcre_callout has been set, the external
1609     function is called. It is provided with data in a structure called
1610     pcre_callout_block, which is defined in pcre.h. If the function returns 0,
1611     matching continues; if it returns a non-zero value, the match at the current
1612     point fails. However, backtracking will occur if possible. [This was changed
1613     later and other features added - see item 49 below.]
1615     29. pcretest is upgraded to test the callout functionality. It provides a
1616     callout function that displays information. By default, it shows the start of
1617     the match and the current position in the text. There are some new data escapes
1618     to vary what happens:
1620     \C+ in addition, show current contents of captured substrings
1621     \C- do not supply a callout function
1622     \C!n return 1 when callout number n is reached
1623     \C!n!m return 1 when callout number n is reached for the mth time
1625     30. If pcregrep was called with the -l option and just a single file name, it
1626     output "<stdin>" if a match was found, instead of the file name.
1628     31. Improve the efficiency of the POSIX API to PCRE. If the number of capturing
1629     slots is less than POSIX_MALLOC_THRESHOLD, use a block on the stack to pass to
1630     pcre_exec(). This saves a malloc/free per call. The default value of
1631     POSIX_MALLOC_THRESHOLD is 10; it can be changed by --with-posix-malloc-threshold
1632     when configuring.
1634     32. The default maximum size of a compiled pattern is 64K. There have been a
1635     few cases of people hitting this limit. The code now uses macros to handle the
1636     storing of links as offsets within the compiled pattern. It defaults to 2-byte
1637     links, but this can be changed to 3 or 4 bytes by --with-link-size when
1638     configuring. Tests 2 and 5 work only with 2-byte links because they output
1639     debugging information about compiled patterns.
1641     33. Internal code re-arrangements:
1643     (a) Moved the debugging function for printing out a compiled regex into
1644     its own source file (printint.c) and used #include to pull it into
1645     pcretest.c and, when DEBUG is defined, into pcre.c, instead of having two
1646     separate copies.
1648     (b) Defined the list of op-code names for debugging as a macro in
1649     internal.h so that it is next to the definition of the opcodes.
1651     (c) Defined a table of op-code lengths for simpler skipping along compiled
1652     code. This is again a macro in internal.h so that it is next to the
1653     definition of the opcodes.
1655     34. Added support for recursive calls to individual subpatterns, along the
1656     lines of Robin Houston's patch (but implemented somewhat differently).
1658     35. Further mods to the Makefile to help Win32. Also, added code to pcregrep to
1659     allow it to read and process whole directories in Win32. This code was
1660     contributed by Lionel Fourquaux; it has not been tested by me.
1662     36. Added support for named subpatterns. The Python syntax (?P<name>...) is
1663     used to name a group. Names consist of alphanumerics and underscores, and must
1664     be unique. Back references use the syntax (?P=name) and recursive calls use
1665     (?P>name) which is a PCRE extension to the Python extension. Groups still have
1666     numbers. The function pcre_fullinfo() can be used after compilation to extract
1667     a name/number map. There are three relevant calls:
1669     PCRE_INFO_NAMEENTRYSIZE yields the size of each entry in the map
1670     PCRE_INFO_NAMECOUNT yields the number of entries
1671     PCRE_INFO_NAMETABLE yields a pointer to the map.
1673     The map is a vector of fixed-size entries. The size of each entry depends on
1674     the length of the longest name used. The first two bytes of each entry are the
1675     group number, most significant byte first. There follows the corresponding
1676     name, zero terminated. The names are in alphabetical order.
1678     37. Make the maximum literal string in the compiled code 250 for the non-UTF-8
1679     case instead of 255. Making it the same both with and without UTF-8 support
1680     means that the same test output works with both.
1682     38. There was a case of malloc(0) in the POSIX testing code in pcretest. Avoid
1683     calling malloc() with a zero argument.
1685     39. Change 25 above had to resort to a heavy-handed test for the .* anchoring
1686     optimization. I've improved things by keeping a bitmap of backreferences with
1687     numbers 1-31 so that if .* occurs inside capturing brackets that are not in
1688     fact referenced, the optimization can be applied. It is unlikely that a
1689     relevant occurrence of .* (i.e. one which might indicate anchoring or forcing
1690     the match to follow \n) will appear inside brackets with a number greater than
1691     31, but if it does, any back reference > 31 suppresses the optimization.
1693     40. Added a new compile-time option PCRE_NO_AUTO_CAPTURE. This has the effect
1694     of disabling numbered capturing parentheses. Any opening parenthesis that is
1695     not followed by ? behaves as if it were followed by ?: but named parentheses
1696     can still be used for capturing (and they will acquire numbers in the usual
1697     way).
1699     41. Redesigned the return codes from the match() function into yes/no/error so
1700     that errors can be passed back from deep inside the nested calls. A malloc
1701     failure while inside a recursive subpattern call now causes the
1702     PCRE_ERROR_NOMEMORY return instead of quietly going wrong.
1704     42. It is now possible to set a limit on the number of times the match()
1705     function is called in a call to pcre_exec(). This facility makes it possible to
1706     limit the amount of recursion and backtracking, though not in a directly
1707     obvious way, because the match() function is used in a number of different
1708     circumstances. The count starts from zero for each position in the subject
1709     string (for non-anchored patterns). The default limit is, for compatibility, a
1710     large number, namely 10 000 000. You can change this in two ways:
1712     (a) When configuring PCRE before making, you can use --with-match-limit=n
1713     to set a default value for the compiled library.
1715     (b) For each call to pcre_exec(), you can pass a pcre_extra block in which
1716     a different value is set. See 45 below.
1718     If the limit is exceeded, pcre_exec() returns PCRE_ERROR_MATCHLIMIT.
1720     43. Added a new function pcre_config(int, void *) to enable run-time extraction
1721     of things that can be changed at compile time. The first argument specifies
1722     what is wanted and the second points to where the information is to be placed.
1723     The current list of available information is:
1727     The output is an integer that is set to one if UTF-8 support is available;
1728     otherwise it is set to zero.
1732     The output is an integer that it set to the value of the code that is used for
1733     newline. It is either LF (10) or CR (13).
1737     The output is an integer that contains the number of bytes used for internal
1738     linkage in compiled expressions. The value is 2, 3, or 4. See item 32 above.
1742     The output is an integer that contains the threshold above which the POSIX
1743     interface uses malloc() for output vectors. See item 31 above.
1747     The output is an unsigned integer that contains the default limit of the number
1748     of match() calls in a pcre_exec() execution. See 42 above.
1750     44. pcretest has been upgraded by the addition of the -C option. This causes it
1751     to extract all the available output from the new pcre_config() function, and to
1752     output it. The program then exits immediately.
1754     45. A need has arisen to pass over additional data with calls to pcre_exec() in
1755     order to support additional features. One way would have been to define
1756     pcre_exec2() (for example) with extra arguments, but this would not have been
1757     extensible, and would also have required all calls to the original function to
1758     be mapped to the new one. Instead, I have chosen to extend the mechanism that
1759     is used for passing in "extra" data from pcre_study().
1761     The pcre_extra structure is now exposed and defined in pcre.h. It currently
1762     contains the following fields:
1764     flags a bitmap indicating which of the following fields are set
1765     study_data opaque data from pcre_study()
1766     match_limit a way of specifying a limit on match() calls for a specific
1767     call to pcre_exec()
1768     callout_data data for callouts (see 49 below)
1770     The flag bits are also defined in pcre.h, and are
1776     The pcre_study() function now returns one of these new pcre_extra blocks, with
1777     the actual study data pointed to by the study_data field, and the
1778     PCRE_EXTRA_STUDY_DATA flag set. This can be passed directly to pcre_exec() as
1779     before. That is, this change is entirely upwards-compatible and requires no
1780     change to existing code.
1782     If you want to pass in additional data to pcre_exec(), you can either place it
1783     in a pcre_extra block provided by pcre_study(), or create your own pcre_extra
1784     block.
1786     46. pcretest has been extended to test the PCRE_EXTRA_MATCH_LIMIT feature. If a
1787     data string contains the escape sequence \M, pcretest calls pcre_exec() several
1788     times with different match limits, until it finds the minimum value needed for
1789     pcre_exec() to complete. The value is then output. This can be instructive; for
1790     most simple matches the number is quite small, but for pathological cases it
1791     gets very large very quickly.
1793     47. There's a new option for pcre_fullinfo() called PCRE_INFO_STUDYSIZE. It
1794     returns the size of the data block pointed to by the study_data field in a
1795     pcre_extra block, that is, the value that was passed as the argument to
1796     pcre_malloc() when PCRE was getting memory in which to place the information
1797     created by pcre_study(). The fourth argument should point to a size_t variable.
1798     pcretest has been extended so that this information is shown after a successful
1799     pcre_study() call when information about the compiled regex is being displayed.
1801     48. Cosmetic change to Makefile: there's no need to have / after $(DESTDIR)
1802     because what follows is always an absolute path. (Later: it turns out that this
1803     is more than cosmetic for MinGW, because it doesn't like empty path
1804     components.)
1806     49. Some changes have been made to the callout feature (see 28 above):
1808     (i) A callout function now has three choices for what it returns:
1810     0 => success, carry on matching
1811     > 0 => failure at this point, but backtrack if possible
1812     < 0 => serious error, return this value from pcre_exec()
1814     Negative values should normally be chosen from the set of PCRE_ERROR_xxx
1815     values. In particular, returning PCRE_ERROR_NOMATCH forces a standard
1816     "match failed" error. The error number PCRE_ERROR_CALLOUT is reserved for
1817     use by callout functions. It will never be used by PCRE itself.
1819     (ii) The pcre_extra structure (see 45 above) has a void * field called
1820     callout_data, with corresponding flag bit PCRE_EXTRA_CALLOUT_DATA. The
1821     pcre_callout_block structure has a field of the same name. The contents of
1822     the field passed in the pcre_extra structure are passed to the callout
1823     function in the corresponding field in the callout block. This makes it
1824     easier to use the same callout-containing regex from multiple threads. For
1825     testing, the pcretest program has a new data escape
1827     \C*n pass the number n (may be negative) as callout_data
1829     If the callout function in pcretest receives a non-zero value as
1830     callout_data, it returns that value.
1832     50. Makefile wasn't handling CFLAGS properly when compiling dftables. Also,
1833     there were some redundant $(CFLAGS) in commands that are now specified as
1834     $(LINK), which already includes $(CFLAGS).
1836     51. Extensions to UTF-8 support are listed below. These all apply when (a) PCRE
1837     has been compiled with UTF-8 support *and* pcre_compile() has been compiled
1838     with the PCRE_UTF8 flag. Patterns that are compiled without that flag assume
1839     one-byte characters throughout. Note that case-insensitive matching applies
1840     only to characters whose values are less than 256. PCRE doesn't support the
1841     notion of cases for higher-valued characters.
1843     (i) A character class whose characters are all within 0-255 is handled as
1844     a bit map, and the map is inverted for negative classes. Previously, a
1845     character > 255 always failed to match such a class; however it should
1846     match if the class was a negative one (e.g. [^ab]). This has been fixed.
1848     (ii) A negated character class with a single character < 255 is coded as
1849     "not this character" (OP_NOT). This wasn't working properly when the test
1850     character was multibyte, either singly or repeated.
1852     (iii) Repeats of multibyte characters are now handled correctly in UTF-8
1853     mode, for example: \x{100}{2,3}.
1855     (iv) The character escapes \b, \B, \d, \D, \s, \S, \w, and \W (either
1856     singly or repeated) now correctly test multibyte characters. However,
1857     PCRE doesn't recognize any characters with values greater than 255 as
1858     digits, spaces, or word characters. Such characters always match \D, \S,
1859     and \W, and never match \d, \s, or \w.
1861     (v) Classes may now contain characters and character ranges with values
1862     greater than 255. For example: [ab\x{100}-\x{400}].
1864     (vi) pcregrep now has a --utf-8 option (synonym -u) which makes it call
1865     PCRE in UTF-8 mode.
1867     52. The info request value PCRE_INFO_FIRSTCHAR has been renamed
1868     PCRE_INFO_FIRSTBYTE because it is a byte value. However, the old name is
1869     retained for backwards compatibility. (Note that LASTLITERAL is also a byte
1870     value.)
1872     53. The single man page has become too large. I have therefore split it up into
1873     a number of separate man pages. These also give rise to individual HTML pages;
1874     these are now put in a separate directory, and there is an index.html page that
1875     lists them all. Some hyperlinking between the pages has been installed.
1877     54. Added convenience functions for handling named capturing parentheses.
1879     55. Unknown escapes inside character classes (e.g. [\M]) and escapes that
1880     aren't interpreted therein (e.g. [\C]) are literals in Perl. This is now also
1881     true in PCRE, except when the PCRE_EXTENDED option is set, in which case they
1882     are faulted.
1884     56. Introduced HOST_CC and HOST_CFLAGS which can be set in the environment when
1885     calling configure. These values are used when compiling the dftables.c program
1886     which is run to generate the source of the default character tables. They
1887     default to the values of CC and CFLAGS. If you are cross-compiling PCRE,
1888     you will need to set these values.
1890     57. Updated the building process for Windows DLL, as provided by Fred Cox.
1893     Version 3.9 02-Jan-02
1894 nigel 61 ---------------------
1896     1. A bit of extraneous text had somehow crept into the pcregrep documentation.
1898     2. If --disable-static was given, the building process failed when trying to
1899     build pcretest and pcregrep. (For some reason it was using libtool to compile
1900     them, which is not right, as they aren't part of the library.)
1903 nigel 59 Version 3.8 18-Dec-01
1904     ---------------------
1906     1. The experimental UTF-8 code was completely screwed up. It was packing the
1907     bytes in the wrong order. How dumb can you get?
1910 nigel 57 Version 3.7 29-Oct-01
1911     ---------------------
1913     1. In updating pcretest to check change 1 of version 3.6, I screwed up.
1914     This caused pcretest, when used on the test data, to segfault. Unfortunately,
1915     this didn't happen under Solaris 8, where I normally test things.
1917 nigel 59 2. The Makefile had to be changed to make it work on BSD systems, where 'make'
1918     doesn't seem to recognize that ./xxx and xxx are the same file. (This entry
1919     isn't in ChangeLog distributed with 3.7 because I forgot when I hastily made
1920     this fix an hour or so after the initial 3.7 release.)
1921 nigel 57
1922 nigel 59
1923 nigel 55 Version 3.6 23-Oct-01
1924     ---------------------
1926     1. Crashed with /(sens|respons)e and \1ibility/ and "sense and sensibility" if
1927     offsets passed as NULL with zero offset count.
1929     2. The config.guess and config.sub files had not been updated when I moved to
1930     the latest autoconf.
1933 nigel 53 Version 3.5 15-Aug-01
1934     ---------------------
1935 nigel 5
1936 nigel 53 1. Added some missing #if !defined NOPOSIX conditionals in pcretest.c that
1937     had been forgotten.
1939     2. By using declared but undefined structures, we can avoid using "void"
1940     definitions in pcre.h while keeping the internal definitions of the structures
1941     private.
1943     3. The distribution is now built using autoconf 2.50 and libtool 1.4. From a
1944     user point of view, this means that both static and shared libraries are built
1945     by default, but this can be individually controlled. More of the work of
1946     handling this static/shared cases is now inside libtool instead of PCRE's make
1947     file.
1949     4. The pcretest utility is now installed along with pcregrep because it is
1950     useful for users (to test regexs) and by doing this, it automatically gets
1951     relinked by libtool. The documentation has been turned into a man page, so
1952     there are now .1, .txt, and .html versions in /doc.
1954     5. Upgrades to pcregrep:
1955     (i) Added long-form option names like gnu grep.
1956     (ii) Added --help to list all options with an explanatory phrase.
1957     (iii) Added -r, --recursive to recurse into sub-directories.
1958     (iv) Added -f, --file to read patterns from a file.
1960     6. pcre_exec() was referring to its "code" argument before testing that
1961     argument for NULL (and giving an error if it was NULL).
1963     7. Upgraded Makefile.in to allow for compiling in a different directory from
1964     the source directory.
1966     8. Tiny buglet in pcretest: when pcre_fullinfo() was called to retrieve the
1967     options bits, the pointer it was passed was to an int instead of to an unsigned
1968     long int. This mattered only on 64-bit systems.
1970     9. Fixed typo (3.4/1) in pcre.h again. Sigh. I had changed pcre.h (which is
1971     generated) instead of pcre.in, which it its source. Also made the same change
1972     in several of the .c files.
1974     10. A new release of gcc defines printf() as a macro, which broke pcretest
1975     because it had an ifdef in the middle of a string argument for printf(). Fixed
1976     by using separate calls to printf().
1978     11. Added --enable-newline-is-cr and --enable-newline-is-lf to the configure
1979     script, to force use of CR or LF instead of \n in the source. On non-Unix
1980     systems, the value can be set in config.h.
1982     12. The limit of 200 on non-capturing parentheses is a _nesting_ limit, not an
1983     absolute limit. Changed the text of the error message to make this clear, and
1984     likewise updated the man page.
1986     13. The limit of 99 on the number of capturing subpatterns has been removed.
1987     The new limit is 65535, which I hope will not be a "real" limit.
1990 nigel 51 Version 3.4 22-Aug-00
1991     ---------------------
1993     1. Fixed typo in pcre.h: unsigned const char * changed to const unsigned char *.
1995     2. Diagnose condition (?(0) as an error instead of crashing on matching.
1998 nigel 49 Version 3.3 01-Aug-00
1999     ---------------------
2001     1. If an octal character was given, but the value was greater than \377, it
2002     was not getting masked to the least significant bits, as documented. This could
2003     lead to crashes in some systems.
2005     2. Perl 5.6 (if not earlier versions) accepts classes like [a-\d] and treats
2006     the hyphen as a literal. PCRE used to give an error; it now behaves like Perl.
2008     3. Added the functions pcre_free_substring() and pcre_free_substring_list().
2009     These just pass their arguments on to (pcre_free)(), but they are provided
2010     because some uses of PCRE bind it to non-C systems that can call its functions,
2011     but cannot call free() or pcre_free() directly.
2013     4. Add "make test" as a synonym for "make check". Corrected some comments in
2014     the Makefile.
2016     5. Add $(DESTDIR)/ in front of all the paths in the "install" target in the
2017     Makefile.
2019     6. Changed the name of pgrep to pcregrep, because Solaris has introduced a
2020     command called pgrep for grepping around the active processes.
2022     7. Added the beginnings of support for UTF-8 character strings.
2024     8. Arranged for the Makefile to pass over the settings of CC, CFLAGS, and
2025     RANLIB to ./ltconfig so that they are used by libtool. I think these are all
2026     the relevant ones. (AR is not passed because ./ltconfig does its own figuring
2027     out for the ar command.)
2030 nigel 47 Version 3.2 12-May-00
2031     ---------------------
2033     This is purely a bug fixing release.
2035     1. If the pattern /((Z)+|A)*/ was matched agained ZABCDEFG it matched Z instead
2036     of ZA. This was just one example of several cases that could provoke this bug,
2037     which was introduced by change 9 of version 2.00. The code for breaking
2038     infinite loops after an iteration that matches an empty string was't working
2039     correctly.
2041     2. The pcretest program was not imitating Perl correctly for the pattern /a*/g
2042     when matched against abbab (for example). After matching an empty string, it
2043     wasn't forcing anchoring when setting PCRE_NOTEMPTY for the next attempt; this
2044     caused it to match further down the string than it should.
2046     3. The code contained an inclusion of sys/types.h. It isn't clear why this
2047     was there because it doesn't seem to be needed, and it causes trouble on some
2048     systems, as it is not a Standard C header. It has been removed.
2050     4. Made 4 silly changes to the source to avoid stupid compiler warnings that
2051     were reported on the Macintosh. The changes were from
2053     while ((c = *(++ptr)) != 0 && c != '\n');
2054     to
2055     while ((c = *(++ptr)) != 0 && c != '\n') ;
2057     Totally extraordinary, but if that's what it takes...
2059     5. PCRE is being used in one environment where neither memmove() nor bcopy() is
2060     available. Added HAVE_BCOPY and an autoconf test for it; if neither
2061     HAVE_MEMMOVE nor HAVE_BCOPY is set, use a built-in emulation function which
2062     assumes the way PCRE uses memmove() (always moving upwards).
2064     6. PCRE is being used in one environment where strchr() is not available. There
2065     was only one use in pcre.c, and writing it out to avoid strchr() probably gives
2066     faster code anyway.
2069 nigel 45 Version 3.1 09-Feb-00
2070     ---------------------
2072     The only change in this release is the fixing of some bugs in Makefile.in for
2073     the "install" target:
2075     (1) It was failing to install pcreposix.h.
2077     (2) It was overwriting the pcre.3 man page with the pcreposix.3 man page.
2080 nigel 43 Version 3.0 01-Feb-00
2081     ---------------------
2082 nigel 41
2083     1. Add support for the /+ modifier to perltest (to output $` like it does in
2084     pcretest).
2086     2. Add support for the /g modifier to perltest.
2088     3. Fix pcretest so that it behaves even more like Perl for /g when the pattern
2089     matches null strings.
2091     4. Fix perltest so that it doesn't do unwanted things when fed an empty
2092     pattern. Perl treats empty patterns specially - it reuses the most recent
2093     pattern, which is not what we want. Replace // by /(?#)/ in order to avoid this
2094     effect.
2096     5. The POSIX interface was broken in that it was just handing over the POSIX
2097     captured string vector to pcre_exec(), but (since release 2.00) PCRE has
2098     required a bigger vector, with some working space on the end. This means that
2099     the POSIX wrapper now has to get and free some memory, and copy the results.
2101 nigel 43 6. Added some simple autoconf support, placing the test data and the
2102     documentation in separate directories, re-organizing some of the
2103     information files, and making it build pcre-config (a GNU standard). Also added
2104     libtool support for building PCRE as a shared library, which is now the
2105     default.
2106 nigel 41
2107 nigel 43 7. Got rid of the leading zero in the definition of PCRE_MINOR because 08 and
2108     09 are not valid octal constants. Single digits will be used for minor values
2109     less than 10.
2111     8. Defined REG_EXTENDED and REG_NOSUB as zero in the POSIX header, so that
2112     existing programs that set these in the POSIX interface can use PCRE without
2113     modification.
2115     9. Added a new function, pcre_fullinfo() with an extensible interface. It can
2116     return all that pcre_info() returns, plus additional data. The pcre_info()
2117     function is retained for compatibility, but is considered to be obsolete.
2119     10. Added experimental recursion feature (?R) to handle one common case that
2120     Perl 5.6 will be able to do with (?p{...}).
2122     11. Added support for POSIX character classes like [:alpha:], which Perl is
2123     adopting.
2126 nigel 39 Version 2.08 31-Aug-99
2127     ----------------------
2129     1. When startoffset was not zero and the pattern began with ".*", PCRE was not
2130     trying to match at the startoffset position, but instead was moving forward to
2131     the next newline as if a previous match had failed.
2133     2. pcretest was not making use of PCRE_NOTEMPTY when repeating for /g and /G,
2134     and could get into a loop if a null string was matched other than at the start
2135     of the subject.
2137     3. Added definitions of PCRE_MAJOR and PCRE_MINOR to pcre.h so the version can
2138     be distinguished at compile time, and for completeness also added PCRE_DATE.
2140     5. Added Paul Sokolovsky's minor changes to make it easy to compile a Win32 DLL
2141     in GnuWin32 environments.
2144 nigel 37 Version 2.07 29-Jul-99
2145     ----------------------
2147     1. The documentation is now supplied in plain text form and HTML as well as in
2148     the form of man page sources.
2150     2. C++ compilers don't like assigning (void *) values to other pointer types.
2151     In particular this affects malloc(). Although there is no problem in Standard
2152     C, I've put in casts to keep C++ compilers happy.
2154     3. Typo on pcretest.c; a cast of (unsigned char *) in the POSIX regexec() call
2155     should be (const char *).
2157     4. If NOPOSIX is defined, pcretest.c compiles without POSIX support. This may
2158     be useful for non-Unix systems who don't want to bother with the POSIX stuff.
2159     However, I haven't made this a standard facility. The documentation doesn't
2160     mention it, and the Makefile doesn't support it.
2162     5. The Makefile now contains an "install" target, with editable destinations at
2163     the top of the file. The pcretest program is not installed.
2165     6. pgrep -V now gives the PCRE version number and date.
2167     7. Fixed bug: a zero repetition after a literal string (e.g. /abcde{0}/) was
2168     causing the entire string to be ignored, instead of just the last character.
2170     8. If a pattern like /"([^\\"]+|\\.)*"/ is applied in the normal way to a
2171     non-matching string, it can take a very, very long time, even for strings of
2172     quite modest length, because of the nested recursion. PCRE now does better in
2173     some of these cases. It does this by remembering the last required literal
2174     character in the pattern, and pre-searching the subject to ensure it is present
2175     before running the real match. In other words, it applies a heuristic to detect
2176     some types of certain failure quickly, and in the above example, if presented
2177     with a string that has no trailing " it gives "no match" very quickly.
2179     9. A new runtime option PCRE_NOTEMPTY causes null string matches to be ignored;
2180     other alternatives are tried instead.
2183 nigel 35 Version 2.06 09-Jun-99
2184     ----------------------
2186     1. Change pcretest's output for amount of store used to show just the code
2187     space, because the remainder (the data block) varies in size between 32-bit and
2188     64-bit systems.
2190     2. Added an extra argument to pcre_exec() to supply an offset in the subject to
2191     start matching at. This allows lookbehinds to work when searching for multiple
2192     occurrences in a string.
2194     3. Added additional options to pcretest for testing multiple occurrences:
2196     /+ outputs the rest of the string that follows a match
2197     /g loops for multiple occurrences, using the new startoffset argument
2198     /G loops for multiple occurrences by passing an incremented pointer
2200     4. PCRE wasn't doing the "first character" optimization for patterns starting
2201     with \b or \B, though it was doing it for other lookbehind assertions. That is,
2202     it wasn't noticing that a match for a pattern such as /\bxyz/ has to start with
2203     the letter 'x'. On long subject strings, this gives a significant speed-up.
2206 nigel 33 Version 2.05 21-Apr-99
2207     ----------------------
2209     1. Changed the type of magic_number from int to long int so that it works
2210     properly on 16-bit systems.
2212     2. Fixed a bug which caused patterns starting with .* not to work correctly
2213     when the subject string contained newline characters. PCRE was assuming
2214     anchoring for such patterns in all cases, which is not correct because .* will
2215     not pass a newline unless PCRE_DOTALL is set. It now assumes anchoring only if
2216     DOTALL is set at top level; otherwise it knows that patterns starting with .*
2217     must be retried after every newline in the subject.
2220 nigel 31 Version 2.04 18-Feb-99
2221     ----------------------
2223     1. For parenthesized subpatterns with repeats whose minimum was zero, the
2224     computation of the store needed to hold the pattern was incorrect (too large).
2225     If such patterns were nested a few deep, this could multiply and become a real
2226     problem.
2228     2. Added /M option to pcretest to show the memory requirement of a specific
2229     pattern. Made -m a synonym of -s (which does this globally) for compatibility.
2231     3. Subpatterns of the form (regex){n,m} (i.e. limited maximum) were being
2232     compiled in such a way that the backtracking after subsequent failure was
2233     pessimal. Something like (a){0,3} was compiled as (a)?(a)?(a)? instead of
2234     ((a)((a)(a)?)?)? with disastrous performance if the maximum was of any size.
2237 nigel 29 Version 2.03 02-Feb-99
2238     ----------------------
2240     1. Fixed typo and small mistake in man page.
2242 nigel 31 2. Added 4th condition (GPL supersedes if conflict) and created separate
2243     LICENCE file containing the conditions.
2244 nigel 29
2245     3. Updated pcretest so that patterns such as /abc\/def/ work like they do in
2246     Perl, that is the internal \ allows the delimiter to be included in the
2247     pattern. Locked out the use of \ as a delimiter. If \ immediately follows
2248     the final delimiter, add \ to the end of the pattern (to test the error).
2250     4. Added the convenience functions for extracting substrings after a successful
2251     match. Updated pcretest to make it able to test these functions.
2254 nigel 27 Version 2.02 14-Jan-99
2255     ----------------------
2257     1. Initialized the working variables associated with each extraction so that
2258     their saving and restoring doesn't refer to uninitialized store.
2260     2. Put dummy code into study.c in order to trick the optimizer of the IBM C
2261     compiler for OS/2 into generating correct code. Apparently IBM isn't going to
2262     fix the problem.
2264     3. Pcretest: the timing code wasn't using LOOPREPEAT for timing execution
2265     calls, and wasn't printing the correct value for compiling calls. Increased the
2266     default value of LOOPREPEAT, and the number of significant figures in the
2267     times.
2269     4. Changed "/bin/rm" in the Makefile to "-rm" so it works on Windows NT.
2271     5. Renamed "deftables" as "dftables" to get it down to 8 characters, to avoid
2272     a building problem on Windows NT with a FAT file system.
2275 nigel 25 Version 2.01 21-Oct-98
2276     ----------------------
2278     1. Changed the API for pcre_compile() to allow for the provision of a pointer
2279     to character tables built by pcre_maketables() in the current locale. If NULL
2280     is passed, the default tables are used.
2283 nigel 23 Version 2.00 24-Sep-98
2284 nigel 21 ----------------------
2286 nigel 23 1. Since the (>?) facility is in Perl 5.005, don't require PCRE_EXTRA to enable
2287     it any more.
2289     2. Allow quantification of (?>) groups, and make it work correctly.
2291     3. The first character computation wasn't working for (?>) groups.
2293     4. Correct the implementation of \Z (it is permitted to match on the \n at the
2294     end of the subject) and add 5.005's \z, which really does match only at the
2295     very end of the subject.
2297     5. Remove the \X "cut" facility; Perl doesn't have it, and (?> is neater.
2299     6. Remove the ability to specify CASELESS, MULTILINE, DOTALL, and
2300     DOLLAR_END_ONLY at runtime, to make it possible to implement the Perl 5.005
2301     localized options. All options to pcre_study() were also removed.
2303     7. Add other new features from 5.005:
2305     $(?<= positive lookbehind
2306     $(?<! negative lookbehind
2307     (?imsx-imsx) added the unsetting capability
2308     such a setting is global if at outer level; local otherwise
2309     (?imsx-imsx:) non-capturing groups with option setting
2310     (?(cond)re|re) conditional pattern matching
2312     A backreference to itself in a repeated group matches the previous
2313     captured string.
2315     8. General tidying up of studying (both automatic and via "study")
2316     consequential on the addition of new assertions.
2318     9. As in 5.005, unlimited repeated groups that could match an empty substring
2319     are no longer faulted at compile time. Instead, the loop is forcibly broken at
2320     runtime if any iteration does actually match an empty substring.
2322     10. Include the RunTest script in the distribution.
2324     11. Added tests from the Perl 5.005_02 distribution. This showed up a few
2325     discrepancies, some of which were old and were also with respect to 5.004. They
2326     have now been fixed.
2329     Version 1.09 28-Apr-98
2330     ----------------------
2332 nigel 21 1. A negated single character class followed by a quantifier with a minimum
2333     value of one (e.g. [^x]{1,6} ) was not compiled correctly. This could lead to
2334     program crashes, or just wrong answers. This did not apply to negated classes
2335     containing more than one character, or to minima other than one.
2338 nigel 19 Version 1.08 27-Mar-98
2339     ----------------------
2341     1. Add PCRE_UNGREEDY to invert the greediness of quantifiers.
2343     2. Add (?U) and (?X) to set PCRE_UNGREEDY and PCRE_EXTRA respectively. The
2344     latter must appear before anything that relies on it in the pattern.
2347 nigel 17 Version 1.07 16-Feb-98
2348     ----------------------
2350     1. A pattern such as /((a)*)*/ was not being diagnosed as in error (unlimited
2351     repeat of a potentially empty string).
2354 nigel 15 Version 1.06 23-Jan-98
2355     ----------------------
2357     1. Added Markus Oberhumer's little patches for C++.
2359     2. Literal strings longer than 255 characters were broken.
2362 nigel 13 Version 1.05 23-Dec-97
2363     ----------------------
2365     1. Negated character classes containing more than one character were failing if
2366     PCRE_CASELESS was set at run time.
2369 nigel 11 Version 1.04 19-Dec-97
2370     ----------------------
2372     1. Corrected the man page, where some "const" qualifiers had been omitted.
2374     2. Made debugging output print "{0,xxx}" instead of just "{,xxx}" to agree with
2375     input syntax.
2377     3. Fixed memory leak which occurred when a regex with back references was
2378     matched with an offsets vector that wasn't big enough. The temporary memory
2379     that is used in this case wasn't being freed if the match failed.
2381     4. Tidied pcretest to ensure it frees memory that it gets.
2383     5. Temporary memory was being obtained in the case where the passed offsets
2384     vector was exactly big enough.
2386     6. Corrected definition of offsetof() from change 5 below.
2388     7. I had screwed up change 6 below and broken the rules for the use of
2389     setjmp(). Now fixed.
2392 nigel 9 Version 1.03 18-Dec-97
2393     ----------------------
2395     1. A erroneous regex with a missing opening parenthesis was correctly
2396     diagnosed, but PCRE attempted to access brastack[-1], which could cause crashes
2397     on some systems.
2399     2. Replaced offsetof(real_pcre, code) by offsetof(real_pcre, code[0]) because
2400     it was reported that one broken compiler failed on the former because "code" is
2401     also an independent variable.
2403     3. The erroneous regex a[]b caused an array overrun reference.
2405     4. A regex ending with a one-character negative class (e.g. /[^k]$/) did not
2406     fail on data ending with that character. (It was going on too far, and checking
2407     the next character, typically a binary zero.) This was specific to the
2408     optimized code for single-character negative classes.
2410     5. Added a contributed patch from the TIN world which does the following:
2412     + Add an undef for memmove, in case the the system defines a macro for it.
2414     + Add a definition of offsetof(), in case there isn't one. (I don't know
2415     the reason behind this - offsetof() is part of the ANSI standard - but
2416     it does no harm).
2418     + Reduce the ifdef's in pcre.c using macro DPRINTF, thereby eliminating
2419     most of the places where whitespace preceded '#'. I have given up and
2420     allowed the remaining 2 cases to be at the margin.
2422     + Rename some variables in pcre to eliminate shadowing. This seems very
2423     pedantic, but does no harm, of course.
2425     6. Moved the call to setjmp() into its own function, to get rid of warnings
2426     from gcc -Wall, and avoided calling it at all unless PCRE_EXTRA is used.
2428     7. Constructs such as \d{8,} were compiling into the equivalent of
2429 nigel 11 \d{8}\d{0,65527} instead of \d{8}\d* which didn't make much difference to the
2430 nigel 9 outcome, but in this particular case used more store than had been allocated,
2431     which caused the bug to be discovered because it threw up an internal error.
2433     8. The debugging code in both pcre and pcretest for outputting the compiled
2434     form of a regex was going wrong in the case of back references followed by
2435     curly-bracketed repeats.
2438 nigel 7 Version 1.02 12-Dec-97
2439     ----------------------
2441     1. Typos in pcre.3 and comments in the source fixed.
2443     2. Applied a contributed patch to get rid of places where it used to remove
2444     'const' from variables, and fixed some signed/unsigned and uninitialized
2445     variable warnings.
2447     3. Added the "runtest" target to Makefile.
2449     4. Set default compiler flag to -O2 rather than just -O.
2452 nigel 5 Version 1.01 19-Nov-97
2453     ----------------------
2455     1. PCRE was failing to diagnose unlimited repeat of empty string for patterns
2456     like /([ab]*)*/, that is, for classes with more than one character in them.
2458     2. Likewise, it wasn't diagnosing patterns with "once-only" subpatterns, such
2459     as /((?>a*))*/ (a PCRE_EXTRA facility).
2462     Version 1.00 18-Nov-97
2463     ----------------------
2465     1. Added compile-time macros to support systems such as SunOS4 which don't have
2466     memmove() or strerror() but have other things that can be used instead.
2468     2. Arranged that "make clean" removes the executables.
2471 nigel 3 Version 0.99 27-Oct-97
2472     ----------------------
2474     1. Fixed bug in code for optimizing classes with only one character. It was
2475     initializing a 32-byte map regardless, which could cause it to run off the end
2476     of the memory it had got.
2478     2. Added, conditional on PCRE_EXTRA, the proposed (?>REGEX) construction.
2481     Version 0.98 22-Oct-97
2482     ----------------------
2484     1. Fixed bug in code for handling temporary memory usage when there are more
2485     back references than supplied space in the ovector. This could cause segfaults.
2488     Version 0.97 21-Oct-97
2489     ----------------------
2491     1. Added the \X "cut" facility, conditional on PCRE_EXTRA.
2493     2. Optimized negated single characters not to use a bit map.
2495     3. Brought error texts together as macro definitions; clarified some of them;
2496     fixed one that was wrong - it said "range out of order" when it meant "invalid
2497     escape sequence".
2499     4. Changed some char * arguments to const char *.
2501     5. Added PCRE_NOTBOL and PCRE_NOTEOL (from POSIX).
2503     6. Added the POSIX-style API wrapper in pcreposix.a and testing facilities in
2504     pcretest.
2507     Version 0.96 16-Oct-97
2508     ----------------------
2510     1. Added a simple "pgrep" utility to the distribution.
2512     2. Fixed an incompatibility with Perl: "{" is now treated as a normal character
2513     unless it appears in one of the precise forms "{ddd}", "{ddd,}", or "{ddd,ddd}"
2514     where "ddd" means "one or more decimal digits".
2516     3. Fixed serious bug. If a pattern had a back reference, but the call to
2517     pcre_exec() didn't supply a large enough ovector to record the related
2518     identifying subpattern, the match always failed. PCRE now remembers the number
2519     of the largest back reference, and gets some temporary memory in which to save
2520     the offsets during matching if necessary, in order to ensure that
2521     backreferences always work.
2523     4. Increased the compatibility with Perl in a number of ways:
2525     (a) . no longer matches \n by default; an option PCRE_DOTALL is provided
2526     to request this handling. The option can be set at compile or exec time.
2528     (b) $ matches before a terminating newline by default; an option
2529     PCRE_DOLLAR_ENDONLY is provided to override this (but not in multiline
2530     mode). The option can be set at compile or exec time.
2532     (c) The handling of \ followed by a digit other than 0 is now supposed to be
2533     the same as Perl's. If the decimal number it represents is less than 10
2534     or there aren't that many previous left capturing parentheses, an octal
2535     escape is read. Inside a character class, it's always an octal escape,
2536     even if it is a single digit.
2538     (d) An escaped but undefined alphabetic character is taken as a literal,
2539     unless PCRE_EXTRA is set. Currently this just reserves the remaining
2540     escapes.
2542     (e) {0} is now permitted. (The previous item is removed from the compiled
2543     pattern).
2545     5. Changed all the names of code files so that the basic parts are no longer
2546     than 10 characters, and abolished the teeny "globals.c" file.
2548     6. Changed the handling of character classes; they are now done with a 32-byte
2549     bit map always.
2551     7. Added the -d and /D options to pcretest to make it possible to look at the
2552     internals of compilation without having to recompile pcre.
2555     Version 0.95 23-Sep-97
2556     ----------------------
2558     1. Fixed bug in pre-pass concerning escaped "normal" characters such as \x5c or
2559     \x20 at the start of a run of normal characters. These were being treated as
2560     real characters, instead of the source characters being re-checked.
2563     Version 0.94 18-Sep-97
2564     ----------------------
2566     1. The functions are now thread-safe, with the caveat that the global variables
2567     containing pointers to malloc() and free() or alternative functions are the
2568     same for all threads.
2570     2. Get pcre_study() to generate a bitmap of initial characters for non-
2571     anchored patterns when this is possible, and use it if passed to pcre_exec().
2574     Version 0.93 15-Sep-97
2575     ----------------------
2577     1. /(b)|(:+)/ was computing an incorrect first character.
2579     2. Add pcre_study() to the API and the passing of pcre_extra to pcre_exec(),
2580     but not actually doing anything yet.
2582     3. Treat "-" characters in classes that cannot be part of ranges as literals,
2583     as Perl does (e.g. [-az] or [az-]).
2585     4. Set the anchored flag if a branch starts with .* or .*? because that tests
2586     all possible positions.
2588     5. Split up into different modules to avoid including unneeded functions in a
2589     compiled binary. However, compile and exec are still in one module. The "study"
2590     function is split off.
2592     6. The character tables are now in a separate module whose source is generated
2593     by an auxiliary program - but can then be edited by hand if required. There are
2594     now no calls to isalnum(), isspace(), isdigit(), isxdigit(), tolower() or
2595     toupper() in the code.
2597     7. Turn the malloc/free funtions variables into pcre_malloc and pcre_free and
2598     make them global. Abolish the function for setting them, as the caller can now
2599     set them directly.
2602     Version 0.92 11-Sep-97
2603     ----------------------
2605     1. A repeat with a fixed maximum and a minimum of 1 for an ordinary character
2606     (e.g. /a{1,3}/) was broken (I mis-optimized it).
2608     2. Caseless matching was not working in character classes if the characters in
2609     the pattern were in upper case.
2611     3. Make ranges like [W-c] work in the same way as Perl for caseless matching.
2613     4. Make PCRE_ANCHORED public and accept as a compile option.
2615     5. Add an options word to pcre_exec() and accept PCRE_ANCHORED and
2616     PCRE_CASELESS at run time. Add escapes \A and \I to pcretest to cause it to
2617     pass them.
2619     6. Give an error if bad option bits passed at compile or run time.
2621     7. Add PCRE_MULTILINE at compile and exec time, and (?m) as well. Add \M to
2622     pcretest to cause it to pass that flag.
2624     8. Add pcre_info(), to get the number of identifying subpatterns, the stored
2625     options, and the first character, if set.
2627     9. Recognize C+ or C{n,m} where n >= 1 as providing a fixed starting character.
2630     Version 0.91 10-Sep-97
2631     ----------------------
2633     1. PCRE was failing to diagnose unlimited repeats of subpatterns that could
2634     match the empty string as in /(a*)*/. It was looping and ultimately crashing.
2636     2. PCRE was looping on encountering an indefinitely repeated back reference to
2637     a subpattern that had matched an empty string, e.g. /(a|)\1*/. It now does what
2638     Perl does - treats the match as successful.
2640     ****


Name Value
svn:eol-style native
svn:keywords "Author Date Id Revision Url"

ViewVC Help
Powered by ViewVC 1.1.12