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

Contents of /code/trunk/ChangeLog

Parent Directory Parent Directory | Revision Log Revision Log


Revision 118 - (hide annotations) (download)
Fri Mar 9 16:17:46 2007 UTC (7 years, 5 months ago) by ph10
File size: 125025 byte(s)
Fix bug in emulated memmove() for downward moves.

1 nigel 3 ChangeLog for PCRE
2     ------------------
3    
4 ph10 98 Version 7.1 05-Mar-07
5     ---------------------
6    
7 ph10 111 1. Applied Bob Rossi and Daniel G's patches to convert the build system to one
8     that is more "standard", making use of automake and other autotools. There
9 ph10 99 is some re-arrangement of the files and adjustment of comments consequent
10     on this.
11 ph10 111
12     2. Part of the patch fixed a problem with the pcregrep tests. The test of -r
13     for recursive directory scanning broke on some systems because the files
14     are not scanned in any specific order and on different systems the order
15     was different. A call to "sort" has been inserted into RunGrepTest for the
16     approprate test as a short-term fix. In the longer term there may be an
17 ph10 100 alternative.
18 ph10 111
19 ph10 100 3. I had an email from Eric Raymond about problems translating some of PCRE's
20 ph10 111 man pages to HTML (despite the fact that I distribute HTML pages, some
21     people do their own conversions for various reasons). The problems
22     concerned the use of low-level troff macros .br and .in. I have therefore
23     removed all such uses from the man pages (some were redundant, some could
24 ph10 113 be replaced by .nf/.fi pairs). The 132html script that I use to generate
25     HTML has been updated to handle .nf/.fi and to complain if it encounters
26     .br or .in.
27 ph10 111
28 ph10 100 4. Updated comments in configure.ac that get placed in config.h.in and also
29 ph10 111 arranged for config.h to be included in the distribution, with the name
30     config.h.generic, for the benefit of those who have to compile without
31     Autotools (compare pcre.h, which is now distributed as pcre.h.generic).
32    
33     5. Updated the support (such as it is) for Virtual Pascal, thanks to Stefan
34     Weber: (1) pcre_internal.h was missing some function renames; (2) updated
35     makevp.bat for the current PCRE, using the additional files !compile.txt,
36 ph10 101 !linklib.txt, and pcregexp.pas.
37 ph10 111
38     6. A Windows user reported a minor discrepancy with test 2, which turned out
39     to be caused by a trailing space on an input line that had got lost in his
40 ph10 102 copy. The trailing space was an accident, so I've just removed it.
41 ph10 111
42 ph10 104 7. Add -Wl,-R... flags in pcre-config.in for *BSD* systems, as I'm told
43 ph10 111 that is needed.
44    
45 ph10 105 8. Mark ucp_table (in ucptable.h) and ucp_gentype (in pcre_ucp_searchfuncs.c)
46     as "const" (a) because they are and (b) because it helps the PHP
47     maintainers who have recently made a script to detect big data structures
48 ph10 111 in the php code that should be moved to the .rodata section. I remembered
49     to update Builducptable as well, so it won't revert if ucptable.h is ever
50 ph10 105 re-created.
51 ph10 111
52     9. Added some extra #ifdef SUPPORT_UTF8 conditionals into pcretest.c,
53 ph10 107 pcre_printint.src, pcre_compile.c, pcre_study.c, and pcre_tables.c, in
54 ph10 111 order to be able to cut out the UTF-8 tables in the latter when UTF-8
55     support is not required. This saves 1.5-2K of code, which is important in
56 ph10 107 some applications.
57 ph10 111
58 ph10 110 Later: more #ifdefs are needed in pcre_ord2utf8.c and pcre_valid_utf8.c
59 ph10 111 so as not to refer to the tables, even though these functions will never be
60     called when UTF-8 support is disabled. Otherwise there are problems with a
61     shared library.
62    
63 ph10 118 10. Fixed two bugs in the emulated memmove() function in pcre_internal.h:
64    
65     (a) It was defining its arguments as char * instead of void *.
66    
67     (b) It was assuming that all moves were upwards in memory; this was true
68     a long time ago when I wrote it, but is no longer the case.
69    
70     The emulated memove() is provided for those environments that have neither
71     memmove() nor bcopy(). I didn't think anyone used it these days, but that
72     is clearly not the case, as these two bugs were recently reported.
73 ph10 108
74 ph10 111 11. The script PrepareRelease is now distributed: it calls 132html, CleanTxt,
75     and Detrail to create the HTML documentation, the .txt form of the man
76     pages, and it removes trailing spaces from listed files. It also creates
77     pcre.h.generic and config.h.generic from pcre.h and config.h. In the latter
78     case, it wraps all the #defines with #ifndefs. This script should be run
79     before "make dist".
80 ph10 115
81     12. Fixed two fairly obscure bugs concerned with quantified caseless matching
82     with Unicode property support.
83    
84     (a) For a maximizing quantifier, if the two different cases of the
85     character were of different lengths in their UTF-8 codings (there are
86     some cases like this - I found 11), and the matching function had to
87     back up over a mixture of the two cases, it incorrectly assumed they
88     were both the same length.
89    
90     (b) When PCRE was configured to use the heap rather than the stack for
91     recursion during matching, it was not correctly preserving the data for
92     the other case of a UTF-8 character when checking ahead for a match
93     while processing a minimizing repeat. If the check also involved
94     matching a wide character, but failed, corruption could cause an
95     erroneous result when trying to check for a repeat of the original
96     character.
97 ph10 116
98     13. Some tidying changes to the testing mechanism:
99 ph10 98
100 ph10 116 (a) The RunTest script now detects the internal link size and whether there
101     is UTF-8 and UCP support by running ./pcretest -C instead of relying on
102     values substituted by "configure". (The RunGrepTest script already did
103     this for UTF-8.) The configure.ac script no longer substitutes the
104     relevant variables.
105    
106     (b) The debugging options /B and /D in pcretest show the compiled bytecode
107     with length and offset values. This means that the output is different
108     for different internal link sizes. Test 2 is skipped for link sizes
109     other than 2 because of this, bypassing the problem. Unfortunately,
110     there was also a test in test 3 (the locale tests) that used /B and
111     failed for link sizes other than 2. Rather than cut the whole test out,
112     I have added a new /Z option to pcretest that replaces the length and
113     offset values with spaces. This is now used to make test 3 independent
114     of link size.
115 ph10 118
116 ph10 111
117 nigel 93 Version 7.0 19-Dec-06
118     ---------------------
119    
120     1. Fixed a signed/unsigned compiler warning in pcre_compile.c, shown up by
121     moving to gcc 4.1.1.
122    
123     2. The -S option for pcretest uses setrlimit(); I had omitted to #include
124     sys/time.h, which is documented as needed for this function. It doesn't
125     seem to matter on Linux, but it showed up on some releases of OS X.
126    
127     3. It seems that there are systems where bytes whose values are greater than
128     127 match isprint() in the "C" locale. The "C" locale should be the
129     default when a C program starts up. In most systems, only ASCII printing
130     characters match isprint(). This difference caused the output from pcretest
131     to vary, making some of the tests fail. I have changed pcretest so that:
132    
133     (a) When it is outputting text in the compiled version of a pattern, bytes
134     other than 32-126 are always shown as hex escapes.
135    
136     (b) When it is outputting text that is a matched part of a subject string,
137     it does the same, unless a different locale has been set for the match
138     (using the /L modifier). In this case, it uses isprint() to decide.
139    
140     4. Fixed a major bug that caused incorrect computation of the amount of memory
141     required for a compiled pattern when options that changed within the
142     pattern affected the logic of the preliminary scan that determines the
143     length. The relevant options are -x, and -i in UTF-8 mode. The result was
144     that the computed length was too small. The symptoms of this bug were
145     either the PCRE error "internal error: code overflow" from pcre_compile(),
146     or a glibc crash with a message such as "pcretest: free(): invalid next
147     size (fast)". Examples of patterns that provoked this bug (shown in
148     pcretest format) are:
149    
150     /(?-x: )/x
151     /(?x)(?-x: \s*#\s*)/
152     /((?i)[\x{c0}])/8
153     /(?i:[\x{c0}])/8
154    
155     HOWEVER: Change 17 below makes this fix obsolete as the memory computation
156     is now done differently.
157    
158     5. Applied patches from Google to: (a) add a QuoteMeta function to the C++
159     wrapper classes; (b) implement a new function in the C++ scanner that is
160     more efficient than the old way of doing things because it avoids levels of
161     recursion in the regex matching; (c) add a paragraph to the documentation
162     for the FullMatch() function.
163    
164     6. The escape sequence \n was being treated as whatever was defined as
165     "newline". Not only was this contrary to the documentation, which states
166     that \n is character 10 (hex 0A), but it also went horribly wrong when
167     "newline" was defined as CRLF. This has been fixed.
168    
169     7. In pcre_dfa_exec.c the value of an unsigned integer (the variable called c)
170     was being set to -1 for the "end of line" case (supposedly a value that no
171     character can have). Though this value is never used (the check for end of
172     line is "zero bytes in current character"), it caused compiler complaints.
173     I've changed it to 0xffffffff.
174    
175     8. In pcre_version.c, the version string was being built by a sequence of
176     C macros that, in the event of PCRE_PRERELEASE being defined as an empty
177     string (as it is for production releases) called a macro with an empty
178     argument. The C standard says the result of this is undefined. The gcc
179     compiler treats it as an empty string (which was what was wanted) but it is
180     reported that Visual C gives an error. The source has been hacked around to
181     avoid this problem.
182    
183     9. On the advice of a Windows user, included <io.h> and <fcntl.h> in Windows
184     builds of pcretest, and changed the call to _setmode() to use _O_BINARY
185     instead of 0x8000. Made all the #ifdefs test both _WIN32 and WIN32 (not all
186     of them did).
187    
188     10. Originally, pcretest opened its input and output without "b"; then I was
189     told that "b" was needed in some environments, so it was added for release
190     5.0 to both the input and output. (It makes no difference on Unix-like
191     systems.) Later I was told that it is wrong for the input on Windows. I've
192     now abstracted the modes into two macros, to make it easier to fiddle with
193     them, and removed "b" from the input mode under Windows.
194    
195     11. Added pkgconfig support for the C++ wrapper library, libpcrecpp.
196    
197     12. Added -help and --help to pcretest as an official way of being reminded
198     of the options.
199    
200     13. Removed some redundant semicolons after macro calls in pcrecpparg.h.in
201     and pcrecpp.cc because they annoy compilers at high warning levels.
202    
203     14. A bit of tidying/refactoring in pcre_exec.c in the main bumpalong loop.
204    
205     15. Fixed an occurrence of == in configure.ac that should have been = (shell
206     scripts are not C programs :-) and which was not noticed because it works
207     on Linux.
208    
209     16. pcretest is supposed to handle any length of pattern and data line (as one
210     line or as a continued sequence of lines) by extending its input buffer if
211     necessary. This feature was broken for very long pattern lines, leading to
212     a string of junk being passed to pcre_compile() if the pattern was longer
213     than about 50K.
214    
215     17. I have done a major re-factoring of the way pcre_compile() computes the
216     amount of memory needed for a compiled pattern. Previously, there was code
217     that made a preliminary scan of the pattern in order to do this. That was
218     OK when PCRE was new, but as the facilities have expanded, it has become
219     harder and harder to keep it in step with the real compile phase, and there
220     have been a number of bugs (see for example, 4 above). I have now found a
221     cunning way of running the real compile function in a "fake" mode that
222     enables it to compute how much memory it would need, while actually only
223     ever using a few hundred bytes of working memory and without too many
224     tests of the mode. This should make future maintenance and development
225     easier. A side effect of this work is that the limit of 200 on the nesting
226     depth of parentheses has been removed (though this was never a serious
227     limitation, I suspect). However, there is a downside: pcre_compile() now
228     runs more slowly than before (30% or more, depending on the pattern). I
229     hope this isn't a big issue. There is no effect on runtime performance.
230    
231     18. Fixed a minor bug in pcretest: if a pattern line was not terminated by a
232     newline (only possible for the last line of a file) and it was a
233     pattern that set a locale (followed by /Lsomething), pcretest crashed.
234    
235     19. Added additional timing features to pcretest. (1) The -tm option now times
236     matching only, not compiling. (2) Both -t and -tm can be followed, as a
237     separate command line item, by a number that specifies the number of
238     repeats to use when timing. The default is 50000; this gives better
239     precision, but takes uncomfortably long for very large patterns.
240    
241     20. Extended pcre_study() to be more clever in cases where a branch of a
242     subpattern has no definite first character. For example, (a*|b*)[cd] would
243     previously give no result from pcre_study(). Now it recognizes that the
244     first character must be a, b, c, or d.
245    
246     21. There was an incorrect error "recursive call could loop indefinitely" if
247     a subpattern (or the entire pattern) that was being tested for matching an
248     empty string contained only one non-empty item after a nested subpattern.
249     For example, the pattern (?>\x{100}*)\d(?R) provoked this error
250     incorrectly, because the \d was being skipped in the check.
251    
252     22. The pcretest program now has a new pattern option /B and a command line
253     option -b, which is equivalent to adding /B to every pattern. This causes
254     it to show the compiled bytecode, without the additional information that
255     -d shows. The effect of -d is now the same as -b with -i (and similarly, /D
256     is the same as /B/I).
257    
258     23. A new optimization is now able automatically to treat some sequences such
259     as a*b as a*+b. More specifically, if something simple (such as a character
260     or a simple class like \d) has an unlimited quantifier, and is followed by
261     something that cannot possibly match the quantified thing, the quantifier
262     is automatically "possessified".
263    
264     24. A recursive reference to a subpattern whose number was greater than 39
265     went wrong under certain circumstances in UTF-8 mode. This bug could also
266     have affected the operation of pcre_study().
267    
268     25. Realized that a little bit of performance could be had by replacing
269     (c & 0xc0) == 0xc0 with c >= 0xc0 when processing UTF-8 characters.
270    
271     26. Timing data from pcretest is now shown to 4 decimal places instead of 3.
272    
273     27. Possessive quantifiers such as a++ were previously implemented by turning
274     them into atomic groups such as ($>a+). Now they have their own opcodes,
275     which improves performance. This includes the automatically created ones
276     from 23 above.
277    
278     28. A pattern such as (?=(\w+))\1: which simulates an atomic group using a
279     lookahead was broken if it was not anchored. PCRE was mistakenly expecting
280     the first matched character to be a colon. This applied both to named and
281     numbered groups.
282    
283     29. The ucpinternal.h header file was missing its idempotency #ifdef.
284    
285     30. I was sent a "project" file called libpcre.a.dev which I understand makes
286     building PCRE on Windows easier, so I have included it in the distribution.
287    
288     31. There is now a check in pcretest against a ridiculously large number being
289     returned by pcre_exec() or pcre_dfa_exec(). If this happens in a /g or /G
290     loop, the loop is abandoned.
291    
292     32. Forward references to subpatterns in conditions such as (?(2)...) where
293     subpattern 2 is defined later cause pcre_compile() to search forwards in
294     the pattern for the relevant set of parentheses. This search went wrong
295     when there were unescaped parentheses in a character class, parentheses
296     escaped with \Q...\E, or parentheses in a #-comment in /x mode.
297    
298     33. "Subroutine" calls and backreferences were previously restricted to
299     referencing subpatterns earlier in the regex. This restriction has now
300     been removed.
301    
302     34. Added a number of extra features that are going to be in Perl 5.10. On the
303     whole, these are just syntactic alternatives for features that PCRE had
304     previously implemented using the Python syntax or my own invention. The
305     other formats are all retained for compatibility.
306    
307     (a) Named groups can now be defined as (?<name>...) or (?'name'...) as well
308     as (?P<name>...). The new forms, as well as being in Perl 5.10, are
309     also .NET compatible.
310    
311     (b) A recursion or subroutine call to a named group can now be defined as
312     (?&name) as well as (?P>name).
313    
314     (c) A backreference to a named group can now be defined as \k<name> or
315     \k'name' as well as (?P=name). The new forms, as well as being in Perl
316     5.10, are also .NET compatible.
317    
318     (d) A conditional reference to a named group can now use the syntax
319     (?(<name>) or (?('name') as well as (?(name).
320    
321     (e) A "conditional group" of the form (?(DEFINE)...) can be used to define
322     groups (named and numbered) that are never evaluated inline, but can be
323     called as "subroutines" from elsewhere. In effect, the DEFINE condition
324     is always false. There may be only one alternative in such a group.
325    
326     (f) A test for recursion can be given as (?(R1).. or (?(R&name)... as well
327     as the simple (?(R). The condition is true only if the most recent
328     recursion is that of the given number or name. It does not search out
329     through the entire recursion stack.
330    
331     (g) The escape \gN or \g{N} has been added, where N is a positive or
332     negative number, specifying an absolute or relative reference.
333    
334     35. Tidied to get rid of some further signed/unsigned compiler warnings and
335     some "unreachable code" warnings.
336    
337     36. Updated the Unicode property tables to Unicode version 5.0.0. Amongst other
338     things, this adds five new scripts.
339    
340     37. Perl ignores orphaned \E escapes completely. PCRE now does the same.
341     There were also incompatibilities regarding the handling of \Q..\E inside
342     character classes, for example with patterns like [\Qa\E-\Qz\E] where the
343     hyphen was adjacent to \Q or \E. I hope I've cleared all this up now.
344    
345     38. Like Perl, PCRE detects when an indefinitely repeated parenthesized group
346     matches an empty string, and forcibly breaks the loop. There were bugs in
347     this code in non-simple cases. For a pattern such as ^(a()*)* matched
348     against aaaa the result was just "a" rather than "aaaa", for example. Two
349     separate and independent bugs (that affected different cases) have been
350     fixed.
351    
352     39. Refactored the code to abolish the use of different opcodes for small
353     capturing bracket numbers. This is a tidy that I avoided doing when I
354     removed the limit on the number of capturing brackets for 3.5 back in 2001.
355     The new approach is not only tidier, it makes it possible to reduce the
356     memory needed to fix the previous bug (38).
357    
358     40. Implemented PCRE_NEWLINE_ANY to recognize any of the Unicode newline
359     sequences (http://unicode.org/unicode/reports/tr18/) as "newline" when
360     processing dot, circumflex, or dollar metacharacters, or #-comments in /x
361     mode.
362    
363     41. Add \R to match any Unicode newline sequence, as suggested in the Unicode
364     report.
365    
366     42. Applied patch, originally from Ari Pollak, modified by Google, to allow
367     copy construction and assignment in the C++ wrapper.
368    
369     43. Updated pcregrep to support "--newline=any". In the process, I fixed a
370     couple of bugs that could have given wrong results in the "--newline=crlf"
371     case.
372    
373     44. Added a number of casts and did some reorganization of signed/unsigned int
374     variables following suggestions from Dair Grant. Also renamed the variable
375     "this" as "item" because it is a C++ keyword.
376    
377     45. Arranged for dftables to add
378    
379     #include "pcre_internal.h"
380    
381     to pcre_chartables.c because without it, gcc 4.x may remove the array
382     definition from the final binary if PCRE is built into a static library and
383     dead code stripping is activated.
384    
385     46. For an unanchored pattern, if a match attempt fails at the start of a
386     newline sequence, and the newline setting is CRLF or ANY, and the next two
387     characters are CRLF, advance by two characters instead of one.
388    
389    
390 nigel 91 Version 6.7 04-Jul-06
391     ---------------------
392    
393     1. In order to handle tests when input lines are enormously long, pcretest has
394     been re-factored so that it automatically extends its buffers when
395     necessary. The code is crude, but this _is_ just a test program. The
396     default size has been increased from 32K to 50K.
397    
398     2. The code in pcre_study() was using the value of the re argument before
399     testing it for NULL. (Of course, in any sensible call of the function, it
400     won't be NULL.)
401    
402     3. The memmove() emulation function in pcre_internal.h, which is used on
403     systems that lack both memmove() and bcopy() - that is, hardly ever -
404     was missing a "static" storage class specifier.
405    
406     4. When UTF-8 mode was not set, PCRE looped when compiling certain patterns
407     containing an extended class (one that cannot be represented by a bitmap
408     because it contains high-valued characters or Unicode property items, e.g.
409     [\pZ]). Almost always one would set UTF-8 mode when processing such a
410     pattern, but PCRE should not loop if you do not (it no longer does).
411     [Detail: two cases were found: (a) a repeated subpattern containing an
412     extended class; (b) a recursive reference to a subpattern that followed a
413     previous extended class. It wasn't skipping over the extended class
414     correctly when UTF-8 mode was not set.]
415    
416     5. A negated single-character class was not being recognized as fixed-length
417     in lookbehind assertions such as (?<=[^f]), leading to an incorrect
418     compile error "lookbehind assertion is not fixed length".
419    
420     6. The RunPerlTest auxiliary script was showing an unexpected difference
421     between PCRE and Perl for UTF-8 tests. It turns out that it is hard to
422     write a Perl script that can interpret lines of an input file either as
423     byte characters or as UTF-8, which is what "perltest" was being required to
424     do for the non-UTF-8 and UTF-8 tests, respectively. Essentially what you
425     can't do is switch easily at run time between having the "use utf8;" pragma
426     or not. In the end, I fudged it by using the RunPerlTest script to insert
427     "use utf8;" explicitly for the UTF-8 tests.
428    
429     7. In multiline (/m) mode, PCRE was matching ^ after a terminating newline at
430     the end of the subject string, contrary to the documentation and to what
431     Perl does. This was true of both matching functions. Now it matches only at
432     the start of the subject and immediately after *internal* newlines.
433    
434     8. A call of pcre_fullinfo() from pcretest to get the option bits was passing
435     a pointer to an int instead of a pointer to an unsigned long int. This
436     caused problems on 64-bit systems.
437    
438     9. Applied a patch from the folks at Google to pcrecpp.cc, to fix "another
439     instance of the 'standard' template library not being so standard".
440    
441     10. There was no check on the number of named subpatterns nor the maximum
442     length of a subpattern name. The product of these values is used to compute
443     the size of the memory block for a compiled pattern. By supplying a very
444     long subpattern name and a large number of named subpatterns, the size
445     computation could be caused to overflow. This is now prevented by limiting
446     the length of names to 32 characters, and the number of named subpatterns
447     to 10,000.
448    
449     11. Subpatterns that are repeated with specific counts have to be replicated in
450     the compiled pattern. The size of memory for this was computed from the
451     length of the subpattern and the repeat count. The latter is limited to
452     65535, but there was no limit on the former, meaning that integer overflow
453     could in principle occur. The compiled length of a repeated subpattern is
454     now limited to 30,000 bytes in order to prevent this.
455    
456     12. Added the optional facility to have named substrings with the same name.
457    
458     13. Added the ability to use a named substring as a condition, using the
459     Python syntax: (?(name)yes|no). This overloads (?(R)... and names that
460     are numbers (not recommended). Forward references are permitted.
461    
462     14. Added forward references in named backreferences (if you see what I mean).
463    
464     15. In UTF-8 mode, with the PCRE_DOTALL option set, a quantified dot in the
465     pattern could run off the end of the subject. For example, the pattern
466     "(?s)(.{1,5})"8 did this with the subject "ab".
467    
468     16. If PCRE_DOTALL or PCRE_MULTILINE were set, pcre_dfa_exec() behaved as if
469     PCRE_CASELESS was set when matching characters that were quantified with ?
470     or *.
471    
472     17. A character class other than a single negated character that had a minimum
473     but no maximum quantifier - for example [ab]{6,} - was not handled
474     correctly by pce_dfa_exec(). It would match only one character.
475    
476     18. A valid (though odd) pattern that looked like a POSIX character
477     class but used an invalid character after [ (for example [[,abc,]]) caused
478     pcre_compile() to give the error "Failed: internal error: code overflow" or
479     in some cases to crash with a glibc free() error. This could even happen if
480     the pattern terminated after [[ but there just happened to be a sequence of
481     letters, a binary zero, and a closing ] in the memory that followed.
482    
483     19. Perl's treatment of octal escapes in the range \400 to \777 has changed
484     over the years. Originally (before any Unicode support), just the bottom 8
485     bits were taken. Thus, for example, \500 really meant \100. Nowadays the
486     output from "man perlunicode" includes this:
487    
488     The regular expression compiler produces polymorphic opcodes. That
489     is, the pattern adapts to the data and automatically switches to
490     the Unicode character scheme when presented with Unicode data--or
491     instead uses a traditional byte scheme when presented with byte
492     data.
493    
494     Sadly, a wide octal escape does not cause a switch, and in a string with
495     no other multibyte characters, these octal escapes are treated as before.
496     Thus, in Perl, the pattern /\500/ actually matches \100 but the pattern
497     /\500|\x{1ff}/ matches \500 or \777 because the whole thing is treated as a
498     Unicode string.
499    
500     I have not perpetrated such confusion in PCRE. Up till now, it took just
501     the bottom 8 bits, as in old Perl. I have now made octal escapes with
502     values greater than \377 illegal in non-UTF-8 mode. In UTF-8 mode they
503     translate to the appropriate multibyte character.
504    
505     29. Applied some refactoring to reduce the number of warnings from Microsoft
506     and Borland compilers. This has included removing the fudge introduced
507     seven years ago for the OS/2 compiler (see 2.02/2 below) because it caused
508     a warning about an unused variable.
509    
510     21. PCRE has not included VT (character 0x0b) in the set of whitespace
511     characters since release 4.0, because Perl (from release 5.004) does not.
512     [Or at least, is documented not to: some releases seem to be in conflict
513     with the documentation.] However, when a pattern was studied with
514     pcre_study() and all its branches started with \s, PCRE still included VT
515     as a possible starting character. Of course, this did no harm; it just
516     caused an unnecessary match attempt.
517    
518     22. Removed a now-redundant internal flag bit that recorded the fact that case
519     dependency changed within the pattern. This was once needed for "required
520     byte" processing, but is no longer used. This recovers a now-scarce options
521     bit. Also moved the least significant internal flag bit to the most-
522     significant bit of the word, which was not previously used (hangover from
523     the days when it was an int rather than a uint) to free up another bit for
524     the future.
525    
526     23. Added support for CRLF line endings as well as CR and LF. As well as the
527     default being selectable at build time, it can now be changed at runtime
528     via the PCRE_NEWLINE_xxx flags. There are now options for pcregrep to
529     specify that it is scanning data with non-default line endings.
530    
531     24. Changed the definition of CXXLINK to make it agree with the definition of
532     LINK in the Makefile, by replacing LDFLAGS to CXXFLAGS.
533    
534     25. Applied Ian Taylor's patches to avoid using another stack frame for tail
535     recursions. This makes a big different to stack usage for some patterns.
536    
537     26. If a subpattern containing a named recursion or subroutine reference such
538     as (?P>B) was quantified, for example (xxx(?P>B)){3}, the calculation of
539     the space required for the compiled pattern went wrong and gave too small a
540     value. Depending on the environment, this could lead to "Failed: internal
541     error: code overflow at offset 49" or "glibc detected double free or
542     corruption" errors.
543    
544     27. Applied patches from Google (a) to support the new newline modes and (b) to
545     advance over multibyte UTF-8 characters in GlobalReplace.
546    
547     28. Change free() to pcre_free() in pcredemo.c. Apparently this makes a
548     difference for some implementation of PCRE in some Windows version.
549    
550     29. Added some extra testing facilities to pcretest:
551    
552     \q<number> in a data line sets the "match limit" value
553     \Q<number> in a data line sets the "match recursion limt" value
554     -S <number> sets the stack size, where <number> is in megabytes
555    
556     The -S option isn't available for Windows.
557    
558    
559 nigel 89 Version 6.6 06-Feb-06
560     ---------------------
561    
562     1. Change 16(a) for 6.5 broke things, because PCRE_DATA_SCOPE was not defined
563     in pcreposix.h. I have copied the definition from pcre.h.
564    
565     2. Change 25 for 6.5 broke compilation in a build directory out-of-tree
566     because pcre.h is no longer a built file.
567    
568     3. Added Jeff Friedl's additional debugging patches to pcregrep. These are
569     not normally included in the compiled code.
570    
571    
572 nigel 87 Version 6.5 01-Feb-06
573     ---------------------
574    
575     1. When using the partial match feature with pcre_dfa_exec(), it was not
576     anchoring the second and subsequent partial matches at the new starting
577     point. This could lead to incorrect results. For example, with the pattern
578     /1234/, partially matching against "123" and then "a4" gave a match.
579    
580     2. Changes to pcregrep:
581    
582     (a) All non-match returns from pcre_exec() were being treated as failures
583     to match the line. Now, unless the error is PCRE_ERROR_NOMATCH, an
584     error message is output. Some extra information is given for the
585     PCRE_ERROR_MATCHLIMIT and PCRE_ERROR_RECURSIONLIMIT errors, which are
586     probably the only errors that are likely to be caused by users (by
587     specifying a regex that has nested indefinite repeats, for instance).
588     If there are more than 20 of these errors, pcregrep is abandoned.
589    
590     (b) A binary zero was treated as data while matching, but terminated the
591     output line if it was written out. This has been fixed: binary zeroes
592     are now no different to any other data bytes.
593    
594     (c) Whichever of the LC_ALL or LC_CTYPE environment variables is set is
595     used to set a locale for matching. The --locale=xxxx long option has
596     been added (no short equivalent) to specify a locale explicitly on the
597     pcregrep command, overriding the environment variables.
598    
599     (d) When -B was used with -n, some line numbers in the output were one less
600     than they should have been.
601    
602     (e) Added the -o (--only-matching) option.
603    
604     (f) If -A or -C was used with -c (count only), some lines of context were
605     accidentally printed for the final match.
606    
607     (g) Added the -H (--with-filename) option.
608    
609     (h) The combination of options -rh failed to suppress file names for files
610     that were found from directory arguments.
611    
612     (i) Added the -D (--devices) and -d (--directories) options.
613    
614     (j) Added the -F (--fixed-strings) option.
615    
616     (k) Allow "-" to be used as a file name for -f as well as for a data file.
617    
618     (l) Added the --colo(u)r option.
619    
620     (m) Added Jeffrey Friedl's -S testing option, but within #ifdefs so that it
621     is not present by default.
622    
623     3. A nasty bug was discovered in the handling of recursive patterns, that is,
624     items such as (?R) or (?1), when the recursion could match a number of
625     alternatives. If it matched one of the alternatives, but subsequently,
626     outside the recursion, there was a failure, the code tried to back up into
627     the recursion. However, because of the way PCRE is implemented, this is not
628     possible, and the result was an incorrect result from the match.
629    
630     In order to prevent this happening, the specification of recursion has
631     been changed so that all such subpatterns are automatically treated as
632     atomic groups. Thus, for example, (?R) is treated as if it were (?>(?R)).
633    
634     4. I had overlooked the fact that, in some locales, there are characters for
635     which isalpha() is true but neither isupper() nor islower() are true. In
636     the fr_FR locale, for instance, the \xAA and \xBA characters (ordmasculine
637     and ordfeminine) are like this. This affected the treatment of \w and \W
638     when they appeared in character classes, but not when they appeared outside
639     a character class. The bit map for "word" characters is now created
640     separately from the results of isalnum() instead of just taking it from the
641     upper, lower, and digit maps. (Plus the underscore character, of course.)
642    
643     5. The above bug also affected the handling of POSIX character classes such as
644     [[:alpha:]] and [[:alnum:]]. These do not have their own bit maps in PCRE's
645     permanent tables. Instead, the bit maps for such a class were previously
646     created as the appropriate unions of the upper, lower, and digit bitmaps.
647     Now they are created by subtraction from the [[:word:]] class, which has
648     its own bitmap.
649    
650     6. The [[:blank:]] character class matches horizontal, but not vertical space.
651     It is created by subtracting the vertical space characters (\x09, \x0a,
652     \x0b, \x0c) from the [[:space:]] bitmap. Previously, however, the
653     subtraction was done in the overall bitmap for a character class, meaning
654     that a class such as [\x0c[:blank:]] was incorrect because \x0c would not
655     be recognized. This bug has been fixed.
656    
657     7. Patches from the folks at Google:
658    
659     (a) pcrecpp.cc: "to handle a corner case that may or may not happen in
660     real life, but is still worth protecting against".
661    
662     (b) pcrecpp.cc: "corrects a bug when negative radixes are used with
663     regular expressions".
664    
665     (c) pcre_scanner.cc: avoid use of std::count() because not all systems
666     have it.
667    
668     (d) Split off pcrecpparg.h from pcrecpp.h and had the former built by
669     "configure" and the latter not, in order to fix a problem somebody had
670     with compiling the Arg class on HP-UX.
671    
672     (e) Improve the error-handling of the C++ wrapper a little bit.
673    
674     (f) New tests for checking recursion limiting.
675    
676     8. The pcre_memmove() function, which is used only if the environment does not
677     have a standard memmove() function (and is therefore rarely compiled),
678     contained two bugs: (a) use of int instead of size_t, and (b) it was not
679     returning a result (though PCRE never actually uses the result).
680    
681     9. In the POSIX regexec() interface, if nmatch is specified as a ridiculously
682     large number - greater than INT_MAX/(3*sizeof(int)) - REG_ESPACE is
683     returned instead of calling malloc() with an overflowing number that would
684     most likely cause subsequent chaos.
685    
686     10. The debugging option of pcretest was not showing the NO_AUTO_CAPTURE flag.
687    
688     11. The POSIX flag REG_NOSUB is now supported. When a pattern that was compiled
689     with this option is matched, the nmatch and pmatch options of regexec() are
690     ignored.
691    
692     12. Added REG_UTF8 to the POSIX interface. This is not defined by POSIX, but is
693     provided in case anyone wants to the the POSIX interface with UTF-8
694     strings.
695    
696     13. Added CXXLDFLAGS to the Makefile parameters to provide settings only on the
697     C++ linking (needed for some HP-UX environments).
698    
699     14. Avoid compiler warnings in get_ucpname() when compiled without UCP support
700     (unused parameter) and in the pcre_printint() function (omitted "default"
701     switch label when the default is to do nothing).
702    
703     15. Added some code to make it possible, when PCRE is compiled as a C++
704     library, to replace subject pointers for pcre_exec() with a smart pointer
705     class, thus making it possible to process discontinuous strings.
706    
707     16. The two macros PCRE_EXPORT and PCRE_DATA_SCOPE are confusing, and perform
708     much the same function. They were added by different people who were trying
709     to make PCRE easy to compile on non-Unix systems. It has been suggested
710     that PCRE_EXPORT be abolished now that there is more automatic apparatus
711     for compiling on Windows systems. I have therefore replaced it with
712     PCRE_DATA_SCOPE. This is set automatically for Windows; if not set it
713     defaults to "extern" for C or "extern C" for C++, which works fine on
714     Unix-like systems. It is now possible to override the value of PCRE_DATA_
715     SCOPE with something explicit in config.h. In addition:
716    
717     (a) pcreposix.h still had just "extern" instead of either of these macros;
718     I have replaced it with PCRE_DATA_SCOPE.
719    
720     (b) Functions such as _pcre_xclass(), which are internal to the library,
721     but external in the C sense, all had PCRE_EXPORT in their definitions.
722     This is apparently wrong for the Windows case, so I have removed it.
723     (It makes no difference on Unix-like systems.)
724    
725     17. Added a new limit, MATCH_LIMIT_RECURSION, which limits the depth of nesting
726     of recursive calls to match(). This is different to MATCH_LIMIT because
727     that limits the total number of calls to match(), not all of which increase
728     the depth of recursion. Limiting the recursion depth limits the amount of
729     stack (or heap if NO_RECURSE is set) that is used. The default can be set
730     when PCRE is compiled, and changed at run time. A patch from Google adds
731     this functionality to the C++ interface.
732    
733     18. Changes to the handling of Unicode character properties:
734    
735     (a) Updated the table to Unicode 4.1.0.
736    
737     (b) Recognize characters that are not in the table as "Cn" (undefined).
738    
739     (c) I revised the way the table is implemented to a much improved format
740     which includes recognition of ranges. It now supports the ranges that
741     are defined in UnicodeData.txt, and it also amalgamates other
742     characters into ranges. This has reduced the number of entries in the
743     table from around 16,000 to around 3,000, thus reducing its size
744     considerably. I realized I did not need to use a tree structure after
745     all - a binary chop search is just as efficient. Having reduced the
746     number of entries, I extended their size from 6 bytes to 8 bytes to
747     allow for more data.
748    
749     (d) Added support for Unicode script names via properties such as \p{Han}.
750    
751     19. In UTF-8 mode, a backslash followed by a non-Ascii character was not
752     matching that character.
753    
754     20. When matching a repeated Unicode property with a minimum greater than zero,
755     (for example \pL{2,}), PCRE could look past the end of the subject if it
756     reached it while seeking the minimum number of characters. This could
757     happen only if some of the characters were more than one byte long, because
758     there is a check for at least the minimum number of bytes.
759    
760     21. Refactored the implementation of \p and \P so as to be more general, to
761     allow for more different types of property in future. This has changed the
762     compiled form incompatibly. Anybody with saved compiled patterns that use
763     \p or \P will have to recompile them.
764    
765     22. Added "Any" and "L&" to the supported property types.
766    
767     23. Recognize \x{...} as a code point specifier, even when not in UTF-8 mode,
768     but give a compile time error if the value is greater than 0xff.
769    
770     24. The man pages for pcrepartial, pcreprecompile, and pcre_compile2 were
771     accidentally not being installed or uninstalled.
772    
773     25. The pcre.h file was built from pcre.h.in, but the only changes that were
774     made were to insert the current release number. This seemed silly, because
775     it made things harder for people building PCRE on systems that don't run
776     "configure". I have turned pcre.h into a distributed file, no longer built
777     by "configure", with the version identification directly included. There is
778     no longer a pcre.h.in file.
779    
780     However, this change necessitated a change to the pcre-config script as
781     well. It is built from pcre-config.in, and one of the substitutions was the
782     release number. I have updated configure.ac so that ./configure now finds
783     the release number by grepping pcre.h.
784    
785     26. Added the ability to run the tests under valgrind.
786    
787    
788 nigel 85 Version 6.4 05-Sep-05
789     ---------------------
790    
791     1. Change 6.0/10/(l) to pcregrep introduced a bug that caused separator lines
792     "--" to be printed when multiple files were scanned, even when none of the
793     -A, -B, or -C options were used. This is not compatible with Gnu grep, so I
794     consider it to be a bug, and have restored the previous behaviour.
795    
796     2. A couple of code tidies to get rid of compiler warnings.
797    
798     3. The pcretest program used to cheat by referring to symbols in the library
799     whose names begin with _pcre_. These are internal symbols that are not
800     really supposed to be visible externally, and in some environments it is
801     possible to suppress them. The cheating is now confined to including
802     certain files from the library's source, which is a bit cleaner.
803    
804     4. Renamed pcre.in as pcre.h.in to go with pcrecpp.h.in; it also makes the
805     file's purpose clearer.
806    
807     5. Reorganized pcre_ucp_findchar().
808    
809    
810 nigel 83 Version 6.3 15-Aug-05
811     ---------------------
812    
813     1. The file libpcre.pc.in did not have general read permission in the tarball.
814    
815     2. There were some problems when building without C++ support:
816    
817     (a) If C++ support was not built, "make install" and "make test" still
818     tried to test it.
819    
820     (b) There were problems when the value of CXX was explicitly set. Some
821     changes have been made to try to fix these, and ...
822    
823     (c) --disable-cpp can now be used to explicitly disable C++ support.
824    
825     (d) The use of @CPP_OBJ@ directly caused a blank line preceded by a
826     backslash in a target when C++ was disabled. This confuses some
827     versions of "make", apparently. Using an intermediate variable solves
828     this. (Same for CPP_LOBJ.)
829    
830     3. $(LINK_FOR_BUILD) now includes $(CFLAGS_FOR_BUILD) and $(LINK)
831     (non-Windows) now includes $(CFLAGS) because these flags are sometimes
832     necessary on certain architectures.
833    
834     4. Added a setting of -export-symbols-regex to the link command to remove
835     those symbols that are exported in the C sense, but actually are local
836     within the library, and not documented. Their names all begin with
837     "_pcre_". This is not a perfect job, because (a) we have to except some
838     symbols that pcretest ("illegally") uses, and (b) the facility isn't always
839     available (and never for static libraries). I have made a note to try to
840     find a way round (a) in the future.
841    
842    
843 nigel 81 Version 6.2 01-Aug-05
844     ---------------------
845    
846     1. There was no test for integer overflow of quantifier values. A construction
847     such as {1111111111111111} would give undefined results. What is worse, if
848     a minimum quantifier for a parenthesized subpattern overflowed and became
849     negative, the calculation of the memory size went wrong. This could have
850     led to memory overwriting.
851    
852     2. Building PCRE using VPATH was broken. Hopefully it is now fixed.
853    
854     3. Added "b" to the 2nd argument of fopen() in dftables.c, for non-Unix-like
855     operating environments where this matters.
856    
857     4. Applied Giuseppe Maxia's patch to add additional features for controlling
858     PCRE options from within the C++ wrapper.
859    
860     5. Named capturing subpatterns were not being correctly counted when a pattern
861     was compiled. This caused two problems: (a) If there were more than 100
862     such subpatterns, the calculation of the memory needed for the whole
863     compiled pattern went wrong, leading to an overflow error. (b) Numerical
864     back references of the form \12, where the number was greater than 9, were
865     not recognized as back references, even though there were sufficient
866     previous subpatterns.
867    
868     6. Two minor patches to pcrecpp.cc in order to allow it to compile on older
869     versions of gcc, e.g. 2.95.4.
870    
871    
872 nigel 79 Version 6.1 21-Jun-05
873     ---------------------
874    
875     1. There was one reference to the variable "posix" in pcretest.c that was not
876     surrounded by "#if !defined NOPOSIX".
877    
878     2. Make it possible to compile pcretest without DFA support, UTF8 support, or
879     the cross-check on the old pcre_info() function, for the benefit of the
880     cut-down version of PCRE that is currently imported into Exim.
881    
882     3. A (silly) pattern starting with (?i)(?-i) caused an internal space
883     allocation error. I've done the easy fix, which wastes 2 bytes for sensible
884     patterns that start (?i) but I don't think that matters. The use of (?i) is
885     just an example; this all applies to the other options as well.
886    
887     4. Since libtool seems to echo the compile commands it is issuing, the output
888     from "make" can be reduced a bit by putting "@" in front of each libtool
889     compile command.
890    
891     5. Patch from the folks at Google for configure.in to be a bit more thorough
892     in checking for a suitable C++ installation before trying to compile the
893     C++ stuff. This should fix a reported problem when a compiler was present,
894     but no suitable headers.
895    
896     6. The man pages all had just "PCRE" as their title. I have changed them to
897     be the relevant file name. I have also arranged that these names are
898     retained in the file doc/pcre.txt, which is a concatenation in text format
899     of all the man pages except the little individual ones for each function.
900    
901     7. The NON-UNIX-USE file had not been updated for the different set of source
902     files that come with release 6. I also added a few comments about the C++
903     wrapper.
904    
905    
906 nigel 77 Version 6.0 07-Jun-05
907     ---------------------
908    
909     1. Some minor internal re-organization to help with my DFA experiments.
910    
911     2. Some missing #ifdef SUPPORT_UCP conditionals in pcretest and printint that
912     didn't matter for the library itself when fully configured, but did matter
913     when compiling without UCP support, or within Exim, where the ucp files are
914     not imported.
915    
916     3. Refactoring of the library code to split up the various functions into
917     different source modules. The addition of the new DFA matching code (see
918     below) to a single monolithic source would have made it really too
919     unwieldy, quite apart from causing all the code to be include in a
920     statically linked application, when only some functions are used. This is
921     relevant even without the DFA addition now that patterns can be compiled in
922     one application and matched in another.
923    
924     The downside of splitting up is that there have to be some external
925     functions and data tables that are used internally in different modules of
926     the library but which are not part of the API. These have all had their
927     names changed to start with "_pcre_" so that they are unlikely to clash
928     with other external names.
929    
930     4. Added an alternate matching function, pcre_dfa_exec(), which matches using
931     a different (DFA) algorithm. Although it is slower than the original
932     function, it does have some advantages for certain types of matching
933     problem.
934    
935     5. Upgrades to pcretest in order to test the features of pcre_dfa_exec(),
936     including restarting after a partial match.
937    
938     6. A patch for pcregrep that defines INVALID_FILE_ATTRIBUTES if it is not
939     defined when compiling for Windows was sent to me. I have put it into the
940     code, though I have no means of testing or verifying it.
941    
942     7. Added the pcre_refcount() auxiliary function.
943    
944     8. Added the PCRE_FIRSTLINE option. This constrains an unanchored pattern to
945     match before or at the first newline in the subject string. In pcretest,
946     the /f option on a pattern can be used to set this.
947    
948     9. A repeated \w when used in UTF-8 mode with characters greater than 256
949     would behave wrongly. This has been present in PCRE since release 4.0.
950    
951     10. A number of changes to the pcregrep command:
952    
953     (a) Refactored how -x works; insert ^(...)$ instead of setting
954     PCRE_ANCHORED and checking the length, in preparation for adding
955     something similar for -w.
956    
957     (b) Added the -w (match as a word) option.
958    
959     (c) Refactored the way lines are read and buffered so as to have more
960     than one at a time available.
961    
962     (d) Implemented a pcregrep test script.
963    
964     (e) Added the -M (multiline match) option. This allows patterns to match
965     over several lines of the subject. The buffering ensures that at least
966     8K, or the rest of the document (whichever is the shorter) is available
967     for matching (and similarly the previous 8K for lookbehind assertions).
968    
969     (f) Changed the --help output so that it now says
970    
971     -w, --word-regex(p)
972    
973     instead of two lines, one with "regex" and the other with "regexp"
974     because that confused at least one person since the short forms are the
975     same. (This required a bit of code, as the output is generated
976     automatically from a table. It wasn't just a text change.)
977    
978     (g) -- can be used to terminate pcregrep options if the next thing isn't an
979     option but starts with a hyphen. Could be a pattern or a path name
980     starting with a hyphen, for instance.
981    
982     (h) "-" can be given as a file name to represent stdin.
983    
984     (i) When file names are being printed, "(standard input)" is used for
985     the standard input, for compatibility with GNU grep. Previously
986     "<stdin>" was used.
987    
988     (j) The option --label=xxx can be used to supply a name to be used for
989     stdin when file names are being printed. There is no short form.
990    
991     (k) Re-factored the options decoding logic because we are going to add
992     two more options that take data. Such options can now be given in four
993     different ways, e.g. "-fname", "-f name", "--file=name", "--file name".
994    
995     (l) Added the -A, -B, and -C options for requesting that lines of context
996     around matches be printed.
997    
998     (m) Added the -L option to print the names of files that do not contain
999     any matching lines, that is, the complement of -l.
1000    
1001     (n) The return code is 2 if any file cannot be opened, but pcregrep does
1002     continue to scan other files.
1003    
1004     (o) The -s option was incorrectly implemented. For compatibility with other
1005     greps, it now suppresses the error message for a non-existent or non-
1006     accessible file (but not the return code). There is a new option called
1007     -q that suppresses the output of matching lines, which was what -s was
1008     previously doing.
1009    
1010     (p) Added --include and --exclude options to specify files for inclusion
1011     and exclusion when recursing.
1012    
1013     11. The Makefile was not using the Autoconf-supported LDFLAGS macro properly.
1014     Hopefully, it now does.
1015    
1016     12. Missing cast in pcre_study().
1017    
1018     13. Added an "uninstall" target to the makefile.
1019    
1020     14. Replaced "extern" in the function prototypes in Makefile.in with
1021     "PCRE_DATA_SCOPE", which defaults to 'extern' or 'extern "C"' in the Unix
1022     world, but is set differently for Windows.
1023    
1024     15. Added a second compiling function called pcre_compile2(). The only
1025     difference is that it has an extra argument, which is a pointer to an
1026     integer error code. When there is a compile-time failure, this is set
1027     non-zero, in addition to the error test pointer being set to point to an
1028     error message. The new argument may be NULL if no error number is required
1029     (but then you may as well call pcre_compile(), which is now just a
1030     wrapper). This facility is provided because some applications need a
1031     numeric error indication, but it has also enabled me to tidy up the way
1032     compile-time errors are handled in the POSIX wrapper.
1033    
1034     16. Added VPATH=.libs to the makefile; this should help when building with one
1035     prefix path and installing with another. (Or so I'm told by someone who
1036     knows more about this stuff than I do.)
1037    
1038     17. Added a new option, REG_DOTALL, to the POSIX function regcomp(). This
1039     passes PCRE_DOTALL to the pcre_compile() function, making the "." character
1040     match everything, including newlines. This is not POSIX-compatible, but
1041     somebody wanted the feature. From pcretest it can be activated by using
1042     both the P and the s flags.
1043    
1044     18. AC_PROG_LIBTOOL appeared twice in Makefile.in. Removed one.
1045    
1046     19. libpcre.pc was being incorrectly installed as executable.
1047    
1048     20. A couple of places in pcretest check for end-of-line by looking for '\n';
1049     it now also looks for '\r' so that it will work unmodified on Windows.
1050    
1051     21. Added Google's contributed C++ wrapper to the distribution.
1052    
1053     22. Added some untidy missing memory free() calls in pcretest, to keep
1054     Electric Fence happy when testing.
1055    
1056    
1057    
1058 nigel 75 Version 5.0 13-Sep-04
1059     ---------------------
1060    
1061     1. Internal change: literal characters are no longer packed up into items
1062     containing multiple characters in a single byte-string. Each character
1063     is now matched using a separate opcode. However, there may be more than one
1064     byte in the character in UTF-8 mode.
1065    
1066     2. The pcre_callout_block structure has two new fields: pattern_position and
1067     next_item_length. These contain the offset in the pattern to the next match
1068     item, and its length, respectively.
1069    
1070     3. The PCRE_AUTO_CALLOUT option for pcre_compile() requests the automatic
1071     insertion of callouts before each pattern item. Added the /C option to
1072     pcretest to make use of this.
1073    
1074     4. On the advice of a Windows user, the lines
1075    
1076     #if defined(_WIN32) || defined(WIN32)
1077     _setmode( _fileno( stdout ), 0x8000 );
1078     #endif /* defined(_WIN32) || defined(WIN32) */
1079    
1080     have been added to the source of pcretest. This apparently does useful
1081     magic in relation to line terminators.
1082    
1083     5. Changed "r" and "w" in the calls to fopen() in pcretest to "rb" and "wb"
1084     for the benefit of those environments where the "b" makes a difference.
1085    
1086     6. The icc compiler has the same options as gcc, but "configure" doesn't seem
1087     to know about it. I have put a hack into configure.in that adds in code
1088     to set GCC=yes if CC=icc. This seems to end up at a point in the
1089     generated configure script that is early enough to affect the setting of
1090     compiler options, which is what is needed, but I have no means of testing
1091     whether it really works. (The user who reported this had patched the
1092     generated configure script, which of course I cannot do.)
1093    
1094     LATER: After change 22 below (new libtool files), the configure script
1095     seems to know about icc (and also ecc). Therefore, I have commented out
1096     this hack in configure.in.
1097    
1098     7. Added support for pkg-config (2 patches were sent in).
1099    
1100     8. Negated POSIX character classes that used a combination of internal tables
1101     were completely broken. These were [[:^alpha:]], [[:^alnum:]], and
1102     [[:^ascii]]. Typically, they would match almost any characters. The other
1103     POSIX classes were not broken in this way.
1104    
1105     9. Matching the pattern "\b.*?" against "ab cd", starting at offset 1, failed
1106     to find the match, as PCRE was deluded into thinking that the match had to
1107     start at the start point or following a newline. The same bug applied to
1108     patterns with negative forward assertions or any backward assertions
1109     preceding ".*" at the start, unless the pattern required a fixed first
1110     character. This was a failing pattern: "(?!.bcd).*". The bug is now fixed.
1111    
1112     10. In UTF-8 mode, when moving forwards in the subject after a failed match
1113     starting at the last subject character, bytes beyond the end of the subject
1114     string were read.
1115    
1116     11. Renamed the variable "class" as "classbits" to make life easier for C++
1117     users. (Previously there was a macro definition, but it apparently wasn't
1118     enough.)
1119    
1120     12. Added the new field "tables" to the extra data so that tables can be passed
1121     in at exec time, or the internal tables can be re-selected. This allows
1122     a compiled regex to be saved and re-used at a later time by a different
1123     program that might have everything at different addresses.
1124    
1125     13. Modified the pcre-config script so that, when run on Solaris, it shows a
1126     -R library as well as a -L library.
1127    
1128     14. The debugging options of pcretest (-d on the command line or D on a
1129     pattern) showed incorrect output for anything following an extended class
1130     that contained multibyte characters and which was followed by a quantifier.
1131    
1132     15. Added optional support for general category Unicode character properties
1133     via the \p, \P, and \X escapes. Unicode property support implies UTF-8
1134     support. It adds about 90K to the size of the library. The meanings of the
1135     inbuilt class escapes such as \d and \s have NOT been changed.
1136    
1137     16. Updated pcredemo.c to include calls to free() to release the memory for the
1138     compiled pattern.
1139    
1140     17. The generated file chartables.c was being created in the source directory
1141     instead of in the building directory. This caused the build to fail if the
1142     source directory was different from the building directory, and was
1143     read-only.
1144    
1145     18. Added some sample Win commands from Mark Tetrode into the NON-UNIX-USE
1146     file. No doubt somebody will tell me if they don't make sense... Also added
1147     Dan Mooney's comments about building on OpenVMS.
1148    
1149     19. Added support for partial matching via the PCRE_PARTIAL option for
1150     pcre_exec() and the \P data escape in pcretest.
1151    
1152     20. Extended pcretest with 3 new pattern features:
1153    
1154     (i) A pattern option of the form ">rest-of-line" causes pcretest to
1155     write the compiled pattern to the file whose name is "rest-of-line".
1156     This is a straight binary dump of the data, with the saved pointer to
1157     the character tables forced to be NULL. The study data, if any, is
1158     written too. After writing, pcretest reads a new pattern.
1159    
1160     (ii) If, instead of a pattern, "<rest-of-line" is given, pcretest reads a
1161     compiled pattern from the given file. There must not be any
1162     occurrences of "<" in the file name (pretty unlikely); if there are,
1163     pcretest will instead treat the initial "<" as a pattern delimiter.
1164     After reading in the pattern, pcretest goes on to read data lines as
1165     usual.
1166    
1167     (iii) The F pattern option causes pcretest to flip the bytes in the 32-bit
1168     and 16-bit fields in a compiled pattern, to simulate a pattern that
1169     was compiled on a host of opposite endianness.
1170    
1171     21. The pcre-exec() function can now cope with patterns that were compiled on
1172     hosts of opposite endianness, with this restriction:
1173    
1174     As for any compiled expression that is saved and used later, the tables
1175     pointer field cannot be preserved; the extra_data field in the arguments
1176     to pcre_exec() should be used to pass in a tables address if a value
1177     other than the default internal tables were used at compile time.
1178    
1179     22. Calling pcre_exec() with a negative value of the "ovecsize" parameter is
1180     now diagnosed as an error. Previously, most of the time, a negative number
1181     would have been treated as zero, but if in addition "ovector" was passed as
1182     NULL, a crash could occur.
1183    
1184     23. Updated the files ltmain.sh, config.sub, config.guess, and aclocal.m4 with
1185     new versions from the libtool 1.5 distribution (the last one is a copy of
1186     a file called libtool.m4). This seems to have fixed the need to patch
1187     "configure" to support Darwin 1.3 (which I used to do). However, I still
1188     had to patch ltmain.sh to ensure that ${SED} is set (it isn't on my
1189     workstation).
1190    
1191     24. Changed the PCRE licence to be the more standard "BSD" licence.
1192    
1193    
1194 nigel 73 Version 4.5 01-Dec-03
1195     ---------------------
1196    
1197     1. There has been some re-arrangement of the code for the match() function so
1198     that it can be compiled in a version that does not call itself recursively.
1199     Instead, it keeps those local variables that need separate instances for
1200     each "recursion" in a frame on the heap, and gets/frees frames whenever it
1201     needs to "recurse". Keeping track of where control must go is done by means
1202     of setjmp/longjmp. The whole thing is implemented by a set of macros that
1203     hide most of the details from the main code, and operates only if
1204     NO_RECURSE is defined while compiling pcre.c. If PCRE is built using the
1205     "configure" mechanism, "--disable-stack-for-recursion" turns on this way of
1206     operating.
1207    
1208     To make it easier for callers to provide specially tailored get/free
1209     functions for this usage, two new functions, pcre_stack_malloc, and
1210     pcre_stack_free, are used. They are always called in strict stacking order,
1211     and the size of block requested is always the same.
1212    
1213     The PCRE_CONFIG_STACKRECURSE info parameter can be used to find out whether
1214     PCRE has been compiled to use the stack or the heap for recursion. The
1215     -C option of pcretest uses this to show which version is compiled.
1216    
1217     A new data escape \S, is added to pcretest; it causes the amounts of store
1218     obtained and freed by both kinds of malloc/free at match time to be added
1219     to the output.
1220    
1221     2. Changed the locale test to use "fr_FR" instead of "fr" because that's
1222     what's available on my current Linux desktop machine.
1223    
1224     3. When matching a UTF-8 string, the test for a valid string at the start has
1225     been extended. If start_offset is not zero, PCRE now checks that it points
1226     to a byte that is the start of a UTF-8 character. If not, it returns
1227     PCRE_ERROR_BADUTF8_OFFSET (-11). Note: the whole string is still checked;
1228     this is necessary because there may be backward assertions in the pattern.
1229     When matching the same subject several times, it may save resources to use
1230     PCRE_NO_UTF8_CHECK on all but the first call if the string is long.
1231    
1232     4. The code for checking the validity of UTF-8 strings has been tightened so
1233     that it rejects (a) strings containing 0xfe or 0xff bytes and (b) strings
1234     containing "overlong sequences".
1235    
1236     5. Fixed a bug (appearing twice) that I could not find any way of exploiting!
1237     I had written "if ((digitab[*p++] && chtab_digit) == 0)" where the "&&"
1238     should have been "&", but it just so happened that all the cases this let
1239     through by mistake were picked up later in the function.
1240    
1241     6. I had used a variable called "isblank" - this is a C99 function, causing
1242     some compilers to warn. To avoid this, I renamed it (as "blankclass").
1243    
1244     7. Cosmetic: (a) only output another newline at the end of pcretest if it is
1245     prompting; (b) run "./pcretest /dev/null" at the start of the test script
1246     so the version is shown; (c) stop "make test" echoing "./RunTest".
1247    
1248     8. Added patches from David Burgess to enable PCRE to run on EBCDIC systems.
1249    
1250     9. The prototype for memmove() for systems that don't have it was using
1251     size_t, but the inclusion of the header that defines size_t was later. I've
1252     moved the #includes for the C headers earlier to avoid this.
1253    
1254     10. Added some adjustments to the code to make it easier to compiler on certain
1255     special systems:
1256    
1257     (a) Some "const" qualifiers were missing.
1258     (b) Added the macro EXPORT before all exported functions; by default this
1259     is defined to be empty.
1260     (c) Changed the dftables auxiliary program (that builds chartables.c) so
1261     that it reads its output file name as an argument instead of writing
1262     to the standard output and assuming this can be redirected.
1263    
1264     11. In UTF-8 mode, if a recursive reference (e.g. (?1)) followed a character
1265     class containing characters with values greater than 255, PCRE compilation
1266     went into a loop.
1267    
1268     12. A recursive reference to a subpattern that was within another subpattern
1269     that had a minimum quantifier of zero caused PCRE to crash. For example,
1270     (x(y(?2))z)? provoked this bug with a subject that got as far as the
1271     recursion. If the recursively-called subpattern itself had a zero repeat,
1272     that was OK.
1273    
1274     13. In pcretest, the buffer for reading a data line was set at 30K, but the
1275     buffer into which it was copied (for escape processing) was still set at
1276     1024, so long lines caused crashes.
1277    
1278     14. A pattern such as /[ab]{1,3}+/ failed to compile, giving the error
1279     "internal error: code overflow...". This applied to any character class
1280     that was followed by a possessive quantifier.
1281    
1282     15. Modified the Makefile to add libpcre.la as a prerequisite for
1283     libpcreposix.la because I was told this is needed for a parallel build to
1284     work.
1285    
1286     16. If a pattern that contained .* following optional items at the start was
1287     studied, the wrong optimizing data was generated, leading to matching
1288     errors. For example, studying /[ab]*.*c/ concluded, erroneously, that any
1289     matching string must start with a or b or c. The correct conclusion for
1290     this pattern is that a match can start with any character.
1291    
1292    
1293 nigel 71 Version 4.4 13-Aug-03
1294     ---------------------
1295    
1296     1. In UTF-8 mode, a character class containing characters with values between
1297     127 and 255 was not handled correctly if the compiled pattern was studied.
1298     In fixing this, I have also improved the studying algorithm for such
1299     classes (slightly).
1300    
1301     2. Three internal functions had redundant arguments passed to them. Removal
1302     might give a very teeny performance improvement.
1303    
1304     3. Documentation bug: the value of the capture_top field in a callout is *one
1305     more than* the number of the hightest numbered captured substring.
1306    
1307     4. The Makefile linked pcretest and pcregrep with -lpcre, which could result
1308     in incorrectly linking with a previously installed version. They now link
1309     explicitly with libpcre.la.
1310    
1311     5. configure.in no longer needs to recognize Cygwin specially.
1312    
1313     6. A problem in pcre.in for Windows platforms is fixed.
1314    
1315     7. If a pattern was successfully studied, and the -d (or /D) flag was given to
1316     pcretest, it used to include the size of the study block as part of its
1317     output. Unfortunately, the structure contains a field that has a different
1318     size on different hardware architectures. This meant that the tests that
1319     showed this size failed. As the block is currently always of a fixed size,
1320     this information isn't actually particularly useful in pcretest output, so
1321     I have just removed it.
1322    
1323     8. Three pre-processor statements accidentally did not start in column 1.
1324     Sadly, there are *still* compilers around that complain, even though
1325     standard C has not required this for well over a decade. Sigh.
1326    
1327     9. In pcretest, the code for checking callouts passed small integers in the
1328     callout_data field, which is a void * field. However, some picky compilers
1329     complained about the casts involved for this on 64-bit systems. Now
1330     pcretest passes the address of the small integer instead, which should get
1331     rid of the warnings.
1332    
1333     10. By default, when in UTF-8 mode, PCRE now checks for valid UTF-8 strings at
1334     both compile and run time, and gives an error if an invalid UTF-8 sequence
1335     is found. There is a option for disabling this check in cases where the
1336     string is known to be correct and/or the maximum performance is wanted.
1337    
1338     11. In response to a bug report, I changed one line in Makefile.in from
1339    
1340     -Wl,--out-implib,.libs/lib@WIN_PREFIX@pcreposix.dll.a \
1341     to
1342     -Wl,--out-implib,.libs/@WIN_PREFIX@libpcreposix.dll.a \
1343    
1344     to look similar to other lines, but I have no way of telling whether this
1345     is the right thing to do, as I do not use Windows. No doubt I'll get told
1346     if it's wrong...
1347    
1348    
1349 nigel 69 Version 4.3 21-May-03
1350     ---------------------
1351    
1352     1. Two instances of @WIN_PREFIX@ omitted from the Windows targets in the
1353     Makefile.
1354    
1355     2. Some refactoring to improve the quality of the code:
1356    
1357     (i) The utf8_table... variables are now declared "const".
1358    
1359     (ii) The code for \cx, which used the "case flipping" table to upper case
1360     lower case letters, now just substracts 32. This is ASCII-specific,
1361     but the whole concept of \cx is ASCII-specific, so it seems
1362     reasonable.
1363    
1364     (iii) PCRE was using its character types table to recognize decimal and
1365     hexadecimal digits in the pattern. This is silly, because it handles
1366     only 0-9, a-f, and A-F, but the character types table is locale-
1367     specific, which means strange things might happen. A private
1368     table is now used for this - though it costs 256 bytes, a table is
1369     much faster than multiple explicit tests. Of course, the standard
1370     character types table is still used for matching digits in subject
1371     strings against \d.
1372    
1373     (iv) Strictly, the identifier ESC_t is reserved by POSIX (all identifiers
1374     ending in _t are). So I've renamed it as ESC_tee.
1375    
1376     3. The first argument for regexec() in the POSIX wrapper should have been
1377     defined as "const".
1378    
1379     4. Changed pcretest to use malloc() for its buffers so that they can be
1380     Electric Fenced for debugging.
1381    
1382     5. There were several places in the code where, in UTF-8 mode, PCRE would try
1383     to read one or more bytes before the start of the subject string. Often this
1384     had no effect on PCRE's behaviour, but in some circumstances it could
1385     provoke a segmentation fault.
1386    
1387     6. A lookbehind at the start of a pattern in UTF-8 mode could also cause PCRE
1388     to try to read one or more bytes before the start of the subject string.
1389    
1390     7. A lookbehind in a pattern matched in non-UTF-8 mode on a PCRE compiled with
1391     UTF-8 support could misbehave in various ways if the subject string
1392     contained bytes with the 0x80 bit set and the 0x40 bit unset in a lookbehind
1393     area. (PCRE was not checking for the UTF-8 mode flag, and trying to move
1394     back over UTF-8 characters.)
1395    
1396    
1397 nigel 67 Version 4.2 14-Apr-03
1398     ---------------------
1399    
1400     1. Typo "#if SUPPORT_UTF8" instead of "#ifdef SUPPORT_UTF8" fixed.
1401    
1402     2. Changes to the building process, supplied by Ronald Landheer-Cieslak
1403     [ON_WINDOWS]: new variable, "#" on non-Windows platforms
1404     [NOT_ON_WINDOWS]: new variable, "#" on Windows platforms
1405     [WIN_PREFIX]: new variable, "cyg" for Cygwin
1406     * Makefile.in: use autoconf substitution for OBJEXT, EXEEXT, BUILD_OBJEXT
1407     and BUILD_EXEEXT
1408     Note: automatic setting of the BUILD variables is not yet working
1409     set CPPFLAGS and BUILD_CPPFLAGS (but don't use yet) - should be used at
1410     compile-time but not at link-time
1411     [LINK]: use for linking executables only
1412     make different versions for Windows and non-Windows
1413     [LINKLIB]: new variable, copy of UNIX-style LINK, used for linking
1414     libraries
1415     [LINK_FOR_BUILD]: new variable
1416     [OBJEXT]: use throughout
1417     [EXEEXT]: use throughout
1418     <winshared>: new target
1419     <wininstall>: new target
1420     <dftables.o>: use native compiler
1421     <dftables>: use native linker
1422     <install>: handle Windows platform correctly
1423     <clean>: ditto
1424     <check>: ditto
1425     copy DLL to top builddir before testing
1426    
1427     As part of these changes, -no-undefined was removed again. This was reported
1428     to give trouble on HP-UX 11.0, so getting rid of it seems like a good idea
1429     in any case.
1430    
1431     3. Some tidies to get rid of compiler warnings:
1432    
1433     . In the match_data structure, match_limit was an unsigned long int, whereas
1434     match_call_count was an int. I've made them both unsigned long ints.
1435    
1436     . In pcretest the fact that a const uschar * doesn't automatically cast to
1437     a void * provoked a warning.
1438    
1439     . Turning on some more compiler warnings threw up some "shadow" variables
1440     and a few more missing casts.
1441    
1442     4. If PCRE was complied with UTF-8 support, but called without the PCRE_UTF8
1443     option, a class that contained a single character with a value between 128
1444     and 255 (e.g. /[\xFF]/) caused PCRE to crash.
1445    
1446     5. If PCRE was compiled with UTF-8 support, but called without the PCRE_UTF8
1447     option, a class that contained several characters, but with at least one
1448     whose value was between 128 and 255 caused PCRE to crash.
1449    
1450    
1451 nigel 65 Version 4.1 12-Mar-03
1452     ---------------------
1453 nigel 63
1454 nigel 65 1. Compiling with gcc -pedantic found a couple of places where casts were
1455     needed, and a string in dftables.c that was longer than standard compilers are
1456     required to support.
1457    
1458     2. Compiling with Sun's compiler found a few more places where the code could
1459     be tidied up in order to avoid warnings.
1460    
1461     3. The variables for cross-compiling were called HOST_CC and HOST_CFLAGS; the
1462     first of these names is deprecated in the latest Autoconf in favour of the name
1463     CC_FOR_BUILD, because "host" is typically used to mean the system on which the
1464     compiled code will be run. I can't find a reference for HOST_CFLAGS, but by
1465     analogy I have changed it to CFLAGS_FOR_BUILD.
1466    
1467     4. Added -no-undefined to the linking command in the Makefile, because this is
1468     apparently helpful for Windows. To make it work, also added "-L. -lpcre" to the
1469     linking step for the pcreposix library.
1470    
1471     5. PCRE was failing to diagnose the case of two named groups with the same
1472     name.
1473    
1474     6. A problem with one of PCRE's optimizations was discovered. PCRE remembers a
1475     literal character that is needed in the subject for a match, and scans along to
1476     ensure that it is present before embarking on the full matching process. This
1477     saves time in cases of nested unlimited repeats that are never going to match.
1478     Problem: the scan can take a lot of time if the subject is very long (e.g.
1479     megabytes), thus penalizing straightforward matches. It is now done only if the
1480     amount of subject to be scanned is less than 1000 bytes.
1481    
1482     7. A lesser problem with the same optimization is that it was recording the
1483     first character of an anchored pattern as "needed", thus provoking a search
1484     right along the subject, even when the first match of the pattern was going to
1485     fail. The "needed" character is now not set for anchored patterns, unless it
1486     follows something in the pattern that is of non-fixed length. Thus, it still
1487     fulfils its original purpose of finding quick non-matches in cases of nested
1488     unlimited repeats, but isn't used for simple anchored patterns such as /^abc/.
1489    
1490    
1491     Version 4.0 17-Feb-03
1492     ---------------------
1493    
1494 nigel 63 1. If a comment in an extended regex that started immediately after a meta-item
1495     extended to the end of string, PCRE compiled incorrect data. This could lead to
1496     all kinds of weird effects. Example: /#/ was bad; /()#/ was bad; /a#/ was not.
1497    
1498     2. Moved to autoconf 2.53 and libtool 1.4.2.
1499    
1500     3. Perl 5.8 no longer needs "use utf8" for doing UTF-8 things. Consequently,
1501     the special perltest8 script is no longer needed - all the tests can be run
1502     from a single perltest script.
1503    
1504     4. From 5.004, Perl has not included the VT character (0x0b) in the set defined
1505     by \s. It has now been removed in PCRE. This means it isn't recognized as
1506     whitespace in /x regexes too, which is the same as Perl. Note that the POSIX
1507     class [:space:] *does* include VT, thereby creating a mess.
1508    
1509     5. Added the class [:blank:] (a GNU extension from Perl 5.8) to match only
1510     space and tab.
1511    
1512     6. Perl 5.005 was a long time ago. It's time to amalgamate the tests that use
1513     its new features into the main test script, reducing the number of scripts.
1514    
1515     7. Perl 5.8 has changed the meaning of patterns like /a(?i)b/. Earlier versions
1516     were backward compatible, and made the (?i) apply to the whole pattern, as if
1517     /i were given. Now it behaves more logically, and applies the option setting
1518     only to what follows. PCRE has been changed to follow suit. However, if it
1519     finds options settings right at the start of the pattern, it extracts them into
1520     the global options, as before. Thus, they show up in the info data.
1521    
1522     8. Added support for the \Q...\E escape sequence. Characters in between are
1523     treated as literals. This is slightly different from Perl in that $ and @ are
1524     also handled as literals inside the quotes. In Perl, they will cause variable
1525     interpolation. Note the following examples:
1526    
1527     Pattern PCRE matches Perl matches
1528    
1529     \Qabc$xyz\E abc$xyz abc followed by the contents of $xyz
1530     \Qabc\$xyz\E abc\$xyz abc\$xyz
1531     \Qabc\E\$\Qxyz\E abc$xyz abc$xyz
1532    
1533     For compatibility with Perl, \Q...\E sequences are recognized inside character
1534     classes as well as outside them.
1535    
1536     9. Re-organized 3 code statements in pcretest to avoid "overflow in
1537     floating-point constant arithmetic" warnings from a Microsoft compiler. Added a
1538     (size_t) cast to one statement in pcretest and one in pcreposix to avoid
1539     signed/unsigned warnings.
1540    
1541     10. SunOS4 doesn't have strtoul(). This was used only for unpicking the -o
1542     option for pcretest, so I've replaced it by a simple function that does just
1543     that job.
1544    
1545     11. pcregrep was ending with code 0 instead of 2 for the commands "pcregrep" or
1546     "pcregrep -".
1547    
1548     12. Added "possessive quantifiers" ?+, *+, ++, and {,}+ which come from Sun's
1549     Java package. This provides some syntactic sugar for simple cases of what my
1550     documentation calls "once-only subpatterns". A pattern such as x*+ is the same
1551     as (?>x*). In other words, if what is inside (?>...) is just a single repeated
1552     item, you can use this simplified notation. Note that only makes sense with
1553     greedy quantifiers. Consequently, the use of the possessive quantifier forces
1554     greediness, whatever the setting of the PCRE_UNGREEDY option.
1555    
1556     13. A change of greediness default within a pattern was not taking effect at
1557     the current level for patterns like /(b+(?U)a+)/. It did apply to parenthesized
1558     subpatterns that followed. Patterns like /b+(?U)a+/ worked because the option
1559     was abstracted outside.
1560    
1561     14. PCRE now supports the \G assertion. It is true when the current matching
1562     position is at the start point of the match. This differs from \A when the
1563     starting offset is non-zero. Used with the /g option of pcretest (or similar
1564     code), it works in the same way as it does for Perl's /g option. If all
1565     alternatives of a regex begin with \G, the expression is anchored to the start
1566     match position, and the "anchored" flag is set in the compiled expression.
1567    
1568     15. Some bugs concerning the handling of certain option changes within patterns
1569     have been fixed. These applied to options other than (?ims). For example,
1570     "a(?x: b c )d" did not match "XabcdY" but did match "Xa b c dY". It should have
1571     been the other way round. Some of this was related to change 7 above.
1572    
1573     16. PCRE now gives errors for /[.x.]/ and /[=x=]/ as unsupported POSIX
1574     features, as Perl does. Previously, PCRE gave the warnings only for /[[.x.]]/
1575     and /[[=x=]]/. PCRE now also gives an error for /[:name:]/ because it supports
1576     POSIX classes only within a class (e.g. /[[:alpha:]]/).
1577    
1578     17. Added support for Perl's \C escape. This matches one byte, even in UTF8
1579     mode. Unlike ".", it always matches newline, whatever the setting of
1580     PCRE_DOTALL. However, PCRE does not permit \C to appear in lookbehind
1581     assertions. Perl allows it, but it doesn't (in general) work because it can't
1582     calculate the length of the lookbehind. At least, that's the case for Perl
1583     5.8.0 - I've been told they are going to document that it doesn't work in
1584     future.
1585    
1586     18. Added an error diagnosis for escapes that PCRE does not support: these are
1587     \L, \l, \N, \P, \p, \U, \u, and \X.
1588    
1589     19. Although correctly diagnosing a missing ']' in a character class, PCRE was
1590     reading past the end of the pattern in cases such as /[abcd/.
1591    
1592     20. PCRE was getting more memory than necessary for patterns with classes that
1593     contained both POSIX named classes and other characters, e.g. /[[:space:]abc/.
1594    
1595     21. Added some code, conditional on #ifdef VPCOMPAT, to make life easier for
1596     compiling PCRE for use with Virtual Pascal.
1597    
1598     22. Small fix to the Makefile to make it work properly if the build is done
1599     outside the source tree.
1600    
1601     23. Added a new extension: a condition to go with recursion. If a conditional
1602     subpattern starts with (?(R) the "true" branch is used if recursion has
1603     happened, whereas the "false" branch is used only at the top level.
1604    
1605     24. When there was a very long string of literal characters (over 255 bytes
1606     without UTF support, over 250 bytes with UTF support), the computation of how
1607     much memory was required could be incorrect, leading to segfaults or other
1608     strange effects.
1609    
1610     25. PCRE was incorrectly assuming anchoring (either to start of subject or to
1611     start of line for a non-DOTALL pattern) when a pattern started with (.*) and
1612     there was a subsequent back reference to those brackets. This meant that, for
1613     example, /(.*)\d+\1/ failed to match "abc123bc". Unfortunately, it isn't
1614     possible to check for precisely this case. All we can do is abandon the
1615     optimization if .* occurs inside capturing brackets when there are any back
1616     references whatsoever. (See below for a better fix that came later.)
1617    
1618     26. The handling of the optimization for finding the first character of a
1619     non-anchored pattern, and for finding a character that is required later in the
1620     match were failing in some cases. This didn't break the matching; it just
1621     failed to optimize when it could. The way this is done has been re-implemented.
1622    
1623     27. Fixed typo in error message for invalid (?R item (it said "(?p").
1624    
1625     28. Added a new feature that provides some of the functionality that Perl
1626     provides with (?{...}). The facility is termed a "callout". The way it is done
1627     in PCRE is for the caller to provide an optional function, by setting
1628     pcre_callout to its entry point. Like pcre_malloc and pcre_free, this is a
1629     global variable. By default it is unset, which disables all calling out. To get
1630     the function called, the regex must include (?C) at appropriate points. This
1631     is, in fact, equivalent to (?C0), and any number <= 255 may be given with (?C).
1632     This provides a means of identifying different callout points. When PCRE
1633     reaches such a point in the regex, if pcre_callout has been set, the external
1634     function is called. It is provided with data in a structure called
1635     pcre_callout_block, which is defined in pcre.h. If the function returns 0,
1636     matching continues; if it returns a non-zero value, the match at the current
1637     point fails. However, backtracking will occur if possible. [This was changed
1638     later and other features added - see item 49 below.]
1639    
1640     29. pcretest is upgraded to test the callout functionality. It provides a
1641     callout function that displays information. By default, it shows the start of
1642     the match and the current position in the text. There are some new data escapes
1643     to vary what happens:
1644    
1645     \C+ in addition, show current contents of captured substrings
1646     \C- do not supply a callout function
1647     \C!n return 1 when callout number n is reached
1648     \C!n!m return 1 when callout number n is reached for the mth time
1649    
1650     30. If pcregrep was called with the -l option and just a single file name, it
1651     output "<stdin>" if a match was found, instead of the file name.
1652    
1653     31. Improve the efficiency of the POSIX API to PCRE. If the number of capturing
1654     slots is less than POSIX_MALLOC_THRESHOLD, use a block on the stack to pass to
1655     pcre_exec(). This saves a malloc/free per call. The default value of
1656     POSIX_MALLOC_THRESHOLD is 10; it can be changed by --with-posix-malloc-threshold
1657     when configuring.
1658    
1659     32. The default maximum size of a compiled pattern is 64K. There have been a
1660     few cases of people hitting this limit. The code now uses macros to handle the
1661     storing of links as offsets within the compiled pattern. It defaults to 2-byte
1662     links, but this can be changed to 3 or 4 bytes by --with-link-size when
1663     configuring. Tests 2 and 5 work only with 2-byte links because they output
1664     debugging information about compiled patterns.
1665    
1666     33. Internal code re-arrangements:
1667    
1668     (a) Moved the debugging function for printing out a compiled regex into
1669     its own source file (printint.c) and used #include to pull it into
1670     pcretest.c and, when DEBUG is defined, into pcre.c, instead of having two
1671     separate copies.
1672    
1673     (b) Defined the list of op-code names for debugging as a macro in
1674     internal.h so that it is next to the definition of the opcodes.
1675    
1676     (c) Defined a table of op-code lengths for simpler skipping along compiled
1677     code. This is again a macro in internal.h so that it is next to the
1678     definition of the opcodes.
1679    
1680     34. Added support for recursive calls to individual subpatterns, along the
1681     lines of Robin Houston's patch (but implemented somewhat differently).
1682    
1683     35. Further mods to the Makefile to help Win32. Also, added code to pcregrep to
1684     allow it to read and process whole directories in Win32. This code was
1685     contributed by Lionel Fourquaux; it has not been tested by me.
1686    
1687     36. Added support for named subpatterns. The Python syntax (?P<name>...) is
1688     used to name a group. Names consist of alphanumerics and underscores, and must
1689     be unique. Back references use the syntax (?P=name) and recursive calls use
1690     (?P>name) which is a PCRE extension to the Python extension. Groups still have
1691     numbers. The function pcre_fullinfo() can be used after compilation to extract
1692     a name/number map. There are three relevant calls:
1693    
1694     PCRE_INFO_NAMEENTRYSIZE yields the size of each entry in the map
1695     PCRE_INFO_NAMECOUNT yields the number of entries
1696     PCRE_INFO_NAMETABLE yields a pointer to the map.
1697    
1698     The map is a vector of fixed-size entries. The size of each entry depends on
1699     the length of the longest name used. The first two bytes of each entry are the
1700     group number, most significant byte first. There follows the corresponding
1701     name, zero terminated. The names are in alphabetical order.
1702    
1703     37. Make the maximum literal string in the compiled code 250 for the non-UTF-8
1704     case instead of 255. Making it the same both with and without UTF-8 support
1705     means that the same test output works with both.
1706    
1707     38. There was a case of malloc(0) in the POSIX testing code in pcretest. Avoid
1708     calling malloc() with a zero argument.
1709    
1710     39. Change 25 above had to resort to a heavy-handed test for the .* anchoring
1711     optimization. I've improved things by keeping a bitmap of backreferences with
1712     numbers 1-31 so that if .* occurs inside capturing brackets that are not in
1713     fact referenced, the optimization can be applied. It is unlikely that a
1714     relevant occurrence of .* (i.e. one which might indicate anchoring or forcing
1715     the match to follow \n) will appear inside brackets with a number greater than
1716     31, but if it does, any back reference > 31 suppresses the optimization.
1717    
1718     40. Added a new compile-time option PCRE_NO_AUTO_CAPTURE. This has the effect
1719     of disabling numbered capturing parentheses. Any opening parenthesis that is
1720     not followed by ? behaves as if it were followed by ?: but named parentheses
1721     can still be used for capturing (and they will acquire numbers in the usual
1722     way).
1723    
1724     41. Redesigned the return codes from the match() function into yes/no/error so
1725     that errors can be passed back from deep inside the nested calls. A malloc
1726     failure while inside a recursive subpattern call now causes the
1727     PCRE_ERROR_NOMEMORY return instead of quietly going wrong.
1728    
1729     42. It is now possible to set a limit on the number of times the match()
1730     function is called in a call to pcre_exec(). This facility makes it possible to
1731     limit the amount of recursion and backtracking, though not in a directly
1732     obvious way, because the match() function is used in a number of different
1733     circumstances. The count starts from zero for each position in the subject
1734     string (for non-anchored patterns). The default limit is, for compatibility, a
1735     large number, namely 10 000 000. You can change this in two ways:
1736    
1737     (a) When configuring PCRE before making, you can use --with-match-limit=n
1738     to set a default value for the compiled library.
1739    
1740     (b) For each call to pcre_exec(), you can pass a pcre_extra block in which
1741     a different value is set. See 45 below.
1742    
1743     If the limit is exceeded, pcre_exec() returns PCRE_ERROR_MATCHLIMIT.
1744    
1745     43. Added a new function pcre_config(int, void *) to enable run-time extraction
1746     of things that can be changed at compile time. The first argument specifies
1747     what is wanted and the second points to where the information is to be placed.
1748     The current list of available information is:
1749    
1750     PCRE_CONFIG_UTF8
1751    
1752     The output is an integer that is set to one if UTF-8 support is available;
1753     otherwise it is set to zero.
1754    
1755     PCRE_CONFIG_NEWLINE
1756    
1757     The output is an integer that it set to the value of the code that is used for
1758     newline. It is either LF (10) or CR (13).
1759    
1760     PCRE_CONFIG_LINK_SIZE
1761    
1762     The output is an integer that contains the number of bytes used for internal
1763     linkage in compiled expressions. The value is 2, 3, or 4. See item 32 above.
1764    
1765     PCRE_CONFIG_POSIX_MALLOC_THRESHOLD
1766    
1767     The output is an integer that contains the threshold above which the POSIX
1768     interface uses malloc() for output vectors. See item 31 above.
1769    
1770     PCRE_CONFIG_MATCH_LIMIT
1771    
1772     The output is an unsigned integer that contains the default limit of the number
1773     of match() calls in a pcre_exec() execution. See 42 above.
1774    
1775     44. pcretest has been upgraded by the addition of the -C option. This causes it
1776     to extract all the available output from the new pcre_config() function, and to
1777     output it. The program then exits immediately.
1778    
1779     45. A need has arisen to pass over additional data with calls to pcre_exec() in
1780     order to support additional features. One way would have been to define
1781     pcre_exec2() (for example) with extra arguments, but this would not have been
1782     extensible, and would also have required all calls to the original function to
1783     be mapped to the new one. Instead, I have chosen to extend the mechanism that
1784     is used for passing in "extra" data from pcre_study().
1785    
1786     The pcre_extra structure is now exposed and defined in pcre.h. It currently
1787     contains the following fields:
1788    
1789     flags a bitmap indicating which of the following fields are set
1790     study_data opaque data from pcre_study()
1791     match_limit a way of specifying a limit on match() calls for a specific
1792     call to pcre_exec()
1793     callout_data data for callouts (see 49 below)
1794    
1795     The flag bits are also defined in pcre.h, and are
1796    
1797     PCRE_EXTRA_STUDY_DATA
1798     PCRE_EXTRA_MATCH_LIMIT
1799     PCRE_EXTRA_CALLOUT_DATA
1800    
1801     The pcre_study() function now returns one of these new pcre_extra blocks, with
1802     the actual study data pointed to by the study_data field, and the
1803     PCRE_EXTRA_STUDY_DATA flag set. This can be passed directly to pcre_exec() as
1804     before. That is, this change is entirely upwards-compatible and requires no
1805     change to existing code.
1806    
1807     If you want to pass in additional data to pcre_exec(), you can either place it
1808     in a pcre_extra block provided by pcre_study(), or create your own pcre_extra
1809     block.
1810    
1811     46. pcretest has been extended to test the PCRE_EXTRA_MATCH_LIMIT feature. If a
1812     data string contains the escape sequence \M, pcretest calls pcre_exec() several
1813     times with different match limits, until it finds the minimum value needed for
1814     pcre_exec() to complete. The value is then output. This can be instructive; for
1815     most simple matches the number is quite small, but for pathological cases it
1816     gets very large very quickly.
1817    
1818     47. There's a new option for pcre_fullinfo() called PCRE_INFO_STUDYSIZE. It
1819     returns the size of the data block pointed to by the study_data field in a
1820     pcre_extra block, that is, the value that was passed as the argument to
1821     pcre_malloc() when PCRE was getting memory in which to place the information
1822     created by pcre_study(). The fourth argument should point to a size_t variable.
1823     pcretest has been extended so that this information is shown after a successful
1824     pcre_study() call when information about the compiled regex is being displayed.
1825    
1826     48. Cosmetic change to Makefile: there's no need to have / after $(DESTDIR)
1827     because what follows is always an absolute path. (Later: it turns out that this
1828     is more than cosmetic for MinGW, because it doesn't like empty path
1829     components.)
1830    
1831     49. Some changes have been made to the callout feature (see 28 above):
1832    
1833     (i) A callout function now has three choices for what it returns:
1834    
1835     0 => success, carry on matching
1836     > 0 => failure at this point, but backtrack if possible
1837     < 0 => serious error, return this value from pcre_exec()
1838    
1839     Negative values should normally be chosen from the set of PCRE_ERROR_xxx
1840     values. In particular, returning PCRE_ERROR_NOMATCH forces a standard
1841     "match failed" error. The error number PCRE_ERROR_CALLOUT is reserved for
1842     use by callout functions. It will never be used by PCRE itself.
1843    
1844     (ii) The pcre_extra structure (see 45 above) has a void * field called
1845     callout_data, with corresponding flag bit PCRE_EXTRA_CALLOUT_DATA. The
1846     pcre_callout_block structure has a field of the same name. The contents of
1847     the field passed in the pcre_extra structure are passed to the callout
1848     function in the corresponding field in the callout block. This makes it
1849     easier to use the same callout-containing regex from multiple threads. For
1850     testing, the pcretest program has a new data escape
1851    
1852     \C*n pass the number n (may be negative) as callout_data
1853    
1854     If the callout function in pcretest receives a non-zero value as
1855     callout_data, it returns that value.
1856    
1857     50. Makefile wasn't handling CFLAGS properly when compiling dftables. Also,
1858     there were some redundant $(CFLAGS) in commands that are now specified as
1859     $(LINK), which already includes $(CFLAGS).
1860    
1861     51. Extensions to UTF-8 support are listed below. These all apply when (a) PCRE
1862     has been compiled with UTF-8 support *and* pcre_compile() has been compiled
1863     with the PCRE_UTF8 flag. Patterns that are compiled without that flag assume
1864     one-byte characters throughout. Note that case-insensitive matching applies
1865     only to characters whose values are less than 256. PCRE doesn't support the
1866     notion of cases for higher-valued characters.
1867    
1868     (i) A character class whose characters are all within 0-255 is handled as
1869     a bit map, and the map is inverted for negative classes. Previously, a
1870     character > 255 always failed to match such a class; however it should
1871     match if the class was a negative one (e.g. [^ab]). This has been fixed.
1872    
1873     (ii) A negated character class with a single character < 255 is coded as
1874     "not this character" (OP_NOT). This wasn't working properly when the test
1875     character was multibyte, either singly or repeated.
1876    
1877     (iii) Repeats of multibyte characters are now handled correctly in UTF-8
1878     mode, for example: \x{100}{2,3}.
1879    
1880     (iv) The character escapes \b, \B, \d, \D, \s, \S, \w, and \W (either
1881     singly or repeated) now correctly test multibyte characters. However,
1882     PCRE doesn't recognize any characters with values greater than 255 as
1883     digits, spaces, or word characters. Such characters always match \D, \S,
1884     and \W, and never match \d, \s, or \w.
1885    
1886     (v) Classes may now contain characters and character ranges with values
1887     greater than 255. For example: [ab\x{100}-\x{400}].
1888    
1889     (vi) pcregrep now has a --utf-8 option (synonym -u) which makes it call
1890     PCRE in UTF-8 mode.
1891    
1892     52. The info request value PCRE_INFO_FIRSTCHAR has been renamed
1893     PCRE_INFO_FIRSTBYTE because it is a byte value. However, the old name is
1894     retained for backwards compatibility. (Note that LASTLITERAL is also a byte
1895     value.)
1896    
1897     53. The single man page has become too large. I have therefore split it up into
1898     a number of separate man pages. These also give rise to individual HTML pages;
1899     these are now put in a separate directory, and there is an index.html page that
1900     lists them all. Some hyperlinking between the pages has been installed.
1901    
1902     54. Added convenience functions for handling named capturing parentheses.
1903    
1904     55. Unknown escapes inside character classes (e.g. [\M]) and escapes that
1905     aren't interpreted therein (e.g. [\C]) are literals in Perl. This is now also
1906     true in PCRE, except when the PCRE_EXTENDED option is set, in which case they
1907     are faulted.
1908    
1909     56. Introduced HOST_CC and HOST_CFLAGS which can be set in the environment when
1910     calling configure. These values are used when compiling the dftables.c program
1911     which is run to generate the source of the default character tables. They
1912     default to the values of CC and CFLAGS. If you are cross-compiling PCRE,
1913     you will need to set these values.
1914    
1915     57. Updated the building process for Windows DLL, as provided by Fred Cox.
1916    
1917    
1918     Version 3.9 02-Jan-02
1919 nigel 61 ---------------------
1920    
1921     1. A bit of extraneous text had somehow crept into the pcregrep documentation.
1922    
1923     2. If --disable-static was given, the building process failed when trying to
1924     build pcretest and pcregrep. (For some reason it was using libtool to compile
1925     them, which is not right, as they aren't part of the library.)
1926    
1927    
1928 nigel 59 Version 3.8 18-Dec-01
1929     ---------------------
1930    
1931     1. The experimental UTF-8 code was completely screwed up. It was packing the
1932     bytes in the wrong order. How dumb can you get?
1933    
1934    
1935 nigel 57 Version 3.7 29-Oct-01
1936     ---------------------
1937    
1938     1. In updating pcretest to check change 1 of version 3.6, I screwed up.
1939     This caused pcretest, when used on the test data, to segfault. Unfortunately,
1940     this didn't happen under Solaris 8, where I normally test things.
1941    
1942 nigel 59 2. The Makefile had to be changed to make it work on BSD systems, where 'make'
1943     doesn't seem to recognize that ./xxx and xxx are the same file. (This entry
1944     isn't in ChangeLog distributed with 3.7 because I forgot when I hastily made
1945     this fix an hour or so after the initial 3.7 release.)
1946 nigel 57
1947 nigel 59
1948 nigel 55 Version 3.6 23-Oct-01
1949     ---------------------
1950    
1951     1. Crashed with /(sens|respons)e and \1ibility/ and "sense and sensibility" if
1952     offsets passed as NULL with zero offset count.
1953    
1954     2. The config.guess and config.sub files had not been updated when I moved to
1955     the latest autoconf.
1956    
1957    
1958 nigel 53 Version 3.5 15-Aug-01
1959     ---------------------
1960 nigel 5
1961 nigel 53 1. Added some missing #if !defined NOPOSIX conditionals in pcretest.c that
1962     had been forgotten.
1963    
1964     2. By using declared but undefined structures, we can avoid using "void"
1965     definitions in pcre.h while keeping the internal definitions of the structures
1966     private.
1967    
1968     3. The distribution is now built using autoconf 2.50 and libtool 1.4. From a
1969     user point of view, this means that both static and shared libraries are built
1970     by default, but this can be individually controlled. More of the work of
1971     handling this static/shared cases is now inside libtool instead of PCRE's make
1972     file.
1973    
1974     4. The pcretest utility is now installed along with pcregrep because it is
1975     useful for users (to test regexs) and by doing this, it automatically gets
1976     relinked by libtool. The documentation has been turned into a man page, so
1977     there are now .1, .txt, and .html versions in /doc.
1978    
1979     5. Upgrades to pcregrep:
1980     (i) Added long-form option names like gnu grep.
1981     (ii) Added --help to list all options with an explanatory phrase.
1982     (iii) Added -r, --recursive to recurse into sub-directories.
1983     (iv) Added -f, --file to read patterns from a file.
1984    
1985     6. pcre_exec() was referring to its "code" argument before testing that
1986     argument for NULL (and giving an error if it was NULL).
1987    
1988     7. Upgraded Makefile.in to allow for compiling in a different directory from
1989     the source directory.
1990    
1991     8. Tiny buglet in pcretest: when pcre_fullinfo() was called to retrieve the
1992     options bits, the pointer it was passed was to an int instead of to an unsigned
1993     long int. This mattered only on 64-bit systems.
1994    
1995     9. Fixed typo (3.4/1) in pcre.h again. Sigh. I had changed pcre.h (which is
1996     generated) instead of pcre.in, which it its source. Also made the same change
1997     in several of the .c files.
1998    
1999     10. A new release of gcc defines printf() as a macro, which broke pcretest
2000     because it had an ifdef in the middle of a string argument for printf(). Fixed
2001     by using separate calls to printf().
2002    
2003     11. Added --enable-newline-is-cr and --enable-newline-is-lf to the configure
2004     script, to force use of CR or LF instead of \n in the source. On non-Unix
2005     systems, the value can be set in config.h.
2006    
2007     12. The limit of 200 on non-capturing parentheses is a _nesting_ limit, not an
2008     absolute limit. Changed the text of the error message to make this clear, and
2009     likewise updated the man page.
2010    
2011     13. The limit of 99 on the number of capturing subpatterns has been removed.
2012     The new limit is 65535, which I hope will not be a "real" limit.
2013    
2014    
2015 nigel 51 Version 3.4 22-Aug-00
2016     ---------------------
2017    
2018     1. Fixed typo in pcre.h: unsigned const char * changed to const unsigned char *.
2019    
2020     2. Diagnose condition (?(0) as an error instead of crashing on matching.
2021    
2022    
2023 nigel 49 Version 3.3 01-Aug-00
2024     ---------------------
2025    
2026     1. If an octal character was given, but the value was greater than \377, it
2027     was not getting masked to the least significant bits, as documented. This could
2028     lead to crashes in some systems.
2029    
2030     2. Perl 5.6 (if not earlier versions) accepts classes like [a-\d] and treats
2031     the hyphen as a literal. PCRE used to give an error; it now behaves like Perl.
2032    
2033     3. Added the functions pcre_free_substring() and pcre_free_substring_list().
2034     These just pass their arguments on to (pcre_free)(), but they are provided
2035     because some uses of PCRE bind it to non-C systems that can call its functions,
2036     but cannot call free() or pcre_free() directly.
2037    
2038     4. Add "make test" as a synonym for "make check". Corrected some comments in
2039     the Makefile.
2040    
2041     5. Add $(DESTDIR)/ in front of all the paths in the "install" target in the
2042     Makefile.
2043    
2044     6. Changed the name of pgrep to pcregrep, because Solaris has introduced a
2045     command called pgrep for grepping around the active processes.
2046    
2047     7. Added the beginnings of support for UTF-8 character strings.
2048    
2049     8. Arranged for the Makefile to pass over the settings of CC, CFLAGS, and
2050     RANLIB to ./ltconfig so that they are used by libtool. I think these are all
2051     the relevant ones. (AR is not passed because ./ltconfig does its own figuring
2052     out for the ar command.)
2053    
2054    
2055 nigel 47 Version 3.2 12-May-00
2056     ---------------------
2057    
2058     This is purely a bug fixing release.
2059    
2060     1. If the pattern /((Z)+|A)*/ was matched agained ZABCDEFG it matched Z instead
2061     of ZA. This was just one example of several cases that could provoke this bug,
2062     which was introduced by change 9 of version 2.00. The code for breaking
2063     infinite loops after an iteration that matches an empty string was't working
2064     correctly.
2065    
2066     2. The pcretest program was not imitating Perl correctly for the pattern /a*/g
2067     when matched against abbab (for example). After matching an empty string, it
2068     wasn't forcing anchoring when setting PCRE_NOTEMPTY for the next attempt; this
2069     caused it to match further down the string than it should.
2070    
2071     3. The code contained an inclusion of sys/types.h. It isn't clear why this
2072     was there because it doesn't seem to be needed, and it causes trouble on some
2073     systems, as it is not a Standard C header. It has been removed.
2074    
2075     4. Made 4 silly changes to the source to avoid stupid compiler warnings that
2076     were reported on the Macintosh. The changes were from
2077    
2078     while ((c = *(++ptr)) != 0 && c != '\n');
2079     to
2080     while ((c = *(++ptr)) != 0 && c != '\n') ;
2081    
2082     Totally extraordinary, but if that's what it takes...
2083    
2084     5. PCRE is being used in one environment where neither memmove() nor bcopy() is
2085     available. Added HAVE_BCOPY and an autoconf test for it; if neither
2086     HAVE_MEMMOVE nor HAVE_BCOPY is set, use a built-in emulation function which
2087     assumes the way PCRE uses memmove() (always moving upwards).
2088    
2089     6. PCRE is being used in one environment where strchr() is not available. There
2090     was only one use in pcre.c, and writing it out to avoid strchr() probably gives
2091     faster code anyway.
2092    
2093    
2094 nigel 45 Version 3.1 09-Feb-00
2095     ---------------------
2096    
2097     The only change in this release is the fixing of some bugs in Makefile.in for
2098     the "install" target:
2099    
2100     (1) It was failing to install pcreposix.h.
2101    
2102     (2) It was overwriting the pcre.3 man page with the pcreposix.3 man page.
2103    
2104    
2105 nigel 43 Version 3.0 01-Feb-00
2106     ---------------------
2107 nigel 41
2108     1. Add support for the /+ modifier to perltest (to output $` like it does in
2109     pcretest).
2110    
2111     2. Add support for the /g modifier to perltest.
2112    
2113     3. Fix pcretest so that it behaves even more like Perl for /g when the pattern
2114     matches null strings.
2115    
2116     4. Fix perltest so that it doesn't do unwanted things when fed an empty
2117     pattern. Perl treats empty patterns specially - it reuses the most recent
2118     pattern, which is not what we want. Replace // by /(?#)/ in order to avoid this
2119     effect.
2120    
2121     5. The POSIX interface was broken in that it was just handing over the POSIX
2122     captured string vector to pcre_exec(), but (since release 2.00) PCRE has
2123     required a bigger vector, with some working space on the end. This means that
2124     the POSIX wrapper now has to get and free some memory, and copy the results.
2125    
2126 nigel 43 6. Added some simple autoconf support, placing the test data and the
2127     documentation in separate directories, re-organizing some of the
2128     information files, and making it build pcre-config (a GNU standard). Also added
2129     libtool support for building PCRE as a shared library, which is now the
2130     default.
2131 nigel 41
2132 nigel 43 7. Got rid of the leading zero in the definition of PCRE_MINOR because 08 and
2133     09 are not valid octal constants. Single digits will be used for minor values
2134     less than 10.
2135    
2136     8. Defined REG_EXTENDED and REG_NOSUB as zero in the POSIX header, so that
2137     existing programs that set these in the POSIX interface can use PCRE without
2138     modification.
2139    
2140     9. Added a new function, pcre_fullinfo() with an extensible interface. It can
2141     return all that pcre_info() returns, plus additional data. The pcre_info()
2142     function is retained for compatibility, but is considered to be obsolete.
2143    
2144     10. Added experimental recursion feature (?R) to handle one common case that
2145     Perl 5.6 will be able to do with (?p{...}).
2146    
2147     11. Added support for POSIX character classes like [:alpha:], which Perl is
2148     adopting.
2149    
2150    
2151 nigel 39 Version 2.08 31-Aug-99
2152     ----------------------
2153    
2154     1. When startoffset was not zero and the pattern began with ".*", PCRE was not
2155     trying to match at the startoffset position, but instead was moving forward to
2156     the next newline as if a previous match had failed.
2157    
2158     2. pcretest was not making use of PCRE_NOTEMPTY when repeating for /g and /G,
2159     and could get into a loop if a null string was matched other than at the start
2160     of the subject.
2161    
2162     3. Added definitions of PCRE_MAJOR and PCRE_MINOR to pcre.h so the version can
2163     be distinguished at compile time, and for completeness also added PCRE_DATE.
2164    
2165     5. Added Paul Sokolovsky's minor changes to make it easy to compile a Win32 DLL
2166     in GnuWin32 environments.
2167    
2168    
2169 nigel 37 Version 2.07 29-Jul-99
2170     ----------------------
2171    
2172     1. The documentation is now supplied in plain text form and HTML as well as in
2173     the form of man page sources.
2174    
2175     2. C++ compilers don't like assigning (void *) values to other pointer types.
2176     In particular this affects malloc(). Although there is no problem in Standard
2177     C, I've put in casts to keep C++ compilers happy.
2178    
2179     3. Typo on pcretest.c; a cast of (unsigned char *) in the POSIX regexec() call
2180     should be (const char *).
2181    
2182     4. If NOPOSIX is defined, pcretest.c compiles without POSIX support. This may
2183     be useful for non-Unix systems who don't want to bother with the POSIX stuff.
2184     However, I haven't made this a standard facility. The documentation doesn't
2185     mention it, and the Makefile doesn't support it.
2186    
2187     5. The Makefile now contains an "install" target, with editable destinations at
2188     the top of the file. The pcretest program is not installed.
2189    
2190     6. pgrep -V now gives the PCRE version number and date.
2191    
2192     7. Fixed bug: a zero repetition after a literal string (e.g. /abcde{0}/) was
2193     causing the entire string to be ignored, instead of just the last character.
2194    
2195     8. If a pattern like /"([^\\"]+|\\.)*"/ is applied in the normal way to a
2196     non-matching string, it can take a very, very long time, even for strings of
2197     quite modest length, because of the nested recursion. PCRE now does better in
2198     some of these cases. It does this by remembering the last required literal
2199     character in the pattern, and pre-searching the subject to ensure it is present
2200     before running the real match. In other words, it applies a heuristic to detect
2201     some types of certain failure quickly, and in the above example, if presented
2202     with a string that has no trailing " it gives "no match" very quickly.
2203    
2204     9. A new runtime option PCRE_NOTEMPTY causes null string matches to be ignored;
2205     other alternatives are tried instead.
2206    
2207    
2208 nigel 35 Version 2.06 09-Jun-99
2209     ----------------------
2210    
2211     1. Change pcretest's output for amount of store used to show just the code
2212     space, because the remainder (the data block) varies in size between 32-bit and
2213     64-bit systems.
2214    
2215     2. Added an extra argument to pcre_exec() to supply an offset in the subject to
2216     start matching at. This allows lookbehinds to work when searching for multiple
2217     occurrences in a string.
2218    
2219     3. Added additional options to pcretest for testing multiple occurrences:
2220    
2221     /+ outputs the rest of the string that follows a match
2222     /g loops for multiple occurrences, using the new startoffset argument
2223     /G loops for multiple occurrences by passing an incremented pointer
2224    
2225     4. PCRE wasn't doing the "first character" optimization for patterns starting
2226     with \b or \B, though it was doing it for other lookbehind assertions. That is,
2227     it wasn't noticing that a match for a pattern such as /\bxyz/ has to start with
2228     the letter 'x'. On long subject strings, this gives a significant speed-up.
2229    
2230    
2231 nigel 33 Version 2.05 21-Apr-99
2232     ----------------------
2233    
2234     1. Changed the type of magic_number from int to long int so that it works
2235     properly on 16-bit systems.
2236    
2237     2. Fixed a bug which caused patterns starting with .* not to work correctly
2238     when the subject string contained newline characters. PCRE was assuming
2239     anchoring for such patterns in all cases, which is not correct because .* will
2240     not pass a newline unless PCRE_DOTALL is set. It now assumes anchoring only if
2241     DOTALL is set at top level; otherwise it knows that patterns starting with .*
2242     must be retried after every newline in the subject.
2243    
2244    
2245 nigel 31 Version 2.04 18-Feb-99
2246     ----------------------
2247    
2248     1. For parenthesized subpatterns with repeats whose minimum was zero, the
2249     computation of the store needed to hold the pattern was incorrect (too large).
2250     If such patterns were nested a few deep, this could multiply and become a real
2251     problem.
2252    
2253     2. Added /M option to pcretest to show the memory requirement of a specific
2254     pattern. Made -m a synonym of -s (which does this globally) for compatibility.
2255    
2256     3. Subpatterns of the form (regex){n,m} (i.e. limited maximum) were being
2257     compiled in such a way that the backtracking after subsequent failure was
2258     pessimal. Something like (a){0,3} was compiled as (a)?(a)?(a)? instead of
2259     ((a)((a)(a)?)?)? with disastrous performance if the maximum was of any size.
2260    
2261    
2262 nigel 29 Version 2.03 02-Feb-99
2263     ----------------------
2264    
2265     1. Fixed typo and small mistake in man page.
2266    
2267 nigel 31 2. Added 4th condition (GPL supersedes if conflict) and created separate
2268     LICENCE file containing the conditions.
2269 nigel 29
2270     3. Updated pcretest so that patterns such as /abc\/def/ work like they do in
2271     Perl, that is the internal \ allows the delimiter to be included in the
2272     pattern. Locked out the use of \ as a delimiter. If \ immediately follows
2273     the final delimiter, add \ to the end of the pattern (to test the error).
2274    
2275     4. Added the convenience functions for extracting substrings after a successful
2276     match. Updated pcretest to make it able to test these functions.
2277    
2278    
2279 nigel 27 Version 2.02 14-Jan-99
2280     ----------------------
2281    
2282     1. Initialized the working variables associated with each extraction so that
2283     their saving and restoring doesn't refer to uninitialized store.
2284    
2285     2. Put dummy code into study.c in order to trick the optimizer of the IBM C
2286     compiler for OS/2 into generating correct code. Apparently IBM isn't going to
2287     fix the problem.
2288    
2289     3. Pcretest: the timing code wasn't using LOOPREPEAT for timing execution
2290     calls, and wasn't printing the correct value for compiling calls. Increased the
2291     default value of LOOPREPEAT, and the number of significant figures in the
2292     times.
2293    
2294     4. Changed "/bin/rm" in the Makefile to "-rm" so it works on Windows NT.
2295    
2296     5. Renamed "deftables" as "dftables" to get it down to 8 characters, to avoid
2297     a building problem on Windows NT with a FAT file system.
2298    
2299    
2300 nigel 25 Version 2.01 21-Oct-98
2301     ----------------------
2302    
2303     1. Changed the API for pcre_compile() to allow for the provision of a pointer
2304     to character tables built by pcre_maketables() in the current locale. If NULL
2305     is passed, the default tables are used.
2306    
2307    
2308 nigel 23 Version 2.00 24-Sep-98
2309 nigel 21 ----------------------
2310    
2311 nigel 23 1. Since the (>?) facility is in Perl 5.005, don't require PCRE_EXTRA to enable
2312     it any more.
2313    
2314     2. Allow quantification of (?>) groups, and make it work correctly.
2315    
2316     3. The first character computation wasn't working for (?>) groups.
2317    
2318     4. Correct the implementation of \Z (it is permitted to match on the \n at the
2319     end of the subject) and add 5.005's \z, which really does match only at the
2320     very end of the subject.
2321    
2322     5. Remove the \X "cut" facility; Perl doesn't have it, and (?> is neater.
2323    
2324     6. Remove the ability to specify CASELESS, MULTILINE, DOTALL, and
2325     DOLLAR_END_ONLY at runtime, to make it possible to implement the Perl 5.005
2326     localized options. All options to pcre_study() were also removed.
2327    
2328     7. Add other new features from 5.005:
2329    
2330     $(?<= positive lookbehind
2331     $(?<! negative lookbehind
2332     (?imsx-imsx) added the unsetting capability
2333     such a setting is global if at outer level; local otherwise
2334     (?imsx-imsx:) non-capturing groups with option setting
2335     (?(cond)re|re) conditional pattern matching
2336    
2337     A backreference to itself in a repeated group matches the previous
2338     captured string.
2339    
2340     8. General tidying up of studying (both automatic and via "study")
2341     consequential on the addition of new assertions.
2342    
2343     9. As in 5.005, unlimited repeated groups that could match an empty substring
2344     are no longer faulted at compile time. Instead, the loop is forcibly broken at
2345     runtime if any iteration does actually match an empty substring.
2346    
2347     10. Include the RunTest script in the distribution.
2348    
2349     11. Added tests from the Perl 5.005_02 distribution. This showed up a few
2350     discrepancies, some of which were old and were also with respect to 5.004. They
2351     have now been fixed.
2352    
2353    
2354     Version 1.09 28-Apr-98
2355     ----------------------
2356    
2357 nigel 21 1. A negated single character class followed by a quantifier with a minimum
2358     value of one (e.g. [^x]{1,6} ) was not compiled correctly. This could lead to
2359     program crashes, or just wrong answers. This did not apply to negated classes
2360     containing more than one character, or to minima other than one.
2361    
2362    
2363 nigel 19 Version 1.08 27-Mar-98
2364     ----------------------
2365    
2366     1. Add PCRE_UNGREEDY to invert the greediness of quantifiers.
2367    
2368     2. Add (?U) and (?X) to set PCRE_UNGREEDY and PCRE_EXTRA respectively. The
2369     latter must appear before anything that relies on it in the pattern.
2370    
2371    
2372 nigel 17 Version 1.07 16-Feb-98
2373     ----------------------
2374    
2375     1. A pattern such as /((a)*)*/ was not being diagnosed as in error (unlimited
2376     repeat of a potentially empty string).
2377    
2378    
2379 nigel 15 Version 1.06 23-Jan-98
2380     ----------------------
2381    
2382     1. Added Markus Oberhumer's little patches for C++.
2383    
2384     2. Literal strings longer than 255 characters were broken.
2385    
2386    
2387 nigel 13 Version 1.05 23-Dec-97
2388     ----------------------
2389    
2390     1. Negated character classes containing more than one character were failing if
2391     PCRE_CASELESS was set at run time.
2392    
2393    
2394 nigel 11 Version 1.04 19-Dec-97
2395     ----------------------
2396    
2397     1. Corrected the man page, where some "const" qualifiers had been omitted.
2398    
2399     2. Made debugging output print "{0,xxx}" instead of just "{,xxx}" to agree with
2400     input syntax.
2401    
2402     3. Fixed memory leak which occurred when a regex with back references was
2403     matched with an offsets vector that wasn't big enough. The temporary memory
2404     that is used in this case wasn't being freed if the match failed.
2405    
2406     4. Tidied pcretest to ensure it frees memory that it gets.
2407    
2408     5. Temporary memory was being obtained in the case where the passed offsets
2409     vector was exactly big enough.
2410    
2411     6. Corrected definition of offsetof() from change 5 below.
2412    
2413     7. I had screwed up change 6 below and broken the rules for the use of
2414     setjmp(). Now fixed.
2415    
2416    
2417 nigel 9 Version 1.03 18-Dec-97
2418     ----------------------
2419    
2420     1. A erroneous regex with a missing opening parenthesis was correctly
2421     diagnosed, but PCRE attempted to access brastack[-1], which could cause crashes
2422     on some systems.
2423    
2424     2. Replaced offsetof(real_pcre, code) by offsetof(real_pcre, code[0]) because
2425     it was reported that one broken compiler failed on the former because "code" is
2426     also an independent variable.
2427    
2428     3. The erroneous regex a[]b caused an array overrun reference.
2429    
2430     4. A regex ending with a one-character negative class (e.g. /[^k]$/) did not
2431     fail on data ending with that character. (It was going on too far, and checking
2432     the next character, typically a binary zero.) This was specific to the
2433     optimized code for single-character negative classes.
2434    
2435     5. Added a contributed patch from the TIN world which does the following:
2436    
2437     + Add an undef for memmove, in case the the system defines a macro for it.
2438    
2439     + Add a definition of offsetof(), in case there isn't one. (I don't know
2440     the reason behind this - offsetof() is part of the ANSI standard - but
2441     it does no harm).
2442    
2443     + Reduce the ifdef's in pcre.c using macro DPRINTF, thereby eliminating
2444     most of the places where whitespace preceded '#'. I have given up and
2445     allowed the remaining 2 cases to be at the margin.
2446    
2447     + Rename some variables in pcre to eliminate shadowing. This seems very
2448     pedantic, but does no harm, of course.
2449    
2450     6. Moved the call to setjmp() into its own function, to get rid of warnings
2451     from gcc -Wall, and avoided calling it at all unless PCRE_EXTRA is used.
2452    
2453     7. Constructs such as \d{8,} were compiling into the equivalent of
2454 nigel 11 \d{8}\d{0,65527} instead of \d{8}\d* which didn't make much difference to the
2455 nigel 9 outcome, but in this particular case used more store than had been allocated,
2456     which caused the bug to be discovered because it threw up an internal error.
2457    
2458     8. The debugging code in both pcre and pcretest for outputting the compiled
2459     form of a regex was going wrong in the case of back references followed by
2460     curly-bracketed repeats.
2461    
2462    
2463 nigel 7 Version 1.02 12-Dec-97
2464     ----------------------
2465    
2466     1. Typos in pcre.3 and comments in the source fixed.
2467    
2468     2. Applied a contributed patch to get rid of places where it used to remove
2469     'const' from variables, and fixed some signed/unsigned and uninitialized
2470     variable warnings.
2471    
2472     3. Added the "runtest" target to Makefile.
2473    
2474     4. Set default compiler flag to -O2 rather than just -O.
2475    
2476    
2477 nigel 5 Version 1.01 19-Nov-97
2478     ----------------------
2479    
2480     1. PCRE was failing to diagnose unlimited repeat of empty string for patterns
2481     like /([ab]*)*/, that is, for classes with more than one character in them.
2482    
2483     2. Likewise, it wasn't diagnosing patterns with "once-only" subpatterns, such
2484     as /((?>a*))*/ (a PCRE_EXTRA facility).
2485    
2486    
2487     Version 1.00 18-Nov-97
2488     ----------------------
2489    
2490     1. Added compile-time macros to support systems such as SunOS4 which don't have
2491     memmove() or strerror() but have other things that can be used instead.
2492    
2493     2. Arranged that "make clean" removes the executables.
2494    
2495    
2496 nigel 3 Version 0.99 27-Oct-97
2497     ----------------------
2498    
2499     1. Fixed bug in code for optimizing classes with only one character. It was
2500     initializing a 32-byte map regardless, which could cause it to run off the end
2501     of the memory it had got.
2502    
2503     2. Added, conditional on PCRE_EXTRA, the proposed (?>REGEX) construction.
2504    
2505    
2506     Version 0.98 22-Oct-97
2507     ----------------------
2508    
2509     1. Fixed bug in code for handling temporary memory usage when there are more
2510     back references than supplied space in the ovector. This could cause segfaults.
2511    
2512    
2513     Version 0.97 21-Oct-97
2514     ----------------------
2515    
2516     1. Added the \X "cut" facility, conditional on PCRE_EXTRA.
2517    
2518     2. Optimized negated single characters not to use a bit map.
2519    
2520     3. Brought error texts together as macro definitions; clarified some of them;
2521     fixed one that was wrong - it said "range out of order" when it meant "invalid
2522     escape sequence".
2523    
2524     4. Changed some char * arguments to const char *.
2525    
2526     5. Added PCRE_NOTBOL and PCRE_NOTEOL (from POSIX).
2527    
2528     6. Added the POSIX-style API wrapper in pcreposix.a and testing facilities in
2529     pcretest.
2530    
2531    
2532     Version 0.96 16-Oct-97
2533     ----------------------
2534    
2535     1. Added a simple "pgrep" utility to the distribution.
2536    
2537     2. Fixed an incompatibility with Perl: "{" is now treated as a normal character
2538     unless it appears in one of the precise forms "{ddd}", "{ddd,}", or "{ddd,ddd}"
2539     where "ddd" means "one or more decimal digits".
2540    
2541     3. Fixed serious bug. If a pattern had a back reference, but the call to
2542     pcre_exec() didn't supply a large enough ovector to record the related
2543     identifying subpattern, the match always failed. PCRE now remembers the number
2544     of the largest back reference, and gets some temporary memory in which to save
2545     the offsets during matching if necessary, in order to ensure that
2546     backreferences always work.
2547    
2548     4. Increased the compatibility with Perl in a number of ways:
2549    
2550     (a) . no longer matches \n by default; an option PCRE_DOTALL is provided
2551     to request this handling. The option can be set at compile or exec time.
2552    
2553     (b) $ matches before a terminating newline by default; an option
2554     PCRE_DOLLAR_ENDONLY is provided to override this (but not in multiline
2555     mode). The option can be set at compile or exec time.
2556    
2557     (c) The handling of \ followed by a digit other than 0 is now supposed to be
2558     the same as Perl's. If the decimal number it represents is less than 10
2559     or there aren't that many previous left capturing parentheses, an octal
2560     escape is read. Inside a character class, it's always an octal escape,
2561     even if it is a single digit.
2562    
2563     (d) An escaped but undefined alphabetic character is taken as a literal,
2564     unless PCRE_EXTRA is set. Currently this just reserves the remaining
2565     escapes.
2566    
2567     (e) {0} is now permitted. (The previous item is removed from the compiled
2568     pattern).
2569    
2570     5. Changed all the names of code files so that the basic parts are no longer
2571     than 10 characters, and abolished the teeny "globals.c" file.
2572    
2573     6. Changed the handling of character classes; they are now done with a 32-byte
2574     bit map always.
2575    
2576     7. Added the -d and /D options to pcretest to make it possible to look at the
2577     internals of compilation without having to recompile pcre.
2578    
2579    
2580     Version 0.95 23-Sep-97
2581     ----------------------
2582    
2583     1. Fixed bug in pre-pass concerning escaped "normal" characters such as \x5c or
2584     \x20 at the start of a run of normal characters. These were being treated as
2585     real characters, instead of the source characters being re-checked.
2586    
2587    
2588     Version 0.94 18-Sep-97
2589     ----------------------
2590    
2591     1. The functions are now thread-safe, with the caveat that the global variables
2592     containing pointers to malloc() and free() or alternative functions are the
2593     same for all threads.
2594    
2595     2. Get pcre_study() to generate a bitmap of initial characters for non-
2596     anchored patterns when this is possible, and use it if passed to pcre_exec().
2597    
2598    
2599     Version 0.93 15-Sep-97
2600     ----------------------
2601    
2602     1. /(b)|(:+)/ was computing an incorrect first character.
2603    
2604     2. Add pcre_study() to the API and the passing of pcre_extra to pcre_exec(),
2605     but not actually doing anything yet.
2606    
2607     3. Treat "-" characters in classes that cannot be part of ranges as literals,
2608     as Perl does (e.g. [-az] or [az-]).
2609    
2610     4. Set the anchored flag if a branch starts with .* or .*? because that tests
2611     all possible positions.
2612    
2613     5. Split up into different modules to avoid including unneeded functions in a
2614     compiled binary. However, compile and exec are still in one module. The "study"
2615     function is split off.
2616    
2617     6. The character tables are now in a separate module whose source is generated
2618     by an auxiliary program - but can then be edited by hand if required. There are
2619     now no calls to isalnum(), isspace(), isdigit(), isxdigit(), tolower() or
2620     toupper() in the code.
2621    
2622     7. Turn the malloc/free funtions variables into pcre_malloc and pcre_free and
2623     make them global. Abolish the function for setting them, as the caller can now
2624     set them directly.
2625    
2626    
2627     Version 0.92 11-Sep-97
2628     ----------------------
2629    
2630     1. A repeat with a fixed maximum and a minimum of 1 for an ordinary character
2631     (e.g. /a{1,3}/) was broken (I mis-optimized it).
2632    
2633     2. Caseless matching was not working in character classes if the characters in
2634     the pattern were in upper case.
2635    
2636     3. Make ranges like [W-c] work in the same way as Perl for caseless matching.
2637    
2638     4. Make PCRE_ANCHORED public and accept as a compile option.
2639    
2640     5. Add an options word to pcre_exec() and accept PCRE_ANCHORED and
2641     PCRE_CASELESS at run time. Add escapes \A and \I to pcretest to cause it to
2642     pass them.
2643    
2644     6. Give an error if bad option bits passed at compile or run time.
2645    
2646     7. Add PCRE_MULTILINE at compile and exec time, and (?m) as well. Add \M to
2647     pcretest to cause it to pass that flag.
2648    
2649     8. Add pcre_info(), to get the number of identifying subpatterns, the stored
2650     options, and the first character, if set.
2651    
2652     9. Recognize C+ or C{n,m} where n >= 1 as providing a fixed starting character.
2653    
2654    
2655     Version 0.91 10-Sep-97
2656     ----------------------
2657    
2658     1. PCRE was failing to diagnose unlimited repeats of subpatterns that could
2659     match the empty string as in /(a*)*/. It was looping and ultimately crashing.
2660    
2661     2. PCRE was looping on encountering an indefinitely repeated back reference to
2662     a subpattern that had matched an empty string, e.g. /(a|)\1*/. It now does what
2663     Perl does - treats the match as successful.
2664    
2665     ****

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