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

Contents of /code/trunk/ChangeLog

Parent Directory Parent Directory | Revision Log Revision Log


Revision 105 - (hide annotations) (download)
Tue Mar 6 16:32:53 2007 UTC (7 years, 8 months ago) by ph10
File size: 121465 byte(s)
Mark ucp_table (in ucptable.h) and ucp_gentype (in pcre_ucp_searchfuncs.c)
as "const".

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 ph10 100 2. Part of the patch fixed a problem with the pcregrep tests. The test of -r
13     for recursive directory scanning broke on some systems because the files
14     are not scanned in any specific order and on different systems the order
15     was different. A call to "sort" has been inserted into RunGrepTest for the
16     approprate test as a short-term fix. In the longer term there may be an
17     alternative.
18    
19     3. I had an email from Eric Raymond about problems translating some of PCRE's
20 ph10 99 man pages to HTML (despite the fact that I distribute HTML pages, some
21     people do their own conversions for various reasons). The problems
22     concerned the use of low-level troff macros .br and .in. I have therefore
23     removed all such uses from the man pages (some were redundant, some could
24     be replaced by .nf/.fi pairs). The maintain/132html script that I use to
25     generate HTML has been updated to handle .nf/.fi and to complain if it
26     encounters .br or .in.
27    
28 ph10 100 4. Updated comments in configure.ac that get placed in config.h.in and also
29     arranged for config.h to be included in the distribution, for the benefit
30     of those who have to compile without Autotools (compare pcre.h).
31    
32 ph10 101 5. Updated the support (such as it is) for Virtual Pascal, thanks to Stefan
33     Weber: (1) pcre_internal.h was missing some function renames; (2) updated
34     makevp.bat for the current PCRE, using the additional files !compile.txt,
35     !linklib.txt, and pcregexp.pas.
36    
37 ph10 102 6. A Windows user reported a minor discrepancy with test 2, which turned out
38     to be caused by a trailing space on an input line that had got lost in his
39     copy. The trailing space was an accident, so I've just removed it.
40    
41 ph10 104 7. Add -Wl,-R... flags in pcre-config.in for *BSD* systems, as I'm told
42     that is needed.
43    
44 ph10 105 8. Mark ucp_table (in ucptable.h) and ucp_gentype (in pcre_ucp_searchfuncs.c)
45     as "const" (a) because they are and (b) because it helps the PHP
46     maintainers who have recently made a script to detect big data structures
47     in the php code that should be moved to the .rodata section. I remembered
48     to update Builducptable as well, so it won't revert if ucptable.h is ever
49     re-created.
50    
51 ph10 98
52 nigel 93 Version 7.0 19-Dec-06
53     ---------------------
54    
55     1. Fixed a signed/unsigned compiler warning in pcre_compile.c, shown up by
56     moving to gcc 4.1.1.
57    
58     2. The -S option for pcretest uses setrlimit(); I had omitted to #include
59     sys/time.h, which is documented as needed for this function. It doesn't
60     seem to matter on Linux, but it showed up on some releases of OS X.
61    
62     3. It seems that there are systems where bytes whose values are greater than
63     127 match isprint() in the "C" locale. The "C" locale should be the
64     default when a C program starts up. In most systems, only ASCII printing
65     characters match isprint(). This difference caused the output from pcretest
66     to vary, making some of the tests fail. I have changed pcretest so that:
67    
68     (a) When it is outputting text in the compiled version of a pattern, bytes
69     other than 32-126 are always shown as hex escapes.
70    
71     (b) When it is outputting text that is a matched part of a subject string,
72     it does the same, unless a different locale has been set for the match
73     (using the /L modifier). In this case, it uses isprint() to decide.
74    
75     4. Fixed a major bug that caused incorrect computation of the amount of memory
76     required for a compiled pattern when options that changed within the
77     pattern affected the logic of the preliminary scan that determines the
78     length. The relevant options are -x, and -i in UTF-8 mode. The result was
79     that the computed length was too small. The symptoms of this bug were
80     either the PCRE error "internal error: code overflow" from pcre_compile(),
81     or a glibc crash with a message such as "pcretest: free(): invalid next
82     size (fast)". Examples of patterns that provoked this bug (shown in
83     pcretest format) are:
84    
85     /(?-x: )/x
86     /(?x)(?-x: \s*#\s*)/
87     /((?i)[\x{c0}])/8
88     /(?i:[\x{c0}])/8
89    
90     HOWEVER: Change 17 below makes this fix obsolete as the memory computation
91     is now done differently.
92    
93     5. Applied patches from Google to: (a) add a QuoteMeta function to the C++
94     wrapper classes; (b) implement a new function in the C++ scanner that is
95     more efficient than the old way of doing things because it avoids levels of
96     recursion in the regex matching; (c) add a paragraph to the documentation
97     for the FullMatch() function.
98    
99     6. The escape sequence \n was being treated as whatever was defined as
100     "newline". Not only was this contrary to the documentation, which states
101     that \n is character 10 (hex 0A), but it also went horribly wrong when
102     "newline" was defined as CRLF. This has been fixed.
103    
104     7. In pcre_dfa_exec.c the value of an unsigned integer (the variable called c)
105     was being set to -1 for the "end of line" case (supposedly a value that no
106     character can have). Though this value is never used (the check for end of
107     line is "zero bytes in current character"), it caused compiler complaints.
108     I've changed it to 0xffffffff.
109    
110     8. In pcre_version.c, the version string was being built by a sequence of
111     C macros that, in the event of PCRE_PRERELEASE being defined as an empty
112     string (as it is for production releases) called a macro with an empty
113     argument. The C standard says the result of this is undefined. The gcc
114     compiler treats it as an empty string (which was what was wanted) but it is
115     reported that Visual C gives an error. The source has been hacked around to
116     avoid this problem.
117    
118     9. On the advice of a Windows user, included <io.h> and <fcntl.h> in Windows
119     builds of pcretest, and changed the call to _setmode() to use _O_BINARY
120     instead of 0x8000. Made all the #ifdefs test both _WIN32 and WIN32 (not all
121     of them did).
122    
123     10. Originally, pcretest opened its input and output without "b"; then I was
124     told that "b" was needed in some environments, so it was added for release
125     5.0 to both the input and output. (It makes no difference on Unix-like
126     systems.) Later I was told that it is wrong for the input on Windows. I've
127     now abstracted the modes into two macros, to make it easier to fiddle with
128     them, and removed "b" from the input mode under Windows.
129    
130     11. Added pkgconfig support for the C++ wrapper library, libpcrecpp.
131    
132     12. Added -help and --help to pcretest as an official way of being reminded
133     of the options.
134    
135     13. Removed some redundant semicolons after macro calls in pcrecpparg.h.in
136     and pcrecpp.cc because they annoy compilers at high warning levels.
137    
138     14. A bit of tidying/refactoring in pcre_exec.c in the main bumpalong loop.
139    
140     15. Fixed an occurrence of == in configure.ac that should have been = (shell
141     scripts are not C programs :-) and which was not noticed because it works
142     on Linux.
143    
144     16. pcretest is supposed to handle any length of pattern and data line (as one
145     line or as a continued sequence of lines) by extending its input buffer if
146     necessary. This feature was broken for very long pattern lines, leading to
147     a string of junk being passed to pcre_compile() if the pattern was longer
148     than about 50K.
149    
150     17. I have done a major re-factoring of the way pcre_compile() computes the
151     amount of memory needed for a compiled pattern. Previously, there was code
152     that made a preliminary scan of the pattern in order to do this. That was
153     OK when PCRE was new, but as the facilities have expanded, it has become
154     harder and harder to keep it in step with the real compile phase, and there
155     have been a number of bugs (see for example, 4 above). I have now found a
156     cunning way of running the real compile function in a "fake" mode that
157     enables it to compute how much memory it would need, while actually only
158     ever using a few hundred bytes of working memory and without too many
159     tests of the mode. This should make future maintenance and development
160     easier. A side effect of this work is that the limit of 200 on the nesting
161     depth of parentheses has been removed (though this was never a serious
162     limitation, I suspect). However, there is a downside: pcre_compile() now
163     runs more slowly than before (30% or more, depending on the pattern). I
164     hope this isn't a big issue. There is no effect on runtime performance.
165    
166     18. Fixed a minor bug in pcretest: if a pattern line was not terminated by a
167     newline (only possible for the last line of a file) and it was a
168     pattern that set a locale (followed by /Lsomething), pcretest crashed.
169    
170     19. Added additional timing features to pcretest. (1) The -tm option now times
171     matching only, not compiling. (2) Both -t and -tm can be followed, as a
172     separate command line item, by a number that specifies the number of
173     repeats to use when timing. The default is 50000; this gives better
174     precision, but takes uncomfortably long for very large patterns.
175    
176     20. Extended pcre_study() to be more clever in cases where a branch of a
177     subpattern has no definite first character. For example, (a*|b*)[cd] would
178     previously give no result from pcre_study(). Now it recognizes that the
179     first character must be a, b, c, or d.
180    
181     21. There was an incorrect error "recursive call could loop indefinitely" if
182     a subpattern (or the entire pattern) that was being tested for matching an
183     empty string contained only one non-empty item after a nested subpattern.
184     For example, the pattern (?>\x{100}*)\d(?R) provoked this error
185     incorrectly, because the \d was being skipped in the check.
186    
187     22. The pcretest program now has a new pattern option /B and a command line
188     option -b, which is equivalent to adding /B to every pattern. This causes
189     it to show the compiled bytecode, without the additional information that
190     -d shows. The effect of -d is now the same as -b with -i (and similarly, /D
191     is the same as /B/I).
192    
193     23. A new optimization is now able automatically to treat some sequences such
194     as a*b as a*+b. More specifically, if something simple (such as a character
195     or a simple class like \d) has an unlimited quantifier, and is followed by
196     something that cannot possibly match the quantified thing, the quantifier
197     is automatically "possessified".
198    
199     24. A recursive reference to a subpattern whose number was greater than 39
200     went wrong under certain circumstances in UTF-8 mode. This bug could also
201     have affected the operation of pcre_study().
202    
203     25. Realized that a little bit of performance could be had by replacing
204     (c & 0xc0) == 0xc0 with c >= 0xc0 when processing UTF-8 characters.
205    
206     26. Timing data from pcretest is now shown to 4 decimal places instead of 3.
207    
208     27. Possessive quantifiers such as a++ were previously implemented by turning
209     them into atomic groups such as ($>a+). Now they have their own opcodes,
210     which improves performance. This includes the automatically created ones
211     from 23 above.
212    
213     28. A pattern such as (?=(\w+))\1: which simulates an atomic group using a
214     lookahead was broken if it was not anchored. PCRE was mistakenly expecting
215     the first matched character to be a colon. This applied both to named and
216     numbered groups.
217    
218     29. The ucpinternal.h header file was missing its idempotency #ifdef.
219    
220     30. I was sent a "project" file called libpcre.a.dev which I understand makes
221     building PCRE on Windows easier, so I have included it in the distribution.
222    
223     31. There is now a check in pcretest against a ridiculously large number being
224     returned by pcre_exec() or pcre_dfa_exec(). If this happens in a /g or /G
225     loop, the loop is abandoned.
226    
227     32. Forward references to subpatterns in conditions such as (?(2)...) where
228     subpattern 2 is defined later cause pcre_compile() to search forwards in
229     the pattern for the relevant set of parentheses. This search went wrong
230     when there were unescaped parentheses in a character class, parentheses
231     escaped with \Q...\E, or parentheses in a #-comment in /x mode.
232    
233     33. "Subroutine" calls and backreferences were previously restricted to
234     referencing subpatterns earlier in the regex. This restriction has now
235     been removed.
236    
237     34. Added a number of extra features that are going to be in Perl 5.10. On the
238     whole, these are just syntactic alternatives for features that PCRE had
239     previously implemented using the Python syntax or my own invention. The
240     other formats are all retained for compatibility.
241    
242     (a) Named groups can now be defined as (?<name>...) or (?'name'...) as well
243     as (?P<name>...). The new forms, as well as being in Perl 5.10, are
244     also .NET compatible.
245    
246     (b) A recursion or subroutine call to a named group can now be defined as
247     (?&name) as well as (?P>name).
248    
249     (c) A backreference to a named group can now be defined as \k<name> or
250     \k'name' as well as (?P=name). The new forms, as well as being in Perl
251     5.10, are also .NET compatible.
252    
253     (d) A conditional reference to a named group can now use the syntax
254     (?(<name>) or (?('name') as well as (?(name).
255    
256     (e) A "conditional group" of the form (?(DEFINE)...) can be used to define
257     groups (named and numbered) that are never evaluated inline, but can be
258     called as "subroutines" from elsewhere. In effect, the DEFINE condition
259     is always false. There may be only one alternative in such a group.
260    
261     (f) A test for recursion can be given as (?(R1).. or (?(R&name)... as well
262     as the simple (?(R). The condition is true only if the most recent
263     recursion is that of the given number or name. It does not search out
264     through the entire recursion stack.
265    
266     (g) The escape \gN or \g{N} has been added, where N is a positive or
267     negative number, specifying an absolute or relative reference.
268    
269     35. Tidied to get rid of some further signed/unsigned compiler warnings and
270     some "unreachable code" warnings.
271    
272     36. Updated the Unicode property tables to Unicode version 5.0.0. Amongst other
273     things, this adds five new scripts.
274    
275     37. Perl ignores orphaned \E escapes completely. PCRE now does the same.
276     There were also incompatibilities regarding the handling of \Q..\E inside
277     character classes, for example with patterns like [\Qa\E-\Qz\E] where the
278     hyphen was adjacent to \Q or \E. I hope I've cleared all this up now.
279    
280     38. Like Perl, PCRE detects when an indefinitely repeated parenthesized group
281     matches an empty string, and forcibly breaks the loop. There were bugs in
282     this code in non-simple cases. For a pattern such as ^(a()*)* matched
283     against aaaa the result was just "a" rather than "aaaa", for example. Two
284     separate and independent bugs (that affected different cases) have been
285     fixed.
286    
287     39. Refactored the code to abolish the use of different opcodes for small
288     capturing bracket numbers. This is a tidy that I avoided doing when I
289     removed the limit on the number of capturing brackets for 3.5 back in 2001.
290     The new approach is not only tidier, it makes it possible to reduce the
291     memory needed to fix the previous bug (38).
292    
293     40. Implemented PCRE_NEWLINE_ANY to recognize any of the Unicode newline
294     sequences (http://unicode.org/unicode/reports/tr18/) as "newline" when
295     processing dot, circumflex, or dollar metacharacters, or #-comments in /x
296     mode.
297    
298     41. Add \R to match any Unicode newline sequence, as suggested in the Unicode
299     report.
300    
301     42. Applied patch, originally from Ari Pollak, modified by Google, to allow
302     copy construction and assignment in the C++ wrapper.
303    
304     43. Updated pcregrep to support "--newline=any". In the process, I fixed a
305     couple of bugs that could have given wrong results in the "--newline=crlf"
306     case.
307    
308     44. Added a number of casts and did some reorganization of signed/unsigned int
309     variables following suggestions from Dair Grant. Also renamed the variable
310     "this" as "item" because it is a C++ keyword.
311    
312     45. Arranged for dftables to add
313    
314     #include "pcre_internal.h"
315    
316     to pcre_chartables.c because without it, gcc 4.x may remove the array
317     definition from the final binary if PCRE is built into a static library and
318     dead code stripping is activated.
319    
320     46. For an unanchored pattern, if a match attempt fails at the start of a
321     newline sequence, and the newline setting is CRLF or ANY, and the next two
322     characters are CRLF, advance by two characters instead of one.
323    
324    
325 nigel 91 Version 6.7 04-Jul-06
326     ---------------------
327    
328     1. In order to handle tests when input lines are enormously long, pcretest has
329     been re-factored so that it automatically extends its buffers when
330     necessary. The code is crude, but this _is_ just a test program. The
331     default size has been increased from 32K to 50K.
332    
333     2. The code in pcre_study() was using the value of the re argument before
334     testing it for NULL. (Of course, in any sensible call of the function, it
335     won't be NULL.)
336    
337     3. The memmove() emulation function in pcre_internal.h, which is used on
338     systems that lack both memmove() and bcopy() - that is, hardly ever -
339     was missing a "static" storage class specifier.
340    
341     4. When UTF-8 mode was not set, PCRE looped when compiling certain patterns
342     containing an extended class (one that cannot be represented by a bitmap
343     because it contains high-valued characters or Unicode property items, e.g.
344     [\pZ]). Almost always one would set UTF-8 mode when processing such a
345     pattern, but PCRE should not loop if you do not (it no longer does).
346     [Detail: two cases were found: (a) a repeated subpattern containing an
347     extended class; (b) a recursive reference to a subpattern that followed a
348     previous extended class. It wasn't skipping over the extended class
349     correctly when UTF-8 mode was not set.]
350    
351     5. A negated single-character class was not being recognized as fixed-length
352     in lookbehind assertions such as (?<=[^f]), leading to an incorrect
353     compile error "lookbehind assertion is not fixed length".
354    
355     6. The RunPerlTest auxiliary script was showing an unexpected difference
356     between PCRE and Perl for UTF-8 tests. It turns out that it is hard to
357     write a Perl script that can interpret lines of an input file either as
358     byte characters or as UTF-8, which is what "perltest" was being required to
359     do for the non-UTF-8 and UTF-8 tests, respectively. Essentially what you
360     can't do is switch easily at run time between having the "use utf8;" pragma
361     or not. In the end, I fudged it by using the RunPerlTest script to insert
362     "use utf8;" explicitly for the UTF-8 tests.
363    
364     7. In multiline (/m) mode, PCRE was matching ^ after a terminating newline at
365     the end of the subject string, contrary to the documentation and to what
366     Perl does. This was true of both matching functions. Now it matches only at
367     the start of the subject and immediately after *internal* newlines.
368    
369     8. A call of pcre_fullinfo() from pcretest to get the option bits was passing
370     a pointer to an int instead of a pointer to an unsigned long int. This
371     caused problems on 64-bit systems.
372    
373     9. Applied a patch from the folks at Google to pcrecpp.cc, to fix "another
374     instance of the 'standard' template library not being so standard".
375    
376     10. There was no check on the number of named subpatterns nor the maximum
377     length of a subpattern name. The product of these values is used to compute
378     the size of the memory block for a compiled pattern. By supplying a very
379     long subpattern name and a large number of named subpatterns, the size
380     computation could be caused to overflow. This is now prevented by limiting
381     the length of names to 32 characters, and the number of named subpatterns
382     to 10,000.
383    
384     11. Subpatterns that are repeated with specific counts have to be replicated in
385     the compiled pattern. The size of memory for this was computed from the
386     length of the subpattern and the repeat count. The latter is limited to
387     65535, but there was no limit on the former, meaning that integer overflow
388     could in principle occur. The compiled length of a repeated subpattern is
389     now limited to 30,000 bytes in order to prevent this.
390    
391     12. Added the optional facility to have named substrings with the same name.
392    
393     13. Added the ability to use a named substring as a condition, using the
394     Python syntax: (?(name)yes|no). This overloads (?(R)... and names that
395     are numbers (not recommended). Forward references are permitted.
396    
397     14. Added forward references in named backreferences (if you see what I mean).
398    
399     15. In UTF-8 mode, with the PCRE_DOTALL option set, a quantified dot in the
400     pattern could run off the end of the subject. For example, the pattern
401     "(?s)(.{1,5})"8 did this with the subject "ab".
402    
403     16. If PCRE_DOTALL or PCRE_MULTILINE were set, pcre_dfa_exec() behaved as if
404     PCRE_CASELESS was set when matching characters that were quantified with ?
405     or *.
406    
407     17. A character class other than a single negated character that had a minimum
408     but no maximum quantifier - for example [ab]{6,} - was not handled
409     correctly by pce_dfa_exec(). It would match only one character.
410    
411     18. A valid (though odd) pattern that looked like a POSIX character
412     class but used an invalid character after [ (for example [[,abc,]]) caused
413     pcre_compile() to give the error "Failed: internal error: code overflow" or
414     in some cases to crash with a glibc free() error. This could even happen if
415     the pattern terminated after [[ but there just happened to be a sequence of
416     letters, a binary zero, and a closing ] in the memory that followed.
417    
418     19. Perl's treatment of octal escapes in the range \400 to \777 has changed
419     over the years. Originally (before any Unicode support), just the bottom 8
420     bits were taken. Thus, for example, \500 really meant \100. Nowadays the
421     output from "man perlunicode" includes this:
422    
423     The regular expression compiler produces polymorphic opcodes. That
424     is, the pattern adapts to the data and automatically switches to
425     the Unicode character scheme when presented with Unicode data--or
426     instead uses a traditional byte scheme when presented with byte
427     data.
428    
429     Sadly, a wide octal escape does not cause a switch, and in a string with
430     no other multibyte characters, these octal escapes are treated as before.
431     Thus, in Perl, the pattern /\500/ actually matches \100 but the pattern
432     /\500|\x{1ff}/ matches \500 or \777 because the whole thing is treated as a
433     Unicode string.
434    
435     I have not perpetrated such confusion in PCRE. Up till now, it took just
436     the bottom 8 bits, as in old Perl. I have now made octal escapes with
437     values greater than \377 illegal in non-UTF-8 mode. In UTF-8 mode they
438     translate to the appropriate multibyte character.
439    
440     29. Applied some refactoring to reduce the number of warnings from Microsoft
441     and Borland compilers. This has included removing the fudge introduced
442     seven years ago for the OS/2 compiler (see 2.02/2 below) because it caused
443     a warning about an unused variable.
444    
445     21. PCRE has not included VT (character 0x0b) in the set of whitespace
446     characters since release 4.0, because Perl (from release 5.004) does not.
447     [Or at least, is documented not to: some releases seem to be in conflict
448     with the documentation.] However, when a pattern was studied with
449     pcre_study() and all its branches started with \s, PCRE still included VT
450     as a possible starting character. Of course, this did no harm; it just
451     caused an unnecessary match attempt.
452    
453     22. Removed a now-redundant internal flag bit that recorded the fact that case
454     dependency changed within the pattern. This was once needed for "required
455     byte" processing, but is no longer used. This recovers a now-scarce options
456     bit. Also moved the least significant internal flag bit to the most-
457     significant bit of the word, which was not previously used (hangover from
458     the days when it was an int rather than a uint) to free up another bit for
459     the future.
460    
461     23. Added support for CRLF line endings as well as CR and LF. As well as the
462     default being selectable at build time, it can now be changed at runtime
463     via the PCRE_NEWLINE_xxx flags. There are now options for pcregrep to
464     specify that it is scanning data with non-default line endings.
465    
466     24. Changed the definition of CXXLINK to make it agree with the definition of
467     LINK in the Makefile, by replacing LDFLAGS to CXXFLAGS.
468    
469     25. Applied Ian Taylor's patches to avoid using another stack frame for tail
470     recursions. This makes a big different to stack usage for some patterns.
471    
472     26. If a subpattern containing a named recursion or subroutine reference such
473     as (?P>B) was quantified, for example (xxx(?P>B)){3}, the calculation of
474     the space required for the compiled pattern went wrong and gave too small a
475     value. Depending on the environment, this could lead to "Failed: internal
476     error: code overflow at offset 49" or "glibc detected double free or
477     corruption" errors.
478    
479     27. Applied patches from Google (a) to support the new newline modes and (b) to
480     advance over multibyte UTF-8 characters in GlobalReplace.
481    
482     28. Change free() to pcre_free() in pcredemo.c. Apparently this makes a
483     difference for some implementation of PCRE in some Windows version.
484    
485     29. Added some extra testing facilities to pcretest:
486    
487     \q<number> in a data line sets the "match limit" value
488     \Q<number> in a data line sets the "match recursion limt" value
489     -S <number> sets the stack size, where <number> is in megabytes
490    
491     The -S option isn't available for Windows.
492    
493    
494 nigel 89 Version 6.6 06-Feb-06
495     ---------------------
496    
497     1. Change 16(a) for 6.5 broke things, because PCRE_DATA_SCOPE was not defined
498     in pcreposix.h. I have copied the definition from pcre.h.
499    
500     2. Change 25 for 6.5 broke compilation in a build directory out-of-tree
501     because pcre.h is no longer a built file.
502    
503     3. Added Jeff Friedl's additional debugging patches to pcregrep. These are
504     not normally included in the compiled code.
505    
506    
507 nigel 87 Version 6.5 01-Feb-06
508     ---------------------
509    
510     1. When using the partial match feature with pcre_dfa_exec(), it was not
511     anchoring the second and subsequent partial matches at the new starting
512     point. This could lead to incorrect results. For example, with the pattern
513     /1234/, partially matching against "123" and then "a4" gave a match.
514    
515     2. Changes to pcregrep:
516    
517     (a) All non-match returns from pcre_exec() were being treated as failures
518     to match the line. Now, unless the error is PCRE_ERROR_NOMATCH, an
519     error message is output. Some extra information is given for the
520     PCRE_ERROR_MATCHLIMIT and PCRE_ERROR_RECURSIONLIMIT errors, which are
521     probably the only errors that are likely to be caused by users (by
522     specifying a regex that has nested indefinite repeats, for instance).
523     If there are more than 20 of these errors, pcregrep is abandoned.
524    
525     (b) A binary zero was treated as data while matching, but terminated the
526     output line if it was written out. This has been fixed: binary zeroes
527     are now no different to any other data bytes.
528    
529     (c) Whichever of the LC_ALL or LC_CTYPE environment variables is set is
530     used to set a locale for matching. The --locale=xxxx long option has
531     been added (no short equivalent) to specify a locale explicitly on the
532     pcregrep command, overriding the environment variables.
533    
534     (d) When -B was used with -n, some line numbers in the output were one less
535     than they should have been.
536    
537     (e) Added the -o (--only-matching) option.
538    
539     (f) If -A or -C was used with -c (count only), some lines of context were
540     accidentally printed for the final match.
541    
542     (g) Added the -H (--with-filename) option.
543    
544     (h) The combination of options -rh failed to suppress file names for files
545     that were found from directory arguments.
546    
547     (i) Added the -D (--devices) and -d (--directories) options.
548    
549     (j) Added the -F (--fixed-strings) option.
550    
551     (k) Allow "-" to be used as a file name for -f as well as for a data file.
552    
553     (l) Added the --colo(u)r option.
554    
555     (m) Added Jeffrey Friedl's -S testing option, but within #ifdefs so that it
556     is not present by default.
557    
558     3. A nasty bug was discovered in the handling of recursive patterns, that is,
559     items such as (?R) or (?1), when the recursion could match a number of
560     alternatives. If it matched one of the alternatives, but subsequently,
561     outside the recursion, there was a failure, the code tried to back up into
562     the recursion. However, because of the way PCRE is implemented, this is not
563     possible, and the result was an incorrect result from the match.
564    
565     In order to prevent this happening, the specification of recursion has
566     been changed so that all such subpatterns are automatically treated as
567     atomic groups. Thus, for example, (?R) is treated as if it were (?>(?R)).
568    
569     4. I had overlooked the fact that, in some locales, there are characters for
570     which isalpha() is true but neither isupper() nor islower() are true. In
571     the fr_FR locale, for instance, the \xAA and \xBA characters (ordmasculine
572     and ordfeminine) are like this. This affected the treatment of \w and \W
573     when they appeared in character classes, but not when they appeared outside
574     a character class. The bit map for "word" characters is now created
575     separately from the results of isalnum() instead of just taking it from the
576     upper, lower, and digit maps. (Plus the underscore character, of course.)
577    
578     5. The above bug also affected the handling of POSIX character classes such as
579     [[:alpha:]] and [[:alnum:]]. These do not have their own bit maps in PCRE's
580     permanent tables. Instead, the bit maps for such a class were previously
581     created as the appropriate unions of the upper, lower, and digit bitmaps.
582     Now they are created by subtraction from the [[:word:]] class, which has
583     its own bitmap.
584    
585     6. The [[:blank:]] character class matches horizontal, but not vertical space.
586     It is created by subtracting the vertical space characters (\x09, \x0a,
587     \x0b, \x0c) from the [[:space:]] bitmap. Previously, however, the
588     subtraction was done in the overall bitmap for a character class, meaning
589     that a class such as [\x0c[:blank:]] was incorrect because \x0c would not
590     be recognized. This bug has been fixed.
591    
592     7. Patches from the folks at Google:
593    
594     (a) pcrecpp.cc: "to handle a corner case that may or may not happen in
595     real life, but is still worth protecting against".
596    
597     (b) pcrecpp.cc: "corrects a bug when negative radixes are used with
598     regular expressions".
599    
600     (c) pcre_scanner.cc: avoid use of std::count() because not all systems
601     have it.
602    
603     (d) Split off pcrecpparg.h from pcrecpp.h and had the former built by
604     "configure" and the latter not, in order to fix a problem somebody had
605     with compiling the Arg class on HP-UX.
606    
607     (e) Improve the error-handling of the C++ wrapper a little bit.
608    
609     (f) New tests for checking recursion limiting.
610    
611     8. The pcre_memmove() function, which is used only if the environment does not
612     have a standard memmove() function (and is therefore rarely compiled),
613     contained two bugs: (a) use of int instead of size_t, and (b) it was not
614     returning a result (though PCRE never actually uses the result).
615    
616     9. In the POSIX regexec() interface, if nmatch is specified as a ridiculously
617     large number - greater than INT_MAX/(3*sizeof(int)) - REG_ESPACE is
618     returned instead of calling malloc() with an overflowing number that would
619     most likely cause subsequent chaos.
620    
621     10. The debugging option of pcretest was not showing the NO_AUTO_CAPTURE flag.
622    
623     11. The POSIX flag REG_NOSUB is now supported. When a pattern that was compiled
624     with this option is matched, the nmatch and pmatch options of regexec() are
625     ignored.
626    
627     12. Added REG_UTF8 to the POSIX interface. This is not defined by POSIX, but is
628     provided in case anyone wants to the the POSIX interface with UTF-8
629     strings.
630    
631     13. Added CXXLDFLAGS to the Makefile parameters to provide settings only on the
632     C++ linking (needed for some HP-UX environments).
633    
634     14. Avoid compiler warnings in get_ucpname() when compiled without UCP support
635     (unused parameter) and in the pcre_printint() function (omitted "default"
636     switch label when the default is to do nothing).
637    
638     15. Added some code to make it possible, when PCRE is compiled as a C++
639     library, to replace subject pointers for pcre_exec() with a smart pointer
640     class, thus making it possible to process discontinuous strings.
641    
642     16. The two macros PCRE_EXPORT and PCRE_DATA_SCOPE are confusing, and perform
643     much the same function. They were added by different people who were trying
644     to make PCRE easy to compile on non-Unix systems. It has been suggested
645     that PCRE_EXPORT be abolished now that there is more automatic apparatus
646     for compiling on Windows systems. I have therefore replaced it with
647     PCRE_DATA_SCOPE. This is set automatically for Windows; if not set it
648     defaults to "extern" for C or "extern C" for C++, which works fine on
649     Unix-like systems. It is now possible to override the value of PCRE_DATA_
650     SCOPE with something explicit in config.h. In addition:
651    
652     (a) pcreposix.h still had just "extern" instead of either of these macros;
653     I have replaced it with PCRE_DATA_SCOPE.
654    
655     (b) Functions such as _pcre_xclass(), which are internal to the library,
656     but external in the C sense, all had PCRE_EXPORT in their definitions.
657     This is apparently wrong for the Windows case, so I have removed it.
658     (It makes no difference on Unix-like systems.)
659    
660     17. Added a new limit, MATCH_LIMIT_RECURSION, which limits the depth of nesting
661     of recursive calls to match(). This is different to MATCH_LIMIT because
662     that limits the total number of calls to match(), not all of which increase
663     the depth of recursion. Limiting the recursion depth limits the amount of
664     stack (or heap if NO_RECURSE is set) that is used. The default can be set
665     when PCRE is compiled, and changed at run time. A patch from Google adds
666     this functionality to the C++ interface.
667    
668     18. Changes to the handling of Unicode character properties:
669    
670     (a) Updated the table to Unicode 4.1.0.
671    
672     (b) Recognize characters that are not in the table as "Cn" (undefined).
673    
674     (c) I revised the way the table is implemented to a much improved format
675     which includes recognition of ranges. It now supports the ranges that
676     are defined in UnicodeData.txt, and it also amalgamates other
677     characters into ranges. This has reduced the number of entries in the
678     table from around 16,000 to around 3,000, thus reducing its size
679     considerably. I realized I did not need to use a tree structure after
680     all - a binary chop search is just as efficient. Having reduced the
681     number of entries, I extended their size from 6 bytes to 8 bytes to
682     allow for more data.
683    
684     (d) Added support for Unicode script names via properties such as \p{Han}.
685    
686     19. In UTF-8 mode, a backslash followed by a non-Ascii character was not
687     matching that character.
688    
689     20. When matching a repeated Unicode property with a minimum greater than zero,
690     (for example \pL{2,}), PCRE could look past the end of the subject if it
691     reached it while seeking the minimum number of characters. This could
692     happen only if some of the characters were more than one byte long, because
693     there is a check for at least the minimum number of bytes.
694    
695     21. Refactored the implementation of \p and \P so as to be more general, to
696     allow for more different types of property in future. This has changed the
697     compiled form incompatibly. Anybody with saved compiled patterns that use
698     \p or \P will have to recompile them.
699    
700     22. Added "Any" and "L&" to the supported property types.
701    
702     23. Recognize \x{...} as a code point specifier, even when not in UTF-8 mode,
703     but give a compile time error if the value is greater than 0xff.
704    
705     24. The man pages for pcrepartial, pcreprecompile, and pcre_compile2 were
706     accidentally not being installed or uninstalled.
707    
708     25. The pcre.h file was built from pcre.h.in, but the only changes that were
709     made were to insert the current release number. This seemed silly, because
710     it made things harder for people building PCRE on systems that don't run
711     "configure". I have turned pcre.h into a distributed file, no longer built
712     by "configure", with the version identification directly included. There is
713     no longer a pcre.h.in file.
714    
715     However, this change necessitated a change to the pcre-config script as
716     well. It is built from pcre-config.in, and one of the substitutions was the
717     release number. I have updated configure.ac so that ./configure now finds
718     the release number by grepping pcre.h.
719    
720     26. Added the ability to run the tests under valgrind.
721    
722    
723 nigel 85 Version 6.4 05-Sep-05
724     ---------------------
725    
726     1. Change 6.0/10/(l) to pcregrep introduced a bug that caused separator lines
727     "--" to be printed when multiple files were scanned, even when none of the
728     -A, -B, or -C options were used. This is not compatible with Gnu grep, so I
729     consider it to be a bug, and have restored the previous behaviour.
730    
731     2. A couple of code tidies to get rid of compiler warnings.
732    
733     3. The pcretest program used to cheat by referring to symbols in the library
734     whose names begin with _pcre_. These are internal symbols that are not
735     really supposed to be visible externally, and in some environments it is
736     possible to suppress them. The cheating is now confined to including
737     certain files from the library's source, which is a bit cleaner.
738    
739     4. Renamed pcre.in as pcre.h.in to go with pcrecpp.h.in; it also makes the
740     file's purpose clearer.
741    
742     5. Reorganized pcre_ucp_findchar().
743    
744    
745 nigel 83 Version 6.3 15-Aug-05
746     ---------------------
747    
748     1. The file libpcre.pc.in did not have general read permission in the tarball.
749    
750     2. There were some problems when building without C++ support:
751    
752     (a) If C++ support was not built, "make install" and "make test" still
753     tried to test it.
754    
755     (b) There were problems when the value of CXX was explicitly set. Some
756     changes have been made to try to fix these, and ...
757    
758     (c) --disable-cpp can now be used to explicitly disable C++ support.
759    
760     (d) The use of @CPP_OBJ@ directly caused a blank line preceded by a
761     backslash in a target when C++ was disabled. This confuses some
762     versions of "make", apparently. Using an intermediate variable solves
763     this. (Same for CPP_LOBJ.)
764    
765     3. $(LINK_FOR_BUILD) now includes $(CFLAGS_FOR_BUILD) and $(LINK)
766     (non-Windows) now includes $(CFLAGS) because these flags are sometimes
767     necessary on certain architectures.
768    
769     4. Added a setting of -export-symbols-regex to the link command to remove
770     those symbols that are exported in the C sense, but actually are local
771     within the library, and not documented. Their names all begin with
772     "_pcre_". This is not a perfect job, because (a) we have to except some
773     symbols that pcretest ("illegally") uses, and (b) the facility isn't always
774     available (and never for static libraries). I have made a note to try to
775     find a way round (a) in the future.
776    
777    
778 nigel 81 Version 6.2 01-Aug-05
779     ---------------------
780    
781     1. There was no test for integer overflow of quantifier values. A construction
782     such as {1111111111111111} would give undefined results. What is worse, if
783     a minimum quantifier for a parenthesized subpattern overflowed and became
784     negative, the calculation of the memory size went wrong. This could have
785     led to memory overwriting.
786    
787     2. Building PCRE using VPATH was broken. Hopefully it is now fixed.
788    
789     3. Added "b" to the 2nd argument of fopen() in dftables.c, for non-Unix-like
790     operating environments where this matters.
791    
792     4. Applied Giuseppe Maxia's patch to add additional features for controlling
793     PCRE options from within the C++ wrapper.
794    
795     5. Named capturing subpatterns were not being correctly counted when a pattern
796     was compiled. This caused two problems: (a) If there were more than 100
797     such subpatterns, the calculation of the memory needed for the whole
798     compiled pattern went wrong, leading to an overflow error. (b) Numerical
799     back references of the form \12, where the number was greater than 9, were
800     not recognized as back references, even though there were sufficient
801     previous subpatterns.
802    
803     6. Two minor patches to pcrecpp.cc in order to allow it to compile on older
804     versions of gcc, e.g. 2.95.4.
805    
806    
807 nigel 79 Version 6.1 21-Jun-05
808     ---------------------
809    
810     1. There was one reference to the variable "posix" in pcretest.c that was not
811     surrounded by "#if !defined NOPOSIX".
812    
813     2. Make it possible to compile pcretest without DFA support, UTF8 support, or
814     the cross-check on the old pcre_info() function, for the benefit of the
815     cut-down version of PCRE that is currently imported into Exim.
816    
817     3. A (silly) pattern starting with (?i)(?-i) caused an internal space
818     allocation error. I've done the easy fix, which wastes 2 bytes for sensible
819     patterns that start (?i) but I don't think that matters. The use of (?i) is
820     just an example; this all applies to the other options as well.
821    
822     4. Since libtool seems to echo the compile commands it is issuing, the output
823     from "make" can be reduced a bit by putting "@" in front of each libtool
824     compile command.
825    
826     5. Patch from the folks at Google for configure.in to be a bit more thorough
827     in checking for a suitable C++ installation before trying to compile the
828     C++ stuff. This should fix a reported problem when a compiler was present,
829     but no suitable headers.
830    
831     6. The man pages all had just "PCRE" as their title. I have changed them to
832     be the relevant file name. I have also arranged that these names are
833     retained in the file doc/pcre.txt, which is a concatenation in text format
834     of all the man pages except the little individual ones for each function.
835    
836     7. The NON-UNIX-USE file had not been updated for the different set of source
837     files that come with release 6. I also added a few comments about the C++
838     wrapper.
839    
840    
841 nigel 77 Version 6.0 07-Jun-05
842     ---------------------
843    
844     1. Some minor internal re-organization to help with my DFA experiments.
845    
846     2. Some missing #ifdef SUPPORT_UCP conditionals in pcretest and printint that
847     didn't matter for the library itself when fully configured, but did matter
848     when compiling without UCP support, or within Exim, where the ucp files are
849     not imported.
850    
851     3. Refactoring of the library code to split up the various functions into
852     different source modules. The addition of the new DFA matching code (see
853     below) to a single monolithic source would have made it really too
854     unwieldy, quite apart from causing all the code to be include in a
855     statically linked application, when only some functions are used. This is
856     relevant even without the DFA addition now that patterns can be compiled in
857     one application and matched in another.
858    
859     The downside of splitting up is that there have to be some external
860     functions and data tables that are used internally in different modules of
861     the library but which are not part of the API. These have all had their
862     names changed to start with "_pcre_" so that they are unlikely to clash
863     with other external names.
864    
865     4. Added an alternate matching function, pcre_dfa_exec(), which matches using
866     a different (DFA) algorithm. Although it is slower than the original
867     function, it does have some advantages for certain types of matching
868     problem.
869    
870     5. Upgrades to pcretest in order to test the features of pcre_dfa_exec(),
871     including restarting after a partial match.
872    
873     6. A patch for pcregrep that defines INVALID_FILE_ATTRIBUTES if it is not
874     defined when compiling for Windows was sent to me. I have put it into the
875     code, though I have no means of testing or verifying it.
876    
877     7. Added the pcre_refcount() auxiliary function.
878    
879     8. Added the PCRE_FIRSTLINE option. This constrains an unanchored pattern to
880     match before or at the first newline in the subject string. In pcretest,
881     the /f option on a pattern can be used to set this.
882    
883     9. A repeated \w when used in UTF-8 mode with characters greater than 256
884     would behave wrongly. This has been present in PCRE since release 4.0.
885    
886     10. A number of changes to the pcregrep command:
887    
888     (a) Refactored how -x works; insert ^(...)$ instead of setting
889     PCRE_ANCHORED and checking the length, in preparation for adding
890     something similar for -w.
891    
892     (b) Added the -w (match as a word) option.
893    
894     (c) Refactored the way lines are read and buffered so as to have more
895     than one at a time available.
896    
897     (d) Implemented a pcregrep test script.
898    
899     (e) Added the -M (multiline match) option. This allows patterns to match
900     over several lines of the subject. The buffering ensures that at least
901     8K, or the rest of the document (whichever is the shorter) is available
902     for matching (and similarly the previous 8K for lookbehind assertions).
903    
904     (f) Changed the --help output so that it now says
905    
906     -w, --word-regex(p)
907    
908     instead of two lines, one with "regex" and the other with "regexp"
909     because that confused at least one person since the short forms are the
910     same. (This required a bit of code, as the output is generated
911     automatically from a table. It wasn't just a text change.)
912    
913     (g) -- can be used to terminate pcregrep options if the next thing isn't an
914     option but starts with a hyphen. Could be a pattern or a path name
915     starting with a hyphen, for instance.
916    
917     (h) "-" can be given as a file name to represent stdin.
918    
919     (i) When file names are being printed, "(standard input)" is used for
920     the standard input, for compatibility with GNU grep. Previously
921     "<stdin>" was used.
922    
923     (j) The option --label=xxx can be used to supply a name to be used for
924     stdin when file names are being printed. There is no short form.
925    
926     (k) Re-factored the options decoding logic because we are going to add
927     two more options that take data. Such options can now be given in four
928     different ways, e.g. "-fname", "-f name", "--file=name", "--file name".
929    
930     (l) Added the -A, -B, and -C options for requesting that lines of context
931     around matches be printed.
932    
933     (m) Added the -L option to print the names of files that do not contain
934     any matching lines, that is, the complement of -l.
935    
936     (n) The return code is 2 if any file cannot be opened, but pcregrep does
937     continue to scan other files.
938    
939     (o) The -s option was incorrectly implemented. For compatibility with other
940     greps, it now suppresses the error message for a non-existent or non-
941     accessible file (but not the return code). There is a new option called
942     -q that suppresses the output of matching lines, which was what -s was
943     previously doing.
944    
945     (p) Added --include and --exclude options to specify files for inclusion
946     and exclusion when recursing.
947    
948     11. The Makefile was not using the Autoconf-supported LDFLAGS macro properly.
949     Hopefully, it now does.
950    
951     12. Missing cast in pcre_study().
952    
953     13. Added an "uninstall" target to the makefile.
954    
955     14. Replaced "extern" in the function prototypes in Makefile.in with
956     "PCRE_DATA_SCOPE", which defaults to 'extern' or 'extern "C"' in the Unix
957     world, but is set differently for Windows.
958    
959     15. Added a second compiling function called pcre_compile2(). The only
960     difference is that it has an extra argument, which is a pointer to an
961     integer error code. When there is a compile-time failure, this is set
962     non-zero, in addition to the error test pointer being set to point to an
963     error message. The new argument may be NULL if no error number is required
964     (but then you may as well call pcre_compile(), which is now just a
965     wrapper). This facility is provided because some applications need a
966     numeric error indication, but it has also enabled me to tidy up the way
967     compile-time errors are handled in the POSIX wrapper.
968    
969     16. Added VPATH=.libs to the makefile; this should help when building with one
970     prefix path and installing with another. (Or so I'm told by someone who
971     knows more about this stuff than I do.)
972    
973     17. Added a new option, REG_DOTALL, to the POSIX function regcomp(). This
974     passes PCRE_DOTALL to the pcre_compile() function, making the "." character
975     match everything, including newlines. This is not POSIX-compatible, but
976     somebody wanted the feature. From pcretest it can be activated by using
977     both the P and the s flags.
978    
979     18. AC_PROG_LIBTOOL appeared twice in Makefile.in. Removed one.
980    
981     19. libpcre.pc was being incorrectly installed as executable.
982    
983     20. A couple of places in pcretest check for end-of-line by looking for '\n';
984     it now also looks for '\r' so that it will work unmodified on Windows.
985    
986     21. Added Google's contributed C++ wrapper to the distribution.
987    
988     22. Added some untidy missing memory free() calls in pcretest, to keep
989     Electric Fence happy when testing.
990    
991    
992    
993 nigel 75 Version 5.0 13-Sep-04
994     ---------------------
995    
996     1. Internal change: literal characters are no longer packed up into items
997     containing multiple characters in a single byte-string. Each character
998     is now matched using a separate opcode. However, there may be more than one
999     byte in the character in UTF-8 mode.
1000    
1001     2. The pcre_callout_block structure has two new fields: pattern_position and
1002     next_item_length. These contain the offset in the pattern to the next match
1003     item, and its length, respectively.
1004    
1005     3. The PCRE_AUTO_CALLOUT option for pcre_compile() requests the automatic
1006     insertion of callouts before each pattern item. Added the /C option to
1007     pcretest to make use of this.
1008    
1009     4. On the advice of a Windows user, the lines
1010    
1011     #if defined(_WIN32) || defined(WIN32)
1012     _setmode( _fileno( stdout ), 0x8000 );
1013     #endif /* defined(_WIN32) || defined(WIN32) */
1014    
1015     have been added to the source of pcretest. This apparently does useful
1016     magic in relation to line terminators.
1017    
1018     5. Changed "r" and "w" in the calls to fopen() in pcretest to "rb" and "wb"
1019     for the benefit of those environments where the "b" makes a difference.
1020    
1021     6. The icc compiler has the same options as gcc, but "configure" doesn't seem
1022     to know about it. I have put a hack into configure.in that adds in code
1023     to set GCC=yes if CC=icc. This seems to end up at a point in the
1024     generated configure script that is early enough to affect the setting of
1025     compiler options, which is what is needed, but I have no means of testing
1026     whether it really works. (The user who reported this had patched the
1027     generated configure script, which of course I cannot do.)
1028    
1029     LATER: After change 22 below (new libtool files), the configure script
1030     seems to know about icc (and also ecc). Therefore, I have commented out
1031     this hack in configure.in.
1032    
1033     7. Added support for pkg-config (2 patches were sent in).
1034    
1035     8. Negated POSIX character classes that used a combination of internal tables
1036     were completely broken. These were [[:^alpha:]], [[:^alnum:]], and
1037     [[:^ascii]]. Typically, they would match almost any characters. The other
1038     POSIX classes were not broken in this way.
1039    
1040     9. Matching the pattern "\b.*?" against "ab cd", starting at offset 1, failed
1041     to find the match, as PCRE was deluded into thinking that the match had to
1042     start at the start point or following a newline. The same bug applied to
1043     patterns with negative forward assertions or any backward assertions
1044     preceding ".*" at the start, unless the pattern required a fixed first
1045     character. This was a failing pattern: "(?!.bcd).*". The bug is now fixed.
1046    
1047     10. In UTF-8 mode, when moving forwards in the subject after a failed match
1048     starting at the last subject character, bytes beyond the end of the subject
1049     string were read.
1050    
1051     11. Renamed the variable "class" as "classbits" to make life easier for C++
1052     users. (Previously there was a macro definition, but it apparently wasn't
1053     enough.)
1054    
1055     12. Added the new field "tables" to the extra data so that tables can be passed
1056     in at exec time, or the internal tables can be re-selected. This allows
1057     a compiled regex to be saved and re-used at a later time by a different
1058     program that might have everything at different addresses.
1059    
1060     13. Modified the pcre-config script so that, when run on Solaris, it shows a
1061     -R library as well as a -L library.
1062    
1063     14. The debugging options of pcretest (-d on the command line or D on a
1064     pattern) showed incorrect output for anything following an extended class
1065     that contained multibyte characters and which was followed by a quantifier.
1066    
1067     15. Added optional support for general category Unicode character properties
1068     via the \p, \P, and \X escapes. Unicode property support implies UTF-8
1069     support. It adds about 90K to the size of the library. The meanings of the
1070     inbuilt class escapes such as \d and \s have NOT been changed.
1071    
1072     16. Updated pcredemo.c to include calls to free() to release the memory for the
1073     compiled pattern.
1074    
1075     17. The generated file chartables.c was being created in the source directory
1076     instead of in the building directory. This caused the build to fail if the
1077     source directory was different from the building directory, and was
1078     read-only.
1079    
1080     18. Added some sample Win commands from Mark Tetrode into the NON-UNIX-USE
1081     file. No doubt somebody will tell me if they don't make sense... Also added
1082     Dan Mooney's comments about building on OpenVMS.
1083    
1084     19. Added support for partial matching via the PCRE_PARTIAL option for
1085     pcre_exec() and the \P data escape in pcretest.
1086    
1087     20. Extended pcretest with 3 new pattern features:
1088    
1089     (i) A pattern option of the form ">rest-of-line" causes pcretest to
1090     write the compiled pattern to the file whose name is "rest-of-line".
1091     This is a straight binary dump of the data, with the saved pointer to
1092     the character tables forced to be NULL. The study data, if any, is
1093     written too. After writing, pcretest reads a new pattern.
1094    
1095     (ii) If, instead of a pattern, "<rest-of-line" is given, pcretest reads a
1096     compiled pattern from the given file. There must not be any
1097     occurrences of "<" in the file name (pretty unlikely); if there are,
1098     pcretest will instead treat the initial "<" as a pattern delimiter.
1099     After reading in the pattern, pcretest goes on to read data lines as
1100     usual.
1101    
1102     (iii) The F pattern option causes pcretest to flip the bytes in the 32-bit
1103     and 16-bit fields in a compiled pattern, to simulate a pattern that
1104     was compiled on a host of opposite endianness.
1105    
1106     21. The pcre-exec() function can now cope with patterns that were compiled on
1107     hosts of opposite endianness, with this restriction:
1108    
1109     As for any compiled expression that is saved and used later, the tables
1110     pointer field cannot be preserved; the extra_data field in the arguments
1111     to pcre_exec() should be used to pass in a tables address if a value
1112     other than the default internal tables were used at compile time.
1113    
1114     22. Calling pcre_exec() with a negative value of the "ovecsize" parameter is
1115     now diagnosed as an error. Previously, most of the time, a negative number
1116     would have been treated as zero, but if in addition "ovector" was passed as
1117     NULL, a crash could occur.
1118    
1119     23. Updated the files ltmain.sh, config.sub, config.guess, and aclocal.m4 with
1120     new versions from the libtool 1.5 distribution (the last one is a copy of
1121     a file called libtool.m4). This seems to have fixed the need to patch
1122     "configure" to support Darwin 1.3 (which I used to do). However, I still
1123     had to patch ltmain.sh to ensure that ${SED} is set (it isn't on my
1124     workstation).
1125    
1126     24. Changed the PCRE licence to be the more standard "BSD" licence.
1127    
1128    
1129 nigel 73 Version 4.5 01-Dec-03
1130     ---------------------
1131    
1132     1. There has been some re-arrangement of the code for the match() function so
1133     that it can be compiled in a version that does not call itself recursively.
1134     Instead, it keeps those local variables that need separate instances for
1135     each "recursion" in a frame on the heap, and gets/frees frames whenever it
1136     needs to "recurse". Keeping track of where control must go is done by means
1137     of setjmp/longjmp. The whole thing is implemented by a set of macros that
1138     hide most of the details from the main code, and operates only if
1139     NO_RECURSE is defined while compiling pcre.c. If PCRE is built using the
1140     "configure" mechanism, "--disable-stack-for-recursion" turns on this way of
1141     operating.
1142    
1143     To make it easier for callers to provide specially tailored get/free
1144     functions for this usage, two new functions, pcre_stack_malloc, and
1145     pcre_stack_free, are used. They are always called in strict stacking order,
1146     and the size of block requested is always the same.
1147    
1148     The PCRE_CONFIG_STACKRECURSE info parameter can be used to find out whether
1149     PCRE has been compiled to use the stack or the heap for recursion. The
1150     -C option of pcretest uses this to show which version is compiled.
1151    
1152     A new data escape \S, is added to pcretest; it causes the amounts of store
1153     obtained and freed by both kinds of malloc/free at match time to be added
1154     to the output.
1155    
1156     2. Changed the locale test to use "fr_FR" instead of "fr" because that's
1157     what's available on my current Linux desktop machine.
1158    
1159     3. When matching a UTF-8 string, the test for a valid string at the start has
1160     been extended. If start_offset is not zero, PCRE now checks that it points
1161     to a byte that is the start of a UTF-8 character. If not, it returns
1162     PCRE_ERROR_BADUTF8_OFFSET (-11). Note: the whole string is still checked;
1163     this is necessary because there may be backward assertions in the pattern.
1164     When matching the same subject several times, it may save resources to use
1165     PCRE_NO_UTF8_CHECK on all but the first call if the string is long.
1166    
1167     4. The code for checking the validity of UTF-8 strings has been tightened so
1168     that it rejects (a) strings containing 0xfe or 0xff bytes and (b) strings
1169     containing "overlong sequences".
1170    
1171     5. Fixed a bug (appearing twice) that I could not find any way of exploiting!
1172     I had written "if ((digitab[*p++] && chtab_digit) == 0)" where the "&&"
1173     should have been "&", but it just so happened that all the cases this let
1174     through by mistake were picked up later in the function.
1175    
1176     6. I had used a variable called "isblank" - this is a C99 function, causing
1177     some compilers to warn. To avoid this, I renamed it (as "blankclass").
1178    
1179     7. Cosmetic: (a) only output another newline at the end of pcretest if it is
1180     prompting; (b) run "./pcretest /dev/null" at the start of the test script
1181     so the version is shown; (c) stop "make test" echoing "./RunTest".
1182    
1183     8. Added patches from David Burgess to enable PCRE to run on EBCDIC systems.
1184    
1185     9. The prototype for memmove() for systems that don't have it was using
1186     size_t, but the inclusion of the header that defines size_t was later. I've
1187     moved the #includes for the C headers earlier to avoid this.
1188    
1189     10. Added some adjustments to the code to make it easier to compiler on certain
1190     special systems:
1191    
1192     (a) Some "const" qualifiers were missing.
1193     (b) Added the macro EXPORT before all exported functions; by default this
1194     is defined to be empty.
1195     (c) Changed the dftables auxiliary program (that builds chartables.c) so
1196     that it reads its output file name as an argument instead of writing
1197     to the standard output and assuming this can be redirected.
1198    
1199     11. In UTF-8 mode, if a recursive reference (e.g. (?1)) followed a character
1200     class containing characters with values greater than 255, PCRE compilation
1201     went into a loop.
1202    
1203     12. A recursive reference to a subpattern that was within another subpattern
1204     that had a minimum quantifier of zero caused PCRE to crash. For example,
1205     (x(y(?2))z)? provoked this bug with a subject that got as far as the
1206     recursion. If the recursively-called subpattern itself had a zero repeat,
1207     that was OK.
1208    
1209     13. In pcretest, the buffer for reading a data line was set at 30K, but the
1210     buffer into which it was copied (for escape processing) was still set at
1211     1024, so long lines caused crashes.
1212    
1213     14. A pattern such as /[ab]{1,3}+/ failed to compile, giving the error
1214     "internal error: code overflow...". This applied to any character class
1215     that was followed by a possessive quantifier.
1216    
1217     15. Modified the Makefile to add libpcre.la as a prerequisite for
1218     libpcreposix.la because I was told this is needed for a parallel build to
1219     work.
1220    
1221     16. If a pattern that contained .* following optional items at the start was
1222     studied, the wrong optimizing data was generated, leading to matching
1223     errors. For example, studying /[ab]*.*c/ concluded, erroneously, that any
1224     matching string must start with a or b or c. The correct conclusion for
1225     this pattern is that a match can start with any character.
1226    
1227    
1228 nigel 71 Version 4.4 13-Aug-03
1229     ---------------------
1230    
1231     1. In UTF-8 mode, a character class containing characters with values between
1232     127 and 255 was not handled correctly if the compiled pattern was studied.
1233     In fixing this, I have also improved the studying algorithm for such
1234     classes (slightly).
1235    
1236     2. Three internal functions had redundant arguments passed to them. Removal
1237     might give a very teeny performance improvement.
1238    
1239     3. Documentation bug: the value of the capture_top field in a callout is *one
1240     more than* the number of the hightest numbered captured substring.
1241    
1242     4. The Makefile linked pcretest and pcregrep with -lpcre, which could result
1243     in incorrectly linking with a previously installed version. They now link
1244     explicitly with libpcre.la.
1245    
1246     5. configure.in no longer needs to recognize Cygwin specially.
1247    
1248     6. A problem in pcre.in for Windows platforms is fixed.
1249    
1250     7. If a pattern was successfully studied, and the -d (or /D) flag was given to
1251     pcretest, it used to include the size of the study block as part of its
1252     output. Unfortunately, the structure contains a field that has a different
1253     size on different hardware architectures. This meant that the tests that
1254     showed this size failed. As the block is currently always of a fixed size,
1255     this information isn't actually particularly useful in pcretest output, so
1256     I have just removed it.
1257    
1258     8. Three pre-processor statements accidentally did not start in column 1.
1259     Sadly, there are *still* compilers around that complain, even though
1260     standard C has not required this for well over a decade. Sigh.
1261    
1262     9. In pcretest, the code for checking callouts passed small integers in the
1263     callout_data field, which is a void * field. However, some picky compilers
1264     complained about the casts involved for this on 64-bit systems. Now
1265     pcretest passes the address of the small integer instead, which should get
1266     rid of the warnings.
1267    
1268     10. By default, when in UTF-8 mode, PCRE now checks for valid UTF-8 strings at
1269     both compile and run time, and gives an error if an invalid UTF-8 sequence
1270     is found. There is a option for disabling this check in cases where the
1271     string is known to be correct and/or the maximum performance is wanted.
1272    
1273     11. In response to a bug report, I changed one line in Makefile.in from
1274    
1275     -Wl,--out-implib,.libs/lib@WIN_PREFIX@pcreposix.dll.a \
1276     to
1277     -Wl,--out-implib,.libs/@WIN_PREFIX@libpcreposix.dll.a \
1278    
1279     to look similar to other lines, but I have no way of telling whether this
1280     is the right thing to do, as I do not use Windows. No doubt I'll get told
1281     if it's wrong...
1282    
1283    
1284 nigel 69 Version 4.3 21-May-03
1285     ---------------------
1286    
1287     1. Two instances of @WIN_PREFIX@ omitted from the Windows targets in the
1288     Makefile.
1289    
1290     2. Some refactoring to improve the quality of the code:
1291    
1292     (i) The utf8_table... variables are now declared "const".
1293    
1294     (ii) The code for \cx, which used the "case flipping" table to upper case
1295     lower case letters, now just substracts 32. This is ASCII-specific,
1296     but the whole concept of \cx is ASCII-specific, so it seems
1297     reasonable.
1298    
1299     (iii) PCRE was using its character types table to recognize decimal and
1300     hexadecimal digits in the pattern. This is silly, because it handles
1301     only 0-9, a-f, and A-F, but the character types table is locale-
1302     specific, which means strange things might happen. A private
1303     table is now used for this - though it costs 256 bytes, a table is
1304     much faster than multiple explicit tests. Of course, the standard
1305     character types table is still used for matching digits in subject
1306     strings against \d.
1307    
1308     (iv) Strictly, the identifier ESC_t is reserved by POSIX (all identifiers
1309     ending in _t are). So I've renamed it as ESC_tee.
1310    
1311     3. The first argument for regexec() in the POSIX wrapper should have been
1312     defined as "const".
1313    
1314     4. Changed pcretest to use malloc() for its buffers so that they can be
1315     Electric Fenced for debugging.
1316    
1317     5. There were several places in the code where, in UTF-8 mode, PCRE would try
1318     to read one or more bytes before the start of the subject string. Often this
1319     had no effect on PCRE's behaviour, but in some circumstances it could
1320     provoke a segmentation fault.
1321    
1322     6. A lookbehind at the start of a pattern in UTF-8 mode could also cause PCRE
1323     to try to read one or more bytes before the start of the subject string.
1324    
1325     7. A lookbehind in a pattern matched in non-UTF-8 mode on a PCRE compiled with
1326     UTF-8 support could misbehave in various ways if the subject string
1327     contained bytes with the 0x80 bit set and the 0x40 bit unset in a lookbehind
1328     area. (PCRE was not checking for the UTF-8 mode flag, and trying to move
1329     back over UTF-8 characters.)
1330    
1331    
1332 nigel 67 Version 4.2 14-Apr-03
1333     ---------------------
1334    
1335     1. Typo "#if SUPPORT_UTF8" instead of "#ifdef SUPPORT_UTF8" fixed.
1336    
1337     2. Changes to the building process, supplied by Ronald Landheer-Cieslak
1338     [ON_WINDOWS]: new variable, "#" on non-Windows platforms
1339     [NOT_ON_WINDOWS]: new variable, "#" on Windows platforms
1340     [WIN_PREFIX]: new variable, "cyg" for Cygwin
1341     * Makefile.in: use autoconf substitution for OBJEXT, EXEEXT, BUILD_OBJEXT
1342     and BUILD_EXEEXT
1343     Note: automatic setting of the BUILD variables is not yet working
1344     set CPPFLAGS and BUILD_CPPFLAGS (but don't use yet) - should be used at
1345     compile-time but not at link-time
1346     [LINK]: use for linking executables only
1347     make different versions for Windows and non-Windows
1348     [LINKLIB]: new variable, copy of UNIX-style LINK, used for linking
1349     libraries
1350     [LINK_FOR_BUILD]: new variable
1351     [OBJEXT]: use throughout
1352     [EXEEXT]: use throughout
1353     <winshared>: new target
1354     <wininstall>: new target
1355     <dftables.o>: use native compiler
1356     <dftables>: use native linker
1357     <install>: handle Windows platform correctly
1358     <clean>: ditto
1359     <check>: ditto
1360     copy DLL to top builddir before testing
1361    
1362     As part of these changes, -no-undefined was removed again. This was reported
1363     to give trouble on HP-UX 11.0, so getting rid of it seems like a good idea
1364     in any case.
1365    
1366     3. Some tidies to get rid of compiler warnings:
1367    
1368     . In the match_data structure, match_limit was an unsigned long int, whereas
1369     match_call_count was an int. I've made them both unsigned long ints.
1370    
1371     . In pcretest the fact that a const uschar * doesn't automatically cast to
1372     a void * provoked a warning.
1373    
1374     . Turning on some more compiler warnings threw up some "shadow" variables
1375     and a few more missing casts.
1376    
1377     4. If PCRE was complied with UTF-8 support, but called without the PCRE_UTF8
1378     option, a class that contained a single character with a value between 128
1379     and 255 (e.g. /[\xFF]/) caused PCRE to crash.
1380    
1381     5. If PCRE was compiled with UTF-8 support, but called without the PCRE_UTF8
1382     option, a class that contained several characters, but with at least one
1383     whose value was between 128 and 255 caused PCRE to crash.
1384    
1385    
1386 nigel 65 Version 4.1 12-Mar-03
1387     ---------------------
1388 nigel 63
1389 nigel 65 1. Compiling with gcc -pedantic found a couple of places where casts were
1390     needed, and a string in dftables.c that was longer than standard compilers are
1391     required to support.
1392    
1393     2. Compiling with Sun's compiler found a few more places where the code could
1394     be tidied up in order to avoid warnings.
1395    
1396     3. The variables for cross-compiling were called HOST_CC and HOST_CFLAGS; the
1397     first of these names is deprecated in the latest Autoconf in favour of the name
1398     CC_FOR_BUILD, because "host" is typically used to mean the system on which the
1399     compiled code will be run. I can't find a reference for HOST_CFLAGS, but by
1400     analogy I have changed it to CFLAGS_FOR_BUILD.
1401    
1402     4. Added -no-undefined to the linking command in the Makefile, because this is
1403     apparently helpful for Windows. To make it work, also added "-L. -lpcre" to the
1404     linking step for the pcreposix library.
1405    
1406     5. PCRE was failing to diagnose the case of two named groups with the same
1407     name.
1408    
1409     6. A problem with one of PCRE's optimizations was discovered. PCRE remembers a
1410     literal character that is needed in the subject for a match, and scans along to
1411     ensure that it is present before embarking on the full matching process. This
1412     saves time in cases of nested unlimited repeats that are never going to match.
1413     Problem: the scan can take a lot of time if the subject is very long (e.g.
1414     megabytes), thus penalizing straightforward matches. It is now done only if the
1415     amount of subject to be scanned is less than 1000 bytes.
1416    
1417     7. A lesser problem with the same optimization is that it was recording the
1418     first character of an anchored pattern as "needed", thus provoking a search
1419     right along the subject, even when the first match of the pattern was going to
1420     fail. The "needed" character is now not set for anchored patterns, unless it
1421     follows something in the pattern that is of non-fixed length. Thus, it still
1422     fulfils its original purpose of finding quick non-matches in cases of nested
1423     unlimited repeats, but isn't used for simple anchored patterns such as /^abc/.
1424    
1425    
1426     Version 4.0 17-Feb-03
1427     ---------------------
1428    
1429 nigel 63 1. If a comment in an extended regex that started immediately after a meta-item
1430     extended to the end of string, PCRE compiled incorrect data. This could lead to
1431     all kinds of weird effects. Example: /#/ was bad; /()#/ was bad; /a#/ was not.
1432    
1433     2. Moved to autoconf 2.53 and libtool 1.4.2.
1434    
1435     3. Perl 5.8 no longer needs "use utf8" for doing UTF-8 things. Consequently,
1436     the special perltest8 script is no longer needed - all the tests can be run
1437     from a single perltest script.
1438    
1439     4. From 5.004, Perl has not included the VT character (0x0b) in the set defined
1440     by \s. It has now been removed in PCRE. This means it isn't recognized as
1441     whitespace in /x regexes too, which is the same as Perl. Note that the POSIX
1442     class [:space:] *does* include VT, thereby creating a mess.
1443    
1444     5. Added the class [:blank:] (a GNU extension from Perl 5.8) to match only
1445     space and tab.
1446    
1447     6. Perl 5.005 was a long time ago. It's time to amalgamate the tests that use
1448     its new features into the main test script, reducing the number of scripts.
1449    
1450     7. Perl 5.8 has changed the meaning of patterns like /a(?i)b/. Earlier versions
1451     were backward compatible, and made the (?i) apply to the whole pattern, as if
1452     /i were given. Now it behaves more logically, and applies the option setting
1453     only to what follows. PCRE has been changed to follow suit. However, if it
1454     finds options settings right at the start of the pattern, it extracts them into
1455     the global options, as before. Thus, they show up in the info data.
1456    
1457     8. Added support for the \Q...\E escape sequence. Characters in between are
1458     treated as literals. This is slightly different from Perl in that $ and @ are
1459     also handled as literals inside the quotes. In Perl, they will cause variable
1460     interpolation. Note the following examples:
1461    
1462     Pattern PCRE matches Perl matches
1463    
1464     \Qabc$xyz\E abc$xyz abc followed by the contents of $xyz
1465     \Qabc\$xyz\E abc\$xyz abc\$xyz
1466     \Qabc\E\$\Qxyz\E abc$xyz abc$xyz
1467    
1468     For compatibility with Perl, \Q...\E sequences are recognized inside character
1469     classes as well as outside them.
1470    
1471     9. Re-organized 3 code statements in pcretest to avoid "overflow in
1472     floating-point constant arithmetic" warnings from a Microsoft compiler. Added a
1473     (size_t) cast to one statement in pcretest and one in pcreposix to avoid
1474     signed/unsigned warnings.
1475    
1476     10. SunOS4 doesn't have strtoul(). This was used only for unpicking the -o
1477     option for pcretest, so I've replaced it by a simple function that does just
1478     that job.
1479    
1480     11. pcregrep was ending with code 0 instead of 2 for the commands "pcregrep" or
1481     "pcregrep -".
1482    
1483     12. Added "possessive quantifiers" ?+, *+, ++, and {,}+ which come from Sun's
1484     Java package. This provides some syntactic sugar for simple cases of what my
1485     documentation calls "once-only subpatterns". A pattern such as x*+ is the same
1486     as (?>x*). In other words, if what is inside (?>...) is just a single repeated
1487     item, you can use this simplified notation. Note that only makes sense with
1488     greedy quantifiers. Consequently, the use of the possessive quantifier forces
1489     greediness, whatever the setting of the PCRE_UNGREEDY option.
1490    
1491     13. A change of greediness default within a pattern was not taking effect at
1492     the current level for patterns like /(b+(?U)a+)/. It did apply to parenthesized
1493     subpatterns that followed. Patterns like /b+(?U)a+/ worked because the option
1494     was abstracted outside.
1495    
1496     14. PCRE now supports the \G assertion. It is true when the current matching
1497     position is at the start point of the match. This differs from \A when the
1498     starting offset is non-zero. Used with the /g option of pcretest (or similar
1499     code), it works in the same way as it does for Perl's /g option. If all
1500     alternatives of a regex begin with \G, the expression is anchored to the start
1501     match position, and the "anchored" flag is set in the compiled expression.
1502    
1503     15. Some bugs concerning the handling of certain option changes within patterns
1504     have been fixed. These applied to options other than (?ims). For example,
1505     "a(?x: b c )d" did not match "XabcdY" but did match "Xa b c dY". It should have
1506     been the other way round. Some of this was related to change 7 above.
1507    
1508     16. PCRE now gives errors for /[.x.]/ and /[=x=]/ as unsupported POSIX
1509     features, as Perl does. Previously, PCRE gave the warnings only for /[[.x.]]/
1510     and /[[=x=]]/. PCRE now also gives an error for /[:name:]/ because it supports
1511     POSIX classes only within a class (e.g. /[[:alpha:]]/).
1512    
1513     17. Added support for Perl's \C escape. This matches one byte, even in UTF8
1514     mode. Unlike ".", it always matches newline, whatever the setting of
1515     PCRE_DOTALL. However, PCRE does not permit \C to appear in lookbehind
1516     assertions. Perl allows it, but it doesn't (in general) work because it can't
1517     calculate the length of the lookbehind. At least, that's the case for Perl
1518     5.8.0 - I've been told they are going to document that it doesn't work in
1519     future.
1520    
1521     18. Added an error diagnosis for escapes that PCRE does not support: these are
1522     \L, \l, \N, \P, \p, \U, \u, and \X.
1523    
1524     19. Although correctly diagnosing a missing ']' in a character class, PCRE was
1525     reading past the end of the pattern in cases such as /[abcd/.
1526    
1527     20. PCRE was getting more memory than necessary for patterns with classes that
1528     contained both POSIX named classes and other characters, e.g. /[[:space:]abc/.
1529    
1530     21. Added some code, conditional on #ifdef VPCOMPAT, to make life easier for
1531     compiling PCRE for use with Virtual Pascal.
1532    
1533     22. Small fix to the Makefile to make it work properly if the build is done
1534     outside the source tree.
1535    
1536     23. Added a new extension: a condition to go with recursion. If a conditional
1537     subpattern starts with (?(R) the "true" branch is used if recursion has
1538     happened, whereas the "false" branch is used only at the top level.
1539    
1540     24. When there was a very long string of literal characters (over 255 bytes
1541     without UTF support, over 250 bytes with UTF support), the computation of how
1542     much memory was required could be incorrect, leading to segfaults or other
1543     strange effects.
1544    
1545     25. PCRE was incorrectly assuming anchoring (either to start of subject or to
1546     start of line for a non-DOTALL pattern) when a pattern started with (.*) and
1547     there was a subsequent back reference to those brackets. This meant that, for
1548     example, /(.*)\d+\1/ failed to match "abc123bc". Unfortunately, it isn't
1549     possible to check for precisely this case. All we can do is abandon the
1550     optimization if .* occurs inside capturing brackets when there are any back
1551     references whatsoever. (See below for a better fix that came later.)
1552    
1553     26. The handling of the optimization for finding the first character of a
1554     non-anchored pattern, and for finding a character that is required later in the
1555     match were failing in some cases. This didn't break the matching; it just
1556     failed to optimize when it could. The way this is done has been re-implemented.
1557    
1558     27. Fixed typo in error message for invalid (?R item (it said "(?p").
1559    
1560     28. Added a new feature that provides some of the functionality that Perl
1561     provides with (?{...}). The facility is termed a "callout". The way it is done
1562     in PCRE is for the caller to provide an optional function, by setting
1563     pcre_callout to its entry point. Like pcre_malloc and pcre_free, this is a
1564     global variable. By default it is unset, which disables all calling out. To get
1565     the function called, the regex must include (?C) at appropriate points. This
1566     is, in fact, equivalent to (?C0), and any number <= 255 may be given with (?C).
1567     This provides a means of identifying different callout points. When PCRE
1568     reaches such a point in the regex, if pcre_callout has been set, the external
1569     function is called. It is provided with data in a structure called
1570     pcre_callout_block, which is defined in pcre.h. If the function returns 0,
1571     matching continues; if it returns a non-zero value, the match at the current
1572     point fails. However, backtracking will occur if possible. [This was changed
1573     later and other features added - see item 49 below.]
1574    
1575     29. pcretest is upgraded to test the callout functionality. It provides a
1576     callout function that displays information. By default, it shows the start of
1577     the match and the current position in the text. There are some new data escapes
1578     to vary what happens:
1579    
1580     \C+ in addition, show current contents of captured substrings
1581     \C- do not supply a callout function
1582     \C!n return 1 when callout number n is reached
1583     \C!n!m return 1 when callout number n is reached for the mth time
1584    
1585     30. If pcregrep was called with the -l option and just a single file name, it
1586     output "<stdin>" if a match was found, instead of the file name.
1587    
1588     31. Improve the efficiency of the POSIX API to PCRE. If the number of capturing
1589     slots is less than POSIX_MALLOC_THRESHOLD, use a block on the stack to pass to
1590     pcre_exec(). This saves a malloc/free per call. The default value of
1591     POSIX_MALLOC_THRESHOLD is 10; it can be changed by --with-posix-malloc-threshold
1592     when configuring.
1593    
1594     32. The default maximum size of a compiled pattern is 64K. There have been a
1595     few cases of people hitting this limit. The code now uses macros to handle the
1596     storing of links as offsets within the compiled pattern. It defaults to 2-byte
1597     links, but this can be changed to 3 or 4 bytes by --with-link-size when
1598     configuring. Tests 2 and 5 work only with 2-byte links because they output
1599     debugging information about compiled patterns.
1600    
1601     33. Internal code re-arrangements:
1602    
1603     (a) Moved the debugging function for printing out a compiled regex into
1604     its own source file (printint.c) and used #include to pull it into
1605     pcretest.c and, when DEBUG is defined, into pcre.c, instead of having two
1606     separate copies.
1607    
1608     (b) Defined the list of op-code names for debugging as a macro in
1609     internal.h so that it is next to the definition of the opcodes.
1610    
1611     (c) Defined a table of op-code lengths for simpler skipping along compiled
1612     code. This is again a macro in internal.h so that it is next to the
1613     definition of the opcodes.
1614    
1615     34. Added support for recursive calls to individual subpatterns, along the
1616     lines of Robin Houston's patch (but implemented somewhat differently).
1617    
1618     35. Further mods to the Makefile to help Win32. Also, added code to pcregrep to
1619     allow it to read and process whole directories in Win32. This code was
1620     contributed by Lionel Fourquaux; it has not been tested by me.
1621    
1622     36. Added support for named subpatterns. The Python syntax (?P<name>...) is
1623     used to name a group. Names consist of alphanumerics and underscores, and must
1624     be unique. Back references use the syntax (?P=name) and recursive calls use
1625     (?P>name) which is a PCRE extension to the Python extension. Groups still have
1626     numbers. The function pcre_fullinfo() can be used after compilation to extract
1627     a name/number map. There are three relevant calls:
1628    
1629     PCRE_INFO_NAMEENTRYSIZE yields the size of each entry in the map
1630     PCRE_INFO_NAMECOUNT yields the number of entries
1631     PCRE_INFO_NAMETABLE yields a pointer to the map.
1632    
1633     The map is a vector of fixed-size entries. The size of each entry depends on
1634     the length of the longest name used. The first two bytes of each entry are the
1635     group number, most significant byte first. There follows the corresponding
1636     name, zero terminated. The names are in alphabetical order.
1637    
1638     37. Make the maximum literal string in the compiled code 250 for the non-UTF-8
1639     case instead of 255. Making it the same both with and without UTF-8 support
1640     means that the same test output works with both.
1641    
1642     38. There was a case of malloc(0) in the POSIX testing code in pcretest. Avoid
1643     calling malloc() with a zero argument.
1644    
1645     39. Change 25 above had to resort to a heavy-handed test for the .* anchoring
1646     optimization. I've improved things by keeping a bitmap of backreferences with
1647     numbers 1-31 so that if .* occurs inside capturing brackets that are not in
1648     fact referenced, the optimization can be applied. It is unlikely that a
1649     relevant occurrence of .* (i.e. one which might indicate anchoring or forcing
1650     the match to follow \n) will appear inside brackets with a number greater than
1651     31, but if it does, any back reference > 31 suppresses the optimization.
1652    
1653     40. Added a new compile-time option PCRE_NO_AUTO_CAPTURE. This has the effect
1654     of disabling numbered capturing parentheses. Any opening parenthesis that is
1655     not followed by ? behaves as if it were followed by ?: but named parentheses
1656     can still be used for capturing (and they will acquire numbers in the usual
1657     way).
1658    
1659     41. Redesigned the return codes from the match() function into yes/no/error so
1660     that errors can be passed back from deep inside the nested calls. A malloc
1661     failure while inside a recursive subpattern call now causes the
1662     PCRE_ERROR_NOMEMORY return instead of quietly going wrong.
1663    
1664     42. It is now possible to set a limit on the number of times the match()
1665     function is called in a call to pcre_exec(). This facility makes it possible to
1666     limit the amount of recursion and backtracking, though not in a directly
1667     obvious way, because the match() function is used in a number of different
1668     circumstances. The count starts from zero for each position in the subject
1669     string (for non-anchored patterns). The default limit is, for compatibility, a
1670     large number, namely 10 000 000. You can change this in two ways:
1671    
1672     (a) When configuring PCRE before making, you can use --with-match-limit=n
1673     to set a default value for the compiled library.
1674    
1675     (b) For each call to pcre_exec(), you can pass a pcre_extra block in which
1676     a different value is set. See 45 below.
1677    
1678     If the limit is exceeded, pcre_exec() returns PCRE_ERROR_MATCHLIMIT.
1679    
1680     43. Added a new function pcre_config(int, void *) to enable run-time extraction
1681     of things that can be changed at compile time. The first argument specifies
1682     what is wanted and the second points to where the information is to be placed.
1683     The current list of available information is:
1684    
1685     PCRE_CONFIG_UTF8
1686    
1687     The output is an integer that is set to one if UTF-8 support is available;
1688     otherwise it is set to zero.
1689    
1690     PCRE_CONFIG_NEWLINE
1691    
1692     The output is an integer that it set to the value of the code that is used for
1693     newline. It is either LF (10) or CR (13).
1694    
1695     PCRE_CONFIG_LINK_SIZE
1696    
1697     The output is an integer that contains the number of bytes used for internal
1698     linkage in compiled expressions. The value is 2, 3, or 4. See item 32 above.
1699    
1700     PCRE_CONFIG_POSIX_MALLOC_THRESHOLD
1701    
1702     The output is an integer that contains the threshold above which the POSIX
1703     interface uses malloc() for output vectors. See item 31 above.
1704    
1705     PCRE_CONFIG_MATCH_LIMIT
1706    
1707     The output is an unsigned integer that contains the default limit of the number
1708     of match() calls in a pcre_exec() execution. See 42 above.
1709    
1710     44. pcretest has been upgraded by the addition of the -C option. This causes it
1711     to extract all the available output from the new pcre_config() function, and to
1712     output it. The program then exits immediately.
1713    
1714     45. A need has arisen to pass over additional data with calls to pcre_exec() in
1715     order to support additional features. One way would have been to define
1716     pcre_exec2() (for example) with extra arguments, but this would not have been
1717     extensible, and would also have required all calls to the original function to
1718     be mapped to the new one. Instead, I have chosen to extend the mechanism that
1719     is used for passing in "extra" data from pcre_study().
1720    
1721     The pcre_extra structure is now exposed and defined in pcre.h. It currently
1722     contains the following fields:
1723    
1724     flags a bitmap indicating which of the following fields are set
1725     study_data opaque data from pcre_study()
1726     match_limit a way of specifying a limit on match() calls for a specific
1727     call to pcre_exec()
1728     callout_data data for callouts (see 49 below)
1729    
1730     The flag bits are also defined in pcre.h, and are
1731    
1732     PCRE_EXTRA_STUDY_DATA
1733     PCRE_EXTRA_MATCH_LIMIT
1734     PCRE_EXTRA_CALLOUT_DATA
1735    
1736     The pcre_study() function now returns one of these new pcre_extra blocks, with
1737     the actual study data pointed to by the study_data field, and the
1738     PCRE_EXTRA_STUDY_DATA flag set. This can be passed directly to pcre_exec() as
1739     before. That is, this change is entirely upwards-compatible and requires no
1740     change to existing code.
1741    
1742     If you want to pass in additional data to pcre_exec(), you can either place it
1743     in a pcre_extra block provided by pcre_study(), or create your own pcre_extra
1744     block.
1745    
1746     46. pcretest has been extended to test the PCRE_EXTRA_MATCH_LIMIT feature. If a
1747     data string contains the escape sequence \M, pcretest calls pcre_exec() several
1748     times with different match limits, until it finds the minimum value needed for
1749     pcre_exec() to complete. The value is then output. This can be instructive; for
1750     most simple matches the number is quite small, but for pathological cases it
1751     gets very large very quickly.
1752    
1753     47. There's a new option for pcre_fullinfo() called PCRE_INFO_STUDYSIZE. It
1754     returns the size of the data block pointed to by the study_data field in a
1755     pcre_extra block, that is, the value that was passed as the argument to
1756     pcre_malloc() when PCRE was getting memory in which to place the information
1757     created by pcre_study(). The fourth argument should point to a size_t variable.
1758     pcretest has been extended so that this information is shown after a successful
1759     pcre_study() call when information about the compiled regex is being displayed.
1760    
1761     48. Cosmetic change to Makefile: there's no need to have / after $(DESTDIR)
1762     because what follows is always an absolute path. (Later: it turns out that this
1763     is more than cosmetic for MinGW, because it doesn't like empty path
1764     components.)
1765    
1766     49. Some changes have been made to the callout feature (see 28 above):
1767    
1768     (i) A callout function now has three choices for what it returns:
1769    
1770     0 => success, carry on matching
1771     > 0 => failure at this point, but backtrack if possible
1772     < 0 => serious error, return this value from pcre_exec()
1773    
1774     Negative values should normally be chosen from the set of PCRE_ERROR_xxx
1775     values. In particular, returning PCRE_ERROR_NOMATCH forces a standard
1776     "match failed" error. The error number PCRE_ERROR_CALLOUT is reserved for
1777     use by callout functions. It will never be used by PCRE itself.
1778    
1779     (ii) The pcre_extra structure (see 45 above) has a void * field called
1780     callout_data, with corresponding flag bit PCRE_EXTRA_CALLOUT_DATA. The
1781     pcre_callout_block structure has a field of the same name. The contents of
1782     the field passed in the pcre_extra structure are passed to the callout
1783     function in the corresponding field in the callout block. This makes it
1784     easier to use the same callout-containing regex from multiple threads. For
1785     testing, the pcretest program has a new data escape
1786    
1787     \C*n pass the number n (may be negative) as callout_data
1788    
1789     If the callout function in pcretest receives a non-zero value as
1790     callout_data, it returns that value.
1791    
1792     50. Makefile wasn't handling CFLAGS properly when compiling dftables. Also,
1793     there were some redundant $(CFLAGS) in commands that are now specified as
1794     $(LINK), which already includes $(CFLAGS).
1795    
1796     51. Extensions to UTF-8 support are listed below. These all apply when (a) PCRE
1797     has been compiled with UTF-8 support *and* pcre_compile() has been compiled
1798     with the PCRE_UTF8 flag. Patterns that are compiled without that flag assume
1799     one-byte characters throughout. Note that case-insensitive matching applies
1800     only to characters whose values are less than 256. PCRE doesn't support the
1801     notion of cases for higher-valued characters.
1802    
1803     (i) A character class whose characters are all within 0-255 is handled as
1804     a bit map, and the map is inverted for negative classes. Previously, a
1805     character > 255 always failed to match such a class; however it should
1806     match if the class was a negative one (e.g. [^ab]). This has been fixed.
1807    
1808     (ii) A negated character class with a single character < 255 is coded as
1809     "not this character" (OP_NOT). This wasn't working properly when the test
1810     character was multibyte, either singly or repeated.
1811    
1812     (iii) Repeats of multibyte characters are now handled correctly in UTF-8
1813     mode, for example: \x{100}{2,3}.
1814    
1815     (iv) The character escapes \b, \B, \d, \D, \s, \S, \w, and \W (either
1816     singly or repeated) now correctly test multibyte characters. However,
1817     PCRE doesn't recognize any characters with values greater than 255 as
1818     digits, spaces, or word characters. Such characters always match \D, \S,
1819     and \W, and never match \d, \s, or \w.
1820    
1821     (v) Classes may now contain characters and character ranges with values
1822     greater than 255. For example: [ab\x{100}-\x{400}].
1823    
1824     (vi) pcregrep now has a --utf-8 option (synonym -u) which makes it call
1825     PCRE in UTF-8 mode.
1826    
1827     52. The info request value PCRE_INFO_FIRSTCHAR has been renamed
1828     PCRE_INFO_FIRSTBYTE because it is a byte value. However, the old name is
1829     retained for backwards compatibility. (Note that LASTLITERAL is also a byte
1830     value.)
1831    
1832     53. The single man page has become too large. I have therefore split it up into
1833     a number of separate man pages. These also give rise to individual HTML pages;
1834     these are now put in a separate directory, and there is an index.html page that
1835     lists them all. Some hyperlinking between the pages has been installed.
1836    
1837     54. Added convenience functions for handling named capturing parentheses.
1838    
1839     55. Unknown escapes inside character classes (e.g. [\M]) and escapes that
1840     aren't interpreted therein (e.g. [\C]) are literals in Perl. This is now also
1841     true in PCRE, except when the PCRE_EXTENDED option is set, in which case they
1842     are faulted.
1843    
1844     56. Introduced HOST_CC and HOST_CFLAGS which can be set in the environment when
1845     calling configure. These values are used when compiling the dftables.c program
1846     which is run to generate the source of the default character tables. They
1847     default to the values of CC and CFLAGS. If you are cross-compiling PCRE,
1848     you will need to set these values.
1849    
1850     57. Updated the building process for Windows DLL, as provided by Fred Cox.
1851    
1852    
1853     Version 3.9 02-Jan-02
1854 nigel 61 ---------------------
1855    
1856     1. A bit of extraneous text had somehow crept into the pcregrep documentation.
1857    
1858     2. If --disable-static was given, the building process failed when trying to
1859     build pcretest and pcregrep. (For some reason it was using libtool to compile
1860     them, which is not right, as they aren't part of the library.)
1861    
1862    
1863 nigel 59 Version 3.8 18-Dec-01
1864     ---------------------
1865    
1866     1. The experimental UTF-8 code was completely screwed up. It was packing the
1867     bytes in the wrong order. How dumb can you get?
1868    
1869    
1870 nigel 57 Version 3.7 29-Oct-01
1871     ---------------------
1872    
1873     1. In updating pcretest to check change 1 of version 3.6, I screwed up.
1874     This caused pcretest, when used on the test data, to segfault. Unfortunately,
1875     this didn't happen under Solaris 8, where I normally test things.
1876    
1877 nigel 59 2. The Makefile had to be changed to make it work on BSD systems, where 'make'
1878     doesn't seem to recognize that ./xxx and xxx are the same file. (This entry
1879     isn't in ChangeLog distributed with 3.7 because I forgot when I hastily made
1880     this fix an hour or so after the initial 3.7 release.)
1881 nigel 57
1882 nigel 59
1883 nigel 55 Version 3.6 23-Oct-01
1884     ---------------------
1885    
1886     1. Crashed with /(sens|respons)e and \1ibility/ and "sense and sensibility" if
1887     offsets passed as NULL with zero offset count.
1888    
1889     2. The config.guess and config.sub files had not been updated when I moved to
1890     the latest autoconf.
1891    
1892    
1893 nigel 53 Version 3.5 15-Aug-01
1894     ---------------------
1895 nigel 5
1896 nigel 53 1. Added some missing #if !defined NOPOSIX conditionals in pcretest.c that
1897     had been forgotten.
1898    
1899     2. By using declared but undefined structures, we can avoid using "void"
1900     definitions in pcre.h while keeping the internal definitions of the structures
1901     private.
1902    
1903     3. The distribution is now built using autoconf 2.50 and libtool 1.4. From a
1904     user point of view, this means that both static and shared libraries are built
1905     by default, but this can be individually controlled. More of the work of
1906     handling this static/shared cases is now inside libtool instead of PCRE's make
1907     file.
1908    
1909     4. The pcretest utility is now installed along with pcregrep because it is
1910     useful for users (to test regexs) and by doing this, it automatically gets
1911     relinked by libtool. The documentation has been turned into a man page, so
1912     there are now .1, .txt, and .html versions in /doc.
1913    
1914     5. Upgrades to pcregrep:
1915     (i) Added long-form option names like gnu grep.
1916     (ii) Added --help to list all options with an explanatory phrase.
1917     (iii) Added -r, --recursive to recurse into sub-directories.
1918     (iv) Added -f, --file to read patterns from a file.
1919    
1920     6. pcre_exec() was referring to its "code" argument before testing that
1921     argument for NULL (and giving an error if it was NULL).
1922    
1923     7. Upgraded Makefile.in to allow for compiling in a different directory from
1924     the source directory.
1925    
1926     8. Tiny buglet in pcretest: when pcre_fullinfo() was called to retrieve the
1927     options bits, the pointer it was passed was to an int instead of to an unsigned
1928     long int. This mattered only on 64-bit systems.
1929    
1930     9. Fixed typo (3.4/1) in pcre.h again. Sigh. I had changed pcre.h (which is
1931     generated) instead of pcre.in, which it its source. Also made the same change
1932     in several of the .c files.
1933    
1934     10. A new release of gcc defines printf() as a macro, which broke pcretest
1935     because it had an ifdef in the middle of a string argument for printf(). Fixed
1936     by using separate calls to printf().
1937    
1938     11. Added --enable-newline-is-cr and --enable-newline-is-lf to the configure
1939     script, to force use of CR or LF instead of \n in the source. On non-Unix
1940     systems, the value can be set in config.h.
1941    
1942     12. The limit of 200 on non-capturing parentheses is a _nesting_ limit, not an
1943     absolute limit. Changed the text of the error message to make this clear, and
1944     likewise updated the man page.
1945    
1946     13. The limit of 99 on the number of capturing subpatterns has been removed.
1947     The new limit is 65535, which I hope will not be a "real" limit.
1948    
1949    
1950 nigel 51 Version 3.4 22-Aug-00
1951     ---------------------
1952    
1953     1. Fixed typo in pcre.h: unsigned const char * changed to const unsigned char *.
1954    
1955     2. Diagnose condition (?(0) as an error instead of crashing on matching.
1956    
1957    
1958 nigel 49 Version 3.3 01-Aug-00
1959     ---------------------
1960    
1961     1. If an octal character was given, but the value was greater than \377, it
1962     was not getting masked to the least significant bits, as documented. This could
1963     lead to crashes in some systems.
1964    
1965     2. Perl 5.6 (if not earlier versions) accepts classes like [a-\d] and treats
1966     the hyphen as a literal. PCRE used to give an error; it now behaves like Perl.
1967    
1968     3. Added the functions pcre_free_substring() and pcre_free_substring_list().
1969     These just pass their arguments on to (pcre_free)(), but they are provided
1970     because some uses of PCRE bind it to non-C systems that can call its functions,
1971     but cannot call free() or pcre_free() directly.
1972    
1973     4. Add "make test" as a synonym for "make check". Corrected some comments in
1974     the Makefile.
1975    
1976     5. Add $(DESTDIR)/ in front of all the paths in the "install" target in the
1977     Makefile.
1978    
1979     6. Changed the name of pgrep to pcregrep, because Solaris has introduced a
1980     command called pgrep for grepping around the active processes.
1981    
1982     7. Added the beginnings of support for UTF-8 character strings.
1983    
1984     8. Arranged for the Makefile to pass over the settings of CC, CFLAGS, and
1985     RANLIB to ./ltconfig so that they are used by libtool. I think these are all
1986     the relevant ones. (AR is not passed because ./ltconfig does its own figuring
1987     out for the ar command.)
1988    
1989    
1990 nigel 47 Version 3.2 12-May-00
1991     ---------------------
1992    
1993     This is purely a bug fixing release.
1994    
1995     1. If the pattern /((Z)+|A)*/ was matched agained ZABCDEFG it matched Z instead
1996     of ZA. This was just one example of several cases that could provoke this bug,
1997     which was introduced by change 9 of version 2.00. The code for breaking
1998     infinite loops after an iteration that matches an empty string was't working
1999     correctly.
2000    
2001     2. The pcretest program was not imitating Perl correctly for the pattern /a*/g
2002     when matched against abbab (for example). After matching an empty string, it
2003     wasn't forcing anchoring when setting PCRE_NOTEMPTY for the next attempt; this
2004     caused it to match further down the string than it should.
2005    
2006     3. The code contained an inclusion of sys/types.h. It isn't clear why this
2007     was there because it doesn't seem to be needed, and it causes trouble on some
2008     systems, as it is not a Standard C header. It has been removed.
2009    
2010     4. Made 4 silly changes to the source to avoid stupid compiler warnings that
2011     were reported on the Macintosh. The changes were from
2012    
2013     while ((c = *(++ptr)) != 0 && c != '\n');
2014     to
2015     while ((c = *(++ptr)) != 0 && c != '\n') ;
2016    
2017     Totally extraordinary, but if that's what it takes...
2018    
2019     5. PCRE is being used in one environment where neither memmove() nor bcopy() is
2020     available. Added HAVE_BCOPY and an autoconf test for it; if neither
2021     HAVE_MEMMOVE nor HAVE_BCOPY is set, use a built-in emulation function which
2022     assumes the way PCRE uses memmove() (always moving upwards).
2023    
2024     6. PCRE is being used in one environment where strchr() is not available. There
2025     was only one use in pcre.c, and writing it out to avoid strchr() probably gives
2026     faster code anyway.
2027    
2028    
2029 nigel 45 Version 3.1 09-Feb-00
2030     ---------------------
2031    
2032     The only change in this release is the fixing of some bugs in Makefile.in for
2033     the "install" target:
2034    
2035     (1) It was failing to install pcreposix.h.
2036    
2037     (2) It was overwriting the pcre.3 man page with the pcreposix.3 man page.
2038    
2039    
2040 nigel 43 Version 3.0 01-Feb-00
2041     ---------------------
2042 nigel 41
2043     1. Add support for the /+ modifier to perltest (to output $` like it does in
2044     pcretest).
2045    
2046     2. Add support for the /g modifier to perltest.
2047    
2048     3. Fix pcretest so that it behaves even more like Perl for /g when the pattern
2049     matches null strings.
2050    
2051     4. Fix perltest so that it doesn't do unwanted things when fed an empty
2052     pattern. Perl treats empty patterns specially - it reuses the most recent
2053     pattern, which is not what we want. Replace // by /(?#)/ in order to avoid this
2054     effect.
2055    
2056     5. The POSIX interface was broken in that it was just handing over the POSIX
2057     captured string vector to pcre_exec(), but (since release 2.00) PCRE has
2058     required a bigger vector, with some working space on the end. This means that
2059     the POSIX wrapper now has to get and free some memory, and copy the results.
2060    
2061 nigel 43 6. Added some simple autoconf support, placing the test data and the
2062     documentation in separate directories, re-organizing some of the
2063     information files, and making it build pcre-config (a GNU standard). Also added
2064     libtool support for building PCRE as a shared library, which is now the
2065     default.
2066 nigel 41
2067 nigel 43 7. Got rid of the leading zero in the definition of PCRE_MINOR because 08 and
2068     09 are not valid octal constants. Single digits will be used for minor values
2069     less than 10.
2070    
2071     8. Defined REG_EXTENDED and REG_NOSUB as zero in the POSIX header, so that
2072     existing programs that set these in the POSIX interface can use PCRE without
2073     modification.
2074    
2075     9. Added a new function, pcre_fullinfo() with an extensible interface. It can
2076     return all that pcre_info() returns, plus additional data. The pcre_info()
2077     function is retained for compatibility, but is considered to be obsolete.
2078    
2079     10. Added experimental recursion feature (?R) to handle one common case that
2080     Perl 5.6 will be able to do with (?p{...}).
2081    
2082     11. Added support for POSIX character classes like [:alpha:], which Perl is
2083     adopting.
2084    
2085    
2086 nigel 39 Version 2.08 31-Aug-99
2087     ----------------------
2088    
2089     1. When startoffset was not zero and the pattern began with ".*", PCRE was not
2090     trying to match at the startoffset position, but instead was moving forward to
2091     the next newline as if a previous match had failed.
2092    
2093     2. pcretest was not making use of PCRE_NOTEMPTY when repeating for /g and /G,
2094     and could get into a loop if a null string was matched other than at the start
2095     of the subject.
2096    
2097     3. Added definitions of PCRE_MAJOR and PCRE_MINOR to pcre.h so the version can
2098     be distinguished at compile time, and for completeness also added PCRE_DATE.
2099    
2100     5. Added Paul Sokolovsky's minor changes to make it easy to compile a Win32 DLL
2101     in GnuWin32 environments.
2102    
2103    
2104 nigel 37 Version 2.07 29-Jul-99
2105     ----------------------
2106    
2107     1. The documentation is now supplied in plain text form and HTML as well as in
2108     the form of man page sources.
2109    
2110     2. C++ compilers don't like assigning (void *) values to other pointer types.
2111     In particular this affects malloc(). Although there is no problem in Standard
2112     C, I've put in casts to keep C++ compilers happy.
2113    
2114     3. Typo on pcretest.c; a cast of (unsigned char *) in the POSIX regexec() call
2115     should be (const char *).
2116    
2117     4. If NOPOSIX is defined, pcretest.c compiles without POSIX support. This may
2118     be useful for non-Unix systems who don't want to bother with the POSIX stuff.
2119     However, I haven't made this a standard facility. The documentation doesn't
2120     mention it, and the Makefile doesn't support it.
2121    
2122     5. The Makefile now contains an "install" target, with editable destinations at
2123     the top of the file. The pcretest program is not installed.
2124    
2125     6. pgrep -V now gives the PCRE version number and date.
2126    
2127     7. Fixed bug: a zero repetition after a literal string (e.g. /abcde{0}/) was
2128     causing the entire string to be ignored, instead of just the last character.
2129    
2130     8. If a pattern like /"([^\\"]+|\\.)*"/ is applied in the normal way to a
2131     non-matching string, it can take a very, very long time, even for strings of
2132     quite modest length, because of the nested recursion. PCRE now does better in
2133     some of these cases. It does this by remembering the last required literal
2134     character in the pattern, and pre-searching the subject to ensure it is present
2135     before running the real match. In other words, it applies a heuristic to detect
2136     some types of certain failure quickly, and in the above example, if presented
2137     with a string that has no trailing " it gives "no match" very quickly.
2138    
2139     9. A new runtime option PCRE_NOTEMPTY causes null string matches to be ignored;
2140     other alternatives are tried instead.
2141    
2142    
2143 nigel 35 Version 2.06 09-Jun-99
2144     ----------------------
2145    
2146     1. Change pcretest's output for amount of store used to show just the code
2147     space, because the remainder (the data block) varies in size between 32-bit and
2148     64-bit systems.
2149    
2150     2. Added an extra argument to pcre_exec() to supply an offset in the subject to
2151     start matching at. This allows lookbehinds to work when searching for multiple
2152     occurrences in a string.
2153    
2154     3. Added additional options to pcretest for testing multiple occurrences:
2155    
2156     /+ outputs the rest of the string that follows a match
2157     /g loops for multiple occurrences, using the new startoffset argument
2158     /G loops for multiple occurrences by passing an incremented pointer
2159    
2160     4. PCRE wasn't doing the "first character" optimization for patterns starting
2161     with \b or \B, though it was doing it for other lookbehind assertions. That is,
2162     it wasn't noticing that a match for a pattern such as /\bxyz/ has to start with
2163     the letter 'x'. On long subject strings, this gives a significant speed-up.
2164    
2165    
2166 nigel 33 Version 2.05 21-Apr-99
2167     ----------------------
2168    
2169     1. Changed the type of magic_number from int to long int so that it works
2170     properly on 16-bit systems.
2171    
2172     2. Fixed a bug which caused patterns starting with .* not to work correctly
2173     when the subject string contained newline characters. PCRE was assuming
2174     anchoring for such patterns in all cases, which is not correct because .* will
2175     not pass a newline unless PCRE_DOTALL is set. It now assumes anchoring only if
2176     DOTALL is set at top level; otherwise it knows that patterns starting with .*
2177     must be retried after every newline in the subject.
2178    
2179    
2180 nigel 31 Version 2.04 18-Feb-99
2181     ----------------------
2182    
2183     1. For parenthesized subpatterns with repeats whose minimum was zero, the
2184     computation of the store needed to hold the pattern was incorrect (too large).
2185     If such patterns were nested a few deep, this could multiply and become a real
2186     problem.
2187    
2188     2. Added /M option to pcretest to show the memory requirement of a specific
2189     pattern. Made -m a synonym of -s (which does this globally) for compatibility.
2190    
2191     3. Subpatterns of the form (regex){n,m} (i.e. limited maximum) were being
2192     compiled in such a way that the backtracking after subsequent failure was
2193     pessimal. Something like (a){0,3} was compiled as (a)?(a)?(a)? instead of
2194     ((a)((a)(a)?)?)? with disastrous performance if the maximum was of any size.
2195    
2196    
2197 nigel 29 Version 2.03 02-Feb-99
2198     ----------------------
2199    
2200     1. Fixed typo and small mistake in man page.
2201    
2202 nigel 31 2. Added 4th condition (GPL supersedes if conflict) and created separate
2203     LICENCE file containing the conditions.
2204 nigel 29
2205     3. Updated pcretest so that patterns such as /abc\/def/ work like they do in
2206     Perl, that is the internal \ allows the delimiter to be included in the
2207     pattern. Locked out the use of \ as a delimiter. If \ immediately follows
2208     the final delimiter, add \ to the end of the pattern (to test the error).
2209    
2210     4. Added the convenience functions for extracting substrings after a successful
2211     match. Updated pcretest to make it able to test these functions.
2212    
2213    
2214 nigel 27 Version 2.02 14-Jan-99
2215     ----------------------
2216    
2217     1. Initialized the working variables associated with each extraction so that
2218     their saving and restoring doesn't refer to uninitialized store.
2219    
2220     2. Put dummy code into study.c in order to trick the optimizer of the IBM C
2221     compiler for OS/2 into generating correct code. Apparently IBM isn't going to
2222     fix the problem.
2223    
2224     3. Pcretest: the timing code wasn't using LOOPREPEAT for timing execution
2225     calls, and wasn't printing the correct value for compiling calls. Increased the
2226     default value of LOOPREPEAT, and the number of significant figures in the
2227     times.
2228    
2229     4. Changed "/bin/rm" in the Makefile to "-rm" so it works on Windows NT.
2230    
2231     5. Renamed "deftables" as "dftables" to get it down to 8 characters, to avoid
2232     a building problem on Windows NT with a FAT file system.
2233    
2234    
2235 nigel 25 Version 2.01 21-Oct-98
2236     ----------------------
2237    
2238     1. Changed the API for pcre_compile() to allow for the provision of a pointer
2239     to character tables built by pcre_maketables() in the current locale. If NULL
2240     is passed, the default tables are used.
2241    
2242    
2243 nigel 23 Version 2.00 24-Sep-98
2244 nigel 21 ----------------------
2245    
2246 nigel 23 1. Since the (>?) facility is in Perl 5.005, don't require PCRE_EXTRA to enable
2247     it any more.
2248    
2249     2. Allow quantification of (?>) groups, and make it work correctly.
2250    
2251     3. The first character computation wasn't working for (?>) groups.
2252    
2253     4. Correct the implementation of \Z (it is permitted to match on the \n at the
2254     end of the subject) and add 5.005's \z, which really does match only at the
2255     very end of the subject.
2256    
2257     5. Remove the \X "cut" facility; Perl doesn't have it, and (?> is neater.
2258    
2259     6. Remove the ability to specify CASELESS, MULTILINE, DOTALL, and
2260     DOLLAR_END_ONLY at runtime, to make it possible to implement the Perl 5.005
2261     localized options. All options to pcre_study() were also removed.
2262    
2263     7. Add other new features from 5.005:
2264    
2265     $(?<= positive lookbehind
2266     $(?<! negative lookbehind
2267     (?imsx-imsx) added the unsetting capability
2268     such a setting is global if at outer level; local otherwise
2269     (?imsx-imsx:) non-capturing groups with option setting
2270     (?(cond)re|re) conditional pattern matching
2271    
2272     A backreference to itself in a repeated group matches the previous
2273     captured string.
2274    
2275     8. General tidying up of studying (both automatic and via "study")
2276     consequential on the addition of new assertions.
2277    
2278     9. As in 5.005, unlimited repeated groups that could match an empty substring
2279     are no longer faulted at compile time. Instead, the loop is forcibly broken at
2280     runtime if any iteration does actually match an empty substring.
2281    
2282     10. Include the RunTest script in the distribution.
2283    
2284     11. Added tests from the Perl 5.005_02 distribution. This showed up a few
2285     discrepancies, some of which were old and were also with respect to 5.004. They
2286     have now been fixed.
2287    
2288    
2289     Version 1.09 28-Apr-98
2290     ----------------------
2291    
2292 nigel 21 1. A negated single character class followed by a quantifier with a minimum
2293     value of one (e.g. [^x]{1,6} ) was not compiled correctly. This could lead to
2294     program crashes, or just wrong answers. This did not apply to negated classes
2295     containing more than one character, or to minima other than one.
2296    
2297    
2298 nigel 19 Version 1.08 27-Mar-98
2299     ----------------------
2300    
2301     1. Add PCRE_UNGREEDY to invert the greediness of quantifiers.
2302    
2303     2. Add (?U) and (?X) to set PCRE_UNGREEDY and PCRE_EXTRA respectively. The
2304     latter must appear before anything that relies on it in the pattern.
2305    
2306    
2307 nigel 17 Version 1.07 16-Feb-98
2308     ----------------------
2309    
2310     1. A pattern such as /((a)*)*/ was not being diagnosed as in error (unlimited
2311     repeat of a potentially empty string).
2312    
2313    
2314 nigel 15 Version 1.06 23-Jan-98
2315     ----------------------
2316    
2317     1. Added Markus Oberhumer's little patches for C++.
2318    
2319     2. Literal strings longer than 255 characters were broken.
2320    
2321    
2322 nigel 13 Version 1.05 23-Dec-97
2323     ----------------------
2324    
2325     1. Negated character classes containing more than one character were failing if
2326     PCRE_CASELESS was set at run time.
2327    
2328    
2329 nigel 11 Version 1.04 19-Dec-97
2330     ----------------------
2331    
2332     1. Corrected the man page, where some "const" qualifiers had been omitted.
2333    
2334     2. Made debugging output print "{0,xxx}" instead of just "{,xxx}" to agree with
2335     input syntax.
2336    
2337     3. Fixed memory leak which occurred when a regex with back references was
2338     matched with an offsets vector that wasn't big enough. The temporary memory
2339     that is used in this case wasn't being freed if the match failed.
2340    
2341     4. Tidied pcretest to ensure it frees memory that it gets.
2342    
2343     5. Temporary memory was being obtained in the case where the passed offsets
2344     vector was exactly big enough.
2345    
2346     6. Corrected definition of offsetof() from change 5 below.
2347    
2348     7. I had screwed up change 6 below and broken the rules for the use of
2349     setjmp(). Now fixed.
2350    
2351    
2352 nigel 9 Version 1.03 18-Dec-97
2353     ----------------------
2354    
2355     1. A erroneous regex with a missing opening parenthesis was correctly
2356     diagnosed, but PCRE attempted to access brastack[-1], which could cause crashes
2357     on some systems.
2358    
2359     2. Replaced offsetof(real_pcre, code) by offsetof(real_pcre, code[0]) because
2360     it was reported that one broken compiler failed on the former because "code" is
2361     also an independent variable.
2362    
2363     3. The erroneous regex a[]b caused an array overrun reference.
2364    
2365     4. A regex ending with a one-character negative class (e.g. /[^k]$/) did not
2366     fail on data ending with that character. (It was going on too far, and checking
2367     the next character, typically a binary zero.) This was specific to the
2368     optimized code for single-character negative classes.
2369    
2370     5. Added a contributed patch from the TIN world which does the following:
2371    
2372     + Add an undef for memmove, in case the the system defines a macro for it.
2373    
2374     + Add a definition of offsetof(), in case there isn't one. (I don't know
2375     the reason behind this - offsetof() is part of the ANSI standard - but
2376     it does no harm).
2377    
2378     + Reduce the ifdef's in pcre.c using macro DPRINTF, thereby eliminating
2379     most of the places where whitespace preceded '#'. I have given up and
2380     allowed the remaining 2 cases to be at the margin.
2381    
2382     + Rename some variables in pcre to eliminate shadowing. This seems very
2383     pedantic, but does no harm, of course.
2384    
2385     6. Moved the call to setjmp() into its own function, to get rid of warnings
2386     from gcc -Wall, and avoided calling it at all unless PCRE_EXTRA is used.
2387    
2388     7. Constructs such as \d{8,} were compiling into the equivalent of
2389 nigel 11 \d{8}\d{0,65527} instead of \d{8}\d* which didn't make much difference to the
2390 nigel 9 outcome, but in this particular case used more store than had been allocated,
2391     which caused the bug to be discovered because it threw up an internal error.
2392    
2393     8. The debugging code in both pcre and pcretest for outputting the compiled
2394     form of a regex was going wrong in the case of back references followed by
2395     curly-bracketed repeats.
2396    
2397    
2398 nigel 7 Version 1.02 12-Dec-97
2399     ----------------------
2400    
2401     1. Typos in pcre.3 and comments in the source fixed.
2402    
2403     2. Applied a contributed patch to get rid of places where it used to remove
2404     'const' from variables, and fixed some signed/unsigned and uninitialized
2405     variable warnings.
2406    
2407     3. Added the "runtest" target to Makefile.
2408    
2409     4. Set default compiler flag to -O2 rather than just -O.
2410    
2411    
2412 nigel 5 Version 1.01 19-Nov-97
2413     ----------------------
2414    
2415     1. PCRE was failing to diagnose unlimited repeat of empty string for patterns
2416     like /([ab]*)*/, that is, for classes with more than one character in them.
2417    
2418     2. Likewise, it wasn't diagnosing patterns with "once-only" subpatterns, such
2419     as /((?>a*))*/ (a PCRE_EXTRA facility).
2420    
2421    
2422     Version 1.00 18-Nov-97
2423     ----------------------
2424    
2425     1. Added compile-time macros to support systems such as SunOS4 which don't have
2426     memmove() or strerror() but have other things that can be used instead.
2427    
2428     2. Arranged that "make clean" removes the executables.
2429    
2430    
2431 nigel 3 Version 0.99 27-Oct-97
2432     ----------------------
2433    
2434     1. Fixed bug in code for optimizing classes with only one character. It was
2435     initializing a 32-byte map regardless, which could cause it to run off the end
2436     of the memory it had got.
2437    
2438     2. Added, conditional on PCRE_EXTRA, the proposed (?>REGEX) construction.
2439    
2440    
2441     Version 0.98 22-Oct-97
2442     ----------------------
2443    
2444     1. Fixed bug in code for handling temporary memory usage when there are more
2445     back references than supplied space in the ovector. This could cause segfaults.
2446    
2447    
2448     Version 0.97 21-Oct-97
2449     ----------------------
2450    
2451     1. Added the \X "cut" facility, conditional on PCRE_EXTRA.
2452    
2453     2. Optimized negated single characters not to use a bit map.
2454    
2455     3. Brought error texts together as macro definitions; clarified some of them;
2456     fixed one that was wrong - it said "range out of order" when it meant "invalid
2457     escape sequence".
2458    
2459     4. Changed some char * arguments to const char *.
2460    
2461     5. Added PCRE_NOTBOL and PCRE_NOTEOL (from POSIX).
2462    
2463     6. Added the POSIX-style API wrapper in pcreposix.a and testing facilities in
2464     pcretest.
2465    
2466    
2467     Version 0.96 16-Oct-97
2468     ----------------------
2469    
2470     1. Added a simple "pgrep" utility to the distribution.
2471    
2472     2. Fixed an incompatibility with Perl: "{" is now treated as a normal character
2473     unless it appears in one of the precise forms "{ddd}", "{ddd,}", or "{ddd,ddd}"
2474     where "ddd" means "one or more decimal digits".
2475    
2476     3. Fixed serious bug. If a pattern had a back reference, but the call to
2477     pcre_exec() didn't supply a large enough ovector to record the related
2478     identifying subpattern, the match always failed. PCRE now remembers the number
2479     of the largest back reference, and gets some temporary memory in which to save
2480     the offsets during matching if necessary, in order to ensure that
2481     backreferences always work.
2482    
2483     4. Increased the compatibility with Perl in a number of ways:
2484    
2485     (a) . no longer matches \n by default; an option PCRE_DOTALL is provided
2486     to request this handling. The option can be set at compile or exec time.
2487    
2488     (b) $ matches before a terminating newline by default; an option
2489     PCRE_DOLLAR_ENDONLY is provided to override this (but not in multiline
2490     mode). The option can be set at compile or exec time.
2491    
2492     (c) The handling of \ followed by a digit other than 0 is now supposed to be
2493     the same as Perl's. If the decimal number it represents is less than 10
2494     or there aren't that many previous left capturing parentheses, an octal
2495     escape is read. Inside a character class, it's always an octal escape,
2496     even if it is a single digit.
2497    
2498     (d) An escaped but undefined alphabetic character is taken as a literal,
2499     unless PCRE_EXTRA is set. Currently this just reserves the remaining
2500     escapes.
2501    
2502     (e) {0} is now permitted. (The previous item is removed from the compiled
2503     pattern).
2504    
2505     5. Changed all the names of code files so that the basic parts are no longer
2506     than 10 characters, and abolished the teeny "globals.c" file.
2507    
2508     6. Changed the handling of character classes; they are now done with a 32-byte
2509     bit map always.
2510    
2511     7. Added the -d and /D options to pcretest to make it possible to look at the
2512     internals of compilation without having to recompile pcre.
2513    
2514    
2515     Version 0.95 23-Sep-97
2516     ----------------------
2517    
2518     1. Fixed bug in pre-pass concerning escaped "normal" characters such as \x5c or
2519     \x20 at the start of a run of normal characters. These were being treated as
2520     real characters, instead of the source characters being re-checked.
2521    
2522    
2523     Version 0.94 18-Sep-97
2524     ----------------------
2525    
2526     1. The functions are now thread-safe, with the caveat that the global variables
2527     containing pointers to malloc() and free() or alternative functions are the
2528     same for all threads.
2529    
2530     2. Get pcre_study() to generate a bitmap of initial characters for non-
2531     anchored patterns when this is possible, and use it if passed to pcre_exec().
2532    
2533    
2534     Version 0.93 15-Sep-97
2535     ----------------------
2536    
2537     1. /(b)|(:+)/ was computing an incorrect first character.
2538    
2539     2. Add pcre_study() to the API and the passing of pcre_extra to pcre_exec(),
2540     but not actually doing anything yet.
2541    
2542     3. Treat "-" characters in classes that cannot be part of ranges as literals,
2543     as Perl does (e.g. [-az] or [az-]).
2544    
2545     4. Set the anchored flag if a branch starts with .* or .*? because that tests
2546     all possible positions.
2547    
2548     5. Split up into different modules to avoid including unneeded functions in a
2549     compiled binary. However, compile and exec are still in one module. The "study"
2550     function is split off.
2551    
2552     6. The character tables are now in a separate module whose source is generated
2553     by an auxiliary program - but can then be edited by hand if required. There are
2554     now no calls to isalnum(), isspace(), isdigit(), isxdigit(), tolower() or
2555     toupper() in the code.
2556    
2557     7. Turn the malloc/free funtions variables into pcre_malloc and pcre_free and
2558     make them global. Abolish the function for setting them, as the caller can now
2559     set them directly.
2560    
2561    
2562     Version 0.92 11-Sep-97
2563     ----------------------
2564    
2565     1. A repeat with a fixed maximum and a minimum of 1 for an ordinary character
2566     (e.g. /a{1,3}/) was broken (I mis-optimized it).
2567    
2568     2. Caseless matching was not working in character classes if the characters in
2569     the pattern were in upper case.
2570    
2571     3. Make ranges like [W-c] work in the same way as Perl for caseless matching.
2572    
2573     4. Make PCRE_ANCHORED public and accept as a compile option.
2574    
2575     5. Add an options word to pcre_exec() and accept PCRE_ANCHORED and
2576     PCRE_CASELESS at run time. Add escapes \A and \I to pcretest to cause it to
2577     pass them.
2578    
2579     6. Give an error if bad option bits passed at compile or run time.
2580    
2581     7. Add PCRE_MULTILINE at compile and exec time, and (?m) as well. Add \M to
2582     pcretest to cause it to pass that flag.
2583    
2584     8. Add pcre_info(), to get the number of identifying subpatterns, the stored
2585     options, and the first character, if set.
2586    
2587     9. Recognize C+ or C{n,m} where n >= 1 as providing a fixed starting character.
2588    
2589    
2590     Version 0.91 10-Sep-97
2591     ----------------------
2592    
2593     1. PCRE was failing to diagnose unlimited repeats of subpatterns that could
2594     match the empty string as in /(a*)*/. It was looping and ultimately crashing.
2595    
2596     2. PCRE was looping on encountering an indefinitely repeated back reference to
2597     a subpattern that had matched an empty string, e.g. /(a|)\1*/. It now does what
2598     Perl does - treats the match as successful.
2599    
2600     ****

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