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

Contents of /code/trunk/ChangeLog

Parent Directory Parent Directory | Revision Log Revision Log


Revision 110 - (hide annotations) (download)
Wed Mar 7 16:31:04 2007 UTC (7 years, 8 months ago) by ph10
File size: 122258 byte(s)
Some more #ifdefs for SUPPORT_UTF8.

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 107 9. Added some extra #ifdef SUPPORT_UTF8 conditionals into pcretest.c,
52     pcre_printint.src, pcre_compile.c, pcre_study.c, and pcre_tables.c, in
53     order to be able to cut out the UTF-8 tables in the latter when UTF-8
54     support is not required. This saves 1.5-2K of code, which is important in
55     some applications.
56    
57 ph10 110 Later: more #ifdefs are needed in pcre_ord2utf8.c and pcre_valid_utf8.c
58     so as not to refer to the tables, even though these functions will never be
59     called when UTF-8 support is disabled. Otherwise there are problems with a
60     shared library.
61    
62 ph10 108 10. The emulated memmove() function in pcre_internal.h (provided for those
63     environments that have neither memmove() nor bcopy()) was defining its
64     arguments as char * instead of void *.
65    
66 ph10 98
67 nigel 93 Version 7.0 19-Dec-06
68     ---------------------
69    
70     1. Fixed a signed/unsigned compiler warning in pcre_compile.c, shown up by
71     moving to gcc 4.1.1.
72    
73     2. The -S option for pcretest uses setrlimit(); I had omitted to #include
74     sys/time.h, which is documented as needed for this function. It doesn't
75     seem to matter on Linux, but it showed up on some releases of OS X.
76    
77     3. It seems that there are systems where bytes whose values are greater than
78     127 match isprint() in the "C" locale. The "C" locale should be the
79     default when a C program starts up. In most systems, only ASCII printing
80     characters match isprint(). This difference caused the output from pcretest
81     to vary, making some of the tests fail. I have changed pcretest so that:
82    
83     (a) When it is outputting text in the compiled version of a pattern, bytes
84     other than 32-126 are always shown as hex escapes.
85    
86     (b) When it is outputting text that is a matched part of a subject string,
87     it does the same, unless a different locale has been set for the match
88     (using the /L modifier). In this case, it uses isprint() to decide.
89    
90     4. Fixed a major bug that caused incorrect computation of the amount of memory
91     required for a compiled pattern when options that changed within the
92     pattern affected the logic of the preliminary scan that determines the
93     length. The relevant options are -x, and -i in UTF-8 mode. The result was
94     that the computed length was too small. The symptoms of this bug were
95     either the PCRE error "internal error: code overflow" from pcre_compile(),
96     or a glibc crash with a message such as "pcretest: free(): invalid next
97     size (fast)". Examples of patterns that provoked this bug (shown in
98     pcretest format) are:
99    
100     /(?-x: )/x
101     /(?x)(?-x: \s*#\s*)/
102     /((?i)[\x{c0}])/8
103     /(?i:[\x{c0}])/8
104    
105     HOWEVER: Change 17 below makes this fix obsolete as the memory computation
106     is now done differently.
107    
108     5. Applied patches from Google to: (a) add a QuoteMeta function to the C++
109     wrapper classes; (b) implement a new function in the C++ scanner that is
110     more efficient than the old way of doing things because it avoids levels of
111     recursion in the regex matching; (c) add a paragraph to the documentation
112     for the FullMatch() function.
113    
114     6. The escape sequence \n was being treated as whatever was defined as
115     "newline". Not only was this contrary to the documentation, which states
116     that \n is character 10 (hex 0A), but it also went horribly wrong when
117     "newline" was defined as CRLF. This has been fixed.
118    
119     7. In pcre_dfa_exec.c the value of an unsigned integer (the variable called c)
120     was being set to -1 for the "end of line" case (supposedly a value that no
121     character can have). Though this value is never used (the check for end of
122     line is "zero bytes in current character"), it caused compiler complaints.
123     I've changed it to 0xffffffff.
124    
125     8. In pcre_version.c, the version string was being built by a sequence of
126     C macros that, in the event of PCRE_PRERELEASE being defined as an empty
127     string (as it is for production releases) called a macro with an empty
128     argument. The C standard says the result of this is undefined. The gcc
129     compiler treats it as an empty string (which was what was wanted) but it is
130     reported that Visual C gives an error. The source has been hacked around to
131     avoid this problem.
132    
133     9. On the advice of a Windows user, included <io.h> and <fcntl.h> in Windows
134     builds of pcretest, and changed the call to _setmode() to use _O_BINARY
135     instead of 0x8000. Made all the #ifdefs test both _WIN32 and WIN32 (not all
136     of them did).
137    
138     10. Originally, pcretest opened its input and output without "b"; then I was
139     told that "b" was needed in some environments, so it was added for release
140     5.0 to both the input and output. (It makes no difference on Unix-like
141     systems.) Later I was told that it is wrong for the input on Windows. I've
142     now abstracted the modes into two macros, to make it easier to fiddle with
143     them, and removed "b" from the input mode under Windows.
144    
145     11. Added pkgconfig support for the C++ wrapper library, libpcrecpp.
146    
147     12. Added -help and --help to pcretest as an official way of being reminded
148     of the options.
149    
150     13. Removed some redundant semicolons after macro calls in pcrecpparg.h.in
151     and pcrecpp.cc because they annoy compilers at high warning levels.
152    
153     14. A bit of tidying/refactoring in pcre_exec.c in the main bumpalong loop.
154    
155     15. Fixed an occurrence of == in configure.ac that should have been = (shell
156     scripts are not C programs :-) and which was not noticed because it works
157     on Linux.
158    
159     16. pcretest is supposed to handle any length of pattern and data line (as one
160     line or as a continued sequence of lines) by extending its input buffer if
161     necessary. This feature was broken for very long pattern lines, leading to
162     a string of junk being passed to pcre_compile() if the pattern was longer
163     than about 50K.
164    
165     17. I have done a major re-factoring of the way pcre_compile() computes the
166     amount of memory needed for a compiled pattern. Previously, there was code
167     that made a preliminary scan of the pattern in order to do this. That was
168     OK when PCRE was new, but as the facilities have expanded, it has become
169     harder and harder to keep it in step with the real compile phase, and there
170     have been a number of bugs (see for example, 4 above). I have now found a
171     cunning way of running the real compile function in a "fake" mode that
172     enables it to compute how much memory it would need, while actually only
173     ever using a few hundred bytes of working memory and without too many
174     tests of the mode. This should make future maintenance and development
175     easier. A side effect of this work is that the limit of 200 on the nesting
176     depth of parentheses has been removed (though this was never a serious
177     limitation, I suspect). However, there is a downside: pcre_compile() now
178     runs more slowly than before (30% or more, depending on the pattern). I
179     hope this isn't a big issue. There is no effect on runtime performance.
180    
181     18. Fixed a minor bug in pcretest: if a pattern line was not terminated by a
182     newline (only possible for the last line of a file) and it was a
183     pattern that set a locale (followed by /Lsomething), pcretest crashed.
184    
185     19. Added additional timing features to pcretest. (1) The -tm option now times
186     matching only, not compiling. (2) Both -t and -tm can be followed, as a
187     separate command line item, by a number that specifies the number of
188     repeats to use when timing. The default is 50000; this gives better
189     precision, but takes uncomfortably long for very large patterns.
190    
191     20. Extended pcre_study() to be more clever in cases where a branch of a
192     subpattern has no definite first character. For example, (a*|b*)[cd] would
193     previously give no result from pcre_study(). Now it recognizes that the
194     first character must be a, b, c, or d.
195    
196     21. There was an incorrect error "recursive call could loop indefinitely" if
197     a subpattern (or the entire pattern) that was being tested for matching an
198     empty string contained only one non-empty item after a nested subpattern.
199     For example, the pattern (?>\x{100}*)\d(?R) provoked this error
200     incorrectly, because the \d was being skipped in the check.
201    
202     22. The pcretest program now has a new pattern option /B and a command line
203     option -b, which is equivalent to adding /B to every pattern. This causes
204     it to show the compiled bytecode, without the additional information that
205     -d shows. The effect of -d is now the same as -b with -i (and similarly, /D
206     is the same as /B/I).
207    
208     23. A new optimization is now able automatically to treat some sequences such
209     as a*b as a*+b. More specifically, if something simple (such as a character
210     or a simple class like \d) has an unlimited quantifier, and is followed by
211     something that cannot possibly match the quantified thing, the quantifier
212     is automatically "possessified".
213    
214     24. A recursive reference to a subpattern whose number was greater than 39
215     went wrong under certain circumstances in UTF-8 mode. This bug could also
216     have affected the operation of pcre_study().
217    
218     25. Realized that a little bit of performance could be had by replacing
219     (c & 0xc0) == 0xc0 with c >= 0xc0 when processing UTF-8 characters.
220    
221     26. Timing data from pcretest is now shown to 4 decimal places instead of 3.
222    
223     27. Possessive quantifiers such as a++ were previously implemented by turning
224     them into atomic groups such as ($>a+). Now they have their own opcodes,
225     which improves performance. This includes the automatically created ones
226     from 23 above.
227    
228     28. A pattern such as (?=(\w+))\1: which simulates an atomic group using a
229     lookahead was broken if it was not anchored. PCRE was mistakenly expecting
230     the first matched character to be a colon. This applied both to named and
231     numbered groups.
232    
233     29. The ucpinternal.h header file was missing its idempotency #ifdef.
234    
235     30. I was sent a "project" file called libpcre.a.dev which I understand makes
236     building PCRE on Windows easier, so I have included it in the distribution.
237    
238     31. There is now a check in pcretest against a ridiculously large number being
239     returned by pcre_exec() or pcre_dfa_exec(). If this happens in a /g or /G
240     loop, the loop is abandoned.
241    
242     32. Forward references to subpatterns in conditions such as (?(2)...) where
243     subpattern 2 is defined later cause pcre_compile() to search forwards in
244     the pattern for the relevant set of parentheses. This search went wrong
245     when there were unescaped parentheses in a character class, parentheses
246     escaped with \Q...\E, or parentheses in a #-comment in /x mode.
247    
248     33. "Subroutine" calls and backreferences were previously restricted to
249     referencing subpatterns earlier in the regex. This restriction has now
250     been removed.
251    
252     34. Added a number of extra features that are going to be in Perl 5.10. On the
253     whole, these are just syntactic alternatives for features that PCRE had
254     previously implemented using the Python syntax or my own invention. The
255     other formats are all retained for compatibility.
256    
257     (a) Named groups can now be defined as (?<name>...) or (?'name'...) as well
258     as (?P<name>...). The new forms, as well as being in Perl 5.10, are
259     also .NET compatible.
260    
261     (b) A recursion or subroutine call to a named group can now be defined as
262     (?&name) as well as (?P>name).
263    
264     (c) A backreference to a named group can now be defined as \k<name> or
265     \k'name' as well as (?P=name). The new forms, as well as being in Perl
266     5.10, are also .NET compatible.
267    
268     (d) A conditional reference to a named group can now use the syntax
269     (?(<name>) or (?('name') as well as (?(name).
270    
271     (e) A "conditional group" of the form (?(DEFINE)...) can be used to define
272     groups (named and numbered) that are never evaluated inline, but can be
273     called as "subroutines" from elsewhere. In effect, the DEFINE condition
274     is always false. There may be only one alternative in such a group.
275    
276     (f) A test for recursion can be given as (?(R1).. or (?(R&name)... as well
277     as the simple (?(R). The condition is true only if the most recent
278     recursion is that of the given number or name. It does not search out
279     through the entire recursion stack.
280    
281     (g) The escape \gN or \g{N} has been added, where N is a positive or
282     negative number, specifying an absolute or relative reference.
283    
284     35. Tidied to get rid of some further signed/unsigned compiler warnings and
285     some "unreachable code" warnings.
286    
287     36. Updated the Unicode property tables to Unicode version 5.0.0. Amongst other
288     things, this adds five new scripts.
289    
290     37. Perl ignores orphaned \E escapes completely. PCRE now does the same.
291     There were also incompatibilities regarding the handling of \Q..\E inside
292     character classes, for example with patterns like [\Qa\E-\Qz\E] where the
293     hyphen was adjacent to \Q or \E. I hope I've cleared all this up now.
294    
295     38. Like Perl, PCRE detects when an indefinitely repeated parenthesized group
296     matches an empty string, and forcibly breaks the loop. There were bugs in
297     this code in non-simple cases. For a pattern such as ^(a()*)* matched
298     against aaaa the result was just "a" rather than "aaaa", for example. Two
299     separate and independent bugs (that affected different cases) have been
300     fixed.
301    
302     39. Refactored the code to abolish the use of different opcodes for small
303     capturing bracket numbers. This is a tidy that I avoided doing when I
304     removed the limit on the number of capturing brackets for 3.5 back in 2001.
305     The new approach is not only tidier, it makes it possible to reduce the
306     memory needed to fix the previous bug (38).
307    
308     40. Implemented PCRE_NEWLINE_ANY to recognize any of the Unicode newline
309     sequences (http://unicode.org/unicode/reports/tr18/) as "newline" when
310     processing dot, circumflex, or dollar metacharacters, or #-comments in /x
311     mode.
312    
313     41. Add \R to match any Unicode newline sequence, as suggested in the Unicode
314     report.
315    
316     42. Applied patch, originally from Ari Pollak, modified by Google, to allow
317     copy construction and assignment in the C++ wrapper.
318    
319     43. Updated pcregrep to support "--newline=any". In the process, I fixed a
320     couple of bugs that could have given wrong results in the "--newline=crlf"
321     case.
322    
323     44. Added a number of casts and did some reorganization of signed/unsigned int
324     variables following suggestions from Dair Grant. Also renamed the variable
325     "this" as "item" because it is a C++ keyword.
326    
327     45. Arranged for dftables to add
328    
329     #include "pcre_internal.h"
330    
331     to pcre_chartables.c because without it, gcc 4.x may remove the array
332     definition from the final binary if PCRE is built into a static library and
333     dead code stripping is activated.
334    
335     46. For an unanchored pattern, if a match attempt fails at the start of a
336     newline sequence, and the newline setting is CRLF or ANY, and the next two
337     characters are CRLF, advance by two characters instead of one.
338    
339    
340 nigel 91 Version 6.7 04-Jul-06
341     ---------------------
342    
343     1. In order to handle tests when input lines are enormously long, pcretest has
344     been re-factored so that it automatically extends its buffers when
345     necessary. The code is crude, but this _is_ just a test program. The
346     default size has been increased from 32K to 50K.
347    
348     2. The code in pcre_study() was using the value of the re argument before
349     testing it for NULL. (Of course, in any sensible call of the function, it
350     won't be NULL.)
351    
352     3. The memmove() emulation function in pcre_internal.h, which is used on
353     systems that lack both memmove() and bcopy() - that is, hardly ever -
354     was missing a "static" storage class specifier.
355    
356     4. When UTF-8 mode was not set, PCRE looped when compiling certain patterns
357     containing an extended class (one that cannot be represented by a bitmap
358     because it contains high-valued characters or Unicode property items, e.g.
359     [\pZ]). Almost always one would set UTF-8 mode when processing such a
360     pattern, but PCRE should not loop if you do not (it no longer does).
361     [Detail: two cases were found: (a) a repeated subpattern containing an
362     extended class; (b) a recursive reference to a subpattern that followed a
363     previous extended class. It wasn't skipping over the extended class
364     correctly when UTF-8 mode was not set.]
365    
366     5. A negated single-character class was not being recognized as fixed-length
367     in lookbehind assertions such as (?<=[^f]), leading to an incorrect
368     compile error "lookbehind assertion is not fixed length".
369    
370     6. The RunPerlTest auxiliary script was showing an unexpected difference
371     between PCRE and Perl for UTF-8 tests. It turns out that it is hard to
372     write a Perl script that can interpret lines of an input file either as
373     byte characters or as UTF-8, which is what "perltest" was being required to
374     do for the non-UTF-8 and UTF-8 tests, respectively. Essentially what you
375     can't do is switch easily at run time between having the "use utf8;" pragma
376     or not. In the end, I fudged it by using the RunPerlTest script to insert
377     "use utf8;" explicitly for the UTF-8 tests.
378    
379     7. In multiline (/m) mode, PCRE was matching ^ after a terminating newline at
380     the end of the subject string, contrary to the documentation and to what
381     Perl does. This was true of both matching functions. Now it matches only at
382     the start of the subject and immediately after *internal* newlines.
383    
384     8. A call of pcre_fullinfo() from pcretest to get the option bits was passing
385     a pointer to an int instead of a pointer to an unsigned long int. This
386     caused problems on 64-bit systems.
387    
388     9. Applied a patch from the folks at Google to pcrecpp.cc, to fix "another
389     instance of the 'standard' template library not being so standard".
390    
391     10. There was no check on the number of named subpatterns nor the maximum
392     length of a subpattern name. The product of these values is used to compute
393     the size of the memory block for a compiled pattern. By supplying a very
394     long subpattern name and a large number of named subpatterns, the size
395     computation could be caused to overflow. This is now prevented by limiting
396     the length of names to 32 characters, and the number of named subpatterns
397     to 10,000.
398    
399     11. Subpatterns that are repeated with specific counts have to be replicated in
400     the compiled pattern. The size of memory for this was computed from the
401     length of the subpattern and the repeat count. The latter is limited to
402     65535, but there was no limit on the former, meaning that integer overflow
403     could in principle occur. The compiled length of a repeated subpattern is
404     now limited to 30,000 bytes in order to prevent this.
405    
406     12. Added the optional facility to have named substrings with the same name.
407    
408     13. Added the ability to use a named substring as a condition, using the
409     Python syntax: (?(name)yes|no). This overloads (?(R)... and names that
410     are numbers (not recommended). Forward references are permitted.
411    
412     14. Added forward references in named backreferences (if you see what I mean).
413    
414     15. In UTF-8 mode, with the PCRE_DOTALL option set, a quantified dot in the
415     pattern could run off the end of the subject. For example, the pattern
416     "(?s)(.{1,5})"8 did this with the subject "ab".
417    
418     16. If PCRE_DOTALL or PCRE_MULTILINE were set, pcre_dfa_exec() behaved as if
419     PCRE_CASELESS was set when matching characters that were quantified with ?
420     or *.
421    
422     17. A character class other than a single negated character that had a minimum
423     but no maximum quantifier - for example [ab]{6,} - was not handled
424     correctly by pce_dfa_exec(). It would match only one character.
425    
426     18. A valid (though odd) pattern that looked like a POSIX character
427     class but used an invalid character after [ (for example [[,abc,]]) caused
428     pcre_compile() to give the error "Failed: internal error: code overflow" or
429     in some cases to crash with a glibc free() error. This could even happen if
430     the pattern terminated after [[ but there just happened to be a sequence of
431     letters, a binary zero, and a closing ] in the memory that followed.
432    
433     19. Perl's treatment of octal escapes in the range \400 to \777 has changed
434     over the years. Originally (before any Unicode support), just the bottom 8
435     bits were taken. Thus, for example, \500 really meant \100. Nowadays the
436     output from "man perlunicode" includes this:
437    
438     The regular expression compiler produces polymorphic opcodes. That
439     is, the pattern adapts to the data and automatically switches to
440     the Unicode character scheme when presented with Unicode data--or
441     instead uses a traditional byte scheme when presented with byte
442     data.
443    
444     Sadly, a wide octal escape does not cause a switch, and in a string with
445     no other multibyte characters, these octal escapes are treated as before.
446     Thus, in Perl, the pattern /\500/ actually matches \100 but the pattern
447     /\500|\x{1ff}/ matches \500 or \777 because the whole thing is treated as a
448     Unicode string.
449    
450     I have not perpetrated such confusion in PCRE. Up till now, it took just
451     the bottom 8 bits, as in old Perl. I have now made octal escapes with
452     values greater than \377 illegal in non-UTF-8 mode. In UTF-8 mode they
453     translate to the appropriate multibyte character.
454    
455     29. Applied some refactoring to reduce the number of warnings from Microsoft
456     and Borland compilers. This has included removing the fudge introduced
457     seven years ago for the OS/2 compiler (see 2.02/2 below) because it caused
458     a warning about an unused variable.
459    
460     21. PCRE has not included VT (character 0x0b) in the set of whitespace
461     characters since release 4.0, because Perl (from release 5.004) does not.
462     [Or at least, is documented not to: some releases seem to be in conflict
463     with the documentation.] However, when a pattern was studied with
464     pcre_study() and all its branches started with \s, PCRE still included VT
465     as a possible starting character. Of course, this did no harm; it just
466     caused an unnecessary match attempt.
467    
468     22. Removed a now-redundant internal flag bit that recorded the fact that case
469     dependency changed within the pattern. This was once needed for "required
470     byte" processing, but is no longer used. This recovers a now-scarce options
471     bit. Also moved the least significant internal flag bit to the most-
472     significant bit of the word, which was not previously used (hangover from
473     the days when it was an int rather than a uint) to free up another bit for
474     the future.
475    
476     23. Added support for CRLF line endings as well as CR and LF. As well as the
477     default being selectable at build time, it can now be changed at runtime
478     via the PCRE_NEWLINE_xxx flags. There are now options for pcregrep to
479     specify that it is scanning data with non-default line endings.
480    
481     24. Changed the definition of CXXLINK to make it agree with the definition of
482     LINK in the Makefile, by replacing LDFLAGS to CXXFLAGS.
483    
484     25. Applied Ian Taylor's patches to avoid using another stack frame for tail
485     recursions. This makes a big different to stack usage for some patterns.
486    
487     26. If a subpattern containing a named recursion or subroutine reference such
488     as (?P>B) was quantified, for example (xxx(?P>B)){3}, the calculation of
489     the space required for the compiled pattern went wrong and gave too small a
490     value. Depending on the environment, this could lead to "Failed: internal
491     error: code overflow at offset 49" or "glibc detected double free or
492     corruption" errors.
493    
494     27. Applied patches from Google (a) to support the new newline modes and (b) to
495     advance over multibyte UTF-8 characters in GlobalReplace.
496    
497     28. Change free() to pcre_free() in pcredemo.c. Apparently this makes a
498     difference for some implementation of PCRE in some Windows version.
499    
500     29. Added some extra testing facilities to pcretest:
501    
502     \q<number> in a data line sets the "match limit" value
503     \Q<number> in a data line sets the "match recursion limt" value
504     -S <number> sets the stack size, where <number> is in megabytes
505    
506     The -S option isn't available for Windows.
507    
508    
509 nigel 89 Version 6.6 06-Feb-06
510     ---------------------
511    
512     1. Change 16(a) for 6.5 broke things, because PCRE_DATA_SCOPE was not defined
513     in pcreposix.h. I have copied the definition from pcre.h.
514    
515     2. Change 25 for 6.5 broke compilation in a build directory out-of-tree
516     because pcre.h is no longer a built file.
517    
518     3. Added Jeff Friedl's additional debugging patches to pcregrep. These are
519     not normally included in the compiled code.
520    
521    
522 nigel 87 Version 6.5 01-Feb-06
523     ---------------------
524    
525     1. When using the partial match feature with pcre_dfa_exec(), it was not
526     anchoring the second and subsequent partial matches at the new starting
527     point. This could lead to incorrect results. For example, with the pattern
528     /1234/, partially matching against "123" and then "a4" gave a match.
529    
530     2. Changes to pcregrep:
531    
532     (a) All non-match returns from pcre_exec() were being treated as failures
533     to match the line. Now, unless the error is PCRE_ERROR_NOMATCH, an
534     error message is output. Some extra information is given for the
535     PCRE_ERROR_MATCHLIMIT and PCRE_ERROR_RECURSIONLIMIT errors, which are
536     probably the only errors that are likely to be caused by users (by
537     specifying a regex that has nested indefinite repeats, for instance).
538     If there are more than 20 of these errors, pcregrep is abandoned.
539    
540     (b) A binary zero was treated as data while matching, but terminated the
541     output line if it was written out. This has been fixed: binary zeroes
542     are now no different to any other data bytes.
543    
544     (c) Whichever of the LC_ALL or LC_CTYPE environment variables is set is
545     used to set a locale for matching. The --locale=xxxx long option has
546     been added (no short equivalent) to specify a locale explicitly on the
547     pcregrep command, overriding the environment variables.
548    
549     (d) When -B was used with -n, some line numbers in the output were one less
550     than they should have been.
551    
552     (e) Added the -o (--only-matching) option.
553    
554     (f) If -A or -C was used with -c (count only), some lines of context were
555     accidentally printed for the final match.
556    
557     (g) Added the -H (--with-filename) option.
558    
559     (h) The combination of options -rh failed to suppress file names for files
560     that were found from directory arguments.
561    
562     (i) Added the -D (--devices) and -d (--directories) options.
563    
564     (j) Added the -F (--fixed-strings) option.
565    
566     (k) Allow "-" to be used as a file name for -f as well as for a data file.
567    
568     (l) Added the --colo(u)r option.
569    
570     (m) Added Jeffrey Friedl's -S testing option, but within #ifdefs so that it
571     is not present by default.
572    
573     3. A nasty bug was discovered in the handling of recursive patterns, that is,
574     items such as (?R) or (?1), when the recursion could match a number of
575     alternatives. If it matched one of the alternatives, but subsequently,
576     outside the recursion, there was a failure, the code tried to back up into
577     the recursion. However, because of the way PCRE is implemented, this is not
578     possible, and the result was an incorrect result from the match.
579    
580     In order to prevent this happening, the specification of recursion has
581     been changed so that all such subpatterns are automatically treated as
582     atomic groups. Thus, for example, (?R) is treated as if it were (?>(?R)).
583    
584     4. I had overlooked the fact that, in some locales, there are characters for
585     which isalpha() is true but neither isupper() nor islower() are true. In
586     the fr_FR locale, for instance, the \xAA and \xBA characters (ordmasculine
587     and ordfeminine) are like this. This affected the treatment of \w and \W
588     when they appeared in character classes, but not when they appeared outside
589     a character class. The bit map for "word" characters is now created
590     separately from the results of isalnum() instead of just taking it from the
591     upper, lower, and digit maps. (Plus the underscore character, of course.)
592    
593     5. The above bug also affected the handling of POSIX character classes such as
594     [[:alpha:]] and [[:alnum:]]. These do not have their own bit maps in PCRE's
595     permanent tables. Instead, the bit maps for such a class were previously
596     created as the appropriate unions of the upper, lower, and digit bitmaps.
597     Now they are created by subtraction from the [[:word:]] class, which has
598     its own bitmap.
599    
600     6. The [[:blank:]] character class matches horizontal, but not vertical space.
601     It is created by subtracting the vertical space characters (\x09, \x0a,
602     \x0b, \x0c) from the [[:space:]] bitmap. Previously, however, the
603     subtraction was done in the overall bitmap for a character class, meaning
604     that a class such as [\x0c[:blank:]] was incorrect because \x0c would not
605     be recognized. This bug has been fixed.
606    
607     7. Patches from the folks at Google:
608    
609     (a) pcrecpp.cc: "to handle a corner case that may or may not happen in
610     real life, but is still worth protecting against".
611    
612     (b) pcrecpp.cc: "corrects a bug when negative radixes are used with
613     regular expressions".
614    
615     (c) pcre_scanner.cc: avoid use of std::count() because not all systems
616     have it.
617    
618     (d) Split off pcrecpparg.h from pcrecpp.h and had the former built by
619     "configure" and the latter not, in order to fix a problem somebody had
620     with compiling the Arg class on HP-UX.
621    
622     (e) Improve the error-handling of the C++ wrapper a little bit.
623    
624     (f) New tests for checking recursion limiting.
625    
626     8. The pcre_memmove() function, which is used only if the environment does not
627     have a standard memmove() function (and is therefore rarely compiled),
628     contained two bugs: (a) use of int instead of size_t, and (b) it was not
629     returning a result (though PCRE never actually uses the result).
630    
631     9. In the POSIX regexec() interface, if nmatch is specified as a ridiculously
632     large number - greater than INT_MAX/(3*sizeof(int)) - REG_ESPACE is
633     returned instead of calling malloc() with an overflowing number that would
634     most likely cause subsequent chaos.
635    
636     10. The debugging option of pcretest was not showing the NO_AUTO_CAPTURE flag.
637    
638     11. The POSIX flag REG_NOSUB is now supported. When a pattern that was compiled
639     with this option is matched, the nmatch and pmatch options of regexec() are
640     ignored.
641    
642     12. Added REG_UTF8 to the POSIX interface. This is not defined by POSIX, but is
643     provided in case anyone wants to the the POSIX interface with UTF-8
644     strings.
645    
646     13. Added CXXLDFLAGS to the Makefile parameters to provide settings only on the
647     C++ linking (needed for some HP-UX environments).
648    
649     14. Avoid compiler warnings in get_ucpname() when compiled without UCP support
650     (unused parameter) and in the pcre_printint() function (omitted "default"
651     switch label when the default is to do nothing).
652    
653     15. Added some code to make it possible, when PCRE is compiled as a C++
654     library, to replace subject pointers for pcre_exec() with a smart pointer
655     class, thus making it possible to process discontinuous strings.
656    
657     16. The two macros PCRE_EXPORT and PCRE_DATA_SCOPE are confusing, and perform
658     much the same function. They were added by different people who were trying
659     to make PCRE easy to compile on non-Unix systems. It has been suggested
660     that PCRE_EXPORT be abolished now that there is more automatic apparatus
661     for compiling on Windows systems. I have therefore replaced it with
662     PCRE_DATA_SCOPE. This is set automatically for Windows; if not set it
663     defaults to "extern" for C or "extern C" for C++, which works fine on
664     Unix-like systems. It is now possible to override the value of PCRE_DATA_
665     SCOPE with something explicit in config.h. In addition:
666    
667     (a) pcreposix.h still had just "extern" instead of either of these macros;
668     I have replaced it with PCRE_DATA_SCOPE.
669    
670     (b) Functions such as _pcre_xclass(), which are internal to the library,
671     but external in the C sense, all had PCRE_EXPORT in their definitions.
672     This is apparently wrong for the Windows case, so I have removed it.
673     (It makes no difference on Unix-like systems.)
674    
675     17. Added a new limit, MATCH_LIMIT_RECURSION, which limits the depth of nesting
676     of recursive calls to match(). This is different to MATCH_LIMIT because
677     that limits the total number of calls to match(), not all of which increase
678     the depth of recursion. Limiting the recursion depth limits the amount of
679     stack (or heap if NO_RECURSE is set) that is used. The default can be set
680     when PCRE is compiled, and changed at run time. A patch from Google adds
681     this functionality to the C++ interface.
682    
683     18. Changes to the handling of Unicode character properties:
684    
685     (a) Updated the table to Unicode 4.1.0.
686    
687     (b) Recognize characters that are not in the table as "Cn" (undefined).
688    
689     (c) I revised the way the table is implemented to a much improved format
690     which includes recognition of ranges. It now supports the ranges that
691     are defined in UnicodeData.txt, and it also amalgamates other
692     characters into ranges. This has reduced the number of entries in the
693     table from around 16,000 to around 3,000, thus reducing its size
694     considerably. I realized I did not need to use a tree structure after
695     all - a binary chop search is just as efficient. Having reduced the
696     number of entries, I extended their size from 6 bytes to 8 bytes to
697     allow for more data.
698    
699     (d) Added support for Unicode script names via properties such as \p{Han}.
700    
701     19. In UTF-8 mode, a backslash followed by a non-Ascii character was not
702     matching that character.
703    
704     20. When matching a repeated Unicode property with a minimum greater than zero,
705     (for example \pL{2,}), PCRE could look past the end of the subject if it
706     reached it while seeking the minimum number of characters. This could
707     happen only if some of the characters were more than one byte long, because
708     there is a check for at least the minimum number of bytes.
709    
710     21. Refactored the implementation of \p and \P so as to be more general, to
711     allow for more different types of property in future. This has changed the
712     compiled form incompatibly. Anybody with saved compiled patterns that use
713     \p or \P will have to recompile them.
714    
715     22. Added "Any" and "L&" to the supported property types.
716    
717     23. Recognize \x{...} as a code point specifier, even when not in UTF-8 mode,
718     but give a compile time error if the value is greater than 0xff.
719    
720     24. The man pages for pcrepartial, pcreprecompile, and pcre_compile2 were
721     accidentally not being installed or uninstalled.
722    
723     25. The pcre.h file was built from pcre.h.in, but the only changes that were
724     made were to insert the current release number. This seemed silly, because
725     it made things harder for people building PCRE on systems that don't run
726     "configure". I have turned pcre.h into a distributed file, no longer built
727     by "configure", with the version identification directly included. There is
728     no longer a pcre.h.in file.
729    
730     However, this change necessitated a change to the pcre-config script as
731     well. It is built from pcre-config.in, and one of the substitutions was the
732     release number. I have updated configure.ac so that ./configure now finds
733     the release number by grepping pcre.h.
734    
735     26. Added the ability to run the tests under valgrind.
736    
737    
738 nigel 85 Version 6.4 05-Sep-05
739     ---------------------
740    
741     1. Change 6.0/10/(l) to pcregrep introduced a bug that caused separator lines
742     "--" to be printed when multiple files were scanned, even when none of the
743     -A, -B, or -C options were used. This is not compatible with Gnu grep, so I
744     consider it to be a bug, and have restored the previous behaviour.
745    
746     2. A couple of code tidies to get rid of compiler warnings.
747    
748     3. The pcretest program used to cheat by referring to symbols in the library
749     whose names begin with _pcre_. These are internal symbols that are not
750     really supposed to be visible externally, and in some environments it is
751     possible to suppress them. The cheating is now confined to including
752     certain files from the library's source, which is a bit cleaner.
753    
754     4. Renamed pcre.in as pcre.h.in to go with pcrecpp.h.in; it also makes the
755     file's purpose clearer.
756    
757     5. Reorganized pcre_ucp_findchar().
758    
759    
760 nigel 83 Version 6.3 15-Aug-05
761     ---------------------
762    
763     1. The file libpcre.pc.in did not have general read permission in the tarball.
764    
765     2. There were some problems when building without C++ support:
766    
767     (a) If C++ support was not built, "make install" and "make test" still
768     tried to test it.
769    
770     (b) There were problems when the value of CXX was explicitly set. Some
771     changes have been made to try to fix these, and ...
772    
773     (c) --disable-cpp can now be used to explicitly disable C++ support.
774    
775     (d) The use of @CPP_OBJ@ directly caused a blank line preceded by a
776     backslash in a target when C++ was disabled. This confuses some
777     versions of "make", apparently. Using an intermediate variable solves
778     this. (Same for CPP_LOBJ.)
779    
780     3. $(LINK_FOR_BUILD) now includes $(CFLAGS_FOR_BUILD) and $(LINK)
781     (non-Windows) now includes $(CFLAGS) because these flags are sometimes
782     necessary on certain architectures.
783    
784     4. Added a setting of -export-symbols-regex to the link command to remove
785     those symbols that are exported in the C sense, but actually are local
786     within the library, and not documented. Their names all begin with
787     "_pcre_". This is not a perfect job, because (a) we have to except some
788     symbols that pcretest ("illegally") uses, and (b) the facility isn't always
789     available (and never for static libraries). I have made a note to try to
790     find a way round (a) in the future.
791    
792    
793 nigel 81 Version 6.2 01-Aug-05
794     ---------------------
795    
796     1. There was no test for integer overflow of quantifier values. A construction
797     such as {1111111111111111} would give undefined results. What is worse, if
798     a minimum quantifier for a parenthesized subpattern overflowed and became
799     negative, the calculation of the memory size went wrong. This could have
800     led to memory overwriting.
801    
802     2. Building PCRE using VPATH was broken. Hopefully it is now fixed.
803    
804     3. Added "b" to the 2nd argument of fopen() in dftables.c, for non-Unix-like
805     operating environments where this matters.
806    
807     4. Applied Giuseppe Maxia's patch to add additional features for controlling
808     PCRE options from within the C++ wrapper.
809    
810     5. Named capturing subpatterns were not being correctly counted when a pattern
811     was compiled. This caused two problems: (a) If there were more than 100
812     such subpatterns, the calculation of the memory needed for the whole
813     compiled pattern went wrong, leading to an overflow error. (b) Numerical
814     back references of the form \12, where the number was greater than 9, were
815     not recognized as back references, even though there were sufficient
816     previous subpatterns.
817    
818     6. Two minor patches to pcrecpp.cc in order to allow it to compile on older
819     versions of gcc, e.g. 2.95.4.
820    
821    
822 nigel 79 Version 6.1 21-Jun-05
823     ---------------------
824    
825     1. There was one reference to the variable "posix" in pcretest.c that was not
826     surrounded by "#if !defined NOPOSIX".
827    
828     2. Make it possible to compile pcretest without DFA support, UTF8 support, or
829     the cross-check on the old pcre_info() function, for the benefit of the
830     cut-down version of PCRE that is currently imported into Exim.
831    
832     3. A (silly) pattern starting with (?i)(?-i) caused an internal space
833     allocation error. I've done the easy fix, which wastes 2 bytes for sensible
834     patterns that start (?i) but I don't think that matters. The use of (?i) is
835     just an example; this all applies to the other options as well.
836    
837     4. Since libtool seems to echo the compile commands it is issuing, the output
838     from "make" can be reduced a bit by putting "@" in front of each libtool
839     compile command.
840    
841     5. Patch from the folks at Google for configure.in to be a bit more thorough
842     in checking for a suitable C++ installation before trying to compile the
843     C++ stuff. This should fix a reported problem when a compiler was present,
844     but no suitable headers.
845    
846     6. The man pages all had just "PCRE" as their title. I have changed them to
847     be the relevant file name. I have also arranged that these names are
848     retained in the file doc/pcre.txt, which is a concatenation in text format
849     of all the man pages except the little individual ones for each function.
850    
851     7. The NON-UNIX-USE file had not been updated for the different set of source
852     files that come with release 6. I also added a few comments about the C++
853     wrapper.
854    
855    
856 nigel 77 Version 6.0 07-Jun-05
857     ---------------------
858    
859     1. Some minor internal re-organization to help with my DFA experiments.
860    
861     2. Some missing #ifdef SUPPORT_UCP conditionals in pcretest and printint that
862     didn't matter for the library itself when fully configured, but did matter
863     when compiling without UCP support, or within Exim, where the ucp files are
864     not imported.
865    
866     3. Refactoring of the library code to split up the various functions into
867     different source modules. The addition of the new DFA matching code (see
868     below) to a single monolithic source would have made it really too
869     unwieldy, quite apart from causing all the code to be include in a
870     statically linked application, when only some functions are used. This is
871     relevant even without the DFA addition now that patterns can be compiled in
872     one application and matched in another.
873    
874     The downside of splitting up is that there have to be some external
875     functions and data tables that are used internally in different modules of
876     the library but which are not part of the API. These have all had their
877     names changed to start with "_pcre_" so that they are unlikely to clash
878     with other external names.
879    
880     4. Added an alternate matching function, pcre_dfa_exec(), which matches using
881     a different (DFA) algorithm. Although it is slower than the original
882     function, it does have some advantages for certain types of matching
883     problem.
884    
885     5. Upgrades to pcretest in order to test the features of pcre_dfa_exec(),
886     including restarting after a partial match.
887    
888     6. A patch for pcregrep that defines INVALID_FILE_ATTRIBUTES if it is not
889     defined when compiling for Windows was sent to me. I have put it into the
890     code, though I have no means of testing or verifying it.
891    
892     7. Added the pcre_refcount() auxiliary function.
893    
894     8. Added the PCRE_FIRSTLINE option. This constrains an unanchored pattern to
895     match before or at the first newline in the subject string. In pcretest,
896     the /f option on a pattern can be used to set this.
897    
898     9. A repeated \w when used in UTF-8 mode with characters greater than 256
899     would behave wrongly. This has been present in PCRE since release 4.0.
900    
901     10. A number of changes to the pcregrep command:
902    
903     (a) Refactored how -x works; insert ^(...)$ instead of setting
904     PCRE_ANCHORED and checking the length, in preparation for adding
905     something similar for -w.
906    
907     (b) Added the -w (match as a word) option.
908    
909     (c) Refactored the way lines are read and buffered so as to have more
910     than one at a time available.
911    
912     (d) Implemented a pcregrep test script.
913    
914     (e) Added the -M (multiline match) option. This allows patterns to match
915     over several lines of the subject. The buffering ensures that at least
916     8K, or the rest of the document (whichever is the shorter) is available
917     for matching (and similarly the previous 8K for lookbehind assertions).
918    
919     (f) Changed the --help output so that it now says
920    
921     -w, --word-regex(p)
922    
923     instead of two lines, one with "regex" and the other with "regexp"
924     because that confused at least one person since the short forms are the
925     same. (This required a bit of code, as the output is generated
926     automatically from a table. It wasn't just a text change.)
927    
928     (g) -- can be used to terminate pcregrep options if the next thing isn't an
929     option but starts with a hyphen. Could be a pattern or a path name
930     starting with a hyphen, for instance.
931    
932     (h) "-" can be given as a file name to represent stdin.
933    
934     (i) When file names are being printed, "(standard input)" is used for
935     the standard input, for compatibility with GNU grep. Previously
936     "<stdin>" was used.
937    
938     (j) The option --label=xxx can be used to supply a name to be used for
939     stdin when file names are being printed. There is no short form.
940    
941     (k) Re-factored the options decoding logic because we are going to add
942     two more options that take data. Such options can now be given in four
943     different ways, e.g. "-fname", "-f name", "--file=name", "--file name".
944    
945     (l) Added the -A, -B, and -C options for requesting that lines of context
946     around matches be printed.
947    
948     (m) Added the -L option to print the names of files that do not contain
949     any matching lines, that is, the complement of -l.
950    
951     (n) The return code is 2 if any file cannot be opened, but pcregrep does
952     continue to scan other files.
953    
954     (o) The -s option was incorrectly implemented. For compatibility with other
955     greps, it now suppresses the error message for a non-existent or non-
956     accessible file (but not the return code). There is a new option called
957     -q that suppresses the output of matching lines, which was what -s was
958     previously doing.
959    
960     (p) Added --include and --exclude options to specify files for inclusion
961     and exclusion when recursing.
962    
963     11. The Makefile was not using the Autoconf-supported LDFLAGS macro properly.
964     Hopefully, it now does.
965    
966     12. Missing cast in pcre_study().
967    
968     13. Added an "uninstall" target to the makefile.
969    
970     14. Replaced "extern" in the function prototypes in Makefile.in with
971     "PCRE_DATA_SCOPE", which defaults to 'extern' or 'extern "C"' in the Unix
972     world, but is set differently for Windows.
973    
974     15. Added a second compiling function called pcre_compile2(). The only
975     difference is that it has an extra argument, which is a pointer to an
976     integer error code. When there is a compile-time failure, this is set
977     non-zero, in addition to the error test pointer being set to point to an
978     error message. The new argument may be NULL if no error number is required
979     (but then you may as well call pcre_compile(), which is now just a
980     wrapper). This facility is provided because some applications need a
981     numeric error indication, but it has also enabled me to tidy up the way
982     compile-time errors are handled in the POSIX wrapper.
983    
984     16. Added VPATH=.libs to the makefile; this should help when building with one
985     prefix path and installing with another. (Or so I'm told by someone who
986     knows more about this stuff than I do.)
987    
988     17. Added a new option, REG_DOTALL, to the POSIX function regcomp(). This
989     passes PCRE_DOTALL to the pcre_compile() function, making the "." character
990     match everything, including newlines. This is not POSIX-compatible, but
991     somebody wanted the feature. From pcretest it can be activated by using
992     both the P and the s flags.
993    
994     18. AC_PROG_LIBTOOL appeared twice in Makefile.in. Removed one.
995    
996     19. libpcre.pc was being incorrectly installed as executable.
997    
998     20. A couple of places in pcretest check for end-of-line by looking for '\n';
999     it now also looks for '\r' so that it will work unmodified on Windows.
1000    
1001     21. Added Google's contributed C++ wrapper to the distribution.
1002    
1003     22. Added some untidy missing memory free() calls in pcretest, to keep
1004     Electric Fence happy when testing.
1005    
1006    
1007    
1008 nigel 75 Version 5.0 13-Sep-04
1009     ---------------------
1010    
1011     1. Internal change: literal characters are no longer packed up into items
1012     containing multiple characters in a single byte-string. Each character
1013     is now matched using a separate opcode. However, there may be more than one
1014     byte in the character in UTF-8 mode.
1015    
1016     2. The pcre_callout_block structure has two new fields: pattern_position and
1017     next_item_length. These contain the offset in the pattern to the next match
1018     item, and its length, respectively.
1019    
1020     3. The PCRE_AUTO_CALLOUT option for pcre_compile() requests the automatic
1021     insertion of callouts before each pattern item. Added the /C option to
1022     pcretest to make use of this.
1023    
1024     4. On the advice of a Windows user, the lines
1025    
1026     #if defined(_WIN32) || defined(WIN32)
1027     _setmode( _fileno( stdout ), 0x8000 );
1028     #endif /* defined(_WIN32) || defined(WIN32) */
1029    
1030     have been added to the source of pcretest. This apparently does useful
1031     magic in relation to line terminators.
1032    
1033     5. Changed "r" and "w" in the calls to fopen() in pcretest to "rb" and "wb"
1034     for the benefit of those environments where the "b" makes a difference.
1035    
1036     6. The icc compiler has the same options as gcc, but "configure" doesn't seem
1037     to know about it. I have put a hack into configure.in that adds in code
1038     to set GCC=yes if CC=icc. This seems to end up at a point in the
1039     generated configure script that is early enough to affect the setting of
1040     compiler options, which is what is needed, but I have no means of testing
1041     whether it really works. (The user who reported this had patched the
1042     generated configure script, which of course I cannot do.)
1043    
1044     LATER: After change 22 below (new libtool files), the configure script
1045     seems to know about icc (and also ecc). Therefore, I have commented out
1046     this hack in configure.in.
1047    
1048     7. Added support for pkg-config (2 patches were sent in).
1049    
1050     8. Negated POSIX character classes that used a combination of internal tables
1051     were completely broken. These were [[:^alpha:]], [[:^alnum:]], and
1052     [[:^ascii]]. Typically, they would match almost any characters. The other
1053     POSIX classes were not broken in this way.
1054    
1055     9. Matching the pattern "\b.*?" against "ab cd", starting at offset 1, failed
1056     to find the match, as PCRE was deluded into thinking that the match had to
1057     start at the start point or following a newline. The same bug applied to
1058     patterns with negative forward assertions or any backward assertions
1059     preceding ".*" at the start, unless the pattern required a fixed first
1060     character. This was a failing pattern: "(?!.bcd).*". The bug is now fixed.
1061    
1062     10. In UTF-8 mode, when moving forwards in the subject after a failed match
1063     starting at the last subject character, bytes beyond the end of the subject
1064     string were read.
1065    
1066     11. Renamed the variable "class" as "classbits" to make life easier for C++
1067     users. (Previously there was a macro definition, but it apparently wasn't
1068     enough.)
1069    
1070     12. Added the new field "tables" to the extra data so that tables can be passed
1071     in at exec time, or the internal tables can be re-selected. This allows
1072     a compiled regex to be saved and re-used at a later time by a different
1073     program that might have everything at different addresses.
1074    
1075     13. Modified the pcre-config script so that, when run on Solaris, it shows a
1076     -R library as well as a -L library.
1077    
1078     14. The debugging options of pcretest (-d on the command line or D on a
1079     pattern) showed incorrect output for anything following an extended class
1080     that contained multibyte characters and which was followed by a quantifier.
1081    
1082     15. Added optional support for general category Unicode character properties
1083     via the \p, \P, and \X escapes. Unicode property support implies UTF-8
1084     support. It adds about 90K to the size of the library. The meanings of the
1085     inbuilt class escapes such as \d and \s have NOT been changed.
1086    
1087     16. Updated pcredemo.c to include calls to free() to release the memory for the
1088     compiled pattern.
1089    
1090     17. The generated file chartables.c was being created in the source directory
1091     instead of in the building directory. This caused the build to fail if the
1092     source directory was different from the building directory, and was
1093     read-only.
1094    
1095     18. Added some sample Win commands from Mark Tetrode into the NON-UNIX-USE
1096     file. No doubt somebody will tell me if they don't make sense... Also added
1097     Dan Mooney's comments about building on OpenVMS.
1098    
1099     19. Added support for partial matching via the PCRE_PARTIAL option for
1100     pcre_exec() and the \P data escape in pcretest.
1101    
1102     20. Extended pcretest with 3 new pattern features:
1103    
1104     (i) A pattern option of the form ">rest-of-line" causes pcretest to
1105     write the compiled pattern to the file whose name is "rest-of-line".
1106     This is a straight binary dump of the data, with the saved pointer to
1107     the character tables forced to be NULL. The study data, if any, is
1108     written too. After writing, pcretest reads a new pattern.
1109    
1110     (ii) If, instead of a pattern, "<rest-of-line" is given, pcretest reads a
1111     compiled pattern from the given file. There must not be any
1112     occurrences of "<" in the file name (pretty unlikely); if there are,
1113     pcretest will instead treat the initial "<" as a pattern delimiter.
1114     After reading in the pattern, pcretest goes on to read data lines as
1115     usual.
1116    
1117     (iii) The F pattern option causes pcretest to flip the bytes in the 32-bit
1118     and 16-bit fields in a compiled pattern, to simulate a pattern that
1119     was compiled on a host of opposite endianness.
1120    
1121     21. The pcre-exec() function can now cope with patterns that were compiled on
1122     hosts of opposite endianness, with this restriction:
1123    
1124     As for any compiled expression that is saved and used later, the tables
1125     pointer field cannot be preserved; the extra_data field in the arguments
1126     to pcre_exec() should be used to pass in a tables address if a value
1127     other than the default internal tables were used at compile time.
1128    
1129     22. Calling pcre_exec() with a negative value of the "ovecsize" parameter is
1130     now diagnosed as an error. Previously, most of the time, a negative number
1131     would have been treated as zero, but if in addition "ovector" was passed as
1132     NULL, a crash could occur.
1133    
1134     23. Updated the files ltmain.sh, config.sub, config.guess, and aclocal.m4 with
1135     new versions from the libtool 1.5 distribution (the last one is a copy of
1136     a file called libtool.m4). This seems to have fixed the need to patch
1137     "configure" to support Darwin 1.3 (which I used to do). However, I still
1138     had to patch ltmain.sh to ensure that ${SED} is set (it isn't on my
1139     workstation).
1140    
1141     24. Changed the PCRE licence to be the more standard "BSD" licence.
1142    
1143    
1144 nigel 73 Version 4.5 01-Dec-03
1145     ---------------------
1146    
1147     1. There has been some re-arrangement of the code for the match() function so
1148     that it can be compiled in a version that does not call itself recursively.
1149     Instead, it keeps those local variables that need separate instances for
1150     each "recursion" in a frame on the heap, and gets/frees frames whenever it
1151     needs to "recurse". Keeping track of where control must go is done by means
1152     of setjmp/longjmp. The whole thing is implemented by a set of macros that
1153     hide most of the details from the main code, and operates only if
1154     NO_RECURSE is defined while compiling pcre.c. If PCRE is built using the
1155     "configure" mechanism, "--disable-stack-for-recursion" turns on this way of
1156     operating.
1157    
1158     To make it easier for callers to provide specially tailored get/free
1159     functions for this usage, two new functions, pcre_stack_malloc, and
1160     pcre_stack_free, are used. They are always called in strict stacking order,
1161     and the size of block requested is always the same.
1162    
1163     The PCRE_CONFIG_STACKRECURSE info parameter can be used to find out whether
1164     PCRE has been compiled to use the stack or the heap for recursion. The
1165     -C option of pcretest uses this to show which version is compiled.
1166    
1167     A new data escape \S, is added to pcretest; it causes the amounts of store
1168     obtained and freed by both kinds of malloc/free at match time to be added
1169     to the output.
1170    
1171     2. Changed the locale test to use "fr_FR" instead of "fr" because that's
1172     what's available on my current Linux desktop machine.
1173    
1174     3. When matching a UTF-8 string, the test for a valid string at the start has
1175     been extended. If start_offset is not zero, PCRE now checks that it points
1176     to a byte that is the start of a UTF-8 character. If not, it returns
1177     PCRE_ERROR_BADUTF8_OFFSET (-11). Note: the whole string is still checked;
1178     this is necessary because there may be backward assertions in the pattern.
1179     When matching the same subject several times, it may save resources to use
1180     PCRE_NO_UTF8_CHECK on all but the first call if the string is long.
1181    
1182     4. The code for checking the validity of UTF-8 strings has been tightened so
1183     that it rejects (a) strings containing 0xfe or 0xff bytes and (b) strings
1184     containing "overlong sequences".
1185    
1186     5. Fixed a bug (appearing twice) that I could not find any way of exploiting!
1187     I had written "if ((digitab[*p++] && chtab_digit) == 0)" where the "&&"
1188     should have been "&", but it just so happened that all the cases this let
1189     through by mistake were picked up later in the function.
1190    
1191     6. I had used a variable called "isblank" - this is a C99 function, causing
1192     some compilers to warn. To avoid this, I renamed it (as "blankclass").
1193    
1194     7. Cosmetic: (a) only output another newline at the end of pcretest if it is
1195     prompting; (b) run "./pcretest /dev/null" at the start of the test script
1196     so the version is shown; (c) stop "make test" echoing "./RunTest".
1197    
1198     8. Added patches from David Burgess to enable PCRE to run on EBCDIC systems.
1199    
1200     9. The prototype for memmove() for systems that don't have it was using
1201     size_t, but the inclusion of the header that defines size_t was later. I've
1202     moved the #includes for the C headers earlier to avoid this.
1203    
1204     10. Added some adjustments to the code to make it easier to compiler on certain
1205     special systems:
1206    
1207     (a) Some "const" qualifiers were missing.
1208     (b) Added the macro EXPORT before all exported functions; by default this
1209     is defined to be empty.
1210     (c) Changed the dftables auxiliary program (that builds chartables.c) so
1211     that it reads its output file name as an argument instead of writing
1212     to the standard output and assuming this can be redirected.
1213    
1214     11. In UTF-8 mode, if a recursive reference (e.g. (?1)) followed a character
1215     class containing characters with values greater than 255, PCRE compilation
1216     went into a loop.
1217    
1218     12. A recursive reference to a subpattern that was within another subpattern
1219     that had a minimum quantifier of zero caused PCRE to crash. For example,
1220     (x(y(?2))z)? provoked this bug with a subject that got as far as the
1221     recursion. If the recursively-called subpattern itself had a zero repeat,
1222     that was OK.
1223    
1224     13. In pcretest, the buffer for reading a data line was set at 30K, but the
1225     buffer into which it was copied (for escape processing) was still set at
1226     1024, so long lines caused crashes.
1227    
1228     14. A pattern such as /[ab]{1,3}+/ failed to compile, giving the error
1229     "internal error: code overflow...". This applied to any character class
1230     that was followed by a possessive quantifier.
1231    
1232     15. Modified the Makefile to add libpcre.la as a prerequisite for
1233     libpcreposix.la because I was told this is needed for a parallel build to
1234     work.
1235    
1236     16. If a pattern that contained .* following optional items at the start was
1237     studied, the wrong optimizing data was generated, leading to matching
1238     errors. For example, studying /[ab]*.*c/ concluded, erroneously, that any
1239     matching string must start with a or b or c. The correct conclusion for
1240     this pattern is that a match can start with any character.
1241    
1242    
1243 nigel 71 Version 4.4 13-Aug-03
1244     ---------------------
1245    
1246     1. In UTF-8 mode, a character class containing characters with values between
1247     127 and 255 was not handled correctly if the compiled pattern was studied.
1248     In fixing this, I have also improved the studying algorithm for such
1249     classes (slightly).
1250    
1251     2. Three internal functions had redundant arguments passed to them. Removal
1252     might give a very teeny performance improvement.
1253    
1254     3. Documentation bug: the value of the capture_top field in a callout is *one
1255     more than* the number of the hightest numbered captured substring.
1256    
1257     4. The Makefile linked pcretest and pcregrep with -lpcre, which could result
1258     in incorrectly linking with a previously installed version. They now link
1259     explicitly with libpcre.la.
1260    
1261     5. configure.in no longer needs to recognize Cygwin specially.
1262    
1263     6. A problem in pcre.in for Windows platforms is fixed.
1264    
1265     7. If a pattern was successfully studied, and the -d (or /D) flag was given to
1266     pcretest, it used to include the size of the study block as part of its
1267     output. Unfortunately, the structure contains a field that has a different
1268     size on different hardware architectures. This meant that the tests that
1269     showed this size failed. As the block is currently always of a fixed size,
1270     this information isn't actually particularly useful in pcretest output, so
1271     I have just removed it.
1272    
1273     8. Three pre-processor statements accidentally did not start in column 1.
1274     Sadly, there are *still* compilers around that complain, even though
1275     standard C has not required this for well over a decade. Sigh.
1276    
1277     9. In pcretest, the code for checking callouts passed small integers in the
1278     callout_data field, which is a void * field. However, some picky compilers
1279     complained about the casts involved for this on 64-bit systems. Now
1280     pcretest passes the address of the small integer instead, which should get
1281     rid of the warnings.
1282    
1283     10. By default, when in UTF-8 mode, PCRE now checks for valid UTF-8 strings at
1284     both compile and run time, and gives an error if an invalid UTF-8 sequence
1285     is found. There is a option for disabling this check in cases where the
1286     string is known to be correct and/or the maximum performance is wanted.
1287    
1288     11. In response to a bug report, I changed one line in Makefile.in from
1289    
1290     -Wl,--out-implib,.libs/lib@WIN_PREFIX@pcreposix.dll.a \
1291     to
1292     -Wl,--out-implib,.libs/@WIN_PREFIX@libpcreposix.dll.a \
1293    
1294     to look similar to other lines, but I have no way of telling whether this
1295     is the right thing to do, as I do not use Windows. No doubt I'll get told
1296     if it's wrong...
1297    
1298    
1299 nigel 69 Version 4.3 21-May-03
1300     ---------------------
1301    
1302     1. Two instances of @WIN_PREFIX@ omitted from the Windows targets in the
1303     Makefile.
1304    
1305     2. Some refactoring to improve the quality of the code:
1306    
1307     (i) The utf8_table... variables are now declared "const".
1308    
1309     (ii) The code for \cx, which used the "case flipping" table to upper case
1310     lower case letters, now just substracts 32. This is ASCII-specific,
1311     but the whole concept of \cx is ASCII-specific, so it seems
1312     reasonable.
1313    
1314     (iii) PCRE was using its character types table to recognize decimal and
1315     hexadecimal digits in the pattern. This is silly, because it handles
1316     only 0-9, a-f, and A-F, but the character types table is locale-
1317     specific, which means strange things might happen. A private
1318     table is now used for this - though it costs 256 bytes, a table is
1319     much faster than multiple explicit tests. Of course, the standard
1320     character types table is still used for matching digits in subject
1321     strings against \d.
1322    
1323     (iv) Strictly, the identifier ESC_t is reserved by POSIX (all identifiers
1324     ending in _t are). So I've renamed it as ESC_tee.
1325    
1326     3. The first argument for regexec() in the POSIX wrapper should have been
1327     defined as "const".
1328    
1329     4. Changed pcretest to use malloc() for its buffers so that they can be
1330     Electric Fenced for debugging.
1331    
1332     5. There were several places in the code where, in UTF-8 mode, PCRE would try
1333     to read one or more bytes before the start of the subject string. Often this
1334     had no effect on PCRE's behaviour, but in some circumstances it could
1335     provoke a segmentation fault.
1336    
1337     6. A lookbehind at the start of a pattern in UTF-8 mode could also cause PCRE
1338     to try to read one or more bytes before the start of the subject string.
1339    
1340     7. A lookbehind in a pattern matched in non-UTF-8 mode on a PCRE compiled with
1341     UTF-8 support could misbehave in various ways if the subject string
1342     contained bytes with the 0x80 bit set and the 0x40 bit unset in a lookbehind
1343     area. (PCRE was not checking for the UTF-8 mode flag, and trying to move
1344     back over UTF-8 characters.)
1345    
1346    
1347 nigel 67 Version 4.2 14-Apr-03
1348     ---------------------
1349    
1350     1. Typo "#if SUPPORT_UTF8" instead of "#ifdef SUPPORT_UTF8" fixed.
1351    
1352     2. Changes to the building process, supplied by Ronald Landheer-Cieslak
1353     [ON_WINDOWS]: new variable, "#" on non-Windows platforms
1354     [NOT_ON_WINDOWS]: new variable, "#" on Windows platforms
1355     [WIN_PREFIX]: new variable, "cyg" for Cygwin
1356     * Makefile.in: use autoconf substitution for OBJEXT, EXEEXT, BUILD_OBJEXT
1357     and BUILD_EXEEXT
1358     Note: automatic setting of the BUILD variables is not yet working
1359     set CPPFLAGS and BUILD_CPPFLAGS (but don't use yet) - should be used at
1360     compile-time but not at link-time
1361     [LINK]: use for linking executables only
1362     make different versions for Windows and non-Windows
1363     [LINKLIB]: new variable, copy of UNIX-style LINK, used for linking
1364     libraries
1365     [LINK_FOR_BUILD]: new variable
1366     [OBJEXT]: use throughout
1367     [EXEEXT]: use throughout
1368     <winshared>: new target
1369     <wininstall>: new target
1370     <dftables.o>: use native compiler
1371     <dftables>: use native linker
1372     <install>: handle Windows platform correctly
1373     <clean>: ditto
1374     <check>: ditto
1375     copy DLL to top builddir before testing
1376    
1377     As part of these changes, -no-undefined was removed again. This was reported
1378     to give trouble on HP-UX 11.0, so getting rid of it seems like a good idea
1379     in any case.
1380    
1381     3. Some tidies to get rid of compiler warnings:
1382    
1383     . In the match_data structure, match_limit was an unsigned long int, whereas
1384     match_call_count was an int. I've made them both unsigned long ints.
1385    
1386     . In pcretest the fact that a const uschar * doesn't automatically cast to
1387     a void * provoked a warning.
1388    
1389     . Turning on some more compiler warnings threw up some "shadow" variables
1390     and a few more missing casts.
1391    
1392     4. If PCRE was complied with UTF-8 support, but called without the PCRE_UTF8
1393     option, a class that contained a single character with a value between 128
1394     and 255 (e.g. /[\xFF]/) caused PCRE to crash.
1395    
1396     5. If PCRE was compiled with UTF-8 support, but called without the PCRE_UTF8
1397     option, a class that contained several characters, but with at least one
1398     whose value was between 128 and 255 caused PCRE to crash.
1399    
1400    
1401 nigel 65 Version 4.1 12-Mar-03
1402     ---------------------
1403 nigel 63
1404 nigel 65 1. Compiling with gcc -pedantic found a couple of places where casts were
1405     needed, and a string in dftables.c that was longer than standard compilers are
1406     required to support.
1407    
1408     2. Compiling with Sun's compiler found a few more places where the code could
1409     be tidied up in order to avoid warnings.
1410    
1411     3. The variables for cross-compiling were called HOST_CC and HOST_CFLAGS; the
1412     first of these names is deprecated in the latest Autoconf in favour of the name
1413     CC_FOR_BUILD, because "host" is typically used to mean the system on which the
1414     compiled code will be run. I can't find a reference for HOST_CFLAGS, but by
1415     analogy I have changed it to CFLAGS_FOR_BUILD.
1416    
1417     4. Added -no-undefined to the linking command in the Makefile, because this is
1418     apparently helpful for Windows. To make it work, also added "-L. -lpcre" to the
1419     linking step for the pcreposix library.
1420    
1421     5. PCRE was failing to diagnose the case of two named groups with the same
1422     name.
1423    
1424     6. A problem with one of PCRE's optimizations was discovered. PCRE remembers a
1425     literal character that is needed in the subject for a match, and scans along to
1426     ensure that it is present before embarking on the full matching process. This
1427     saves time in cases of nested unlimited repeats that are never going to match.
1428     Problem: the scan can take a lot of time if the subject is very long (e.g.
1429     megabytes), thus penalizing straightforward matches. It is now done only if the
1430     amount of subject to be scanned is less than 1000 bytes.
1431    
1432     7. A lesser problem with the same optimization is that it was recording the
1433     first character of an anchored pattern as "needed", thus provoking a search
1434     right along the subject, even when the first match of the pattern was going to
1435     fail. The "needed" character is now not set for anchored patterns, unless it
1436     follows something in the pattern that is of non-fixed length. Thus, it still
1437     fulfils its original purpose of finding quick non-matches in cases of nested
1438     unlimited repeats, but isn't used for simple anchored patterns such as /^abc/.
1439    
1440    
1441     Version 4.0 17-Feb-03
1442     ---------------------
1443    
1444 nigel 63 1. If a comment in an extended regex that started immediately after a meta-item
1445     extended to the end of string, PCRE compiled incorrect data. This could lead to
1446     all kinds of weird effects. Example: /#/ was bad; /()#/ was bad; /a#/ was not.
1447    
1448     2. Moved to autoconf 2.53 and libtool 1.4.2.
1449    
1450     3. Perl 5.8 no longer needs "use utf8" for doing UTF-8 things. Consequently,
1451     the special perltest8 script is no longer needed - all the tests can be run
1452     from a single perltest script.
1453    
1454     4. From 5.004, Perl has not included the VT character (0x0b) in the set defined
1455     by \s. It has now been removed in PCRE. This means it isn't recognized as
1456     whitespace in /x regexes too, which is the same as Perl. Note that the POSIX
1457     class [:space:] *does* include VT, thereby creating a mess.
1458    
1459     5. Added the class [:blank:] (a GNU extension from Perl 5.8) to match only
1460     space and tab.
1461    
1462     6. Perl 5.005 was a long time ago. It's time to amalgamate the tests that use
1463     its new features into the main test script, reducing the number of scripts.
1464    
1465     7. Perl 5.8 has changed the meaning of patterns like /a(?i)b/. Earlier versions
1466     were backward compatible, and made the (?i) apply to the whole pattern, as if
1467     /i were given. Now it behaves more logically, and applies the option setting
1468     only to what follows. PCRE has been changed to follow suit. However, if it
1469     finds options settings right at the start of the pattern, it extracts them into
1470     the global options, as before. Thus, they show up in the info data.
1471    
1472     8. Added support for the \Q...\E escape sequence. Characters in between are
1473     treated as literals. This is slightly different from Perl in that $ and @ are
1474     also handled as literals inside the quotes. In Perl, they will cause variable
1475     interpolation. Note the following examples:
1476    
1477     Pattern PCRE matches Perl matches
1478    
1479     \Qabc$xyz\E abc$xyz abc followed by the contents of $xyz
1480     \Qabc\$xyz\E abc\$xyz abc\$xyz
1481     \Qabc\E\$\Qxyz\E abc$xyz abc$xyz
1482    
1483     For compatibility with Perl, \Q...\E sequences are recognized inside character
1484     classes as well as outside them.
1485    
1486     9. Re-organized 3 code statements in pcretest to avoid "overflow in
1487     floating-point constant arithmetic" warnings from a Microsoft compiler. Added a
1488     (size_t) cast to one statement in pcretest and one in pcreposix to avoid
1489     signed/unsigned warnings.
1490    
1491     10. SunOS4 doesn't have strtoul(). This was used only for unpicking the -o
1492     option for pcretest, so I've replaced it by a simple function that does just
1493     that job.
1494    
1495     11. pcregrep was ending with code 0 instead of 2 for the commands "pcregrep" or
1496     "pcregrep -".
1497    
1498     12. Added "possessive quantifiers" ?+, *+, ++, and {,}+ which come from Sun's
1499     Java package. This provides some syntactic sugar for simple cases of what my
1500     documentation calls "once-only subpatterns". A pattern such as x*+ is the same
1501     as (?>x*). In other words, if what is inside (?>...) is just a single repeated
1502     item, you can use this simplified notation. Note that only makes sense with
1503     greedy quantifiers. Consequently, the use of the possessive quantifier forces
1504     greediness, whatever the setting of the PCRE_UNGREEDY option.
1505    
1506     13. A change of greediness default within a pattern was not taking effect at
1507     the current level for patterns like /(b+(?U)a+)/. It did apply to parenthesized
1508     subpatterns that followed. Patterns like /b+(?U)a+/ worked because the option
1509     was abstracted outside.
1510    
1511     14. PCRE now supports the \G assertion. It is true when the current matching
1512     position is at the start point of the match. This differs from \A when the
1513     starting offset is non-zero. Used with the /g option of pcretest (or similar
1514     code), it works in the same way as it does for Perl's /g option. If all
1515     alternatives of a regex begin with \G, the expression is anchored to the start
1516     match position, and the "anchored" flag is set in the compiled expression.
1517    
1518     15. Some bugs concerning the handling of certain option changes within patterns
1519     have been fixed. These applied to options other than (?ims). For example,
1520     "a(?x: b c )d" did not match "XabcdY" but did match "Xa b c dY". It should have
1521     been the other way round. Some of this was related to change 7 above.
1522    
1523     16. PCRE now gives errors for /[.x.]/ and /[=x=]/ as unsupported POSIX
1524     features, as Perl does. Previously, PCRE gave the warnings only for /[[.x.]]/
1525     and /[[=x=]]/. PCRE now also gives an error for /[:name:]/ because it supports
1526     POSIX classes only within a class (e.g. /[[:alpha:]]/).
1527    
1528     17. Added support for Perl's \C escape. This matches one byte, even in UTF8
1529     mode. Unlike ".", it always matches newline, whatever the setting of
1530     PCRE_DOTALL. However, PCRE does not permit \C to appear in lookbehind
1531     assertions. Perl allows it, but it doesn't (in general) work because it can't
1532     calculate the length of the lookbehind. At least, that's the case for Perl
1533     5.8.0 - I've been told they are going to document that it doesn't work in
1534     future.
1535    
1536     18. Added an error diagnosis for escapes that PCRE does not support: these are
1537     \L, \l, \N, \P, \p, \U, \u, and \X.
1538    
1539     19. Although correctly diagnosing a missing ']' in a character class, PCRE was
1540     reading past the end of the pattern in cases such as /[abcd/.
1541    
1542     20. PCRE was getting more memory than necessary for patterns with classes that
1543     contained both POSIX named classes and other characters, e.g. /[[:space:]abc/.
1544    
1545     21. Added some code, conditional on #ifdef VPCOMPAT, to make life easier for
1546     compiling PCRE for use with Virtual Pascal.
1547    
1548     22. Small fix to the Makefile to make it work properly if the build is done
1549     outside the source tree.
1550    
1551     23. Added a new extension: a condition to go with recursion. If a conditional
1552     subpattern starts with (?(R) the "true" branch is used if recursion has
1553     happened, whereas the "false" branch is used only at the top level.
1554    
1555     24. When there was a very long string of literal characters (over 255 bytes
1556     without UTF support, over 250 bytes with UTF support), the computation of how
1557     much memory was required could be incorrect, leading to segfaults or other
1558     strange effects.
1559    
1560     25. PCRE was incorrectly assuming anchoring (either to start of subject or to
1561     start of line for a non-DOTALL pattern) when a pattern started with (.*) and
1562     there was a subsequent back reference to those brackets. This meant that, for
1563     example, /(.*)\d+\1/ failed to match "abc123bc". Unfortunately, it isn't
1564     possible to check for precisely this case. All we can do is abandon the
1565     optimization if .* occurs inside capturing brackets when there are any back
1566     references whatsoever. (See below for a better fix that came later.)
1567    
1568     26. The handling of the optimization for finding the first character of a
1569     non-anchored pattern, and for finding a character that is required later in the
1570     match were failing in some cases. This didn't break the matching; it just
1571     failed to optimize when it could. The way this is done has been re-implemented.
1572    
1573     27. Fixed typo in error message for invalid (?R item (it said "(?p").
1574    
1575     28. Added a new feature that provides some of the functionality that Perl
1576     provides with (?{...}). The facility is termed a "callout". The way it is done
1577     in PCRE is for the caller to provide an optional function, by setting
1578     pcre_callout to its entry point. Like pcre_malloc and pcre_free, this is a
1579     global variable. By default it is unset, which disables all calling out. To get
1580     the function called, the regex must include (?C) at appropriate points. This
1581     is, in fact, equivalent to (?C0), and any number <= 255 may be given with (?C).
1582     This provides a means of identifying different callout points. When PCRE
1583     reaches such a point in the regex, if pcre_callout has been set, the external
1584     function is called. It is provided with data in a structure called
1585     pcre_callout_block, which is defined in pcre.h. If the function returns 0,
1586     matching continues; if it returns a non-zero value, the match at the current
1587     point fails. However, backtracking will occur if possible. [This was changed
1588     later and other features added - see item 49 below.]
1589    
1590     29. pcretest is upgraded to test the callout functionality. It provides a
1591     callout function that displays information. By default, it shows the start of
1592     the match and the current position in the text. There are some new data escapes
1593     to vary what happens:
1594    
1595     \C+ in addition, show current contents of captured substrings
1596     \C- do not supply a callout function
1597     \C!n return 1 when callout number n is reached
1598     \C!n!m return 1 when callout number n is reached for the mth time
1599    
1600     30. If pcregrep was called with the -l option and just a single file name, it
1601     output "<stdin>" if a match was found, instead of the file name.
1602    
1603     31. Improve the efficiency of the POSIX API to PCRE. If the number of capturing
1604     slots is less than POSIX_MALLOC_THRESHOLD, use a block on the stack to pass to
1605     pcre_exec(). This saves a malloc/free per call. The default value of
1606     POSIX_MALLOC_THRESHOLD is 10; it can be changed by --with-posix-malloc-threshold
1607     when configuring.
1608    
1609     32. The default maximum size of a compiled pattern is 64K. There have been a
1610     few cases of people hitting this limit. The code now uses macros to handle the
1611     storing of links as offsets within the compiled pattern. It defaults to 2-byte
1612     links, but this can be changed to 3 or 4 bytes by --with-link-size when
1613     configuring. Tests 2 and 5 work only with 2-byte links because they output
1614     debugging information about compiled patterns.
1615    
1616     33. Internal code re-arrangements:
1617    
1618     (a) Moved the debugging function for printing out a compiled regex into
1619     its own source file (printint.c) and used #include to pull it into
1620     pcretest.c and, when DEBUG is defined, into pcre.c, instead of having two
1621     separate copies.
1622    
1623     (b) Defined the list of op-code names for debugging as a macro in
1624     internal.h so that it is next to the definition of the opcodes.
1625    
1626     (c) Defined a table of op-code lengths for simpler skipping along compiled
1627     code. This is again a macro in internal.h so that it is next to the
1628     definition of the opcodes.
1629    
1630     34. Added support for recursive calls to individual subpatterns, along the
1631     lines of Robin Houston's patch (but implemented somewhat differently).
1632    
1633     35. Further mods to the Makefile to help Win32. Also, added code to pcregrep to
1634     allow it to read and process whole directories in Win32. This code was
1635     contributed by Lionel Fourquaux; it has not been tested by me.
1636    
1637     36. Added support for named subpatterns. The Python syntax (?P<name>...) is
1638     used to name a group. Names consist of alphanumerics and underscores, and must
1639     be unique. Back references use the syntax (?P=name) and recursive calls use
1640     (?P>name) which is a PCRE extension to the Python extension. Groups still have
1641     numbers. The function pcre_fullinfo() can be used after compilation to extract
1642     a name/number map. There are three relevant calls:
1643    
1644     PCRE_INFO_NAMEENTRYSIZE yields the size of each entry in the map
1645     PCRE_INFO_NAMECOUNT yields the number of entries
1646     PCRE_INFO_NAMETABLE yields a pointer to the map.
1647    
1648     The map is a vector of fixed-size entries. The size of each entry depends on
1649     the length of the longest name used. The first two bytes of each entry are the
1650     group number, most significant byte first. There follows the corresponding
1651     name, zero terminated. The names are in alphabetical order.
1652    
1653     37. Make the maximum literal string in the compiled code 250 for the non-UTF-8
1654     case instead of 255. Making it the same both with and without UTF-8 support
1655     means that the same test output works with both.
1656    
1657     38. There was a case of malloc(0) in the POSIX testing code in pcretest. Avoid
1658     calling malloc() with a zero argument.
1659    
1660     39. Change 25 above had to resort to a heavy-handed test for the .* anchoring
1661     optimization. I've improved things by keeping a bitmap of backreferences with
1662     numbers 1-31 so that if .* occurs inside capturing brackets that are not in
1663     fact referenced, the optimization can be applied. It is unlikely that a
1664     relevant occurrence of .* (i.e. one which might indicate anchoring or forcing
1665     the match to follow \n) will appear inside brackets with a number greater than
1666     31, but if it does, any back reference > 31 suppresses the optimization.
1667    
1668     40. Added a new compile-time option PCRE_NO_AUTO_CAPTURE. This has the effect
1669     of disabling numbered capturing parentheses. Any opening parenthesis that is
1670     not followed by ? behaves as if it were followed by ?: but named parentheses
1671     can still be used for capturing (and they will acquire numbers in the usual
1672     way).
1673    
1674     41. Redesigned the return codes from the match() function into yes/no/error so
1675     that errors can be passed back from deep inside the nested calls. A malloc
1676     failure while inside a recursive subpattern call now causes the
1677     PCRE_ERROR_NOMEMORY return instead of quietly going wrong.
1678    
1679     42. It is now possible to set a limit on the number of times the match()
1680     function is called in a call to pcre_exec(). This facility makes it possible to
1681     limit the amount of recursion and backtracking, though not in a directly
1682     obvious way, because the match() function is used in a number of different
1683     circumstances. The count starts from zero for each position in the subject
1684     string (for non-anchored patterns). The default limit is, for compatibility, a
1685     large number, namely 10 000 000. You can change this in two ways:
1686    
1687     (a) When configuring PCRE before making, you can use --with-match-limit=n
1688     to set a default value for the compiled library.
1689    
1690     (b) For each call to pcre_exec(), you can pass a pcre_extra block in which
1691     a different value is set. See 45 below.
1692    
1693     If the limit is exceeded, pcre_exec() returns PCRE_ERROR_MATCHLIMIT.
1694    
1695     43. Added a new function pcre_config(int, void *) to enable run-time extraction
1696     of things that can be changed at compile time. The first argument specifies
1697     what is wanted and the second points to where the information is to be placed.
1698     The current list of available information is:
1699    
1700     PCRE_CONFIG_UTF8
1701    
1702     The output is an integer that is set to one if UTF-8 support is available;
1703     otherwise it is set to zero.
1704    
1705     PCRE_CONFIG_NEWLINE
1706    
1707     The output is an integer that it set to the value of the code that is used for
1708     newline. It is either LF (10) or CR (13).
1709    
1710     PCRE_CONFIG_LINK_SIZE
1711    
1712     The output is an integer that contains the number of bytes used for internal
1713     linkage in compiled expressions. The value is 2, 3, or 4. See item 32 above.
1714    
1715     PCRE_CONFIG_POSIX_MALLOC_THRESHOLD
1716    
1717     The output is an integer that contains the threshold above which the POSIX
1718     interface uses malloc() for output vectors. See item 31 above.
1719    
1720     PCRE_CONFIG_MATCH_LIMIT
1721    
1722     The output is an unsigned integer that contains the default limit of the number
1723     of match() calls in a pcre_exec() execution. See 42 above.
1724    
1725     44. pcretest has been upgraded by the addition of the -C option. This causes it
1726     to extract all the available output from the new pcre_config() function, and to
1727     output it. The program then exits immediately.
1728    
1729     45. A need has arisen to pass over additional data with calls to pcre_exec() in
1730     order to support additional features. One way would have been to define
1731     pcre_exec2() (for example) with extra arguments, but this would not have been
1732     extensible, and would also have required all calls to the original function to
1733     be mapped to the new one. Instead, I have chosen to extend the mechanism that
1734     is used for passing in "extra" data from pcre_study().
1735    
1736     The pcre_extra structure is now exposed and defined in pcre.h. It currently
1737     contains the following fields:
1738    
1739     flags a bitmap indicating which of the following fields are set
1740     study_data opaque data from pcre_study()
1741     match_limit a way of specifying a limit on match() calls for a specific
1742     call to pcre_exec()
1743     callout_data data for callouts (see 49 below)
1744    
1745     The flag bits are also defined in pcre.h, and are
1746    
1747     PCRE_EXTRA_STUDY_DATA
1748     PCRE_EXTRA_MATCH_LIMIT
1749     PCRE_EXTRA_CALLOUT_DATA
1750    
1751     The pcre_study() function now returns one of these new pcre_extra blocks, with
1752     the actual study data pointed to by the study_data field, and the
1753     PCRE_EXTRA_STUDY_DATA flag set. This can be passed directly to pcre_exec() as
1754     before. That is, this change is entirely upwards-compatible and requires no
1755     change to existing code.
1756    
1757     If you want to pass in additional data to pcre_exec(), you can either place it
1758     in a pcre_extra block provided by pcre_study(), or create your own pcre_extra
1759     block.
1760    
1761     46. pcretest has been extended to test the PCRE_EXTRA_MATCH_LIMIT feature. If a
1762     data string contains the escape sequence \M, pcretest calls pcre_exec() several
1763     times with different match limits, until it finds the minimum value needed for
1764     pcre_exec() to complete. The value is then output. This can be instructive; for
1765     most simple matches the number is quite small, but for pathological cases it
1766     gets very large very quickly.
1767    
1768     47. There's a new option for pcre_fullinfo() called PCRE_INFO_STUDYSIZE. It
1769     returns the size of the data block pointed to by the study_data field in a
1770     pcre_extra block, that is, the value that was passed as the argument to
1771     pcre_malloc() when PCRE was getting memory in which to place the information
1772     created by pcre_study(). The fourth argument should point to a size_t variable.
1773     pcretest has been extended so that this information is shown after a successful
1774     pcre_study() call when information about the compiled regex is being displayed.
1775    
1776     48. Cosmetic change to Makefile: there's no need to have / after $(DESTDIR)
1777     because what follows is always an absolute path. (Later: it turns out that this
1778     is more than cosmetic for MinGW, because it doesn't like empty path
1779     components.)
1780    
1781     49. Some changes have been made to the callout feature (see 28 above):
1782    
1783     (i) A callout function now has three choices for what it returns:
1784    
1785     0 => success, carry on matching
1786     > 0 => failure at this point, but backtrack if possible
1787     < 0 => serious error, return this value from pcre_exec()
1788    
1789     Negative values should normally be chosen from the set of PCRE_ERROR_xxx
1790     values. In particular, returning PCRE_ERROR_NOMATCH forces a standard
1791     "match failed" error. The error number PCRE_ERROR_CALLOUT is reserved for
1792     use by callout functions. It will never be used by PCRE itself.
1793    
1794     (ii) The pcre_extra structure (see 45 above) has a void * field called
1795     callout_data, with corresponding flag bit PCRE_EXTRA_CALLOUT_DATA. The
1796     pcre_callout_block structure has a field of the same name. The contents of
1797     the field passed in the pcre_extra structure are passed to the callout
1798     function in the corresponding field in the callout block. This makes it
1799     easier to use the same callout-containing regex from multiple threads. For
1800     testing, the pcretest program has a new data escape
1801    
1802     \C*n pass the number n (may be negative) as callout_data
1803    
1804     If the callout function in pcretest receives a non-zero value as
1805     callout_data, it returns that value.
1806    
1807     50. Makefile wasn't handling CFLAGS properly when compiling dftables. Also,
1808     there were some redundant $(CFLAGS) in commands that are now specified as
1809     $(LINK), which already includes $(CFLAGS).
1810    
1811     51. Extensions to UTF-8 support are listed below. These all apply when (a) PCRE
1812     has been compiled with UTF-8 support *and* pcre_compile() has been compiled
1813     with the PCRE_UTF8 flag. Patterns that are compiled without that flag assume
1814     one-byte characters throughout. Note that case-insensitive matching applies
1815     only to characters whose values are less than 256. PCRE doesn't support the
1816     notion of cases for higher-valued characters.
1817    
1818     (i) A character class whose characters are all within 0-255 is handled as
1819     a bit map, and the map is inverted for negative classes. Previously, a
1820     character > 255 always failed to match such a class; however it should
1821     match if the class was a negative one (e.g. [^ab]). This has been fixed.
1822    
1823     (ii) A negated character class with a single character < 255 is coded as
1824     "not this character" (OP_NOT). This wasn't working properly when the test
1825     character was multibyte, either singly or repeated.
1826    
1827     (iii) Repeats of multibyte characters are now handled correctly in UTF-8
1828     mode, for example: \x{100}{2,3}.
1829    
1830     (iv) The character escapes \b, \B, \d, \D, \s, \S, \w, and \W (either
1831     singly or repeated) now correctly test multibyte characters. However,
1832     PCRE doesn't recognize any characters with values greater than 255 as
1833     digits, spaces, or word characters. Such characters always match \D, \S,
1834     and \W, and never match \d, \s, or \w.
1835    
1836     (v) Classes may now contain characters and character ranges with values
1837     greater than 255. For example: [ab\x{100}-\x{400}].
1838    
1839     (vi) pcregrep now has a --utf-8 option (synonym -u) which makes it call
1840     PCRE in UTF-8 mode.
1841    
1842     52. The info request value PCRE_INFO_FIRSTCHAR has been renamed
1843     PCRE_INFO_FIRSTBYTE because it is a byte value. However, the old name is
1844     retained for backwards compatibility. (Note that LASTLITERAL is also a byte
1845     value.)
1846    
1847     53. The single man page has become too large. I have therefore split it up into
1848     a number of separate man pages. These also give rise to individual HTML pages;
1849     these are now put in a separate directory, and there is an index.html page that
1850     lists them all. Some hyperlinking between the pages has been installed.
1851    
1852     54. Added convenience functions for handling named capturing parentheses.
1853    
1854     55. Unknown escapes inside character classes (e.g. [\M]) and escapes that
1855     aren't interpreted therein (e.g. [\C]) are literals in Perl. This is now also
1856     true in PCRE, except when the PCRE_EXTENDED option is set, in which case they
1857     are faulted.
1858    
1859     56. Introduced HOST_CC and HOST_CFLAGS which can be set in the environment when
1860     calling configure. These values are used when compiling the dftables.c program
1861     which is run to generate the source of the default character tables. They
1862     default to the values of CC and CFLAGS. If you are cross-compiling PCRE,
1863     you will need to set these values.
1864    
1865     57. Updated the building process for Windows DLL, as provided by Fred Cox.
1866    
1867    
1868     Version 3.9 02-Jan-02
1869 nigel 61 ---------------------
1870    
1871     1. A bit of extraneous text had somehow crept into the pcregrep documentation.
1872    
1873     2. If --disable-static was given, the building process failed when trying to
1874     build pcretest and pcregrep. (For some reason it was using libtool to compile
1875     them, which is not right, as they aren't part of the library.)
1876    
1877    
1878 nigel 59 Version 3.8 18-Dec-01
1879     ---------------------
1880    
1881     1. The experimental UTF-8 code was completely screwed up. It was packing the
1882     bytes in the wrong order. How dumb can you get?
1883    
1884    
1885 nigel 57 Version 3.7 29-Oct-01
1886     ---------------------
1887    
1888     1. In updating pcretest to check change 1 of version 3.6, I screwed up.
1889     This caused pcretest, when used on the test data, to segfault. Unfortunately,
1890     this didn't happen under Solaris 8, where I normally test things.
1891    
1892 nigel 59 2. The Makefile had to be changed to make it work on BSD systems, where 'make'
1893     doesn't seem to recognize that ./xxx and xxx are the same file. (This entry
1894     isn't in ChangeLog distributed with 3.7 because I forgot when I hastily made
1895     this fix an hour or so after the initial 3.7 release.)
1896 nigel 57
1897 nigel 59
1898 nigel 55 Version 3.6 23-Oct-01
1899     ---------------------
1900    
1901     1. Crashed with /(sens|respons)e and \1ibility/ and "sense and sensibility" if
1902     offsets passed as NULL with zero offset count.
1903    
1904     2. The config.guess and config.sub files had not been updated when I moved to
1905     the latest autoconf.
1906    
1907    
1908 nigel 53 Version 3.5 15-Aug-01
1909     ---------------------
1910 nigel 5
1911 nigel 53 1. Added some missing #if !defined NOPOSIX conditionals in pcretest.c that
1912     had been forgotten.
1913    
1914     2. By using declared but undefined structures, we can avoid using "void"
1915     definitions in pcre.h while keeping the internal definitions of the structures
1916     private.
1917    
1918     3. The distribution is now built using autoconf 2.50 and libtool 1.4. From a
1919     user point of view, this means that both static and shared libraries are built
1920     by default, but this can be individually controlled. More of the work of
1921     handling this static/shared cases is now inside libtool instead of PCRE's make
1922     file.
1923    
1924     4. The pcretest utility is now installed along with pcregrep because it is
1925     useful for users (to test regexs) and by doing this, it automatically gets
1926     relinked by libtool. The documentation has been turned into a man page, so
1927     there are now .1, .txt, and .html versions in /doc.
1928    
1929     5. Upgrades to pcregrep:
1930     (i) Added long-form option names like gnu grep.
1931     (ii) Added --help to list all options with an explanatory phrase.
1932     (iii) Added -r, --recursive to recurse into sub-directories.
1933     (iv) Added -f, --file to read patterns from a file.
1934    
1935     6. pcre_exec() was referring to its "code" argument before testing that
1936     argument for NULL (and giving an error if it was NULL).
1937    
1938     7. Upgraded Makefile.in to allow for compiling in a different directory from
1939     the source directory.
1940    
1941     8. Tiny buglet in pcretest: when pcre_fullinfo() was called to retrieve the
1942     options bits, the pointer it was passed was to an int instead of to an unsigned
1943     long int. This mattered only on 64-bit systems.
1944    
1945     9. Fixed typo (3.4/1) in pcre.h again. Sigh. I had changed pcre.h (which is
1946     generated) instead of pcre.in, which it its source. Also made the same change
1947     in several of the .c files.
1948    
1949     10. A new release of gcc defines printf() as a macro, which broke pcretest
1950     because it had an ifdef in the middle of a string argument for printf(). Fixed
1951     by using separate calls to printf().
1952    
1953     11. Added --enable-newline-is-cr and --enable-newline-is-lf to the configure
1954     script, to force use of CR or LF instead of \n in the source. On non-Unix
1955     systems, the value can be set in config.h.
1956    
1957     12. The limit of 200 on non-capturing parentheses is a _nesting_ limit, not an
1958     absolute limit. Changed the text of the error message to make this clear, and
1959     likewise updated the man page.
1960    
1961     13. The limit of 99 on the number of capturing subpatterns has been removed.
1962     The new limit is 65535, which I hope will not be a "real" limit.
1963    
1964    
1965 nigel 51 Version 3.4 22-Aug-00
1966     ---------------------
1967    
1968     1. Fixed typo in pcre.h: unsigned const char * changed to const unsigned char *.
1969    
1970     2. Diagnose condition (?(0) as an error instead of crashing on matching.
1971    
1972    
1973 nigel 49 Version 3.3 01-Aug-00
1974     ---------------------
1975    
1976     1. If an octal character was given, but the value was greater than \377, it
1977     was not getting masked to the least significant bits, as documented. This could
1978     lead to crashes in some systems.
1979    
1980     2. Perl 5.6 (if not earlier versions) accepts classes like [a-\d] and treats
1981     the hyphen as a literal. PCRE used to give an error; it now behaves like Perl.
1982    
1983     3. Added the functions pcre_free_substring() and pcre_free_substring_list().
1984     These just pass their arguments on to (pcre_free)(), but they are provided
1985     because some uses of PCRE bind it to non-C systems that can call its functions,
1986     but cannot call free() or pcre_free() directly.
1987    
1988     4. Add "make test" as a synonym for "make check". Corrected some comments in
1989     the Makefile.
1990    
1991     5. Add $(DESTDIR)/ in front of all the paths in the "install" target in the
1992     Makefile.
1993    
1994     6. Changed the name of pgrep to pcregrep, because Solaris has introduced a
1995     command called pgrep for grepping around the active processes.
1996    
1997     7. Added the beginnings of support for UTF-8 character strings.
1998    
1999     8. Arranged for the Makefile to pass over the settings of CC, CFLAGS, and
2000     RANLIB to ./ltconfig so that they are used by libtool. I think these are all
2001     the relevant ones. (AR is not passed because ./ltconfig does its own figuring
2002     out for the ar command.)
2003    
2004    
2005 nigel 47 Version 3.2 12-May-00
2006     ---------------------
2007    
2008     This is purely a bug fixing release.
2009    
2010     1. If the pattern /((Z)+|A)*/ was matched agained ZABCDEFG it matched Z instead
2011     of ZA. This was just one example of several cases that could provoke this bug,
2012     which was introduced by change 9 of version 2.00. The code for breaking
2013     infinite loops after an iteration that matches an empty string was't working
2014     correctly.
2015    
2016     2. The pcretest program was not imitating Perl correctly for the pattern /a*/g
2017     when matched against abbab (for example). After matching an empty string, it
2018     wasn't forcing anchoring when setting PCRE_NOTEMPTY for the next attempt; this
2019     caused it to match further down the string than it should.
2020    
2021     3. The code contained an inclusion of sys/types.h. It isn't clear why this
2022     was there because it doesn't seem to be needed, and it causes trouble on some
2023     systems, as it is not a Standard C header. It has been removed.
2024    
2025     4. Made 4 silly changes to the source to avoid stupid compiler warnings that
2026     were reported on the Macintosh. The changes were from
2027    
2028     while ((c = *(++ptr)) != 0 && c != '\n');
2029     to
2030     while ((c = *(++ptr)) != 0 && c != '\n') ;
2031    
2032     Totally extraordinary, but if that's what it takes...
2033    
2034     5. PCRE is being used in one environment where neither memmove() nor bcopy() is
2035     available. Added HAVE_BCOPY and an autoconf test for it; if neither
2036     HAVE_MEMMOVE nor HAVE_BCOPY is set, use a built-in emulation function which
2037     assumes the way PCRE uses memmove() (always moving upwards).
2038    
2039     6. PCRE is being used in one environment where strchr() is not available. There
2040     was only one use in pcre.c, and writing it out to avoid strchr() probably gives
2041     faster code anyway.
2042    
2043    
2044 nigel 45 Version 3.1 09-Feb-00
2045     ---------------------
2046    
2047     The only change in this release is the fixing of some bugs in Makefile.in for
2048     the "install" target:
2049    
2050     (1) It was failing to install pcreposix.h.
2051    
2052     (2) It was overwriting the pcre.3 man page with the pcreposix.3 man page.
2053    
2054    
2055 nigel 43 Version 3.0 01-Feb-00
2056     ---------------------
2057 nigel 41
2058     1. Add support for the /+ modifier to perltest (to output $` like it does in
2059     pcretest).
2060    
2061     2. Add support for the /g modifier to perltest.
2062    
2063     3. Fix pcretest so that it behaves even more like Perl for /g when the pattern
2064     matches null strings.
2065    
2066     4. Fix perltest so that it doesn't do unwanted things when fed an empty
2067     pattern. Perl treats empty patterns specially - it reuses the most recent
2068     pattern, which is not what we want. Replace // by /(?#)/ in order to avoid this
2069     effect.
2070    
2071     5. The POSIX interface was broken in that it was just handing over the POSIX
2072     captured string vector to pcre_exec(), but (since release 2.00) PCRE has
2073     required a bigger vector, with some working space on the end. This means that
2074     the POSIX wrapper now has to get and free some memory, and copy the results.
2075    
2076 nigel 43 6. Added some simple autoconf support, placing the test data and the
2077     documentation in separate directories, re-organizing some of the
2078     information files, and making it build pcre-config (a GNU standard). Also added
2079     libtool support for building PCRE as a shared library, which is now the
2080     default.
2081 nigel 41
2082 nigel 43 7. Got rid of the leading zero in the definition of PCRE_MINOR because 08 and
2083     09 are not valid octal constants. Single digits will be used for minor values
2084     less than 10.
2085    
2086     8. Defined REG_EXTENDED and REG_NOSUB as zero in the POSIX header, so that
2087     existing programs that set these in the POSIX interface can use PCRE without
2088     modification.
2089    
2090     9. Added a new function, pcre_fullinfo() with an extensible interface. It can
2091     return all that pcre_info() returns, plus additional data. The pcre_info()
2092     function is retained for compatibility, but is considered to be obsolete.
2093    
2094     10. Added experimental recursion feature (?R) to handle one common case that
2095     Perl 5.6 will be able to do with (?p{...}).
2096    
2097     11. Added support for POSIX character classes like [:alpha:], which Perl is
2098     adopting.
2099    
2100    
2101 nigel 39 Version 2.08 31-Aug-99
2102     ----------------------
2103    
2104     1. When startoffset was not zero and the pattern began with ".*", PCRE was not
2105     trying to match at the startoffset position, but instead was moving forward to
2106     the next newline as if a previous match had failed.
2107    
2108     2. pcretest was not making use of PCRE_NOTEMPTY when repeating for /g and /G,
2109     and could get into a loop if a null string was matched other than at the start
2110     of the subject.
2111    
2112     3. Added definitions of PCRE_MAJOR and PCRE_MINOR to pcre.h so the version can
2113     be distinguished at compile time, and for completeness also added PCRE_DATE.
2114    
2115     5. Added Paul Sokolovsky's minor changes to make it easy to compile a Win32 DLL
2116     in GnuWin32 environments.
2117    
2118    
2119 nigel 37 Version 2.07 29-Jul-99
2120     ----------------------
2121    
2122     1. The documentation is now supplied in plain text form and HTML as well as in
2123     the form of man page sources.
2124    
2125     2. C++ compilers don't like assigning (void *) values to other pointer types.
2126     In particular this affects malloc(). Although there is no problem in Standard
2127     C, I've put in casts to keep C++ compilers happy.
2128    
2129     3. Typo on pcretest.c; a cast of (unsigned char *) in the POSIX regexec() call
2130     should be (const char *).
2131    
2132     4. If NOPOSIX is defined, pcretest.c compiles without POSIX support. This may
2133     be useful for non-Unix systems who don't want to bother with the POSIX stuff.
2134     However, I haven't made this a standard facility. The documentation doesn't
2135     mention it, and the Makefile doesn't support it.
2136    
2137     5. The Makefile now contains an "install" target, with editable destinations at
2138     the top of the file. The pcretest program is not installed.
2139    
2140     6. pgrep -V now gives the PCRE version number and date.
2141    
2142     7. Fixed bug: a zero repetition after a literal string (e.g. /abcde{0}/) was
2143     causing the entire string to be ignored, instead of just the last character.
2144    
2145     8. If a pattern like /"([^\\"]+|\\.)*"/ is applied in the normal way to a
2146     non-matching string, it can take a very, very long time, even for strings of
2147     quite modest length, because of the nested recursion. PCRE now does better in
2148     some of these cases. It does this by remembering the last required literal
2149     character in the pattern, and pre-searching the subject to ensure it is present
2150     before running the real match. In other words, it applies a heuristic to detect
2151     some types of certain failure quickly, and in the above example, if presented
2152     with a string that has no trailing " it gives "no match" very quickly.
2153    
2154     9. A new runtime option PCRE_NOTEMPTY causes null string matches to be ignored;
2155     other alternatives are tried instead.
2156    
2157    
2158 nigel 35 Version 2.06 09-Jun-99
2159     ----------------------
2160    
2161     1. Change pcretest's output for amount of store used to show just the code
2162     space, because the remainder (the data block) varies in size between 32-bit and
2163     64-bit systems.
2164    
2165     2. Added an extra argument to pcre_exec() to supply an offset in the subject to
2166     start matching at. This allows lookbehinds to work when searching for multiple
2167     occurrences in a string.
2168    
2169     3. Added additional options to pcretest for testing multiple occurrences:
2170    
2171     /+ outputs the rest of the string that follows a match
2172     /g loops for multiple occurrences, using the new startoffset argument
2173     /G loops for multiple occurrences by passing an incremented pointer
2174    
2175     4. PCRE wasn't doing the "first character" optimization for patterns starting
2176     with \b or \B, though it was doing it for other lookbehind assertions. That is,
2177     it wasn't noticing that a match for a pattern such as /\bxyz/ has to start with
2178     the letter 'x'. On long subject strings, this gives a significant speed-up.
2179    
2180    
2181 nigel 33 Version 2.05 21-Apr-99
2182     ----------------------
2183    
2184     1. Changed the type of magic_number from int to long int so that it works
2185     properly on 16-bit systems.
2186    
2187     2. Fixed a bug which caused patterns starting with .* not to work correctly
2188     when the subject string contained newline characters. PCRE was assuming
2189     anchoring for such patterns in all cases, which is not correct because .* will
2190     not pass a newline unless PCRE_DOTALL is set. It now assumes anchoring only if
2191     DOTALL is set at top level; otherwise it knows that patterns starting with .*
2192     must be retried after every newline in the subject.
2193    
2194    
2195 nigel 31 Version 2.04 18-Feb-99
2196     ----------------------
2197    
2198     1. For parenthesized subpatterns with repeats whose minimum was zero, the
2199     computation of the store needed to hold the pattern was incorrect (too large).
2200     If such patterns were nested a few deep, this could multiply and become a real
2201     problem.
2202    
2203     2. Added /M option to pcretest to show the memory requirement of a specific
2204     pattern. Made -m a synonym of -s (which does this globally) for compatibility.
2205    
2206     3. Subpatterns of the form (regex){n,m} (i.e. limited maximum) were being
2207     compiled in such a way that the backtracking after subsequent failure was
2208     pessimal. Something like (a){0,3} was compiled as (a)?(a)?(a)? instead of
2209     ((a)((a)(a)?)?)? with disastrous performance if the maximum was of any size.
2210    
2211    
2212 nigel 29 Version 2.03 02-Feb-99
2213     ----------------------
2214    
2215     1. Fixed typo and small mistake in man page.
2216    
2217 nigel 31 2. Added 4th condition (GPL supersedes if conflict) and created separate
2218     LICENCE file containing the conditions.
2219 nigel 29
2220     3. Updated pcretest so that patterns such as /abc\/def/ work like they do in
2221     Perl, that is the internal \ allows the delimiter to be included in the
2222     pattern. Locked out the use of \ as a delimiter. If \ immediately follows
2223     the final delimiter, add \ to the end of the pattern (to test the error).
2224    
2225     4. Added the convenience functions for extracting substrings after a successful
2226     match. Updated pcretest to make it able to test these functions.
2227    
2228    
2229 nigel 27 Version 2.02 14-Jan-99
2230     ----------------------
2231    
2232     1. Initialized the working variables associated with each extraction so that
2233     their saving and restoring doesn't refer to uninitialized store.
2234    
2235     2. Put dummy code into study.c in order to trick the optimizer of the IBM C
2236     compiler for OS/2 into generating correct code. Apparently IBM isn't going to
2237     fix the problem.
2238    
2239     3. Pcretest: the timing code wasn't using LOOPREPEAT for timing execution
2240     calls, and wasn't printing the correct value for compiling calls. Increased the
2241     default value of LOOPREPEAT, and the number of significant figures in the
2242     times.
2243    
2244     4. Changed "/bin/rm" in the Makefile to "-rm" so it works on Windows NT.
2245    
2246     5. Renamed "deftables" as "dftables" to get it down to 8 characters, to avoid
2247     a building problem on Windows NT with a FAT file system.
2248    
2249    
2250 nigel 25 Version 2.01 21-Oct-98
2251     ----------------------
2252    
2253     1. Changed the API for pcre_compile() to allow for the provision of a pointer
2254     to character tables built by pcre_maketables() in the current locale. If NULL
2255     is passed, the default tables are used.
2256    
2257    
2258 nigel 23 Version 2.00 24-Sep-98
2259 nigel 21 ----------------------
2260    
2261 nigel 23 1. Since the (>?) facility is in Perl 5.005, don't require PCRE_EXTRA to enable
2262     it any more.
2263    
2264     2. Allow quantification of (?>) groups, and make it work correctly.
2265    
2266     3. The first character computation wasn't working for (?>) groups.
2267    
2268     4. Correct the implementation of \Z (it is permitted to match on the \n at the
2269     end of the subject) and add 5.005's \z, which really does match only at the
2270     very end of the subject.
2271    
2272     5. Remove the \X "cut" facility; Perl doesn't have it, and (?> is neater.
2273    
2274     6. Remove the ability to specify CASELESS, MULTILINE, DOTALL, and
2275     DOLLAR_END_ONLY at runtime, to make it possible to implement the Perl 5.005
2276     localized options. All options to pcre_study() were also removed.
2277    
2278     7. Add other new features from 5.005:
2279    
2280     $(?<= positive lookbehind
2281     $(?<! negative lookbehind
2282     (?imsx-imsx) added the unsetting capability
2283     such a setting is global if at outer level; local otherwise
2284     (?imsx-imsx:) non-capturing groups with option setting
2285     (?(cond)re|re) conditional pattern matching
2286    
2287     A backreference to itself in a repeated group matches the previous
2288     captured string.
2289    
2290     8. General tidying up of studying (both automatic and via "study")
2291     consequential on the addition of new assertions.
2292    
2293     9. As in 5.005, unlimited repeated groups that could match an empty substring
2294     are no longer faulted at compile time. Instead, the loop is forcibly broken at
2295     runtime if any iteration does actually match an empty substring.
2296    
2297     10. Include the RunTest script in the distribution.
2298    
2299     11. Added tests from the Perl 5.005_02 distribution. This showed up a few
2300     discrepancies, some of which were old and were also with respect to 5.004. They
2301     have now been fixed.
2302    
2303    
2304     Version 1.09 28-Apr-98
2305     ----------------------
2306    
2307 nigel 21 1. A negated single character class followed by a quantifier with a minimum
2308     value of one (e.g. [^x]{1,6} ) was not compiled correctly. This could lead to
2309     program crashes, or just wrong answers. This did not apply to negated classes
2310     containing more than one character, or to minima other than one.
2311    
2312    
2313 nigel 19 Version 1.08 27-Mar-98
2314     ----------------------
2315    
2316     1. Add PCRE_UNGREEDY to invert the greediness of quantifiers.
2317    
2318     2. Add (?U) and (?X) to set PCRE_UNGREEDY and PCRE_EXTRA respectively. The
2319     latter must appear before anything that relies on it in the pattern.
2320    
2321    
2322 nigel 17 Version 1.07 16-Feb-98
2323     ----------------------
2324    
2325     1. A pattern such as /((a)*)*/ was not being diagnosed as in error (unlimited
2326     repeat of a potentially empty string).
2327    
2328    
2329 nigel 15 Version 1.06 23-Jan-98
2330     ----------------------
2331    
2332     1. Added Markus Oberhumer's little patches for C++.
2333    
2334     2. Literal strings longer than 255 characters were broken.
2335    
2336    
2337 nigel 13 Version 1.05 23-Dec-97
2338     ----------------------
2339    
2340     1. Negated character classes containing more than one character were failing if
2341     PCRE_CASELESS was set at run time.
2342    
2343    
2344 nigel 11 Version 1.04 19-Dec-97
2345     ----------------------
2346    
2347     1. Corrected the man page, where some "const" qualifiers had been omitted.
2348    
2349     2. Made debugging output print "{0,xxx}" instead of just "{,xxx}" to agree with
2350     input syntax.
2351    
2352     3. Fixed memory leak which occurred when a regex with back references was
2353     matched with an offsets vector that wasn't big enough. The temporary memory
2354     that is used in this case wasn't being freed if the match failed.
2355    
2356     4. Tidied pcretest to ensure it frees memory that it gets.
2357    
2358     5. Temporary memory was being obtained in the case where the passed offsets
2359     vector was exactly big enough.
2360    
2361     6. Corrected definition of offsetof() from change 5 below.
2362    
2363     7. I had screwed up change 6 below and broken the rules for the use of
2364     setjmp(). Now fixed.
2365    
2366    
2367 nigel 9 Version 1.03 18-Dec-97
2368     ----------------------
2369    
2370     1. A erroneous regex with a missing opening parenthesis was correctly
2371     diagnosed, but PCRE attempted to access brastack[-1], which could cause crashes
2372     on some systems.
2373    
2374     2. Replaced offsetof(real_pcre, code) by offsetof(real_pcre, code[0]) because
2375     it was reported that one broken compiler failed on the former because "code" is
2376     also an independent variable.
2377    
2378     3. The erroneous regex a[]b caused an array overrun reference.
2379    
2380     4. A regex ending with a one-character negative class (e.g. /[^k]$/) did not
2381     fail on data ending with that character. (It was going on too far, and checking
2382     the next character, typically a binary zero.) This was specific to the
2383     optimized code for single-character negative classes.
2384    
2385     5. Added a contributed patch from the TIN world which does the following:
2386    
2387     + Add an undef for memmove, in case the the system defines a macro for it.
2388    
2389     + Add a definition of offsetof(), in case there isn't one. (I don't know
2390     the reason behind this - offsetof() is part of the ANSI standard - but
2391     it does no harm).
2392    
2393     + Reduce the ifdef's in pcre.c using macro DPRINTF, thereby eliminating
2394     most of the places where whitespace preceded '#'. I have given up and
2395     allowed the remaining 2 cases to be at the margin.
2396    
2397     + Rename some variables in pcre to eliminate shadowing. This seems very
2398     pedantic, but does no harm, of course.
2399    
2400     6. Moved the call to setjmp() into its own function, to get rid of warnings
2401     from gcc -Wall, and avoided calling it at all unless PCRE_EXTRA is used.
2402    
2403     7. Constructs such as \d{8,} were compiling into the equivalent of
2404 nigel 11 \d{8}\d{0,65527} instead of \d{8}\d* which didn't make much difference to the
2405 nigel 9 outcome, but in this particular case used more store than had been allocated,
2406     which caused the bug to be discovered because it threw up an internal error.
2407    
2408     8. The debugging code in both pcre and pcretest for outputting the compiled
2409     form of a regex was going wrong in the case of back references followed by
2410     curly-bracketed repeats.
2411    
2412    
2413 nigel 7 Version 1.02 12-Dec-97
2414     ----------------------
2415    
2416     1. Typos in pcre.3 and comments in the source fixed.
2417    
2418     2. Applied a contributed patch to get rid of places where it used to remove
2419     'const' from variables, and fixed some signed/unsigned and uninitialized
2420     variable warnings.
2421    
2422     3. Added the "runtest" target to Makefile.
2423    
2424     4. Set default compiler flag to -O2 rather than just -O.
2425    
2426    
2427 nigel 5 Version 1.01 19-Nov-97
2428     ----------------------
2429    
2430     1. PCRE was failing to diagnose unlimited repeat of empty string for patterns
2431     like /([ab]*)*/, that is, for classes with more than one character in them.
2432    
2433     2. Likewise, it wasn't diagnosing patterns with "once-only" subpatterns, such
2434     as /((?>a*))*/ (a PCRE_EXTRA facility).
2435    
2436    
2437     Version 1.00 18-Nov-97
2438     ----------------------
2439    
2440     1. Added compile-time macros to support systems such as SunOS4 which don't have
2441     memmove() or strerror() but have other things that can be used instead.
2442    
2443     2. Arranged that "make clean" removes the executables.
2444    
2445    
2446 nigel 3 Version 0.99 27-Oct-97
2447     ----------------------
2448    
2449     1. Fixed bug in code for optimizing classes with only one character. It was
2450     initializing a 32-byte map regardless, which could cause it to run off the end
2451     of the memory it had got.
2452    
2453     2. Added, conditional on PCRE_EXTRA, the proposed (?>REGEX) construction.
2454    
2455    
2456     Version 0.98 22-Oct-97
2457     ----------------------
2458    
2459     1. Fixed bug in code for handling temporary memory usage when there are more
2460     back references than supplied space in the ovector. This could cause segfaults.
2461    
2462    
2463     Version 0.97 21-Oct-97
2464     ----------------------
2465    
2466     1. Added the \X "cut" facility, conditional on PCRE_EXTRA.
2467    
2468     2. Optimized negated single characters not to use a bit map.
2469    
2470     3. Brought error texts together as macro definitions; clarified some of them;
2471     fixed one that was wrong - it said "range out of order" when it meant "invalid
2472     escape sequence".
2473    
2474     4. Changed some char * arguments to const char *.
2475    
2476     5. Added PCRE_NOTBOL and PCRE_NOTEOL (from POSIX).
2477    
2478     6. Added the POSIX-style API wrapper in pcreposix.a and testing facilities in
2479     pcretest.
2480    
2481    
2482     Version 0.96 16-Oct-97
2483     ----------------------
2484    
2485     1. Added a simple "pgrep" utility to the distribution.
2486    
2487     2. Fixed an incompatibility with Perl: "{" is now treated as a normal character
2488     unless it appears in one of the precise forms "{ddd}", "{ddd,}", or "{ddd,ddd}"
2489     where "ddd" means "one or more decimal digits".
2490    
2491     3. Fixed serious bug. If a pattern had a back reference, but the call to
2492     pcre_exec() didn't supply a large enough ovector to record the related
2493     identifying subpattern, the match always failed. PCRE now remembers the number
2494     of the largest back reference, and gets some temporary memory in which to save
2495     the offsets during matching if necessary, in order to ensure that
2496     backreferences always work.
2497    
2498     4. Increased the compatibility with Perl in a number of ways:
2499    
2500     (a) . no longer matches \n by default; an option PCRE_DOTALL is provided
2501     to request this handling. The option can be set at compile or exec time.
2502    
2503     (b) $ matches before a terminating newline by default; an option
2504     PCRE_DOLLAR_ENDONLY is provided to override this (but not in multiline
2505     mode). The option can be set at compile or exec time.
2506    
2507     (c) The handling of \ followed by a digit other than 0 is now supposed to be
2508     the same as Perl's. If the decimal number it represents is less than 10
2509     or there aren't that many previous left capturing parentheses, an octal
2510     escape is read. Inside a character class, it's always an octal escape,
2511     even if it is a single digit.
2512    
2513     (d) An escaped but undefined alphabetic character is taken as a literal,
2514     unless PCRE_EXTRA is set. Currently this just reserves the remaining
2515     escapes.
2516    
2517     (e) {0} is now permitted. (The previous item is removed from the compiled
2518     pattern).
2519    
2520     5. Changed all the names of code files so that the basic parts are no longer
2521     than 10 characters, and abolished the teeny "globals.c" file.
2522    
2523     6. Changed the handling of character classes; they are now done with a 32-byte
2524     bit map always.
2525    
2526     7. Added the -d and /D options to pcretest to make it possible to look at the
2527     internals of compilation without having to recompile pcre.
2528    
2529    
2530     Version 0.95 23-Sep-97
2531     ----------------------
2532    
2533     1. Fixed bug in pre-pass concerning escaped "normal" characters such as \x5c or
2534     \x20 at the start of a run of normal characters. These were being treated as
2535     real characters, instead of the source characters being re-checked.
2536    
2537    
2538     Version 0.94 18-Sep-97
2539     ----------------------
2540    
2541     1. The functions are now thread-safe, with the caveat that the global variables
2542     containing pointers to malloc() and free() or alternative functions are the
2543     same for all threads.
2544    
2545     2. Get pcre_study() to generate a bitmap of initial characters for non-
2546     anchored patterns when this is possible, and use it if passed to pcre_exec().
2547    
2548    
2549     Version 0.93 15-Sep-97
2550     ----------------------
2551    
2552     1. /(b)|(:+)/ was computing an incorrect first character.
2553    
2554     2. Add pcre_study() to the API and the passing of pcre_extra to pcre_exec(),
2555     but not actually doing anything yet.
2556    
2557     3. Treat "-" characters in classes that cannot be part of ranges as literals,
2558     as Perl does (e.g. [-az] or [az-]).
2559    
2560     4. Set the anchored flag if a branch starts with .* or .*? because that tests
2561     all possible positions.
2562    
2563     5. Split up into different modules to avoid including unneeded functions in a
2564     compiled binary. However, compile and exec are still in one module. The "study"
2565     function is split off.
2566    
2567     6. The character tables are now in a separate module whose source is generated
2568     by an auxiliary program - but can then be edited by hand if required. There are
2569     now no calls to isalnum(), isspace(), isdigit(), isxdigit(), tolower() or
2570     toupper() in the code.
2571    
2572     7. Turn the malloc/free funtions variables into pcre_malloc and pcre_free and
2573     make them global. Abolish the function for setting them, as the caller can now
2574     set them directly.
2575    
2576    
2577     Version 0.92 11-Sep-97
2578     ----------------------
2579    
2580     1. A repeat with a fixed maximum and a minimum of 1 for an ordinary character
2581     (e.g. /a{1,3}/) was broken (I mis-optimized it).
2582    
2583     2. Caseless matching was not working in character classes if the characters in
2584     the pattern were in upper case.
2585    
2586     3. Make ranges like [W-c] work in the same way as Perl for caseless matching.
2587    
2588     4. Make PCRE_ANCHORED public and accept as a compile option.
2589    
2590     5. Add an options word to pcre_exec() and accept PCRE_ANCHORED and
2591     PCRE_CASELESS at run time. Add escapes \A and \I to pcretest to cause it to
2592     pass them.
2593    
2594     6. Give an error if bad option bits passed at compile or run time.
2595    
2596     7. Add PCRE_MULTILINE at compile and exec time, and (?m) as well. Add \M to
2597     pcretest to cause it to pass that flag.
2598    
2599     8. Add pcre_info(), to get the number of identifying subpatterns, the stored
2600     options, and the first character, if set.
2601    
2602     9. Recognize C+ or C{n,m} where n >= 1 as providing a fixed starting character.
2603    
2604    
2605     Version 0.91 10-Sep-97
2606     ----------------------
2607    
2608     1. PCRE was failing to diagnose unlimited repeats of subpatterns that could
2609     match the empty string as in /(a*)*/. It was looping and ultimately crashing.
2610    
2611     2. PCRE was looping on encountering an indefinitely repeated back reference to
2612     a subpattern that had matched an empty string, e.g. /(a|)\1*/. It now does what
2613     Perl does - treats the match as successful.
2614    
2615     ****

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