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

Contents of /code/trunk/ChangeLog

Parent Directory Parent Directory | Revision Log Revision Log


Revision 108 - (hide annotations) (download)
Wed Mar 7 11:11:23 2007 UTC (7 years, 7 months ago) by ph10
File size: 121994 byte(s)
Arguments for the emulated memmove() should be void *, not char *.

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

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