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

Contents of /code/trunk/ChangeLog

Parent Directory Parent Directory | Revision Log Revision Log


Revision 99 - (hide annotations) (download)
Tue Mar 6 12:27:42 2007 UTC (7 years, 7 months ago) by ph10
File size: 119802 byte(s)
1. Move the comment about version numbers from pcre.h.in to configure.ac 
because that's where they are now set.
2. Update all the man pages to remove the use of .br and .in because this
causes trouble for some HTML converters. Also standardised the final sections 
giving author information and revision date.
3. Update the maintain/132html man page converter to handle .nf/.fi and to barf 
at .br/.in.

1 nigel 3 ChangeLog for PCRE
2     ------------------
3    
4 ph10 98 Version 7.1 05-Mar-07
5     ---------------------
6    
7     1. Applied Bob Rossi and Daniel G's patches to convert the build system to one
8 ph10 99 that is more "standard", making use of automake and other autotools. There
9     is some re-arrangement of the files and adjustment of comments consequent
10     on this.
11    
12     2. I had an email from Eric Raymond about problems translating some of PCRE's
13     man pages to HTML (despite the fact that I distribute HTML pages, some
14     people do their own conversions for various reasons). The problems
15     concerned the use of low-level troff macros .br and .in. I have therefore
16     removed all such uses from the man pages (some were redundant, some could
17     be replaced by .nf/.fi pairs). The maintain/132html script that I use to
18     generate HTML has been updated to handle .nf/.fi and to complain if it
19     encounters .br or .in.
20    
21 ph10 98
22 nigel 93 Version 7.0 19-Dec-06
23     ---------------------
24    
25     1. Fixed a signed/unsigned compiler warning in pcre_compile.c, shown up by
26     moving to gcc 4.1.1.
27    
28     2. The -S option for pcretest uses setrlimit(); I had omitted to #include
29     sys/time.h, which is documented as needed for this function. It doesn't
30     seem to matter on Linux, but it showed up on some releases of OS X.
31    
32     3. It seems that there are systems where bytes whose values are greater than
33     127 match isprint() in the "C" locale. The "C" locale should be the
34     default when a C program starts up. In most systems, only ASCII printing
35     characters match isprint(). This difference caused the output from pcretest
36     to vary, making some of the tests fail. I have changed pcretest so that:
37    
38     (a) When it is outputting text in the compiled version of a pattern, bytes
39     other than 32-126 are always shown as hex escapes.
40    
41     (b) When it is outputting text that is a matched part of a subject string,
42     it does the same, unless a different locale has been set for the match
43     (using the /L modifier). In this case, it uses isprint() to decide.
44    
45     4. Fixed a major bug that caused incorrect computation of the amount of memory
46     required for a compiled pattern when options that changed within the
47     pattern affected the logic of the preliminary scan that determines the
48     length. The relevant options are -x, and -i in UTF-8 mode. The result was
49     that the computed length was too small. The symptoms of this bug were
50     either the PCRE error "internal error: code overflow" from pcre_compile(),
51     or a glibc crash with a message such as "pcretest: free(): invalid next
52     size (fast)". Examples of patterns that provoked this bug (shown in
53     pcretest format) are:
54    
55     /(?-x: )/x
56     /(?x)(?-x: \s*#\s*)/
57     /((?i)[\x{c0}])/8
58     /(?i:[\x{c0}])/8
59    
60     HOWEVER: Change 17 below makes this fix obsolete as the memory computation
61     is now done differently.
62    
63     5. Applied patches from Google to: (a) add a QuoteMeta function to the C++
64     wrapper classes; (b) implement a new function in the C++ scanner that is
65     more efficient than the old way of doing things because it avoids levels of
66     recursion in the regex matching; (c) add a paragraph to the documentation
67     for the FullMatch() function.
68    
69     6. The escape sequence \n was being treated as whatever was defined as
70     "newline". Not only was this contrary to the documentation, which states
71     that \n is character 10 (hex 0A), but it also went horribly wrong when
72     "newline" was defined as CRLF. This has been fixed.
73    
74     7. In pcre_dfa_exec.c the value of an unsigned integer (the variable called c)
75     was being set to -1 for the "end of line" case (supposedly a value that no
76     character can have). Though this value is never used (the check for end of
77     line is "zero bytes in current character"), it caused compiler complaints.
78     I've changed it to 0xffffffff.
79    
80     8. In pcre_version.c, the version string was being built by a sequence of
81     C macros that, in the event of PCRE_PRERELEASE being defined as an empty
82     string (as it is for production releases) called a macro with an empty
83     argument. The C standard says the result of this is undefined. The gcc
84     compiler treats it as an empty string (which was what was wanted) but it is
85     reported that Visual C gives an error. The source has been hacked around to
86     avoid this problem.
87    
88     9. On the advice of a Windows user, included <io.h> and <fcntl.h> in Windows
89     builds of pcretest, and changed the call to _setmode() to use _O_BINARY
90     instead of 0x8000. Made all the #ifdefs test both _WIN32 and WIN32 (not all
91     of them did).
92    
93     10. Originally, pcretest opened its input and output without "b"; then I was
94     told that "b" was needed in some environments, so it was added for release
95     5.0 to both the input and output. (It makes no difference on Unix-like
96     systems.) Later I was told that it is wrong for the input on Windows. I've
97     now abstracted the modes into two macros, to make it easier to fiddle with
98     them, and removed "b" from the input mode under Windows.
99    
100     11. Added pkgconfig support for the C++ wrapper library, libpcrecpp.
101    
102     12. Added -help and --help to pcretest as an official way of being reminded
103     of the options.
104    
105     13. Removed some redundant semicolons after macro calls in pcrecpparg.h.in
106     and pcrecpp.cc because they annoy compilers at high warning levels.
107    
108     14. A bit of tidying/refactoring in pcre_exec.c in the main bumpalong loop.
109    
110     15. Fixed an occurrence of == in configure.ac that should have been = (shell
111     scripts are not C programs :-) and which was not noticed because it works
112     on Linux.
113    
114     16. pcretest is supposed to handle any length of pattern and data line (as one
115     line or as a continued sequence of lines) by extending its input buffer if
116     necessary. This feature was broken for very long pattern lines, leading to
117     a string of junk being passed to pcre_compile() if the pattern was longer
118     than about 50K.
119    
120     17. I have done a major re-factoring of the way pcre_compile() computes the
121     amount of memory needed for a compiled pattern. Previously, there was code
122     that made a preliminary scan of the pattern in order to do this. That was
123     OK when PCRE was new, but as the facilities have expanded, it has become
124     harder and harder to keep it in step with the real compile phase, and there
125     have been a number of bugs (see for example, 4 above). I have now found a
126     cunning way of running the real compile function in a "fake" mode that
127     enables it to compute how much memory it would need, while actually only
128     ever using a few hundred bytes of working memory and without too many
129     tests of the mode. This should make future maintenance and development
130     easier. A side effect of this work is that the limit of 200 on the nesting
131     depth of parentheses has been removed (though this was never a serious
132     limitation, I suspect). However, there is a downside: pcre_compile() now
133     runs more slowly than before (30% or more, depending on the pattern). I
134     hope this isn't a big issue. There is no effect on runtime performance.
135    
136     18. Fixed a minor bug in pcretest: if a pattern line was not terminated by a
137     newline (only possible for the last line of a file) and it was a
138     pattern that set a locale (followed by /Lsomething), pcretest crashed.
139    
140     19. Added additional timing features to pcretest. (1) The -tm option now times
141     matching only, not compiling. (2) Both -t and -tm can be followed, as a
142     separate command line item, by a number that specifies the number of
143     repeats to use when timing. The default is 50000; this gives better
144     precision, but takes uncomfortably long for very large patterns.
145    
146     20. Extended pcre_study() to be more clever in cases where a branch of a
147     subpattern has no definite first character. For example, (a*|b*)[cd] would
148     previously give no result from pcre_study(). Now it recognizes that the
149     first character must be a, b, c, or d.
150    
151     21. There was an incorrect error "recursive call could loop indefinitely" if
152     a subpattern (or the entire pattern) that was being tested for matching an
153     empty string contained only one non-empty item after a nested subpattern.
154     For example, the pattern (?>\x{100}*)\d(?R) provoked this error
155     incorrectly, because the \d was being skipped in the check.
156    
157     22. The pcretest program now has a new pattern option /B and a command line
158     option -b, which is equivalent to adding /B to every pattern. This causes
159     it to show the compiled bytecode, without the additional information that
160     -d shows. The effect of -d is now the same as -b with -i (and similarly, /D
161     is the same as /B/I).
162    
163     23. A new optimization is now able automatically to treat some sequences such
164     as a*b as a*+b. More specifically, if something simple (such as a character
165     or a simple class like \d) has an unlimited quantifier, and is followed by
166     something that cannot possibly match the quantified thing, the quantifier
167     is automatically "possessified".
168    
169     24. A recursive reference to a subpattern whose number was greater than 39
170     went wrong under certain circumstances in UTF-8 mode. This bug could also
171     have affected the operation of pcre_study().
172    
173     25. Realized that a little bit of performance could be had by replacing
174     (c & 0xc0) == 0xc0 with c >= 0xc0 when processing UTF-8 characters.
175    
176     26. Timing data from pcretest is now shown to 4 decimal places instead of 3.
177    
178     27. Possessive quantifiers such as a++ were previously implemented by turning
179     them into atomic groups such as ($>a+). Now they have their own opcodes,
180     which improves performance. This includes the automatically created ones
181     from 23 above.
182    
183     28. A pattern such as (?=(\w+))\1: which simulates an atomic group using a
184     lookahead was broken if it was not anchored. PCRE was mistakenly expecting
185     the first matched character to be a colon. This applied both to named and
186     numbered groups.
187    
188     29. The ucpinternal.h header file was missing its idempotency #ifdef.
189    
190     30. I was sent a "project" file called libpcre.a.dev which I understand makes
191     building PCRE on Windows easier, so I have included it in the distribution.
192    
193     31. There is now a check in pcretest against a ridiculously large number being
194     returned by pcre_exec() or pcre_dfa_exec(). If this happens in a /g or /G
195     loop, the loop is abandoned.
196    
197     32. Forward references to subpatterns in conditions such as (?(2)...) where
198     subpattern 2 is defined later cause pcre_compile() to search forwards in
199     the pattern for the relevant set of parentheses. This search went wrong
200     when there were unescaped parentheses in a character class, parentheses
201     escaped with \Q...\E, or parentheses in a #-comment in /x mode.
202    
203     33. "Subroutine" calls and backreferences were previously restricted to
204     referencing subpatterns earlier in the regex. This restriction has now
205     been removed.
206    
207     34. Added a number of extra features that are going to be in Perl 5.10. On the
208     whole, these are just syntactic alternatives for features that PCRE had
209     previously implemented using the Python syntax or my own invention. The
210     other formats are all retained for compatibility.
211    
212     (a) Named groups can now be defined as (?<name>...) or (?'name'...) as well
213     as (?P<name>...). The new forms, as well as being in Perl 5.10, are
214     also .NET compatible.
215    
216     (b) A recursion or subroutine call to a named group can now be defined as
217     (?&name) as well as (?P>name).
218    
219     (c) A backreference to a named group can now be defined as \k<name> or
220     \k'name' as well as (?P=name). The new forms, as well as being in Perl
221     5.10, are also .NET compatible.
222    
223     (d) A conditional reference to a named group can now use the syntax
224     (?(<name>) or (?('name') as well as (?(name).
225    
226     (e) A "conditional group" of the form (?(DEFINE)...) can be used to define
227     groups (named and numbered) that are never evaluated inline, but can be
228     called as "subroutines" from elsewhere. In effect, the DEFINE condition
229     is always false. There may be only one alternative in such a group.
230    
231     (f) A test for recursion can be given as (?(R1).. or (?(R&name)... as well
232     as the simple (?(R). The condition is true only if the most recent
233     recursion is that of the given number or name. It does not search out
234     through the entire recursion stack.
235    
236     (g) The escape \gN or \g{N} has been added, where N is a positive or
237     negative number, specifying an absolute or relative reference.
238    
239     35. Tidied to get rid of some further signed/unsigned compiler warnings and
240     some "unreachable code" warnings.
241    
242     36. Updated the Unicode property tables to Unicode version 5.0.0. Amongst other
243     things, this adds five new scripts.
244    
245     37. Perl ignores orphaned \E escapes completely. PCRE now does the same.
246     There were also incompatibilities regarding the handling of \Q..\E inside
247     character classes, for example with patterns like [\Qa\E-\Qz\E] where the
248     hyphen was adjacent to \Q or \E. I hope I've cleared all this up now.
249    
250     38. Like Perl, PCRE detects when an indefinitely repeated parenthesized group
251     matches an empty string, and forcibly breaks the loop. There were bugs in
252     this code in non-simple cases. For a pattern such as ^(a()*)* matched
253     against aaaa the result was just "a" rather than "aaaa", for example. Two
254     separate and independent bugs (that affected different cases) have been
255     fixed.
256    
257     39. Refactored the code to abolish the use of different opcodes for small
258     capturing bracket numbers. This is a tidy that I avoided doing when I
259     removed the limit on the number of capturing brackets for 3.5 back in 2001.
260     The new approach is not only tidier, it makes it possible to reduce the
261     memory needed to fix the previous bug (38).
262    
263     40. Implemented PCRE_NEWLINE_ANY to recognize any of the Unicode newline
264     sequences (http://unicode.org/unicode/reports/tr18/) as "newline" when
265     processing dot, circumflex, or dollar metacharacters, or #-comments in /x
266     mode.
267    
268     41. Add \R to match any Unicode newline sequence, as suggested in the Unicode
269     report.
270    
271     42. Applied patch, originally from Ari Pollak, modified by Google, to allow
272     copy construction and assignment in the C++ wrapper.
273    
274     43. Updated pcregrep to support "--newline=any". In the process, I fixed a
275     couple of bugs that could have given wrong results in the "--newline=crlf"
276     case.
277    
278     44. Added a number of casts and did some reorganization of signed/unsigned int
279     variables following suggestions from Dair Grant. Also renamed the variable
280     "this" as "item" because it is a C++ keyword.
281    
282     45. Arranged for dftables to add
283    
284     #include "pcre_internal.h"
285    
286     to pcre_chartables.c because without it, gcc 4.x may remove the array
287     definition from the final binary if PCRE is built into a static library and
288     dead code stripping is activated.
289    
290     46. For an unanchored pattern, if a match attempt fails at the start of a
291     newline sequence, and the newline setting is CRLF or ANY, and the next two
292     characters are CRLF, advance by two characters instead of one.
293    
294    
295 nigel 91 Version 6.7 04-Jul-06
296     ---------------------
297    
298     1. In order to handle tests when input lines are enormously long, pcretest has
299     been re-factored so that it automatically extends its buffers when
300     necessary. The code is crude, but this _is_ just a test program. The
301     default size has been increased from 32K to 50K.
302    
303     2. The code in pcre_study() was using the value of the re argument before
304     testing it for NULL. (Of course, in any sensible call of the function, it
305     won't be NULL.)
306    
307     3. The memmove() emulation function in pcre_internal.h, which is used on
308     systems that lack both memmove() and bcopy() - that is, hardly ever -
309     was missing a "static" storage class specifier.
310    
311     4. When UTF-8 mode was not set, PCRE looped when compiling certain patterns
312     containing an extended class (one that cannot be represented by a bitmap
313     because it contains high-valued characters or Unicode property items, e.g.
314     [\pZ]). Almost always one would set UTF-8 mode when processing such a
315     pattern, but PCRE should not loop if you do not (it no longer does).
316     [Detail: two cases were found: (a) a repeated subpattern containing an
317     extended class; (b) a recursive reference to a subpattern that followed a
318     previous extended class. It wasn't skipping over the extended class
319     correctly when UTF-8 mode was not set.]
320    
321     5. A negated single-character class was not being recognized as fixed-length
322     in lookbehind assertions such as (?<=[^f]), leading to an incorrect
323     compile error "lookbehind assertion is not fixed length".
324    
325     6. The RunPerlTest auxiliary script was showing an unexpected difference
326     between PCRE and Perl for UTF-8 tests. It turns out that it is hard to
327     write a Perl script that can interpret lines of an input file either as
328     byte characters or as UTF-8, which is what "perltest" was being required to
329     do for the non-UTF-8 and UTF-8 tests, respectively. Essentially what you
330     can't do is switch easily at run time between having the "use utf8;" pragma
331     or not. In the end, I fudged it by using the RunPerlTest script to insert
332     "use utf8;" explicitly for the UTF-8 tests.
333    
334     7. In multiline (/m) mode, PCRE was matching ^ after a terminating newline at
335     the end of the subject string, contrary to the documentation and to what
336     Perl does. This was true of both matching functions. Now it matches only at
337     the start of the subject and immediately after *internal* newlines.
338    
339     8. A call of pcre_fullinfo() from pcretest to get the option bits was passing
340     a pointer to an int instead of a pointer to an unsigned long int. This
341     caused problems on 64-bit systems.
342    
343     9. Applied a patch from the folks at Google to pcrecpp.cc, to fix "another
344     instance of the 'standard' template library not being so standard".
345    
346     10. There was no check on the number of named subpatterns nor the maximum
347     length of a subpattern name. The product of these values is used to compute
348     the size of the memory block for a compiled pattern. By supplying a very
349     long subpattern name and a large number of named subpatterns, the size
350     computation could be caused to overflow. This is now prevented by limiting
351     the length of names to 32 characters, and the number of named subpatterns
352     to 10,000.
353    
354     11. Subpatterns that are repeated with specific counts have to be replicated in
355     the compiled pattern. The size of memory for this was computed from the
356     length of the subpattern and the repeat count. The latter is limited to
357     65535, but there was no limit on the former, meaning that integer overflow
358     could in principle occur. The compiled length of a repeated subpattern is
359     now limited to 30,000 bytes in order to prevent this.
360    
361     12. Added the optional facility to have named substrings with the same name.
362    
363     13. Added the ability to use a named substring as a condition, using the
364     Python syntax: (?(name)yes|no). This overloads (?(R)... and names that
365     are numbers (not recommended). Forward references are permitted.
366    
367     14. Added forward references in named backreferences (if you see what I mean).
368    
369     15. In UTF-8 mode, with the PCRE_DOTALL option set, a quantified dot in the
370     pattern could run off the end of the subject. For example, the pattern
371     "(?s)(.{1,5})"8 did this with the subject "ab".
372    
373     16. If PCRE_DOTALL or PCRE_MULTILINE were set, pcre_dfa_exec() behaved as if
374     PCRE_CASELESS was set when matching characters that were quantified with ?
375     or *.
376    
377     17. A character class other than a single negated character that had a minimum
378     but no maximum quantifier - for example [ab]{6,} - was not handled
379     correctly by pce_dfa_exec(). It would match only one character.
380    
381     18. A valid (though odd) pattern that looked like a POSIX character
382     class but used an invalid character after [ (for example [[,abc,]]) caused
383     pcre_compile() to give the error "Failed: internal error: code overflow" or
384     in some cases to crash with a glibc free() error. This could even happen if
385     the pattern terminated after [[ but there just happened to be a sequence of
386     letters, a binary zero, and a closing ] in the memory that followed.
387    
388     19. Perl's treatment of octal escapes in the range \400 to \777 has changed
389     over the years. Originally (before any Unicode support), just the bottom 8
390     bits were taken. Thus, for example, \500 really meant \100. Nowadays the
391     output from "man perlunicode" includes this:
392    
393     The regular expression compiler produces polymorphic opcodes. That
394     is, the pattern adapts to the data and automatically switches to
395     the Unicode character scheme when presented with Unicode data--or
396     instead uses a traditional byte scheme when presented with byte
397     data.
398    
399     Sadly, a wide octal escape does not cause a switch, and in a string with
400     no other multibyte characters, these octal escapes are treated as before.
401     Thus, in Perl, the pattern /\500/ actually matches \100 but the pattern
402     /\500|\x{1ff}/ matches \500 or \777 because the whole thing is treated as a
403     Unicode string.
404    
405     I have not perpetrated such confusion in PCRE. Up till now, it took just
406     the bottom 8 bits, as in old Perl. I have now made octal escapes with
407     values greater than \377 illegal in non-UTF-8 mode. In UTF-8 mode they
408     translate to the appropriate multibyte character.
409    
410     29. Applied some refactoring to reduce the number of warnings from Microsoft
411     and Borland compilers. This has included removing the fudge introduced
412     seven years ago for the OS/2 compiler (see 2.02/2 below) because it caused
413     a warning about an unused variable.
414    
415     21. PCRE has not included VT (character 0x0b) in the set of whitespace
416     characters since release 4.0, because Perl (from release 5.004) does not.
417     [Or at least, is documented not to: some releases seem to be in conflict
418     with the documentation.] However, when a pattern was studied with
419     pcre_study() and all its branches started with \s, PCRE still included VT
420     as a possible starting character. Of course, this did no harm; it just
421     caused an unnecessary match attempt.
422    
423     22. Removed a now-redundant internal flag bit that recorded the fact that case
424     dependency changed within the pattern. This was once needed for "required
425     byte" processing, but is no longer used. This recovers a now-scarce options
426     bit. Also moved the least significant internal flag bit to the most-
427     significant bit of the word, which was not previously used (hangover from
428     the days when it was an int rather than a uint) to free up another bit for
429     the future.
430    
431     23. Added support for CRLF line endings as well as CR and LF. As well as the
432     default being selectable at build time, it can now be changed at runtime
433     via the PCRE_NEWLINE_xxx flags. There are now options for pcregrep to
434     specify that it is scanning data with non-default line endings.
435    
436     24. Changed the definition of CXXLINK to make it agree with the definition of
437     LINK in the Makefile, by replacing LDFLAGS to CXXFLAGS.
438    
439     25. Applied Ian Taylor's patches to avoid using another stack frame for tail
440     recursions. This makes a big different to stack usage for some patterns.
441    
442     26. If a subpattern containing a named recursion or subroutine reference such
443     as (?P>B) was quantified, for example (xxx(?P>B)){3}, the calculation of
444     the space required for the compiled pattern went wrong and gave too small a
445     value. Depending on the environment, this could lead to "Failed: internal
446     error: code overflow at offset 49" or "glibc detected double free or
447     corruption" errors.
448    
449     27. Applied patches from Google (a) to support the new newline modes and (b) to
450     advance over multibyte UTF-8 characters in GlobalReplace.
451    
452     28. Change free() to pcre_free() in pcredemo.c. Apparently this makes a
453     difference for some implementation of PCRE in some Windows version.
454    
455     29. Added some extra testing facilities to pcretest:
456    
457     \q<number> in a data line sets the "match limit" value
458     \Q<number> in a data line sets the "match recursion limt" value
459     -S <number> sets the stack size, where <number> is in megabytes
460    
461     The -S option isn't available for Windows.
462    
463    
464 nigel 89 Version 6.6 06-Feb-06
465     ---------------------
466    
467     1. Change 16(a) for 6.5 broke things, because PCRE_DATA_SCOPE was not defined
468     in pcreposix.h. I have copied the definition from pcre.h.
469    
470     2. Change 25 for 6.5 broke compilation in a build directory out-of-tree
471     because pcre.h is no longer a built file.
472    
473     3. Added Jeff Friedl's additional debugging patches to pcregrep. These are
474     not normally included in the compiled code.
475    
476    
477 nigel 87 Version 6.5 01-Feb-06
478     ---------------------
479    
480     1. When using the partial match feature with pcre_dfa_exec(), it was not
481     anchoring the second and subsequent partial matches at the new starting
482     point. This could lead to incorrect results. For example, with the pattern
483     /1234/, partially matching against "123" and then "a4" gave a match.
484    
485     2. Changes to pcregrep:
486    
487     (a) All non-match returns from pcre_exec() were being treated as failures
488     to match the line. Now, unless the error is PCRE_ERROR_NOMATCH, an
489     error message is output. Some extra information is given for the
490     PCRE_ERROR_MATCHLIMIT and PCRE_ERROR_RECURSIONLIMIT errors, which are
491     probably the only errors that are likely to be caused by users (by
492     specifying a regex that has nested indefinite repeats, for instance).
493     If there are more than 20 of these errors, pcregrep is abandoned.
494    
495     (b) A binary zero was treated as data while matching, but terminated the
496     output line if it was written out. This has been fixed: binary zeroes
497     are now no different to any other data bytes.
498    
499     (c) Whichever of the LC_ALL or LC_CTYPE environment variables is set is
500     used to set a locale for matching. The --locale=xxxx long option has
501     been added (no short equivalent) to specify a locale explicitly on the
502     pcregrep command, overriding the environment variables.
503    
504     (d) When -B was used with -n, some line numbers in the output were one less
505     than they should have been.
506    
507     (e) Added the -o (--only-matching) option.
508    
509     (f) If -A or -C was used with -c (count only), some lines of context were
510     accidentally printed for the final match.
511    
512     (g) Added the -H (--with-filename) option.
513    
514     (h) The combination of options -rh failed to suppress file names for files
515     that were found from directory arguments.
516    
517     (i) Added the -D (--devices) and -d (--directories) options.
518    
519     (j) Added the -F (--fixed-strings) option.
520    
521     (k) Allow "-" to be used as a file name for -f as well as for a data file.
522    
523     (l) Added the --colo(u)r option.
524    
525     (m) Added Jeffrey Friedl's -S testing option, but within #ifdefs so that it
526     is not present by default.
527    
528     3. A nasty bug was discovered in the handling of recursive patterns, that is,
529     items such as (?R) or (?1), when the recursion could match a number of
530     alternatives. If it matched one of the alternatives, but subsequently,
531     outside the recursion, there was a failure, the code tried to back up into
532     the recursion. However, because of the way PCRE is implemented, this is not
533     possible, and the result was an incorrect result from the match.
534    
535     In order to prevent this happening, the specification of recursion has
536     been changed so that all such subpatterns are automatically treated as
537     atomic groups. Thus, for example, (?R) is treated as if it were (?>(?R)).
538    
539     4. I had overlooked the fact that, in some locales, there are characters for
540     which isalpha() is true but neither isupper() nor islower() are true. In
541     the fr_FR locale, for instance, the \xAA and \xBA characters (ordmasculine
542     and ordfeminine) are like this. This affected the treatment of \w and \W
543     when they appeared in character classes, but not when they appeared outside
544     a character class. The bit map for "word" characters is now created
545     separately from the results of isalnum() instead of just taking it from the
546     upper, lower, and digit maps. (Plus the underscore character, of course.)
547    
548     5. The above bug also affected the handling of POSIX character classes such as
549     [[:alpha:]] and [[:alnum:]]. These do not have their own bit maps in PCRE's
550     permanent tables. Instead, the bit maps for such a class were previously
551     created as the appropriate unions of the upper, lower, and digit bitmaps.
552     Now they are created by subtraction from the [[:word:]] class, which has
553     its own bitmap.
554    
555     6. The [[:blank:]] character class matches horizontal, but not vertical space.
556     It is created by subtracting the vertical space characters (\x09, \x0a,
557     \x0b, \x0c) from the [[:space:]] bitmap. Previously, however, the
558     subtraction was done in the overall bitmap for a character class, meaning
559     that a class such as [\x0c[:blank:]] was incorrect because \x0c would not
560     be recognized. This bug has been fixed.
561    
562     7. Patches from the folks at Google:
563    
564     (a) pcrecpp.cc: "to handle a corner case that may or may not happen in
565     real life, but is still worth protecting against".
566    
567     (b) pcrecpp.cc: "corrects a bug when negative radixes are used with
568     regular expressions".
569    
570     (c) pcre_scanner.cc: avoid use of std::count() because not all systems
571     have it.
572    
573     (d) Split off pcrecpparg.h from pcrecpp.h and had the former built by
574     "configure" and the latter not, in order to fix a problem somebody had
575     with compiling the Arg class on HP-UX.
576    
577     (e) Improve the error-handling of the C++ wrapper a little bit.
578    
579     (f) New tests for checking recursion limiting.
580    
581     8. The pcre_memmove() function, which is used only if the environment does not
582     have a standard memmove() function (and is therefore rarely compiled),
583     contained two bugs: (a) use of int instead of size_t, and (b) it was not
584     returning a result (though PCRE never actually uses the result).
585    
586     9. In the POSIX regexec() interface, if nmatch is specified as a ridiculously
587     large number - greater than INT_MAX/(3*sizeof(int)) - REG_ESPACE is
588     returned instead of calling malloc() with an overflowing number that would
589     most likely cause subsequent chaos.
590    
591     10. The debugging option of pcretest was not showing the NO_AUTO_CAPTURE flag.
592    
593     11. The POSIX flag REG_NOSUB is now supported. When a pattern that was compiled
594     with this option is matched, the nmatch and pmatch options of regexec() are
595     ignored.
596    
597     12. Added REG_UTF8 to the POSIX interface. This is not defined by POSIX, but is
598     provided in case anyone wants to the the POSIX interface with UTF-8
599     strings.
600    
601     13. Added CXXLDFLAGS to the Makefile parameters to provide settings only on the
602     C++ linking (needed for some HP-UX environments).
603    
604     14. Avoid compiler warnings in get_ucpname() when compiled without UCP support
605     (unused parameter) and in the pcre_printint() function (omitted "default"
606     switch label when the default is to do nothing).
607    
608     15. Added some code to make it possible, when PCRE is compiled as a C++
609     library, to replace subject pointers for pcre_exec() with a smart pointer
610     class, thus making it possible to process discontinuous strings.
611    
612     16. The two macros PCRE_EXPORT and PCRE_DATA_SCOPE are confusing, and perform
613     much the same function. They were added by different people who were trying
614     to make PCRE easy to compile on non-Unix systems. It has been suggested
615     that PCRE_EXPORT be abolished now that there is more automatic apparatus
616     for compiling on Windows systems. I have therefore replaced it with
617     PCRE_DATA_SCOPE. This is set automatically for Windows; if not set it
618     defaults to "extern" for C or "extern C" for C++, which works fine on
619     Unix-like systems. It is now possible to override the value of PCRE_DATA_
620     SCOPE with something explicit in config.h. In addition:
621    
622     (a) pcreposix.h still had just "extern" instead of either of these macros;
623     I have replaced it with PCRE_DATA_SCOPE.
624    
625     (b) Functions such as _pcre_xclass(), which are internal to the library,
626     but external in the C sense, all had PCRE_EXPORT in their definitions.
627     This is apparently wrong for the Windows case, so I have removed it.
628     (It makes no difference on Unix-like systems.)
629    
630     17. Added a new limit, MATCH_LIMIT_RECURSION, which limits the depth of nesting
631     of recursive calls to match(). This is different to MATCH_LIMIT because
632     that limits the total number of calls to match(), not all of which increase
633     the depth of recursion. Limiting the recursion depth limits the amount of
634     stack (or heap if NO_RECURSE is set) that is used. The default can be set
635     when PCRE is compiled, and changed at run time. A patch from Google adds
636     this functionality to the C++ interface.
637    
638     18. Changes to the handling of Unicode character properties:
639    
640     (a) Updated the table to Unicode 4.1.0.
641    
642     (b) Recognize characters that are not in the table as "Cn" (undefined).
643    
644     (c) I revised the way the table is implemented to a much improved format
645     which includes recognition of ranges. It now supports the ranges that
646     are defined in UnicodeData.txt, and it also amalgamates other
647     characters into ranges. This has reduced the number of entries in the
648     table from around 16,000 to around 3,000, thus reducing its size
649     considerably. I realized I did not need to use a tree structure after
650     all - a binary chop search is just as efficient. Having reduced the
651     number of entries, I extended their size from 6 bytes to 8 bytes to
652     allow for more data.
653    
654     (d) Added support for Unicode script names via properties such as \p{Han}.
655    
656     19. In UTF-8 mode, a backslash followed by a non-Ascii character was not
657     matching that character.
658    
659     20. When matching a repeated Unicode property with a minimum greater than zero,
660     (for example \pL{2,}), PCRE could look past the end of the subject if it
661     reached it while seeking the minimum number of characters. This could
662     happen only if some of the characters were more than one byte long, because
663     there is a check for at least the minimum number of bytes.
664    
665     21. Refactored the implementation of \p and \P so as to be more general, to
666     allow for more different types of property in future. This has changed the
667     compiled form incompatibly. Anybody with saved compiled patterns that use
668     \p or \P will have to recompile them.
669    
670     22. Added "Any" and "L&" to the supported property types.
671    
672     23. Recognize \x{...} as a code point specifier, even when not in UTF-8 mode,
673     but give a compile time error if the value is greater than 0xff.
674    
675     24. The man pages for pcrepartial, pcreprecompile, and pcre_compile2 were
676     accidentally not being installed or uninstalled.
677    
678     25. The pcre.h file was built from pcre.h.in, but the only changes that were
679     made were to insert the current release number. This seemed silly, because
680     it made things harder for people building PCRE on systems that don't run
681     "configure". I have turned pcre.h into a distributed file, no longer built
682     by "configure", with the version identification directly included. There is
683     no longer a pcre.h.in file.
684    
685     However, this change necessitated a change to the pcre-config script as
686     well. It is built from pcre-config.in, and one of the substitutions was the
687     release number. I have updated configure.ac so that ./configure now finds
688     the release number by grepping pcre.h.
689    
690     26. Added the ability to run the tests under valgrind.
691    
692    
693 nigel 85 Version 6.4 05-Sep-05
694     ---------------------
695    
696     1. Change 6.0/10/(l) to pcregrep introduced a bug that caused separator lines
697     "--" to be printed when multiple files were scanned, even when none of the
698     -A, -B, or -C options were used. This is not compatible with Gnu grep, so I
699     consider it to be a bug, and have restored the previous behaviour.
700    
701     2. A couple of code tidies to get rid of compiler warnings.
702    
703     3. The pcretest program used to cheat by referring to symbols in the library
704     whose names begin with _pcre_. These are internal symbols that are not
705     really supposed to be visible externally, and in some environments it is
706     possible to suppress them. The cheating is now confined to including
707     certain files from the library's source, which is a bit cleaner.
708    
709     4. Renamed pcre.in as pcre.h.in to go with pcrecpp.h.in; it also makes the
710     file's purpose clearer.
711    
712     5. Reorganized pcre_ucp_findchar().
713    
714    
715 nigel 83 Version 6.3 15-Aug-05
716     ---------------------
717    
718     1. The file libpcre.pc.in did not have general read permission in the tarball.
719    
720     2. There were some problems when building without C++ support:
721    
722     (a) If C++ support was not built, "make install" and "make test" still
723     tried to test it.
724    
725     (b) There were problems when the value of CXX was explicitly set. Some
726     changes have been made to try to fix these, and ...
727    
728     (c) --disable-cpp can now be used to explicitly disable C++ support.
729    
730     (d) The use of @CPP_OBJ@ directly caused a blank line preceded by a
731     backslash in a target when C++ was disabled. This confuses some
732     versions of "make", apparently. Using an intermediate variable solves
733     this. (Same for CPP_LOBJ.)
734    
735     3. $(LINK_FOR_BUILD) now includes $(CFLAGS_FOR_BUILD) and $(LINK)
736     (non-Windows) now includes $(CFLAGS) because these flags are sometimes
737     necessary on certain architectures.
738    
739     4. Added a setting of -export-symbols-regex to the link command to remove
740     those symbols that are exported in the C sense, but actually are local
741     within the library, and not documented. Their names all begin with
742     "_pcre_". This is not a perfect job, because (a) we have to except some
743     symbols that pcretest ("illegally") uses, and (b) the facility isn't always
744     available (and never for static libraries). I have made a note to try to
745     find a way round (a) in the future.
746    
747    
748 nigel 81 Version 6.2 01-Aug-05
749     ---------------------
750    
751     1. There was no test for integer overflow of quantifier values. A construction
752     such as {1111111111111111} would give undefined results. What is worse, if
753     a minimum quantifier for a parenthesized subpattern overflowed and became
754     negative, the calculation of the memory size went wrong. This could have
755     led to memory overwriting.
756    
757     2. Building PCRE using VPATH was broken. Hopefully it is now fixed.
758    
759     3. Added "b" to the 2nd argument of fopen() in dftables.c, for non-Unix-like
760     operating environments where this matters.
761    
762     4. Applied Giuseppe Maxia's patch to add additional features for controlling
763     PCRE options from within the C++ wrapper.
764    
765     5. Named capturing subpatterns were not being correctly counted when a pattern
766     was compiled. This caused two problems: (a) If there were more than 100
767     such subpatterns, the calculation of the memory needed for the whole
768     compiled pattern went wrong, leading to an overflow error. (b) Numerical
769     back references of the form \12, where the number was greater than 9, were
770     not recognized as back references, even though there were sufficient
771     previous subpatterns.
772    
773     6. Two minor patches to pcrecpp.cc in order to allow it to compile on older
774     versions of gcc, e.g. 2.95.4.
775    
776    
777 nigel 79 Version 6.1 21-Jun-05
778     ---------------------
779    
780     1. There was one reference to the variable "posix" in pcretest.c that was not
781     surrounded by "#if !defined NOPOSIX".
782    
783     2. Make it possible to compile pcretest without DFA support, UTF8 support, or
784     the cross-check on the old pcre_info() function, for the benefit of the
785     cut-down version of PCRE that is currently imported into Exim.
786    
787     3. A (silly) pattern starting with (?i)(?-i) caused an internal space
788     allocation error. I've done the easy fix, which wastes 2 bytes for sensible
789     patterns that start (?i) but I don't think that matters. The use of (?i) is
790     just an example; this all applies to the other options as well.
791    
792     4. Since libtool seems to echo the compile commands it is issuing, the output
793     from "make" can be reduced a bit by putting "@" in front of each libtool
794     compile command.
795    
796     5. Patch from the folks at Google for configure.in to be a bit more thorough
797     in checking for a suitable C++ installation before trying to compile the
798     C++ stuff. This should fix a reported problem when a compiler was present,
799     but no suitable headers.
800    
801     6. The man pages all had just "PCRE" as their title. I have changed them to
802     be the relevant file name. I have also arranged that these names are
803     retained in the file doc/pcre.txt, which is a concatenation in text format
804     of all the man pages except the little individual ones for each function.
805    
806     7. The NON-UNIX-USE file had not been updated for the different set of source
807     files that come with release 6. I also added a few comments about the C++
808     wrapper.
809    
810    
811 nigel 77 Version 6.0 07-Jun-05
812     ---------------------
813    
814     1. Some minor internal re-organization to help with my DFA experiments.
815    
816     2. Some missing #ifdef SUPPORT_UCP conditionals in pcretest and printint that
817     didn't matter for the library itself when fully configured, but did matter
818     when compiling without UCP support, or within Exim, where the ucp files are
819     not imported.
820    
821     3. Refactoring of the library code to split up the various functions into
822     different source modules. The addition of the new DFA matching code (see
823     below) to a single monolithic source would have made it really too
824     unwieldy, quite apart from causing all the code to be include in a
825     statically linked application, when only some functions are used. This is
826     relevant even without the DFA addition now that patterns can be compiled in
827     one application and matched in another.
828    
829     The downside of splitting up is that there have to be some external
830     functions and data tables that are used internally in different modules of
831     the library but which are not part of the API. These have all had their
832     names changed to start with "_pcre_" so that they are unlikely to clash
833     with other external names.
834    
835     4. Added an alternate matching function, pcre_dfa_exec(), which matches using
836     a different (DFA) algorithm. Although it is slower than the original
837     function, it does have some advantages for certain types of matching
838     problem.
839    
840     5. Upgrades to pcretest in order to test the features of pcre_dfa_exec(),
841     including restarting after a partial match.
842    
843     6. A patch for pcregrep that defines INVALID_FILE_ATTRIBUTES if it is not
844     defined when compiling for Windows was sent to me. I have put it into the
845     code, though I have no means of testing or verifying it.
846    
847     7. Added the pcre_refcount() auxiliary function.
848    
849     8. Added the PCRE_FIRSTLINE option. This constrains an unanchored pattern to
850     match before or at the first newline in the subject string. In pcretest,
851     the /f option on a pattern can be used to set this.
852    
853     9. A repeated \w when used in UTF-8 mode with characters greater than 256
854     would behave wrongly. This has been present in PCRE since release 4.0.
855    
856     10. A number of changes to the pcregrep command:
857    
858     (a) Refactored how -x works; insert ^(...)$ instead of setting
859     PCRE_ANCHORED and checking the length, in preparation for adding
860     something similar for -w.
861    
862     (b) Added the -w (match as a word) option.
863    
864     (c) Refactored the way lines are read and buffered so as to have more
865     than one at a time available.
866    
867     (d) Implemented a pcregrep test script.
868    
869     (e) Added the -M (multiline match) option. This allows patterns to match
870     over several lines of the subject. The buffering ensures that at least
871     8K, or the rest of the document (whichever is the shorter) is available
872     for matching (and similarly the previous 8K for lookbehind assertions).
873    
874     (f) Changed the --help output so that it now says
875    
876     -w, --word-regex(p)
877    
878     instead of two lines, one with "regex" and the other with "regexp"
879     because that confused at least one person since the short forms are the
880     same. (This required a bit of code, as the output is generated
881     automatically from a table. It wasn't just a text change.)
882    
883     (g) -- can be used to terminate pcregrep options if the next thing isn't an
884     option but starts with a hyphen. Could be a pattern or a path name
885     starting with a hyphen, for instance.
886    
887     (h) "-" can be given as a file name to represent stdin.
888    
889     (i) When file names are being printed, "(standard input)" is used for
890     the standard input, for compatibility with GNU grep. Previously
891     "<stdin>" was used.
892    
893     (j) The option --label=xxx can be used to supply a name to be used for
894     stdin when file names are being printed. There is no short form.
895    
896     (k) Re-factored the options decoding logic because we are going to add
897     two more options that take data. Such options can now be given in four
898     different ways, e.g. "-fname", "-f name", "--file=name", "--file name".
899    
900     (l) Added the -A, -B, and -C options for requesting that lines of context
901     around matches be printed.
902    
903     (m) Added the -L option to print the names of files that do not contain
904     any matching lines, that is, the complement of -l.
905    
906     (n) The return code is 2 if any file cannot be opened, but pcregrep does
907     continue to scan other files.
908    
909     (o) The -s option was incorrectly implemented. For compatibility with other
910     greps, it now suppresses the error message for a non-existent or non-
911     accessible file (but not the return code). There is a new option called
912     -q that suppresses the output of matching lines, which was what -s was
913     previously doing.
914    
915     (p) Added --include and --exclude options to specify files for inclusion
916     and exclusion when recursing.
917    
918     11. The Makefile was not using the Autoconf-supported LDFLAGS macro properly.
919     Hopefully, it now does.
920    
921     12. Missing cast in pcre_study().
922    
923     13. Added an "uninstall" target to the makefile.
924    
925     14. Replaced "extern" in the function prototypes in Makefile.in with
926     "PCRE_DATA_SCOPE", which defaults to 'extern' or 'extern "C"' in the Unix
927     world, but is set differently for Windows.
928    
929     15. Added a second compiling function called pcre_compile2(). The only
930     difference is that it has an extra argument, which is a pointer to an
931     integer error code. When there is a compile-time failure, this is set
932     non-zero, in addition to the error test pointer being set to point to an
933     error message. The new argument may be NULL if no error number is required
934     (but then you may as well call pcre_compile(), which is now just a
935     wrapper). This facility is provided because some applications need a
936     numeric error indication, but it has also enabled me to tidy up the way
937     compile-time errors are handled in the POSIX wrapper.
938    
939     16. Added VPATH=.libs to the makefile; this should help when building with one
940     prefix path and installing with another. (Or so I'm told by someone who
941     knows more about this stuff than I do.)
942    
943     17. Added a new option, REG_DOTALL, to the POSIX function regcomp(). This
944     passes PCRE_DOTALL to the pcre_compile() function, making the "." character
945     match everything, including newlines. This is not POSIX-compatible, but
946     somebody wanted the feature. From pcretest it can be activated by using
947     both the P and the s flags.
948    
949     18. AC_PROG_LIBTOOL appeared twice in Makefile.in. Removed one.
950    
951     19. libpcre.pc was being incorrectly installed as executable.
952    
953     20. A couple of places in pcretest check for end-of-line by looking for '\n';
954     it now also looks for '\r' so that it will work unmodified on Windows.
955    
956     21. Added Google's contributed C++ wrapper to the distribution.
957    
958     22. Added some untidy missing memory free() calls in pcretest, to keep
959     Electric Fence happy when testing.
960    
961    
962    
963 nigel 75 Version 5.0 13-Sep-04
964     ---------------------
965    
966     1. Internal change: literal characters are no longer packed up into items
967     containing multiple characters in a single byte-string. Each character
968     is now matched using a separate opcode. However, there may be more than one
969     byte in the character in UTF-8 mode.
970    
971     2. The pcre_callout_block structure has two new fields: pattern_position and
972     next_item_length. These contain the offset in the pattern to the next match
973     item, and its length, respectively.
974    
975     3. The PCRE_AUTO_CALLOUT option for pcre_compile() requests the automatic
976     insertion of callouts before each pattern item. Added the /C option to
977     pcretest to make use of this.
978    
979     4. On the advice of a Windows user, the lines
980    
981     #if defined(_WIN32) || defined(WIN32)
982     _setmode( _fileno( stdout ), 0x8000 );
983     #endif /* defined(_WIN32) || defined(WIN32) */
984    
985     have been added to the source of pcretest. This apparently does useful
986     magic in relation to line terminators.
987    
988     5. Changed "r" and "w" in the calls to fopen() in pcretest to "rb" and "wb"
989     for the benefit of those environments where the "b" makes a difference.
990    
991     6. The icc compiler has the same options as gcc, but "configure" doesn't seem
992     to know about it. I have put a hack into configure.in that adds in code
993     to set GCC=yes if CC=icc. This seems to end up at a point in the
994     generated configure script that is early enough to affect the setting of
995     compiler options, which is what is needed, but I have no means of testing
996     whether it really works. (The user who reported this had patched the
997     generated configure script, which of course I cannot do.)
998    
999     LATER: After change 22 below (new libtool files), the configure script
1000     seems to know about icc (and also ecc). Therefore, I have commented out
1001     this hack in configure.in.
1002    
1003     7. Added support for pkg-config (2 patches were sent in).
1004    
1005     8. Negated POSIX character classes that used a combination of internal tables
1006     were completely broken. These were [[:^alpha:]], [[:^alnum:]], and
1007     [[:^ascii]]. Typically, they would match almost any characters. The other
1008     POSIX classes were not broken in this way.
1009    
1010     9. Matching the pattern "\b.*?" against "ab cd", starting at offset 1, failed
1011     to find the match, as PCRE was deluded into thinking that the match had to
1012     start at the start point or following a newline. The same bug applied to
1013     patterns with negative forward assertions or any backward assertions
1014     preceding ".*" at the start, unless the pattern required a fixed first
1015     character. This was a failing pattern: "(?!.bcd).*". The bug is now fixed.
1016    
1017     10. In UTF-8 mode, when moving forwards in the subject after a failed match
1018     starting at the last subject character, bytes beyond the end of the subject
1019     string were read.
1020    
1021     11. Renamed the variable "class" as "classbits" to make life easier for C++
1022     users. (Previously there was a macro definition, but it apparently wasn't
1023     enough.)
1024    
1025     12. Added the new field "tables" to the extra data so that tables can be passed
1026     in at exec time, or the internal tables can be re-selected. This allows
1027     a compiled regex to be saved and re-used at a later time by a different
1028     program that might have everything at different addresses.
1029    
1030     13. Modified the pcre-config script so that, when run on Solaris, it shows a
1031     -R library as well as a -L library.
1032    
1033     14. The debugging options of pcretest (-d on the command line or D on a
1034     pattern) showed incorrect output for anything following an extended class
1035     that contained multibyte characters and which was followed by a quantifier.
1036    
1037     15. Added optional support for general category Unicode character properties
1038     via the \p, \P, and \X escapes. Unicode property support implies UTF-8
1039     support. It adds about 90K to the size of the library. The meanings of the
1040     inbuilt class escapes such as \d and \s have NOT been changed.
1041    
1042     16. Updated pcredemo.c to include calls to free() to release the memory for the
1043     compiled pattern.
1044    
1045     17. The generated file chartables.c was being created in the source directory
1046     instead of in the building directory. This caused the build to fail if the
1047     source directory was different from the building directory, and was
1048     read-only.
1049    
1050     18. Added some sample Win commands from Mark Tetrode into the NON-UNIX-USE
1051     file. No doubt somebody will tell me if they don't make sense... Also added
1052     Dan Mooney's comments about building on OpenVMS.
1053    
1054     19. Added support for partial matching via the PCRE_PARTIAL option for
1055     pcre_exec() and the \P data escape in pcretest.
1056    
1057     20. Extended pcretest with 3 new pattern features:
1058    
1059     (i) A pattern option of the form ">rest-of-line" causes pcretest to
1060     write the compiled pattern to the file whose name is "rest-of-line".
1061     This is a straight binary dump of the data, with the saved pointer to
1062     the character tables forced to be NULL. The study data, if any, is
1063     written too. After writing, pcretest reads a new pattern.
1064    
1065     (ii) If, instead of a pattern, "<rest-of-line" is given, pcretest reads a
1066     compiled pattern from the given file. There must not be any
1067     occurrences of "<" in the file name (pretty unlikely); if there are,
1068     pcretest will instead treat the initial "<" as a pattern delimiter.
1069     After reading in the pattern, pcretest goes on to read data lines as
1070     usual.
1071    
1072     (iii) The F pattern option causes pcretest to flip the bytes in the 32-bit
1073     and 16-bit fields in a compiled pattern, to simulate a pattern that
1074     was compiled on a host of opposite endianness.
1075    
1076     21. The pcre-exec() function can now cope with patterns that were compiled on
1077     hosts of opposite endianness, with this restriction:
1078    
1079     As for any compiled expression that is saved and used later, the tables
1080     pointer field cannot be preserved; the extra_data field in the arguments
1081     to pcre_exec() should be used to pass in a tables address if a value
1082     other than the default internal tables were used at compile time.
1083    
1084     22. Calling pcre_exec() with a negative value of the "ovecsize" parameter is
1085     now diagnosed as an error. Previously, most of the time, a negative number
1086     would have been treated as zero, but if in addition "ovector" was passed as
1087     NULL, a crash could occur.
1088    
1089     23. Updated the files ltmain.sh, config.sub, config.guess, and aclocal.m4 with
1090     new versions from the libtool 1.5 distribution (the last one is a copy of
1091     a file called libtool.m4). This seems to have fixed the need to patch
1092     "configure" to support Darwin 1.3 (which I used to do). However, I still
1093     had to patch ltmain.sh to ensure that ${SED} is set (it isn't on my
1094     workstation).
1095    
1096     24. Changed the PCRE licence to be the more standard "BSD" licence.
1097    
1098    
1099 nigel 73 Version 4.5 01-Dec-03
1100     ---------------------
1101    
1102     1. There has been some re-arrangement of the code for the match() function so
1103     that it can be compiled in a version that does not call itself recursively.
1104     Instead, it keeps those local variables that need separate instances for
1105     each "recursion" in a frame on the heap, and gets/frees frames whenever it
1106     needs to "recurse". Keeping track of where control must go is done by means
1107     of setjmp/longjmp. The whole thing is implemented by a set of macros that
1108     hide most of the details from the main code, and operates only if
1109     NO_RECURSE is defined while compiling pcre.c. If PCRE is built using the
1110     "configure" mechanism, "--disable-stack-for-recursion" turns on this way of
1111     operating.
1112    
1113     To make it easier for callers to provide specially tailored get/free
1114     functions for this usage, two new functions, pcre_stack_malloc, and
1115     pcre_stack_free, are used. They are always called in strict stacking order,
1116     and the size of block requested is always the same.
1117    
1118     The PCRE_CONFIG_STACKRECURSE info parameter can be used to find out whether
1119     PCRE has been compiled to use the stack or the heap for recursion. The
1120     -C option of pcretest uses this to show which version is compiled.
1121    
1122     A new data escape \S, is added to pcretest; it causes the amounts of store
1123     obtained and freed by both kinds of malloc/free at match time to be added
1124     to the output.
1125    
1126     2. Changed the locale test to use "fr_FR" instead of "fr" because that's
1127     what's available on my current Linux desktop machine.
1128    
1129     3. When matching a UTF-8 string, the test for a valid string at the start has
1130     been extended. If start_offset is not zero, PCRE now checks that it points
1131     to a byte that is the start of a UTF-8 character. If not, it returns
1132     PCRE_ERROR_BADUTF8_OFFSET (-11). Note: the whole string is still checked;
1133     this is necessary because there may be backward assertions in the pattern.
1134     When matching the same subject several times, it may save resources to use
1135     PCRE_NO_UTF8_CHECK on all but the first call if the string is long.
1136    
1137     4. The code for checking the validity of UTF-8 strings has been tightened so
1138     that it rejects (a) strings containing 0xfe or 0xff bytes and (b) strings
1139     containing "overlong sequences".
1140    
1141     5. Fixed a bug (appearing twice) that I could not find any way of exploiting!
1142     I had written "if ((digitab[*p++] && chtab_digit) == 0)" where the "&&"
1143     should have been "&", but it just so happened that all the cases this let
1144     through by mistake were picked up later in the function.
1145    
1146     6. I had used a variable called "isblank" - this is a C99 function, causing
1147     some compilers to warn. To avoid this, I renamed it (as "blankclass").
1148    
1149     7. Cosmetic: (a) only output another newline at the end of pcretest if it is
1150     prompting; (b) run "./pcretest /dev/null" at the start of the test script
1151     so the version is shown; (c) stop "make test" echoing "./RunTest".
1152    
1153     8. Added patches from David Burgess to enable PCRE to run on EBCDIC systems.
1154    
1155     9. The prototype for memmove() for systems that don't have it was using
1156     size_t, but the inclusion of the header that defines size_t was later. I've
1157     moved the #includes for the C headers earlier to avoid this.
1158    
1159     10. Added some adjustments to the code to make it easier to compiler on certain
1160     special systems:
1161    
1162     (a) Some "const" qualifiers were missing.
1163     (b) Added the macro EXPORT before all exported functions; by default this
1164     is defined to be empty.
1165     (c) Changed the dftables auxiliary program (that builds chartables.c) so
1166     that it reads its output file name as an argument instead of writing
1167     to the standard output and assuming this can be redirected.
1168    
1169     11. In UTF-8 mode, if a recursive reference (e.g. (?1)) followed a character
1170     class containing characters with values greater than 255, PCRE compilation
1171     went into a loop.
1172    
1173     12. A recursive reference to a subpattern that was within another subpattern
1174     that had a minimum quantifier of zero caused PCRE to crash. For example,
1175     (x(y(?2))z)? provoked this bug with a subject that got as far as the
1176     recursion. If the recursively-called subpattern itself had a zero repeat,
1177     that was OK.
1178    
1179     13. In pcretest, the buffer for reading a data line was set at 30K, but the
1180     buffer into which it was copied (for escape processing) was still set at
1181     1024, so long lines caused crashes.
1182    
1183     14. A pattern such as /[ab]{1,3}+/ failed to compile, giving the error
1184     "internal error: code overflow...". This applied to any character class
1185     that was followed by a possessive quantifier.
1186    
1187     15. Modified the Makefile to add libpcre.la as a prerequisite for
1188     libpcreposix.la because I was told this is needed for a parallel build to
1189     work.
1190    
1191     16. If a pattern that contained .* following optional items at the start was
1192     studied, the wrong optimizing data was generated, leading to matching
1193     errors. For example, studying /[ab]*.*c/ concluded, erroneously, that any
1194     matching string must start with a or b or c. The correct conclusion for
1195     this pattern is that a match can start with any character.
1196    
1197    
1198 nigel 71 Version 4.4 13-Aug-03
1199     ---------------------
1200    
1201     1. In UTF-8 mode, a character class containing characters with values between
1202     127 and 255 was not handled correctly if the compiled pattern was studied.
1203     In fixing this, I have also improved the studying algorithm for such
1204     classes (slightly).
1205    
1206     2. Three internal functions had redundant arguments passed to them. Removal
1207     might give a very teeny performance improvement.
1208    
1209     3. Documentation bug: the value of the capture_top field in a callout is *one
1210     more than* the number of the hightest numbered captured substring.
1211    
1212     4. The Makefile linked pcretest and pcregrep with -lpcre, which could result
1213     in incorrectly linking with a previously installed version. They now link
1214     explicitly with libpcre.la.
1215    
1216     5. configure.in no longer needs to recognize Cygwin specially.
1217    
1218     6. A problem in pcre.in for Windows platforms is fixed.
1219    
1220     7. If a pattern was successfully studied, and the -d (or /D) flag was given to
1221     pcretest, it used to include the size of the study block as part of its
1222     output. Unfortunately, the structure contains a field that has a different
1223     size on different hardware architectures. This meant that the tests that
1224     showed this size failed. As the block is currently always of a fixed size,
1225     this information isn't actually particularly useful in pcretest output, so
1226     I have just removed it.
1227    
1228     8. Three pre-processor statements accidentally did not start in column 1.
1229     Sadly, there are *still* compilers around that complain, even though
1230     standard C has not required this for well over a decade. Sigh.
1231    
1232     9. In pcretest, the code for checking callouts passed small integers in the
1233     callout_data field, which is a void * field. However, some picky compilers
1234     complained about the casts involved for this on 64-bit systems. Now
1235     pcretest passes the address of the small integer instead, which should get
1236     rid of the warnings.
1237    
1238     10. By default, when in UTF-8 mode, PCRE now checks for valid UTF-8 strings at
1239     both compile and run time, and gives an error if an invalid UTF-8 sequence
1240     is found. There is a option for disabling this check in cases where the
1241     string is known to be correct and/or the maximum performance is wanted.
1242    
1243     11. In response to a bug report, I changed one line in Makefile.in from
1244    
1245     -Wl,--out-implib,.libs/lib@WIN_PREFIX@pcreposix.dll.a \
1246     to
1247     -Wl,--out-implib,.libs/@WIN_PREFIX@libpcreposix.dll.a \
1248    
1249     to look similar to other lines, but I have no way of telling whether this
1250     is the right thing to do, as I do not use Windows. No doubt I'll get told
1251     if it's wrong...
1252    
1253    
1254 nigel 69 Version 4.3 21-May-03
1255     ---------------------
1256    
1257     1. Two instances of @WIN_PREFIX@ omitted from the Windows targets in the
1258     Makefile.
1259    
1260     2. Some refactoring to improve the quality of the code:
1261    
1262     (i) The utf8_table... variables are now declared "const".
1263    
1264     (ii) The code for \cx, which used the "case flipping" table to upper case
1265     lower case letters, now just substracts 32. This is ASCII-specific,
1266     but the whole concept of \cx is ASCII-specific, so it seems
1267     reasonable.
1268    
1269     (iii) PCRE was using its character types table to recognize decimal and
1270     hexadecimal digits in the pattern. This is silly, because it handles
1271     only 0-9, a-f, and A-F, but the character types table is locale-
1272     specific, which means strange things might happen. A private
1273     table is now used for this - though it costs 256 bytes, a table is
1274     much faster than multiple explicit tests. Of course, the standard
1275     character types table is still used for matching digits in subject
1276     strings against \d.
1277    
1278     (iv) Strictly, the identifier ESC_t is reserved by POSIX (all identifiers
1279     ending in _t are). So I've renamed it as ESC_tee.
1280    
1281     3. The first argument for regexec() in the POSIX wrapper should have been
1282     defined as "const".
1283    
1284     4. Changed pcretest to use malloc() for its buffers so that they can be
1285     Electric Fenced for debugging.
1286    
1287     5. There were several places in the code where, in UTF-8 mode, PCRE would try
1288     to read one or more bytes before the start of the subject string. Often this
1289     had no effect on PCRE's behaviour, but in some circumstances it could
1290     provoke a segmentation fault.
1291    
1292     6. A lookbehind at the start of a pattern in UTF-8 mode could also cause PCRE
1293     to try to read one or more bytes before the start of the subject string.
1294    
1295     7. A lookbehind in a pattern matched in non-UTF-8 mode on a PCRE compiled with
1296     UTF-8 support could misbehave in various ways if the subject string
1297     contained bytes with the 0x80 bit set and the 0x40 bit unset in a lookbehind
1298     area. (PCRE was not checking for the UTF-8 mode flag, and trying to move
1299     back over UTF-8 characters.)
1300    
1301    
1302 nigel 67 Version 4.2 14-Apr-03
1303     ---------------------
1304    
1305     1. Typo "#if SUPPORT_UTF8" instead of "#ifdef SUPPORT_UTF8" fixed.
1306    
1307     2. Changes to the building process, supplied by Ronald Landheer-Cieslak
1308     [ON_WINDOWS]: new variable, "#" on non-Windows platforms
1309     [NOT_ON_WINDOWS]: new variable, "#" on Windows platforms
1310     [WIN_PREFIX]: new variable, "cyg" for Cygwin
1311     * Makefile.in: use autoconf substitution for OBJEXT, EXEEXT, BUILD_OBJEXT
1312     and BUILD_EXEEXT
1313     Note: automatic setting of the BUILD variables is not yet working
1314     set CPPFLAGS and BUILD_CPPFLAGS (but don't use yet) - should be used at
1315     compile-time but not at link-time
1316     [LINK]: use for linking executables only
1317     make different versions for Windows and non-Windows
1318     [LINKLIB]: new variable, copy of UNIX-style LINK, used for linking
1319     libraries
1320     [LINK_FOR_BUILD]: new variable
1321     [OBJEXT]: use throughout
1322     [EXEEXT]: use throughout
1323     <winshared>: new target
1324     <wininstall>: new target
1325     <dftables.o>: use native compiler
1326     <dftables>: use native linker
1327     <install>: handle Windows platform correctly
1328     <clean>: ditto
1329     <check>: ditto
1330     copy DLL to top builddir before testing
1331    
1332     As part of these changes, -no-undefined was removed again. This was reported
1333     to give trouble on HP-UX 11.0, so getting rid of it seems like a good idea
1334     in any case.
1335    
1336     3. Some tidies to get rid of compiler warnings:
1337    
1338     . In the match_data structure, match_limit was an unsigned long int, whereas
1339     match_call_count was an int. I've made them both unsigned long ints.
1340    
1341     . In pcretest the fact that a const uschar * doesn't automatically cast to
1342     a void * provoked a warning.
1343    
1344     . Turning on some more compiler warnings threw up some "shadow" variables
1345     and a few more missing casts.
1346    
1347     4. If PCRE was complied with UTF-8 support, but called without the PCRE_UTF8
1348     option, a class that contained a single character with a value between 128
1349     and 255 (e.g. /[\xFF]/) caused PCRE to crash.
1350    
1351     5. If PCRE was compiled with UTF-8 support, but called without the PCRE_UTF8
1352     option, a class that contained several characters, but with at least one
1353     whose value was between 128 and 255 caused PCRE to crash.
1354    
1355    
1356 nigel 65 Version 4.1 12-Mar-03
1357     ---------------------
1358 nigel 63
1359 nigel 65 1. Compiling with gcc -pedantic found a couple of places where casts were
1360     needed, and a string in dftables.c that was longer than standard compilers are
1361     required to support.
1362    
1363     2. Compiling with Sun's compiler found a few more places where the code could
1364     be tidied up in order to avoid warnings.
1365    
1366     3. The variables for cross-compiling were called HOST_CC and HOST_CFLAGS; the
1367     first of these names is deprecated in the latest Autoconf in favour of the name
1368     CC_FOR_BUILD, because "host" is typically used to mean the system on which the
1369     compiled code will be run. I can't find a reference for HOST_CFLAGS, but by
1370     analogy I have changed it to CFLAGS_FOR_BUILD.
1371    
1372     4. Added -no-undefined to the linking command in the Makefile, because this is
1373     apparently helpful for Windows. To make it work, also added "-L. -lpcre" to the
1374     linking step for the pcreposix library.
1375    
1376     5. PCRE was failing to diagnose the case of two named groups with the same
1377     name.
1378    
1379     6. A problem with one of PCRE's optimizations was discovered. PCRE remembers a
1380     literal character that is needed in the subject for a match, and scans along to
1381     ensure that it is present before embarking on the full matching process. This
1382     saves time in cases of nested unlimited repeats that are never going to match.
1383     Problem: the scan can take a lot of time if the subject is very long (e.g.
1384     megabytes), thus penalizing straightforward matches. It is now done only if the
1385     amount of subject to be scanned is less than 1000 bytes.
1386    
1387     7. A lesser problem with the same optimization is that it was recording the
1388     first character of an anchored pattern as "needed", thus provoking a search
1389     right along the subject, even when the first match of the pattern was going to
1390     fail. The "needed" character is now not set for anchored patterns, unless it
1391     follows something in the pattern that is of non-fixed length. Thus, it still
1392     fulfils its original purpose of finding quick non-matches in cases of nested
1393     unlimited repeats, but isn't used for simple anchored patterns such as /^abc/.
1394    
1395    
1396     Version 4.0 17-Feb-03
1397     ---------------------
1398    
1399 nigel 63 1. If a comment in an extended regex that started immediately after a meta-item
1400     extended to the end of string, PCRE compiled incorrect data. This could lead to
1401     all kinds of weird effects. Example: /#/ was bad; /()#/ was bad; /a#/ was not.
1402    
1403     2. Moved to autoconf 2.53 and libtool 1.4.2.
1404    
1405     3. Perl 5.8 no longer needs "use utf8" for doing UTF-8 things. Consequently,
1406     the special perltest8 script is no longer needed - all the tests can be run
1407     from a single perltest script.
1408    
1409     4. From 5.004, Perl has not included the VT character (0x0b) in the set defined
1410     by \s. It has now been removed in PCRE. This means it isn't recognized as
1411     whitespace in /x regexes too, which is the same as Perl. Note that the POSIX
1412     class [:space:] *does* include VT, thereby creating a mess.
1413    
1414     5. Added the class [:blank:] (a GNU extension from Perl 5.8) to match only
1415     space and tab.
1416    
1417     6. Perl 5.005 was a long time ago. It's time to amalgamate the tests that use
1418     its new features into the main test script, reducing the number of scripts.
1419    
1420     7. Perl 5.8 has changed the meaning of patterns like /a(?i)b/. Earlier versions
1421     were backward compatible, and made the (?i) apply to the whole pattern, as if
1422     /i were given. Now it behaves more logically, and applies the option setting
1423     only to what follows. PCRE has been changed to follow suit. However, if it
1424     finds options settings right at the start of the pattern, it extracts them into
1425     the global options, as before. Thus, they show up in the info data.
1426    
1427     8. Added support for the \Q...\E escape sequence. Characters in between are
1428     treated as literals. This is slightly different from Perl in that $ and @ are
1429     also handled as literals inside the quotes. In Perl, they will cause variable
1430     interpolation. Note the following examples:
1431    
1432     Pattern PCRE matches Perl matches
1433    
1434     \Qabc$xyz\E abc$xyz abc followed by the contents of $xyz
1435     \Qabc\$xyz\E abc\$xyz abc\$xyz
1436     \Qabc\E\$\Qxyz\E abc$xyz abc$xyz
1437    
1438     For compatibility with Perl, \Q...\E sequences are recognized inside character
1439     classes as well as outside them.
1440    
1441     9. Re-organized 3 code statements in pcretest to avoid "overflow in
1442     floating-point constant arithmetic" warnings from a Microsoft compiler. Added a
1443     (size_t) cast to one statement in pcretest and one in pcreposix to avoid
1444     signed/unsigned warnings.
1445    
1446     10. SunOS4 doesn't have strtoul(). This was used only for unpicking the -o
1447     option for pcretest, so I've replaced it by a simple function that does just
1448     that job.
1449    
1450     11. pcregrep was ending with code 0 instead of 2 for the commands "pcregrep" or
1451     "pcregrep -".
1452    
1453     12. Added "possessive quantifiers" ?+, *+, ++, and {,}+ which come from Sun's
1454     Java package. This provides some syntactic sugar for simple cases of what my
1455     documentation calls "once-only subpatterns". A pattern such as x*+ is the same
1456     as (?>x*). In other words, if what is inside (?>...) is just a single repeated
1457     item, you can use this simplified notation. Note that only makes sense with
1458     greedy quantifiers. Consequently, the use of the possessive quantifier forces
1459     greediness, whatever the setting of the PCRE_UNGREEDY option.
1460    
1461     13. A change of greediness default within a pattern was not taking effect at
1462     the current level for patterns like /(b+(?U)a+)/. It did apply to parenthesized
1463     subpatterns that followed. Patterns like /b+(?U)a+/ worked because the option
1464     was abstracted outside.
1465    
1466     14. PCRE now supports the \G assertion. It is true when the current matching
1467     position is at the start point of the match. This differs from \A when the
1468     starting offset is non-zero. Used with the /g option of pcretest (or similar
1469     code), it works in the same way as it does for Perl's /g option. If all
1470     alternatives of a regex begin with \G, the expression is anchored to the start
1471     match position, and the "anchored" flag is set in the compiled expression.
1472    
1473     15. Some bugs concerning the handling of certain option changes within patterns
1474     have been fixed. These applied to options other than (?ims). For example,
1475     "a(?x: b c )d" did not match "XabcdY" but did match "Xa b c dY". It should have
1476     been the other way round. Some of this was related to change 7 above.
1477    
1478     16. PCRE now gives errors for /[.x.]/ and /[=x=]/ as unsupported POSIX
1479     features, as Perl does. Previously, PCRE gave the warnings only for /[[.x.]]/
1480     and /[[=x=]]/. PCRE now also gives an error for /[:name:]/ because it supports
1481     POSIX classes only within a class (e.g. /[[:alpha:]]/).
1482    
1483     17. Added support for Perl's \C escape. This matches one byte, even in UTF8
1484     mode. Unlike ".", it always matches newline, whatever the setting of
1485     PCRE_DOTALL. However, PCRE does not permit \C to appear in lookbehind
1486     assertions. Perl allows it, but it doesn't (in general) work because it can't
1487     calculate the length of the lookbehind. At least, that's the case for Perl
1488     5.8.0 - I've been told they are going to document that it doesn't work in
1489     future.
1490    
1491     18. Added an error diagnosis for escapes that PCRE does not support: these are
1492     \L, \l, \N, \P, \p, \U, \u, and \X.
1493    
1494     19. Although correctly diagnosing a missing ']' in a character class, PCRE was
1495     reading past the end of the pattern in cases such as /[abcd/.
1496    
1497     20. PCRE was getting more memory than necessary for patterns with classes that
1498     contained both POSIX named classes and other characters, e.g. /[[:space:]abc/.
1499    
1500     21. Added some code, conditional on #ifdef VPCOMPAT, to make life easier for
1501     compiling PCRE for use with Virtual Pascal.
1502    
1503     22. Small fix to the Makefile to make it work properly if the build is done
1504     outside the source tree.
1505    
1506     23. Added a new extension: a condition to go with recursion. If a conditional
1507     subpattern starts with (?(R) the "true" branch is used if recursion has
1508     happened, whereas the "false" branch is used only at the top level.
1509    
1510     24. When there was a very long string of literal characters (over 255 bytes
1511     without UTF support, over 250 bytes with UTF support), the computation of how
1512     much memory was required could be incorrect, leading to segfaults or other
1513     strange effects.
1514    
1515     25. PCRE was incorrectly assuming anchoring (either to start of subject or to
1516     start of line for a non-DOTALL pattern) when a pattern started with (.*) and
1517     there was a subsequent back reference to those brackets. This meant that, for
1518     example, /(.*)\d+\1/ failed to match "abc123bc". Unfortunately, it isn't
1519     possible to check for precisely this case. All we can do is abandon the
1520     optimization if .* occurs inside capturing brackets when there are any back
1521     references whatsoever. (See below for a better fix that came later.)
1522    
1523     26. The handling of the optimization for finding the first character of a
1524     non-anchored pattern, and for finding a character that is required later in the
1525     match were failing in some cases. This didn't break the matching; it just
1526     failed to optimize when it could. The way this is done has been re-implemented.
1527    
1528     27. Fixed typo in error message for invalid (?R item (it said "(?p").
1529    
1530     28. Added a new feature that provides some of the functionality that Perl
1531     provides with (?{...}). The facility is termed a "callout". The way it is done
1532     in PCRE is for the caller to provide an optional function, by setting
1533     pcre_callout to its entry point. Like pcre_malloc and pcre_free, this is a
1534     global variable. By default it is unset, which disables all calling out. To get
1535     the function called, the regex must include (?C) at appropriate points. This
1536     is, in fact, equivalent to (?C0), and any number <= 255 may be given with (?C).
1537     This provides a means of identifying different callout points. When PCRE
1538     reaches such a point in the regex, if pcre_callout has been set, the external
1539     function is called. It is provided with data in a structure called
1540     pcre_callout_block, which is defined in pcre.h. If the function returns 0,
1541     matching continues; if it returns a non-zero value, the match at the current
1542     point fails. However, backtracking will occur if possible. [This was changed
1543     later and other features added - see item 49 below.]
1544    
1545     29. pcretest is upgraded to test the callout functionality. It provides a
1546     callout function that displays information. By default, it shows the start of
1547     the match and the current position in the text. There are some new data escapes
1548     to vary what happens:
1549    
1550     \C+ in addition, show current contents of captured substrings
1551     \C- do not supply a callout function
1552     \C!n return 1 when callout number n is reached
1553     \C!n!m return 1 when callout number n is reached for the mth time
1554    
1555     30. If pcregrep was called with the -l option and just a single file name, it
1556     output "<stdin>" if a match was found, instead of the file name.
1557    
1558     31. Improve the efficiency of the POSIX API to PCRE. If the number of capturing
1559     slots is less than POSIX_MALLOC_THRESHOLD, use a block on the stack to pass to
1560     pcre_exec(). This saves a malloc/free per call. The default value of
1561     POSIX_MALLOC_THRESHOLD is 10; it can be changed by --with-posix-malloc-threshold
1562     when configuring.
1563    
1564     32. The default maximum size of a compiled pattern is 64K. There have been a
1565     few cases of people hitting this limit. The code now uses macros to handle the
1566     storing of links as offsets within the compiled pattern. It defaults to 2-byte
1567     links, but this can be changed to 3 or 4 bytes by --with-link-size when
1568     configuring. Tests 2 and 5 work only with 2-byte links because they output
1569     debugging information about compiled patterns.
1570    
1571     33. Internal code re-arrangements:
1572    
1573     (a) Moved the debugging function for printing out a compiled regex into
1574     its own source file (printint.c) and used #include to pull it into
1575     pcretest.c and, when DEBUG is defined, into pcre.c, instead of having two
1576     separate copies.
1577    
1578     (b) Defined the list of op-code names for debugging as a macro in
1579     internal.h so that it is next to the definition of the opcodes.
1580    
1581     (c) Defined a table of op-code lengths for simpler skipping along compiled
1582     code. This is again a macro in internal.h so that it is next to the
1583     definition of the opcodes.
1584    
1585     34. Added support for recursive calls to individual subpatterns, along the
1586     lines of Robin Houston's patch (but implemented somewhat differently).
1587    
1588     35. Further mods to the Makefile to help Win32. Also, added code to pcregrep to
1589     allow it to read and process whole directories in Win32. This code was
1590     contributed by Lionel Fourquaux; it has not been tested by me.
1591    
1592     36. Added support for named subpatterns. The Python syntax (?P<name>...) is
1593     used to name a group. Names consist of alphanumerics and underscores, and must
1594     be unique. Back references use the syntax (?P=name) and recursive calls use
1595     (?P>name) which is a PCRE extension to the Python extension. Groups still have
1596     numbers. The function pcre_fullinfo() can be used after compilation to extract
1597     a name/number map. There are three relevant calls:
1598    
1599     PCRE_INFO_NAMEENTRYSIZE yields the size of each entry in the map
1600     PCRE_INFO_NAMECOUNT yields the number of entries
1601     PCRE_INFO_NAMETABLE yields a pointer to the map.
1602    
1603     The map is a vector of fixed-size entries. The size of each entry depends on
1604     the length of the longest name used. The first two bytes of each entry are the
1605     group number, most significant byte first. There follows the corresponding
1606     name, zero terminated. The names are in alphabetical order.
1607    
1608     37. Make the maximum literal string in the compiled code 250 for the non-UTF-8
1609     case instead of 255. Making it the same both with and without UTF-8 support
1610     means that the same test output works with both.
1611    
1612     38. There was a case of malloc(0) in the POSIX testing code in pcretest. Avoid
1613     calling malloc() with a zero argument.
1614    
1615     39. Change 25 above had to resort to a heavy-handed test for the .* anchoring
1616     optimization. I've improved things by keeping a bitmap of backreferences with
1617     numbers 1-31 so that if .* occurs inside capturing brackets that are not in
1618     fact referenced, the optimization can be applied. It is unlikely that a
1619     relevant occurrence of .* (i.e. one which might indicate anchoring or forcing
1620     the match to follow \n) will appear inside brackets with a number greater than
1621     31, but if it does, any back reference > 31 suppresses the optimization.
1622    
1623     40. Added a new compile-time option PCRE_NO_AUTO_CAPTURE. This has the effect
1624     of disabling numbered capturing parentheses. Any opening parenthesis that is
1625     not followed by ? behaves as if it were followed by ?: but named parentheses
1626     can still be used for capturing (and they will acquire numbers in the usual
1627     way).
1628    
1629     41. Redesigned the return codes from the match() function into yes/no/error so
1630     that errors can be passed back from deep inside the nested calls. A malloc
1631     failure while inside a recursive subpattern call now causes the
1632     PCRE_ERROR_NOMEMORY return instead of quietly going wrong.
1633    
1634     42. It is now possible to set a limit on the number of times the match()
1635     function is called in a call to pcre_exec(). This facility makes it possible to
1636     limit the amount of recursion and backtracking, though not in a directly
1637     obvious way, because the match() function is used in a number of different
1638     circumstances. The count starts from zero for each position in the subject
1639     string (for non-anchored patterns). The default limit is, for compatibility, a
1640     large number, namely 10 000 000. You can change this in two ways:
1641    
1642     (a) When configuring PCRE before making, you can use --with-match-limit=n
1643     to set a default value for the compiled library.
1644    
1645     (b) For each call to pcre_exec(), you can pass a pcre_extra block in which
1646     a different value is set. See 45 below.
1647    
1648     If the limit is exceeded, pcre_exec() returns PCRE_ERROR_MATCHLIMIT.
1649    
1650     43. Added a new function pcre_config(int, void *) to enable run-time extraction
1651     of things that can be changed at compile time. The first argument specifies
1652     what is wanted and the second points to where the information is to be placed.
1653     The current list of available information is:
1654    
1655     PCRE_CONFIG_UTF8
1656    
1657     The output is an integer that is set to one if UTF-8 support is available;
1658     otherwise it is set to zero.
1659    
1660     PCRE_CONFIG_NEWLINE
1661    
1662     The output is an integer that it set to the value of the code that is used for
1663     newline. It is either LF (10) or CR (13).
1664    
1665     PCRE_CONFIG_LINK_SIZE
1666    
1667     The output is an integer that contains the number of bytes used for internal
1668     linkage in compiled expressions. The value is 2, 3, or 4. See item 32 above.
1669    
1670     PCRE_CONFIG_POSIX_MALLOC_THRESHOLD
1671    
1672     The output is an integer that contains the threshold above which the POSIX
1673     interface uses malloc() for output vectors. See item 31 above.
1674    
1675     PCRE_CONFIG_MATCH_LIMIT
1676    
1677     The output is an unsigned integer that contains the default limit of the number
1678     of match() calls in a pcre_exec() execution. See 42 above.
1679    
1680     44. pcretest has been upgraded by the addition of the -C option. This causes it
1681     to extract all the available output from the new pcre_config() function, and to
1682     output it. The program then exits immediately.
1683    
1684     45. A need has arisen to pass over additional data with calls to pcre_exec() in
1685     order to support additional features. One way would have been to define
1686     pcre_exec2() (for example) with extra arguments, but this would not have been
1687     extensible, and would also have required all calls to the original function to
1688     be mapped to the new one. Instead, I have chosen to extend the mechanism that
1689     is used for passing in "extra" data from pcre_study().
1690    
1691     The pcre_extra structure is now exposed and defined in pcre.h. It currently
1692     contains the following fields:
1693    
1694     flags a bitmap indicating which of the following fields are set
1695     study_data opaque data from pcre_study()
1696     match_limit a way of specifying a limit on match() calls for a specific
1697     call to pcre_exec()
1698     callout_data data for callouts (see 49 below)
1699    
1700     The flag bits are also defined in pcre.h, and are
1701    
1702     PCRE_EXTRA_STUDY_DATA
1703     PCRE_EXTRA_MATCH_LIMIT
1704     PCRE_EXTRA_CALLOUT_DATA
1705    
1706     The pcre_study() function now returns one of these new pcre_extra blocks, with
1707     the actual study data pointed to by the study_data field, and the
1708     PCRE_EXTRA_STUDY_DATA flag set. This can be passed directly to pcre_exec() as
1709     before. That is, this change is entirely upwards-compatible and requires no
1710     change to existing code.
1711    
1712     If you want to pass in additional data to pcre_exec(), you can either place it
1713     in a pcre_extra block provided by pcre_study(), or create your own pcre_extra
1714     block.
1715    
1716     46. pcretest has been extended to test the PCRE_EXTRA_MATCH_LIMIT feature. If a
1717     data string contains the escape sequence \M, pcretest calls pcre_exec() several
1718     times with different match limits, until it finds the minimum value needed for
1719     pcre_exec() to complete. The value is then output. This can be instructive; for
1720     most simple matches the number is quite small, but for pathological cases it
1721     gets very large very quickly.
1722    
1723     47. There's a new option for pcre_fullinfo() called PCRE_INFO_STUDYSIZE. It
1724     returns the size of the data block pointed to by the study_data field in a
1725     pcre_extra block, that is, the value that was passed as the argument to
1726     pcre_malloc() when PCRE was getting memory in which to place the information
1727     created by pcre_study(). The fourth argument should point to a size_t variable.
1728     pcretest has been extended so that this information is shown after a successful
1729     pcre_study() call when information about the compiled regex is being displayed.
1730    
1731     48. Cosmetic change to Makefile: there's no need to have / after $(DESTDIR)
1732     because what follows is always an absolute path. (Later: it turns out that this
1733     is more than cosmetic for MinGW, because it doesn't like empty path
1734     components.)
1735    
1736     49. Some changes have been made to the callout feature (see 28 above):
1737    
1738     (i) A callout function now has three choices for what it returns:
1739    
1740     0 => success, carry on matching
1741     > 0 => failure at this point, but backtrack if possible
1742     < 0 => serious error, return this value from pcre_exec()
1743    
1744     Negative values should normally be chosen from the set of PCRE_ERROR_xxx
1745     values. In particular, returning PCRE_ERROR_NOMATCH forces a standard
1746     "match failed" error. The error number PCRE_ERROR_CALLOUT is reserved for
1747     use by callout functions. It will never be used by PCRE itself.
1748    
1749     (ii) The pcre_extra structure (see 45 above) has a void * field called
1750     callout_data, with corresponding flag bit PCRE_EXTRA_CALLOUT_DATA. The
1751     pcre_callout_block structure has a field of the same name. The contents of
1752     the field passed in the pcre_extra structure are passed to the callout
1753     function in the corresponding field in the callout block. This makes it
1754     easier to use the same callout-containing regex from multiple threads. For
1755     testing, the pcretest program has a new data escape
1756    
1757     \C*n pass the number n (may be negative) as callout_data
1758    
1759     If the callout function in pcretest receives a non-zero value as
1760     callout_data, it returns that value.
1761    
1762     50. Makefile wasn't handling CFLAGS properly when compiling dftables. Also,
1763     there were some redundant $(CFLAGS) in commands that are now specified as
1764     $(LINK), which already includes $(CFLAGS).
1765    
1766     51. Extensions to UTF-8 support are listed below. These all apply when (a) PCRE
1767     has been compiled with UTF-8 support *and* pcre_compile() has been compiled
1768     with the PCRE_UTF8 flag. Patterns that are compiled without that flag assume
1769     one-byte characters throughout. Note that case-insensitive matching applies
1770     only to characters whose values are less than 256. PCRE doesn't support the
1771     notion of cases for higher-valued characters.
1772    
1773     (i) A character class whose characters are all within 0-255 is handled as
1774     a bit map, and the map is inverted for negative classes. Previously, a
1775     character > 255 always failed to match such a class; however it should
1776     match if the class was a negative one (e.g. [^ab]). This has been fixed.
1777    
1778     (ii) A negated character class with a single character < 255 is coded as
1779     "not this character" (OP_NOT). This wasn't working properly when the test
1780     character was multibyte, either singly or repeated.
1781    
1782     (iii) Repeats of multibyte characters are now handled correctly in UTF-8
1783     mode, for example: \x{100}{2,3}.
1784    
1785     (iv) The character escapes \b, \B, \d, \D, \s, \S, \w, and \W (either
1786     singly or repeated) now correctly test multibyte characters. However,
1787     PCRE doesn't recognize any characters with values greater than 255 as
1788     digits, spaces, or word characters. Such characters always match \D, \S,
1789     and \W, and never match \d, \s, or \w.
1790    
1791     (v) Classes may now contain characters and character ranges with values
1792     greater than 255. For example: [ab\x{100}-\x{400}].
1793    
1794     (vi) pcregrep now has a --utf-8 option (synonym -u) which makes it call
1795     PCRE in UTF-8 mode.
1796    
1797     52. The info request value PCRE_INFO_FIRSTCHAR has been renamed
1798     PCRE_INFO_FIRSTBYTE because it is a byte value. However, the old name is
1799     retained for backwards compatibility. (Note that LASTLITERAL is also a byte
1800     value.)
1801    
1802     53. The single man page has become too large. I have therefore split it up into
1803     a number of separate man pages. These also give rise to individual HTML pages;
1804     these are now put in a separate directory, and there is an index.html page that
1805     lists them all. Some hyperlinking between the pages has been installed.
1806    
1807     54. Added convenience functions for handling named capturing parentheses.
1808    
1809     55. Unknown escapes inside character classes (e.g. [\M]) and escapes that
1810     aren't interpreted therein (e.g. [\C]) are literals in Perl. This is now also
1811     true in PCRE, except when the PCRE_EXTENDED option is set, in which case they
1812     are faulted.
1813    
1814     56. Introduced HOST_CC and HOST_CFLAGS which can be set in the environment when
1815     calling configure. These values are used when compiling the dftables.c program
1816     which is run to generate the source of the default character tables. They
1817     default to the values of CC and CFLAGS. If you are cross-compiling PCRE,
1818     you will need to set these values.
1819    
1820     57. Updated the building process for Windows DLL, as provided by Fred Cox.
1821    
1822    
1823     Version 3.9 02-Jan-02
1824 nigel 61 ---------------------
1825    
1826     1. A bit of extraneous text had somehow crept into the pcregrep documentation.
1827    
1828     2. If --disable-static was given, the building process failed when trying to
1829     build pcretest and pcregrep. (For some reason it was using libtool to compile
1830     them, which is not right, as they aren't part of the library.)
1831    
1832    
1833 nigel 59 Version 3.8 18-Dec-01
1834     ---------------------
1835    
1836     1. The experimental UTF-8 code was completely screwed up. It was packing the
1837     bytes in the wrong order. How dumb can you get?
1838    
1839    
1840 nigel 57 Version 3.7 29-Oct-01
1841     ---------------------
1842    
1843     1. In updating pcretest to check change 1 of version 3.6, I screwed up.
1844     This caused pcretest, when used on the test data, to segfault. Unfortunately,
1845     this didn't happen under Solaris 8, where I normally test things.
1846    
1847 nigel 59 2. The Makefile had to be changed to make it work on BSD systems, where 'make'
1848     doesn't seem to recognize that ./xxx and xxx are the same file. (This entry
1849     isn't in ChangeLog distributed with 3.7 because I forgot when I hastily made
1850     this fix an hour or so after the initial 3.7 release.)
1851 nigel 57
1852 nigel 59
1853 nigel 55 Version 3.6 23-Oct-01
1854     ---------------------
1855    
1856     1. Crashed with /(sens|respons)e and \1ibility/ and "sense and sensibility" if
1857     offsets passed as NULL with zero offset count.
1858    
1859     2. The config.guess and config.sub files had not been updated when I moved to
1860     the latest autoconf.
1861    
1862    
1863 nigel 53 Version 3.5 15-Aug-01
1864     ---------------------
1865 nigel 5
1866 nigel 53 1. Added some missing #if !defined NOPOSIX conditionals in pcretest.c that
1867     had been forgotten.
1868    
1869     2. By using declared but undefined structures, we can avoid using "void"
1870     definitions in pcre.h while keeping the internal definitions of the structures
1871     private.
1872    
1873     3. The distribution is now built using autoconf 2.50 and libtool 1.4. From a
1874     user point of view, this means that both static and shared libraries are built
1875     by default, but this can be individually controlled. More of the work of
1876     handling this static/shared cases is now inside libtool instead of PCRE's make
1877     file.
1878    
1879     4. The pcretest utility is now installed along with pcregrep because it is
1880     useful for users (to test regexs) and by doing this, it automatically gets
1881     relinked by libtool. The documentation has been turned into a man page, so
1882     there are now .1, .txt, and .html versions in /doc.
1883    
1884     5. Upgrades to pcregrep:
1885     (i) Added long-form option names like gnu grep.
1886     (ii) Added --help to list all options with an explanatory phrase.
1887     (iii) Added -r, --recursive to recurse into sub-directories.
1888     (iv) Added -f, --file to read patterns from a file.
1889    
1890     6. pcre_exec() was referring to its "code" argument before testing that
1891     argument for NULL (and giving an error if it was NULL).
1892    
1893     7. Upgraded Makefile.in to allow for compiling in a different directory from
1894     the source directory.
1895    
1896     8. Tiny buglet in pcretest: when pcre_fullinfo() was called to retrieve the
1897     options bits, the pointer it was passed was to an int instead of to an unsigned
1898     long int. This mattered only on 64-bit systems.
1899    
1900     9. Fixed typo (3.4/1) in pcre.h again. Sigh. I had changed pcre.h (which is
1901     generated) instead of pcre.in, which it its source. Also made the same change
1902     in several of the .c files.
1903    
1904     10. A new release of gcc defines printf() as a macro, which broke pcretest
1905     because it had an ifdef in the middle of a string argument for printf(). Fixed
1906     by using separate calls to printf().
1907    
1908     11. Added --enable-newline-is-cr and --enable-newline-is-lf to the configure
1909     script, to force use of CR or LF instead of \n in the source. On non-Unix
1910     systems, the value can be set in config.h.
1911    
1912     12. The limit of 200 on non-capturing parentheses is a _nesting_ limit, not an
1913     absolute limit. Changed the text of the error message to make this clear, and
1914     likewise updated the man page.
1915    
1916     13. The limit of 99 on the number of capturing subpatterns has been removed.
1917     The new limit is 65535, which I hope will not be a "real" limit.
1918    
1919    
1920 nigel 51 Version 3.4 22-Aug-00
1921     ---------------------
1922    
1923     1. Fixed typo in pcre.h: unsigned const char * changed to const unsigned char *.
1924    
1925     2. Diagnose condition (?(0) as an error instead of crashing on matching.
1926    
1927    
1928 nigel 49 Version 3.3 01-Aug-00
1929     ---------------------
1930    
1931     1. If an octal character was given, but the value was greater than \377, it
1932     was not getting masked to the least significant bits, as documented. This could
1933     lead to crashes in some systems.
1934    
1935     2. Perl 5.6 (if not earlier versions) accepts classes like [a-\d] and treats
1936     the hyphen as a literal. PCRE used to give an error; it now behaves like Perl.
1937    
1938     3. Added the functions pcre_free_substring() and pcre_free_substring_list().
1939     These just pass their arguments on to (pcre_free)(), but they are provided
1940     because some uses of PCRE bind it to non-C systems that can call its functions,
1941     but cannot call free() or pcre_free() directly.
1942    
1943     4. Add "make test" as a synonym for "make check". Corrected some comments in
1944     the Makefile.
1945    
1946     5. Add $(DESTDIR)/ in front of all the paths in the "install" target in the
1947     Makefile.
1948    
1949     6. Changed the name of pgrep to pcregrep, because Solaris has introduced a
1950     command called pgrep for grepping around the active processes.
1951    
1952     7. Added the beginnings of support for UTF-8 character strings.
1953    
1954     8. Arranged for the Makefile to pass over the settings of CC, CFLAGS, and
1955     RANLIB to ./ltconfig so that they are used by libtool. I think these are all
1956     the relevant ones. (AR is not passed because ./ltconfig does its own figuring
1957     out for the ar command.)
1958    
1959    
1960 nigel 47 Version 3.2 12-May-00
1961     ---------------------
1962    
1963     This is purely a bug fixing release.
1964    
1965     1. If the pattern /((Z)+|A)*/ was matched agained ZABCDEFG it matched Z instead
1966     of ZA. This was just one example of several cases that could provoke this bug,
1967     which was introduced by change 9 of version 2.00. The code for breaking
1968     infinite loops after an iteration that matches an empty string was't working
1969     correctly.
1970    
1971     2. The pcretest program was not imitating Perl correctly for the pattern /a*/g
1972     when matched against abbab (for example). After matching an empty string, it
1973     wasn't forcing anchoring when setting PCRE_NOTEMPTY for the next attempt; this
1974     caused it to match further down the string than it should.
1975    
1976     3. The code contained an inclusion of sys/types.h. It isn't clear why this
1977     was there because it doesn't seem to be needed, and it causes trouble on some
1978     systems, as it is not a Standard C header. It has been removed.
1979    
1980     4. Made 4 silly changes to the source to avoid stupid compiler warnings that
1981     were reported on the Macintosh. The changes were from
1982    
1983     while ((c = *(++ptr)) != 0 && c != '\n');
1984     to
1985     while ((c = *(++ptr)) != 0 && c != '\n') ;
1986    
1987     Totally extraordinary, but if that's what it takes...
1988    
1989     5. PCRE is being used in one environment where neither memmove() nor bcopy() is
1990     available. Added HAVE_BCOPY and an autoconf test for it; if neither
1991     HAVE_MEMMOVE nor HAVE_BCOPY is set, use a built-in emulation function which
1992     assumes the way PCRE uses memmove() (always moving upwards).
1993    
1994     6. PCRE is being used in one environment where strchr() is not available. There
1995     was only one use in pcre.c, and writing it out to avoid strchr() probably gives
1996     faster code anyway.
1997    
1998    
1999 nigel 45 Version 3.1 09-Feb-00
2000     ---------------------
2001    
2002     The only change in this release is the fixing of some bugs in Makefile.in for
2003     the "install" target:
2004    
2005     (1) It was failing to install pcreposix.h.
2006    
2007     (2) It was overwriting the pcre.3 man page with the pcreposix.3 man page.
2008    
2009    
2010 nigel 43 Version 3.0 01-Feb-00
2011     ---------------------
2012 nigel 41
2013     1. Add support for the /+ modifier to perltest (to output $` like it does in
2014     pcretest).
2015    
2016     2. Add support for the /g modifier to perltest.
2017    
2018     3. Fix pcretest so that it behaves even more like Perl for /g when the pattern
2019     matches null strings.
2020    
2021     4. Fix perltest so that it doesn't do unwanted things when fed an empty
2022     pattern. Perl treats empty patterns specially - it reuses the most recent
2023     pattern, which is not what we want. Replace // by /(?#)/ in order to avoid this
2024     effect.
2025    
2026     5. The POSIX interface was broken in that it was just handing over the POSIX
2027     captured string vector to pcre_exec(), but (since release 2.00) PCRE has
2028     required a bigger vector, with some working space on the end. This means that
2029     the POSIX wrapper now has to get and free some memory, and copy the results.
2030    
2031 nigel 43 6. Added some simple autoconf support, placing the test data and the
2032     documentation in separate directories, re-organizing some of the
2033     information files, and making it build pcre-config (a GNU standard). Also added
2034     libtool support for building PCRE as a shared library, which is now the
2035     default.
2036 nigel 41
2037 nigel 43 7. Got rid of the leading zero in the definition of PCRE_MINOR because 08 and
2038     09 are not valid octal constants. Single digits will be used for minor values
2039     less than 10.
2040    
2041     8. Defined REG_EXTENDED and REG_NOSUB as zero in the POSIX header, so that
2042     existing programs that set these in the POSIX interface can use PCRE without
2043     modification.
2044    
2045     9. Added a new function, pcre_fullinfo() with an extensible interface. It can
2046     return all that pcre_info() returns, plus additional data. The pcre_info()
2047     function is retained for compatibility, but is considered to be obsolete.
2048    
2049     10. Added experimental recursion feature (?R) to handle one common case that
2050     Perl 5.6 will be able to do with (?p{...}).
2051    
2052     11. Added support for POSIX character classes like [:alpha:], which Perl is
2053     adopting.
2054    
2055    
2056 nigel 39 Version 2.08 31-Aug-99
2057     ----------------------
2058    
2059     1. When startoffset was not zero and the pattern began with ".*", PCRE was not
2060     trying to match at the startoffset position, but instead was moving forward to
2061     the next newline as if a previous match had failed.
2062    
2063     2. pcretest was not making use of PCRE_NOTEMPTY when repeating for /g and /G,
2064     and could get into a loop if a null string was matched other than at the start
2065     of the subject.
2066    
2067     3. Added definitions of PCRE_MAJOR and PCRE_MINOR to pcre.h so the version can
2068     be distinguished at compile time, and for completeness also added PCRE_DATE.
2069    
2070     5. Added Paul Sokolovsky's minor changes to make it easy to compile a Win32 DLL
2071     in GnuWin32 environments.
2072    
2073    
2074 nigel 37 Version 2.07 29-Jul-99
2075     ----------------------
2076    
2077     1. The documentation is now supplied in plain text form and HTML as well as in
2078     the form of man page sources.
2079    
2080     2. C++ compilers don't like assigning (void *) values to other pointer types.
2081     In particular this affects malloc(). Although there is no problem in Standard
2082     C, I've put in casts to keep C++ compilers happy.
2083    
2084     3. Typo on pcretest.c; a cast of (unsigned char *) in the POSIX regexec() call
2085     should be (const char *).
2086    
2087     4. If NOPOSIX is defined, pcretest.c compiles without POSIX support. This may
2088     be useful for non-Unix systems who don't want to bother with the POSIX stuff.
2089     However, I haven't made this a standard facility. The documentation doesn't
2090     mention it, and the Makefile doesn't support it.
2091    
2092     5. The Makefile now contains an "install" target, with editable destinations at
2093     the top of the file. The pcretest program is not installed.
2094    
2095     6. pgrep -V now gives the PCRE version number and date.
2096    
2097     7. Fixed bug: a zero repetition after a literal string (e.g. /abcde{0}/) was
2098     causing the entire string to be ignored, instead of just the last character.
2099    
2100     8. If a pattern like /"([^\\"]+|\\.)*"/ is applied in the normal way to a
2101     non-matching string, it can take a very, very long time, even for strings of
2102     quite modest length, because of the nested recursion. PCRE now does better in
2103     some of these cases. It does this by remembering the last required literal
2104     character in the pattern, and pre-searching the subject to ensure it is present
2105     before running the real match. In other words, it applies a heuristic to detect
2106     some types of certain failure quickly, and in the above example, if presented
2107     with a string that has no trailing " it gives "no match" very quickly.
2108    
2109     9. A new runtime option PCRE_NOTEMPTY causes null string matches to be ignored;
2110     other alternatives are tried instead.
2111    
2112    
2113 nigel 35 Version 2.06 09-Jun-99
2114     ----------------------
2115    
2116     1. Change pcretest's output for amount of store used to show just the code
2117     space, because the remainder (the data block) varies in size between 32-bit and
2118     64-bit systems.
2119    
2120     2. Added an extra argument to pcre_exec() to supply an offset in the subject to
2121     start matching at. This allows lookbehinds to work when searching for multiple
2122     occurrences in a string.
2123    
2124     3. Added additional options to pcretest for testing multiple occurrences:
2125    
2126     /+ outputs the rest of the string that follows a match
2127     /g loops for multiple occurrences, using the new startoffset argument
2128     /G loops for multiple occurrences by passing an incremented pointer
2129    
2130     4. PCRE wasn't doing the "first character" optimization for patterns starting
2131     with \b or \B, though it was doing it for other lookbehind assertions. That is,
2132     it wasn't noticing that a match for a pattern such as /\bxyz/ has to start with
2133     the letter 'x'. On long subject strings, this gives a significant speed-up.
2134    
2135    
2136 nigel 33 Version 2.05 21-Apr-99
2137     ----------------------
2138    
2139     1. Changed the type of magic_number from int to long int so that it works
2140     properly on 16-bit systems.
2141    
2142     2. Fixed a bug which caused patterns starting with .* not to work correctly
2143     when the subject string contained newline characters. PCRE was assuming
2144     anchoring for such patterns in all cases, which is not correct because .* will
2145     not pass a newline unless PCRE_DOTALL is set. It now assumes anchoring only if
2146     DOTALL is set at top level; otherwise it knows that patterns starting with .*
2147     must be retried after every newline in the subject.
2148    
2149    
2150 nigel 31 Version 2.04 18-Feb-99
2151     ----------------------
2152    
2153     1. For parenthesized subpatterns with repeats whose minimum was zero, the
2154     computation of the store needed to hold the pattern was incorrect (too large).
2155     If such patterns were nested a few deep, this could multiply and become a real
2156     problem.
2157    
2158     2. Added /M option to pcretest to show the memory requirement of a specific
2159     pattern. Made -m a synonym of -s (which does this globally) for compatibility.
2160    
2161     3. Subpatterns of the form (regex){n,m} (i.e. limited maximum) were being
2162     compiled in such a way that the backtracking after subsequent failure was
2163     pessimal. Something like (a){0,3} was compiled as (a)?(a)?(a)? instead of
2164     ((a)((a)(a)?)?)? with disastrous performance if the maximum was of any size.
2165    
2166    
2167 nigel 29 Version 2.03 02-Feb-99
2168     ----------------------
2169    
2170     1. Fixed typo and small mistake in man page.
2171    
2172 nigel 31 2. Added 4th condition (GPL supersedes if conflict) and created separate
2173     LICENCE file containing the conditions.
2174 nigel 29
2175     3. Updated pcretest so that patterns such as /abc\/def/ work like they do in
2176     Perl, that is the internal \ allows the delimiter to be included in the
2177     pattern. Locked out the use of \ as a delimiter. If \ immediately follows
2178     the final delimiter, add \ to the end of the pattern (to test the error).
2179    
2180     4. Added the convenience functions for extracting substrings after a successful
2181     match. Updated pcretest to make it able to test these functions.
2182    
2183    
2184 nigel 27 Version 2.02 14-Jan-99
2185     ----------------------
2186    
2187     1. Initialized the working variables associated with each extraction so that
2188     their saving and restoring doesn't refer to uninitialized store.
2189    
2190     2. Put dummy code into study.c in order to trick the optimizer of the IBM C
2191     compiler for OS/2 into generating correct code. Apparently IBM isn't going to
2192     fix the problem.
2193    
2194     3. Pcretest: the timing code wasn't using LOOPREPEAT for timing execution
2195     calls, and wasn't printing the correct value for compiling calls. Increased the
2196     default value of LOOPREPEAT, and the number of significant figures in the
2197     times.
2198    
2199     4. Changed "/bin/rm" in the Makefile to "-rm" so it works on Windows NT.
2200    
2201     5. Renamed "deftables" as "dftables" to get it down to 8 characters, to avoid
2202     a building problem on Windows NT with a FAT file system.
2203    
2204    
2205 nigel 25 Version 2.01 21-Oct-98
2206     ----------------------
2207    
2208     1. Changed the API for pcre_compile() to allow for the provision of a pointer
2209     to character tables built by pcre_maketables() in the current locale. If NULL
2210     is passed, the default tables are used.
2211    
2212    
2213 nigel 23 Version 2.00 24-Sep-98
2214 nigel 21 ----------------------
2215    
2216 nigel 23 1. Since the (>?) facility is in Perl 5.005, don't require PCRE_EXTRA to enable
2217     it any more.
2218    
2219     2. Allow quantification of (?>) groups, and make it work correctly.
2220    
2221     3. The first character computation wasn't working for (?>) groups.
2222    
2223     4. Correct the implementation of \Z (it is permitted to match on the \n at the
2224     end of the subject) and add 5.005's \z, which really does match only at the
2225     very end of the subject.
2226    
2227     5. Remove the \X "cut" facility; Perl doesn't have it, and (?> is neater.
2228    
2229     6. Remove the ability to specify CASELESS, MULTILINE, DOTALL, and
2230     DOLLAR_END_ONLY at runtime, to make it possible to implement the Perl 5.005
2231     localized options. All options to pcre_study() were also removed.
2232    
2233     7. Add other new features from 5.005:
2234    
2235     $(?<= positive lookbehind
2236     $(?<! negative lookbehind
2237     (?imsx-imsx) added the unsetting capability
2238     such a setting is global if at outer level; local otherwise
2239     (?imsx-imsx:) non-capturing groups with option setting
2240     (?(cond)re|re) conditional pattern matching
2241    
2242     A backreference to itself in a repeated group matches the previous
2243     captured string.
2244    
2245     8. General tidying up of studying (both automatic and via "study")
2246     consequential on the addition of new assertions.
2247    
2248     9. As in 5.005, unlimited repeated groups that could match an empty substring
2249     are no longer faulted at compile time. Instead, the loop is forcibly broken at
2250     runtime if any iteration does actually match an empty substring.
2251    
2252     10. Include the RunTest script in the distribution.
2253    
2254     11. Added tests from the Perl 5.005_02 distribution. This showed up a few
2255     discrepancies, some of which were old and were also with respect to 5.004. They
2256     have now been fixed.
2257    
2258    
2259     Version 1.09 28-Apr-98
2260     ----------------------
2261    
2262 nigel 21 1. A negated single character class followed by a quantifier with a minimum
2263     value of one (e.g. [^x]{1,6} ) was not compiled correctly. This could lead to
2264     program crashes, or just wrong answers. This did not apply to negated classes
2265     containing more than one character, or to minima other than one.
2266    
2267    
2268 nigel 19 Version 1.08 27-Mar-98
2269     ----------------------
2270    
2271     1. Add PCRE_UNGREEDY to invert the greediness of quantifiers.
2272    
2273     2. Add (?U) and (?X) to set PCRE_UNGREEDY and PCRE_EXTRA respectively. The
2274     latter must appear before anything that relies on it in the pattern.
2275    
2276    
2277 nigel 17 Version 1.07 16-Feb-98
2278     ----------------------
2279    
2280     1. A pattern such as /((a)*)*/ was not being diagnosed as in error (unlimited
2281     repeat of a potentially empty string).
2282    
2283    
2284 nigel 15 Version 1.06 23-Jan-98
2285     ----------------------
2286    
2287     1. Added Markus Oberhumer's little patches for C++.
2288    
2289     2. Literal strings longer than 255 characters were broken.
2290    
2291    
2292 nigel 13 Version 1.05 23-Dec-97
2293     ----------------------
2294    
2295     1. Negated character classes containing more than one character were failing if
2296     PCRE_CASELESS was set at run time.
2297    
2298    
2299 nigel 11 Version 1.04 19-Dec-97
2300     ----------------------
2301    
2302     1. Corrected the man page, where some "const" qualifiers had been omitted.
2303    
2304     2. Made debugging output print "{0,xxx}" instead of just "{,xxx}" to agree with
2305     input syntax.
2306    
2307     3. Fixed memory leak which occurred when a regex with back references was
2308     matched with an offsets vector that wasn't big enough. The temporary memory
2309     that is used in this case wasn't being freed if the match failed.
2310    
2311     4. Tidied pcretest to ensure it frees memory that it gets.
2312    
2313     5. Temporary memory was being obtained in the case where the passed offsets
2314     vector was exactly big enough.
2315    
2316     6. Corrected definition of offsetof() from change 5 below.
2317    
2318     7. I had screwed up change 6 below and broken the rules for the use of
2319     setjmp(). Now fixed.
2320    
2321    
2322 nigel 9 Version 1.03 18-Dec-97
2323     ----------------------
2324    
2325     1. A erroneous regex with a missing opening parenthesis was correctly
2326     diagnosed, but PCRE attempted to access brastack[-1], which could cause crashes
2327     on some systems.
2328    
2329     2. Replaced offsetof(real_pcre, code) by offsetof(real_pcre, code[0]) because
2330     it was reported that one broken compiler failed on the former because "code" is
2331     also an independent variable.
2332    
2333     3. The erroneous regex a[]b caused an array overrun reference.
2334    
2335     4. A regex ending with a one-character negative class (e.g. /[^k]$/) did not
2336     fail on data ending with that character. (It was going on too far, and checking
2337     the next character, typically a binary zero.) This was specific to the
2338     optimized code for single-character negative classes.
2339    
2340     5. Added a contributed patch from the TIN world which does the following:
2341    
2342     + Add an undef for memmove, in case the the system defines a macro for it.
2343    
2344     + Add a definition of offsetof(), in case there isn't one. (I don't know
2345     the reason behind this - offsetof() is part of the ANSI standard - but
2346     it does no harm).
2347    
2348     + Reduce the ifdef's in pcre.c using macro DPRINTF, thereby eliminating
2349     most of the places where whitespace preceded '#'. I have given up and
2350     allowed the remaining 2 cases to be at the margin.
2351    
2352     + Rename some variables in pcre to eliminate shadowing. This seems very
2353     pedantic, but does no harm, of course.
2354    
2355     6. Moved the call to setjmp() into its own function, to get rid of warnings
2356     from gcc -Wall, and avoided calling it at all unless PCRE_EXTRA is used.
2357    
2358     7. Constructs such as \d{8,} were compiling into the equivalent of
2359 nigel 11 \d{8}\d{0,65527} instead of \d{8}\d* which didn't make much difference to the
2360 nigel 9 outcome, but in this particular case used more store than had been allocated,
2361     which caused the bug to be discovered because it threw up an internal error.
2362    
2363     8. The debugging code in both pcre and pcretest for outputting the compiled
2364     form of a regex was going wrong in the case of back references followed by
2365     curly-bracketed repeats.
2366    
2367    
2368 nigel 7 Version 1.02 12-Dec-97
2369     ----------------------
2370    
2371     1. Typos in pcre.3 and comments in the source fixed.
2372    
2373     2. Applied a contributed patch to get rid of places where it used to remove
2374     'const' from variables, and fixed some signed/unsigned and uninitialized
2375     variable warnings.
2376    
2377     3. Added the "runtest" target to Makefile.
2378    
2379     4. Set default compiler flag to -O2 rather than just -O.
2380    
2381    
2382 nigel 5 Version 1.01 19-Nov-97
2383     ----------------------
2384    
2385     1. PCRE was failing to diagnose unlimited repeat of empty string for patterns
2386     like /([ab]*)*/, that is, for classes with more than one character in them.
2387    
2388     2. Likewise, it wasn't diagnosing patterns with "once-only" subpatterns, such
2389     as /((?>a*))*/ (a PCRE_EXTRA facility).
2390    
2391    
2392     Version 1.00 18-Nov-97
2393     ----------------------
2394    
2395     1. Added compile-time macros to support systems such as SunOS4 which don't have
2396     memmove() or strerror() but have other things that can be used instead.
2397    
2398     2. Arranged that "make clean" removes the executables.
2399    
2400    
2401 nigel 3 Version 0.99 27-Oct-97
2402     ----------------------
2403    
2404     1. Fixed bug in code for optimizing classes with only one character. It was
2405     initializing a 32-byte map regardless, which could cause it to run off the end
2406     of the memory it had got.
2407    
2408     2. Added, conditional on PCRE_EXTRA, the proposed (?>REGEX) construction.
2409    
2410    
2411     Version 0.98 22-Oct-97
2412     ----------------------
2413    
2414     1. Fixed bug in code for handling temporary memory usage when there are more
2415     back references than supplied space in the ovector. This could cause segfaults.
2416    
2417    
2418     Version 0.97 21-Oct-97
2419     ----------------------
2420    
2421     1. Added the \X "cut" facility, conditional on PCRE_EXTRA.
2422    
2423     2. Optimized negated single characters not to use a bit map.
2424    
2425     3. Brought error texts together as macro definitions; clarified some of them;
2426     fixed one that was wrong - it said "range out of order" when it meant "invalid
2427     escape sequence".
2428    
2429     4. Changed some char * arguments to const char *.
2430    
2431     5. Added PCRE_NOTBOL and PCRE_NOTEOL (from POSIX).
2432    
2433     6. Added the POSIX-style API wrapper in pcreposix.a and testing facilities in
2434     pcretest.
2435    
2436    
2437     Version 0.96 16-Oct-97
2438     ----------------------
2439    
2440     1. Added a simple "pgrep" utility to the distribution.
2441    
2442     2. Fixed an incompatibility with Perl: "{" is now treated as a normal character
2443     unless it appears in one of the precise forms "{ddd}", "{ddd,}", or "{ddd,ddd}"
2444     where "ddd" means "one or more decimal digits".
2445    
2446     3. Fixed serious bug. If a pattern had a back reference, but the call to
2447     pcre_exec() didn't supply a large enough ovector to record the related
2448     identifying subpattern, the match always failed. PCRE now remembers the number
2449     of the largest back reference, and gets some temporary memory in which to save
2450     the offsets during matching if necessary, in order to ensure that
2451     backreferences always work.
2452    
2453     4. Increased the compatibility with Perl in a number of ways:
2454    
2455     (a) . no longer matches \n by default; an option PCRE_DOTALL is provided
2456     to request this handling. The option can be set at compile or exec time.
2457    
2458     (b) $ matches before a terminating newline by default; an option
2459     PCRE_DOLLAR_ENDONLY is provided to override this (but not in multiline
2460     mode). The option can be set at compile or exec time.
2461    
2462     (c) The handling of \ followed by a digit other than 0 is now supposed to be
2463     the same as Perl's. If the decimal number it represents is less than 10
2464     or there aren't that many previous left capturing parentheses, an octal
2465     escape is read. Inside a character class, it's always an octal escape,
2466     even if it is a single digit.
2467    
2468     (d) An escaped but undefined alphabetic character is taken as a literal,
2469     unless PCRE_EXTRA is set. Currently this just reserves the remaining
2470     escapes.
2471    
2472     (e) {0} is now permitted. (The previous item is removed from the compiled
2473     pattern).
2474    
2475     5. Changed all the names of code files so that the basic parts are no longer
2476     than 10 characters, and abolished the teeny "globals.c" file.
2477    
2478     6. Changed the handling of character classes; they are now done with a 32-byte
2479     bit map always.
2480    
2481     7. Added the -d and /D options to pcretest to make it possible to look at the
2482     internals of compilation without having to recompile pcre.
2483    
2484    
2485     Version 0.95 23-Sep-97
2486     ----------------------
2487    
2488     1. Fixed bug in pre-pass concerning escaped "normal" characters such as \x5c or
2489     \x20 at the start of a run of normal characters. These were being treated as
2490     real characters, instead of the source characters being re-checked.
2491    
2492    
2493     Version 0.94 18-Sep-97
2494     ----------------------
2495    
2496     1. The functions are now thread-safe, with the caveat that the global variables
2497     containing pointers to malloc() and free() or alternative functions are the
2498     same for all threads.
2499    
2500     2. Get pcre_study() to generate a bitmap of initial characters for non-
2501     anchored patterns when this is possible, and use it if passed to pcre_exec().
2502    
2503    
2504     Version 0.93 15-Sep-97
2505     ----------------------
2506    
2507     1. /(b)|(:+)/ was computing an incorrect first character.
2508    
2509     2. Add pcre_study() to the API and the passing of pcre_extra to pcre_exec(),
2510     but not actually doing anything yet.
2511    
2512     3. Treat "-" characters in classes that cannot be part of ranges as literals,
2513     as Perl does (e.g. [-az] or [az-]).
2514    
2515     4. Set the anchored flag if a branch starts with .* or .*? because that tests
2516     all possible positions.
2517    
2518     5. Split up into different modules to avoid including unneeded functions in a
2519     compiled binary. However, compile and exec are still in one module. The "study"
2520     function is split off.
2521    
2522     6. The character tables are now in a separate module whose source is generated
2523     by an auxiliary program - but can then be edited by hand if required. There are
2524     now no calls to isalnum(), isspace(), isdigit(), isxdigit(), tolower() or
2525     toupper() in the code.
2526    
2527     7. Turn the malloc/free funtions variables into pcre_malloc and pcre_free and
2528     make them global. Abolish the function for setting them, as the caller can now
2529     set them directly.
2530    
2531    
2532     Version 0.92 11-Sep-97
2533     ----------------------
2534    
2535     1. A repeat with a fixed maximum and a minimum of 1 for an ordinary character
2536     (e.g. /a{1,3}/) was broken (I mis-optimized it).
2537    
2538     2. Caseless matching was not working in character classes if the characters in
2539     the pattern were in upper case.
2540    
2541     3. Make ranges like [W-c] work in the same way as Perl for caseless matching.
2542    
2543     4. Make PCRE_ANCHORED public and accept as a compile option.
2544    
2545     5. Add an options word to pcre_exec() and accept PCRE_ANCHORED and
2546     PCRE_CASELESS at run time. Add escapes \A and \I to pcretest to cause it to
2547     pass them.
2548    
2549     6. Give an error if bad option bits passed at compile or run time.
2550    
2551     7. Add PCRE_MULTILINE at compile and exec time, and (?m) as well. Add \M to
2552     pcretest to cause it to pass that flag.
2553    
2554     8. Add pcre_info(), to get the number of identifying subpatterns, the stored
2555     options, and the first character, if set.
2556    
2557     9. Recognize C+ or C{n,m} where n >= 1 as providing a fixed starting character.
2558    
2559    
2560     Version 0.91 10-Sep-97
2561     ----------------------
2562    
2563     1. PCRE was failing to diagnose unlimited repeats of subpatterns that could
2564     match the empty string as in /(a*)*/. It was looping and ultimately crashing.
2565    
2566     2. PCRE was looping on encountering an indefinitely repeated back reference to
2567     a subpattern that had matched an empty string, e.g. /(a|)\1*/. It now does what
2568     Perl does - treats the match as successful.
2569    
2570     ****

Properties

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

webmaster@exim.org
ViewVC Help
Powered by ViewVC 1.1.12