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

Contents of /code/trunk/ChangeLog

Parent Directory Parent Directory | Revision Log Revision Log


Revision 532 - (hide annotations) (download)
Tue Jun 1 16:21:42 2010 UTC (3 years, 10 months ago) by ph10
File size: 188309 byte(s)
Extend auto-possessify to handle some Unicode properties.

1 nigel 3 ChangeLog for PCRE
2     ------------------
3    
4 ph10 513 Version 8.10 03 May-2010
5 ph10 510 ------------------------
6    
7 ph10 512 1. Added support for (*MARK:ARG) and for ARG additions to PRUNE, SKIP, and
8 ph10 510 THEN.
9    
10 ph10 512 2. (*ACCEPT) was not working when inside an atomic group.
11 ph10 510
12 ph10 513 3. Inside a character class, \B is treated as a literal by default, but
13     faulted if PCRE_EXTRA is set. This mimics Perl's behaviour (the -w option
14     causes the error). The code is unchanged, but I tidied the documentation.
15    
16     4. Inside a character class, PCRE always treated \R and \X as literals,
17     whereas Perl faults them if its -w option is set. I have changed PCRE so
18     that it faults them when PCRE_EXTRA is set.
19 ph10 514
20     5. Added support for \N, which always matches any character other than
21     newline. (It is the same as "." when PCRE_DOTALL is not set.)
22 ph10 515
23     6. When compiling pcregrep with newer versions of gcc which may have
24     FORTIFY_SOURCE set, several warnings "ignoring return value of 'fwrite',
25     declared with attribute warn_unused_result" were given. Just casting the
26     result to (void) does not stop the warnings; a more elaborate fudge is
27     needed. I've used a macro to implement this.
28 ph10 516
29     7. Minor change to pcretest.c to avoid a compiler warning.
30 ph10 512
31 ph10 517 8. Added four artifical Unicode properties to help with an option to make
32 ph10 518 \s etc use properties (see next item). The new properties are: Xan
33     (alphanumeric), Xsp (Perl space), Xps (POSIX space), and Xwd (word).
34    
35     9. Added PCRE_UCP to make \b, \d, \s, \w, and certain POSIX character classes
36     use Unicode properties. (*UCP) at the start of a pattern can be used to set
37     this option. Modified pcretest to add /W to test this facility. Added
38     REG_UCP to make it available via the POSIX interface.
39 ph10 519
40     10. Added --line-buffered to pcregrep.
41 ph10 513
42 ph10 520 11. In UTF-8 mode, if a pattern that was compiled with PCRE_CASELESS was
43     studied, and the match started with a letter with a code point greater than
44     127 whose first byte was different to the first byte of the other case of
45     the letter, the other case of this starting letter was not recognized.
46 ph10 521
47     12. pcreposix.c included pcre.h before including pcre_internal.h. This caused a
48     conflict in the definition of PCRE_EXP_DECL. I have removed the include of
49     pcre.h as pcre_internal.h includes pcre.h itself. (This may be a bit of
50     historical tidying that never got done.)
51 ph10 523
52     13. If a pattern that was studied started with a repeated Unicode property
53     test, for example, \p{Nd}+, there was the theoretical possibility of
54     setting up an incorrect bitmap of starting bytes, but fortunately it could
55     not have actually happened in practice until change 8 above was made (it
56     added property types that matched character-matching opcodes).
57 ph10 524
58     14. pcre_study() now recognizes \h, \v, and \R when constructing a bit map of
59     possible starting bytes for non-anchored patterns.
60 ph10 528
61 ph10 532 15. Extended the "auto-possessify" feature of pcre_compile(). It now recognizes
62     \R, and also a number of cases that involve Unicode properties, both
63     explicit and implicit when PCRE_UCP is set.
64 ph10 529
65     16. If a repeated Unicode property match (e.g. \p{Lu}*) was used with non-UTF-8
66     input, it could crash or give wrong results if characters with values
67     greater than 0xc0 were present in the subject string. (Detail: it assumed
68     UTF-8 input when processing these items.)
69 ph10 530
70     17. Added a lot of (int) casts to avoid compiler warnings in systems where
71     size_t is 64-bit.
72 ph10 531
73     18. Added a check for running out of memory when PCRE is compiled with
74     --disable-stack-for-recursion.
75 ph10 532
76 ph10 517
77 ph10 520
78 ph10 508 Version 8.02 19-Mar-2010
79 ph10 491 ------------------------
80    
81     1. The Unicode data tables have been updated to Unicode 5.2.0.
82    
83 ph10 492 2. Added the option --libs-cpp to pcre-config, but only when C++ support is
84     configured.
85 ph10 507
86 ph10 493 3. Updated the licensing terms in the pcregexp.pas file, as agreed with the
87 ph10 507 original author of that file, following a query about its status.
88    
89     4. On systems that do not have stdint.h (e.g. Solaris), check for and include
90 ph10 494 inttypes.h instead. This fixes a bug that was introduced by change 8.01/8.
91 ph10 507
92 ph10 496 5. A pattern such as (?&t)*+(?(DEFINE)(?<t>.)) which has a possessive
93     quantifier applied to a forward-referencing subroutine call, could compile
94     incorrect code or give the error "internal error: previously-checked
95     referenced subpattern not found".
96 ph10 507
97     6. Both MS Visual Studio and Symbian OS have problems with initializing
98 ph10 497 variables to point to external functions. For these systems, therefore,
99     pcre_malloc etc. are now initialized to local functions that call the
100     relevant global functions.
101 ph10 507
102 ph10 498 7. There were two entries missing in the vectors called coptable and poptable
103     in pcre_dfa_exec.c. This could lead to memory accesses outsize the vectors.
104 ph10 507 I've fixed the data, and added a kludgy way of testing at compile time that
105     the lengths are correct (equal to the number of opcodes).
106    
107     8. Following on from 7, I added a similar kludge to check the length of the
108     eint vector in pcreposix.c.
109    
110     9. Error texts for pcre_compile() are held as one long string to avoid too
111     much relocation at load time. To find a text, the string is searched,
112 ph10 499 counting zeros. There was no check for running off the end of the string,
113     which could happen if a new error number was added without updating the
114 ph10 507 string.
115    
116     10. \K gave a compile-time error if it appeared in a lookbehind assersion.
117    
118 ph10 500 11. \K was not working if it appeared in an atomic group or in a group that
119 ph10 507 was called as a "subroutine", or in an assertion. Perl 5.11 documents that
120     \K is "not well defined" if used in an assertion. PCRE now accepts it if
121     the assertion is positive, but not if it is negative.
122    
123 ph10 501 12. Change 11 fortuitously reduced the size of the stack frame used in the
124 ph10 507 "match()" function of pcre_exec.c by one pointer. Forthcoming
125     implementation of support for (*MARK) will need an extra pointer on the
126 ph10 501 stack; I have reserved it now, so that the stack frame size does not
127     decrease.
128 ph10 507
129     13. A pattern such as (?P<L1>(?P<L2>0)|(?P>L2)(?P>L1)) in which the only other
130     item in branch that calls a recursion is a subroutine call - as in the
131 ph10 503 second branch in the above example - was incorrectly given the compile-
132     time error "recursive call could loop indefinitely" because pcre_compile()
133 ph10 507 was not correctly checking the subroutine for matching a non-empty string.
134    
135 ph10 505 14. The checks for overrunning compiling workspace could trigger after an
136 ph10 507 overrun had occurred. This is a "should never occur" error, but it can be
137 ph10 505 triggered by pathological patterns such as hundreds of nested parentheses.
138 ph10 507 The checks now trigger 100 bytes before the end of the workspace.
139 ph10 491
140 ph10 507 15. Fix typo in configure.ac: "srtoq" should be "strtoq".
141 ph10 492
142 ph10 507
143 ph10 489 Version 8.01 19-Jan-2010
144     ------------------------
145 ph10 471
146     1. If a pattern contained a conditional subpattern with only one branch (in
147 ph10 489 particular, this includes all (*DEFINE) patterns), a call to pcre_study()
148 ph10 473 computed the wrong minimum data length (which is of course zero for such
149 ph10 489 subpatterns). This could cause incorrect "no match" results.
150 ph10 473
151     2. For patterns such as (?i)a(?-i)b|c where an option setting at the start of
152 ph10 472 the pattern is reset in the first branch, pcre_compile() failed with
153     "internal error: code overflow at offset...". This happened only when
154 ph10 473 the reset was to the original external option setting. (An optimization
155     abstracts leading options settings into an external setting, which was the
156 ph10 472 cause of this.)
157 ph10 471
158 ph10 473 3. A pattern such as ^(?!a(*SKIP)b) where a negative assertion contained one
159     of the verbs SKIP, PRUNE, or COMMIT, did not work correctly. When the
160     assertion pattern did not match (meaning that the assertion was true), it
161     was incorrectly treated as false if the SKIP had been reached during the
162     matching. This also applied to assertions used as conditions.
163 ph10 471
164 ph10 473 4. If an item that is not supported by pcre_dfa_exec() was encountered in an
165     assertion subpattern, including such a pattern used as a condition,
166     unpredictable results occurred, instead of the error return
167     PCRE_ERROR_DFA_UITEM.
168 ph10 478
169 ph10 474 5. The C++ GlobalReplace function was not working like Perl for the special
170     situation when an empty string is matched. It now does the fancy magic
171 ph10 478 stuff that is necessary.
172    
173     6. In pcre_internal.h, obsolete includes to setjmp.h and stdarg.h have been
174 ph10 475 removed. (These were left over from very, very early versions of PCRE.)
175 ph10 478
176 ph10 475 7. Some cosmetic changes to the code to make life easier when compiling it
177     as part of something else:
178 ph10 478
179     (a) Change DEBUG to PCRE_DEBUG.
180    
181     (b) In pcre_compile(), rename the member of the "branch_chain" structure
182     called "current" as "current_branch", to prevent a collision with the
183 ph10 475 Linux macro when compiled as a kernel module.
184 ph10 478
185     (c) In pcre_study(), rename the function set_bit() as set_table_bit(), to
186     prevent a collision with the Linux macro when compiled as a kernel
187 ph10 475 module.
188 ph10 478
189 ph10 475 8. In pcre_compile() there are some checks for integer overflows that used to
190     cast potentially large values to (double). This has been changed to that
191 ph10 478 when building, a check for int64_t is made, and if it is found, it is used
192     instead, thus avoiding the use of floating point arithmetic. (There is no
193     other use of FP in PCRE.) If int64_t is not found, the fallback is to
194     double.
195    
196     9. Added two casts to avoid signed/unsigned warnings from VS Studio Express
197 ph10 476 2005 (difference between two addresses compared to an unsigned value).
198 ph10 473
199 ph10 478 10. Change the standard AC_CHECK_LIB test for libbz2 in configure.ac to a
200     custom one, because of the following reported problem in Windows:
201 ph10 473
202 ph10 478 - libbz2 uses the Pascal calling convention (WINAPI) for the functions
203 ph10 489 under Win32.
204 ph10 478 - The standard autoconf AC_CHECK_LIB fails to include "bzlib.h",
205 ph10 489 therefore missing the function definition.
206 ph10 478 - The compiler thus generates a "C" signature for the test function.
207     - The linker fails to find the "C" function.
208     - PCRE fails to configure if asked to do so against libbz2.
209 ph10 486
210 ph10 479 11. When running libtoolize from libtool-2.2.6b as part of autogen.sh, these
211     messages were output:
212 ph10 478
213 ph10 479 Consider adding `AC_CONFIG_MACRO_DIR([m4])' to configure.ac and
214     rerunning libtoolize, to keep the correct libtool macros in-tree.
215     Consider adding `-I m4' to ACLOCAL_AMFLAGS in Makefile.am.
216 ph10 486
217 ph10 479 I have done both of these things.
218 ph10 486
219 ph10 480 12. Although pcre_dfa_exec() does not use nearly as much stack as pcre_exec()
220 ph10 486 most of the time, it *can* run out if it is given a pattern that contains a
221     runaway infinite recursion. I updated the discussion in the pcrestack man
222     page.
223    
224     13. Now that we have gone to the x.xx style of version numbers, the minor
225     version may start with zero. Using 08 or 09 is a bad idea because users
226     might check the value of PCRE_MINOR in their code, and 08 or 09 may be
227     interpreted as invalid octal numbers. I've updated the previous comment in
228     configure.ac, and also added a check that gives an error if 08 or 09 are
229 ph10 481 used.
230 ph10 486
231 ph10 482 14. Change 8.00/11 was not quite complete: code had been accidentally omitted,
232 ph10 483 causing partial matching to fail when the end of the subject matched \W
233     in a UTF-8 pattern where \W was quantified with a minimum of 3.
234 ph10 486
235 ph10 483 15. There were some discrepancies between the declarations in pcre_internal.h
236 ph10 486 of _pcre_is_newline(), _pcre_was_newline(), and _pcre_valid_utf8() and
237 ph10 483 their definitions. The declarations used "const uschar *" and the
238     definitions used USPTR. Even though USPTR is normally defined as "const
239     unsigned char *" (and uschar is typedeffed as "unsigned char"), it was
240     reported that: "This difference in casting confuses some C++ compilers, for
241     example, SunCC recognizes above declarations as different functions and
242 ph10 486 generates broken code for hbpcre." I have changed the declarations to use
243 ph10 483 USPTR.
244 ph10 486
245     16. GNU libtool is named differently on some systems. The autogen.sh script now
246 ph10 484 tries several variants such as glibtoolize (MacOSX) and libtoolize1x
247     (FreeBSD).
248 ph10 478
249 ph10 486 17. Applied Craig's patch that fixes an HP aCC compile error in pcre 8.00
250     (strtoXX undefined when compiling pcrecpp.cc). The patch contains this
251     comment: "Figure out how to create a longlong from a string: strtoll and
252     equivalent. It's not enough to call AC_CHECK_FUNCS: hpux has a strtoll, for
253     instance, but it only takes 2 args instead of 3!"
254 ph10 479
255 ph10 488 18. A subtle bug concerned with back references has been fixed by a change of
256     specification, with a corresponding code fix. A pattern such as
257     ^(xa|=?\1a)+$ which contains a back reference inside the group to which it
258     refers, was giving matches when it shouldn't. For example, xa=xaaa would
259     match that pattern. Interestingly, Perl (at least up to 5.11.3) has the
260     same bug. Such groups have to be quantified to be useful, or contained
261     inside another quantified group. (If there's no repetition, the reference
262     can never match.) The problem arises because, having left the group and
263     moved on to the rest of the pattern, a later failure that backtracks into
264     the group uses the captured value from the final iteration of the group
265     rather than the correct earlier one. I have fixed this in PCRE by forcing
266     any group that contains a reference to itself to be an atomic group; that
267     is, there cannot be any backtracking into it once it has completed. This is
268     similar to recursive and subroutine calls.
269 ph10 486
270 ph10 488
271 ph10 469 Version 8.00 19-Oct-09
272 ph10 418 ----------------------
273    
274     1. The table for translating pcre_compile() error codes into POSIX error codes
275 ph10 461 was out-of-date, and there was no check on the pcre_compile() error code
276     being within the table. This could lead to an OK return being given in
277 ph10 418 error.
278 ph10 461
279     2. Changed the call to open a subject file in pcregrep from fopen(pathname,
280     "r") to fopen(pathname, "rb"), which fixed a problem with some of the tests
281     in a Windows environment.
282    
283 ph10 420 3. The pcregrep --count option prints the count for each file even when it is
284     zero, as does GNU grep. However, pcregrep was also printing all files when
285     --files-with-matches was added. Now, when both options are given, it prints
286     counts only for those files that have at least one match. (GNU grep just
287 ph10 461 prints the file name in this circumstance, but including the count seems
288     more useful - otherwise, why use --count?) Also ensured that the
289 ph10 420 combination -clh just lists non-zero counts, with no names.
290 ph10 461
291     4. The long form of the pcregrep -F option was incorrectly implemented as
292     --fixed_strings instead of --fixed-strings. This is an incompatible change,
293     but it seems right to fix it, and I didn't think it was worth preserving
294     the old behaviour.
295    
296     5. The command line items --regex=pattern and --regexp=pattern were not
297 ph10 422 recognized by pcregrep, which required --regex pattern or --regexp pattern
298 ph10 461 (with a space rather than an '='). The man page documented the '=' forms,
299 ph10 422 which are compatible with GNU grep; these now work.
300 ph10 461
301     6. No libpcreposix.pc file was created for pkg-config; there was just
302 ph10 423 libpcre.pc and libpcrecpp.pc. The omission has been rectified.
303 ph10 461
304 ph10 425 7. Added #ifndef SUPPORT_UCP into the pcre_ucd.c module, to reduce its size
305 ph10 461 when UCP support is not needed, by modifying the Python script that
306 ph10 425 generates it from Unicode data files. This should not matter if the module
307     is correctly used as a library, but I received one complaint about 50K of
308     unwanted data. My guess is that the person linked everything into his
309     program rather than using a library. Anyway, it does no harm.
310 ph10 461
311 ph10 426 8. A pattern such as /\x{123}{2,2}+/8 was incorrectly compiled; the trigger
312 ph10 461 was a minimum greater than 1 for a wide character in a possessive
313 ph10 431 repetition. The same bug could also affect patterns like /(\x{ff}{0,2})*/8
314     which had an unlimited repeat of a nested, fixed maximum repeat of a wide
315     character. Chaos in the form of incorrect output or a compiling loop could
316     result.
317 ph10 461
318 ph10 426 9. The restrictions on what a pattern can contain when partial matching is
319 ph10 461 requested for pcre_exec() have been removed. All patterns can now be
320 ph10 426 partially matched by this function. In addition, if there are at least two
321 ph10 435 slots in the offset vector, the offset of the earliest inspected character
322     for the match and the offset of the end of the subject are set in them when
323 ph10 461 PCRE_ERROR_PARTIAL is returned.
324    
325 ph10 428 10. Partial matching has been split into two forms: PCRE_PARTIAL_SOFT, which is
326     synonymous with PCRE_PARTIAL, for backwards compatibility, and
327     PCRE_PARTIAL_HARD, which causes a partial match to supersede a full match,
328 ph10 462 and may be more useful for multi-segment matching.
329 ph10 461
330     11. Partial matching with pcre_exec() is now more intuitive. A partial match
331     used to be given if ever the end of the subject was reached; now it is
332     given only if matching could not proceed because another character was
333     needed. This makes a difference in some odd cases such as Z(*FAIL) with the
334     string "Z", which now yields "no match" instead of "partial match". In the
335     case of pcre_dfa_exec(), "no match" is given if every matching path for the
336     final character ended with (*FAIL).
337    
338 ph10 428 12. Restarting a match using pcre_dfa_exec() after a partial match did not work
339 ph10 461 if the pattern had a "must contain" character that was already found in the
340 ph10 428 earlier partial match, unless partial matching was again requested. For
341     example, with the pattern /dog.(body)?/, the "must contain" character is
342     "g". If the first part-match was for the string "dog", restarting with
343 ph10 435 "sbody" failed. This bug has been fixed.
344 ph10 461
345     13. The string returned by pcre_dfa_exec() after a partial match has been
346     changed so that it starts at the first inspected character rather than the
347     first character of the match. This makes a difference only if the pattern
348     starts with a lookbehind assertion or \b or \B (\K is not supported by
349     pcre_dfa_exec()). It's an incompatible change, but it makes the two
350 ph10 435 matching functions compatible, and I think it's the right thing to do.
351 ph10 461
352 ph10 435 14. Added a pcredemo man page, created automatically from the pcredemo.c file,
353 ph10 461 so that the demonstration program is easily available in environments where
354     PCRE has not been installed from source.
355    
356 ph10 435 15. Arranged to add -DPCRE_STATIC to cflags in libpcre.pc, libpcreposix.cp,
357 ph10 430 libpcrecpp.pc and pcre-config when PCRE is not compiled as a shared
358     library.
359 ph10 461
360 ph10 435 16. Added REG_UNGREEDY to the pcreposix interface, at the request of a user.
361 ph10 432 It maps to PCRE_UNGREEDY. It is not, of course, POSIX-compatible, but it
362 ph10 461 is not the first non-POSIX option to be added. Clearly some people find
363 ph10 433 these options useful.
364 ph10 461
365     17. If a caller to the POSIX matching function regexec() passes a non-zero
366 ph10 438 value for nmatch with a NULL value for pmatch, the value of
367 ph10 461 nmatch is forced to zero.
368    
369 ph10 437 18. RunGrepTest did not have a test for the availability of the -u option of
370 ph10 461 the diff command, as RunTest does. It now checks in the same way as
371 ph10 437 RunTest, and also checks for the -b option.
372 ph10 461
373 ph10 438 19. If an odd number of negated classes containing just a single character
374     interposed, within parentheses, between a forward reference to a named
375 ph10 461 subpattern and the definition of the subpattern, compilation crashed with
376     an internal error, complaining that it could not find the referenced
377 ph10 438 subpattern. An example of a crashing pattern is /(?&A)(([^m])(?<A>))/.
378 ph10 461 [The bug was that it was starting one character too far in when skipping
379     over the character class, thus treating the ] as data rather than
380     terminating the class. This meant it could skip too much.]
381    
382 ph10 442 20. Added PCRE_NOTEMPTY_ATSTART in order to be able to correctly implement the
383 ph10 461 /g option in pcretest when the pattern contains \K, which makes it possible
384 ph10 442 to have an empty string match not at the start, even when the pattern is
385 ph10 461 anchored. Updated pcretest and pcredemo to use this option.
386    
387 ph10 446 21. If the maximum number of capturing subpatterns in a recursion was greater
388 ph10 461 than the maximum at the outer level, the higher number was returned, but
389     with unset values at the outer level. The correct (outer level) value is
390 ph10 446 now given.
391 ph10 461
392 ph10 447 22. If (*ACCEPT) appeared inside capturing parentheses, previous releases of
393     PCRE did not set those parentheses (unlike Perl). I have now found a way to
394     make it do so. The string so far is captured, making this feature
395     compatible with Perl.
396 ph10 461
397     23. The tests have been re-organized, adding tests 11 and 12, to make it
398 ph10 448 possible to check the Perl 5.10 features against Perl 5.10.
399 ph10 461
400 ph10 454 24. Perl 5.10 allows subroutine calls in lookbehinds, as long as the subroutine
401 ph10 461 pattern matches a fixed length string. PCRE did not allow this; now it
402     does. Neither allows recursion.
403    
404     25. I finally figured out how to implement a request to provide the minimum
405     length of subject string that was needed in order to match a given pattern.
406     (It was back references and recursion that I had previously got hung up
407     on.) This code has now been added to pcre_study(); it finds a lower bound
408 ph10 455 to the length of subject needed. It is not necessarily the greatest lower
409     bound, but using it to avoid searching strings that are too short does give
410     some useful speed-ups. The value is available to calling programs via
411     pcre_fullinfo().
412 ph10 461
413 ph10 455 26. While implementing 25, I discovered to my embarrassment that pcretest had
414     not been passing the result of pcre_study() to pcre_dfa_exec(), so the
415     study optimizations had never been tested with that matching function.
416     Oops. What is worse, even when it was passed study data, there was a bug in
417     pcre_dfa_exec() that meant it never actually used it. Double oops. There
418     were also very few tests of studied patterns with pcre_dfa_exec().
419 ph10 461
420 ph10 457 27. If (?| is used to create subpatterns with duplicate numbers, they are now
421     allowed to have the same name, even if PCRE_DUPNAMES is not set. However,
422     on the other side of the coin, they are no longer allowed to have different
423     names, because these cannot be distinguished in PCRE, and this has caused
424     confusion. (This is a difference from Perl.)
425 ph10 461
426     28. When duplicate subpattern names are present (necessarily with different
427     numbers, as required by 27 above), and a test is made by name in a
428     conditional pattern, either for a subpattern having been matched, or for
429     recursion in such a pattern, all the associated numbered subpatterns are
430 ph10 459 tested, and the overall condition is true if the condition is true for any
431     one of them. This is the way Perl works, and is also more like the way
432     testing by number works.
433 ph10 418
434 ph10 461
435 ph10 412 Version 7.9 11-Apr-09
436 ph10 376 ---------------------
437    
438 ph10 392 1. When building with support for bzlib/zlib (pcregrep) and/or readline
439 ph10 376 (pcretest), all targets were linked against these libraries. This included
440     libpcre, libpcreposix, and libpcrecpp, even though they do not use these
441     libraries. This caused unwanted dependencies to be created. This problem
442 ph10 392 has been fixed, and now only pcregrep is linked with bzlib/zlib and only
443 ph10 376 pcretest is linked with readline.
444 ph10 392
445 ph10 376 2. The "typedef int BOOL" in pcre_internal.h that was included inside the
446     "#ifndef FALSE" condition by an earlier change (probably 7.8/18) has been
447     moved outside it again, because FALSE and TRUE are already defined in AIX,
448     but BOOL is not.
449 ph10 392
450     3. The pcre_config() function was treating the PCRE_MATCH_LIMIT and
451 ph10 409 PCRE_MATCH_LIMIT_RECURSION values as ints, when they should be long ints.
452 ph10 392
453 ph10 376 4. The pcregrep documentation said spaces were inserted as well as colons (or
454     hyphens) following file names and line numbers when outputting matching
455 ph10 392 lines. This is not true; no spaces are inserted. I have also clarified the
456 ph10 376 wording for the --colour (or --color) option.
457 ph10 392
458 ph10 378 5. In pcregrep, when --colour was used with -o, the list of matching strings
459     was not coloured; this is different to GNU grep, so I have changed it to be
460     the same.
461 ph10 392
462 ph10 378 6. When --colo(u)r was used in pcregrep, only the first matching substring in
463 ph10 392 each matching line was coloured. Now it goes on to look for further matches
464     of any of the test patterns, which is the same behaviour as GNU grep.
465    
466     7. A pattern that could match an empty string could cause pcregrep to loop; it
467     doesn't make sense to accept an empty string match in pcregrep, so I have
468 ph10 379 locked it out (using PCRE's PCRE_NOTEMPTY option). By experiment, this
469     seems to be how GNU grep behaves.
470 ph10 392
471 ph10 380 8. The pattern (?(?=.*b)b|^) was incorrectly compiled as "match must be at
472 ph10 392 start or after a newline", because the conditional assertion was not being
473     correctly handled. The rule now is that both the assertion and what follows
474     in the first alternative must satisfy the test.
475    
476 ph10 405 9. If auto-callout was enabled in a pattern with a conditional group whose
477 ph10 399 condition was an assertion, PCRE could crash during matching, both with
478     pcre_exec() and pcre_dfa_exec().
479 ph10 392
480     10. The PCRE_DOLLAR_ENDONLY option was not working when pcre_dfa_exec() was
481     used for matching.
482    
483     11. Unicode property support in character classes was not working for
484 ph10 385 characters (bytes) greater than 127 when not in UTF-8 mode.
485 ph10 388
486 ph10 392 12. Added the -M command line option to pcretest.
487    
488 ph10 388 14. Added the non-standard REG_NOTEMPTY option to the POSIX interface.
489 ph10 389
490     15. Added the PCRE_NO_START_OPTIMIZE match-time option.
491 ph10 390
492 ph10 392 16. Added comments and documentation about mis-use of no_arg in the C++
493 ph10 390 wrapper.
494 ph10 392
495 ph10 391 17. Implemented support for UTF-8 encoding in EBCDIC environments, a patch
496 ph10 392 from Martin Jerabek that uses macro names for all relevant character and
497 ph10 391 string constants.
498 ph10 405
499 ph10 393 18. Added to pcre_internal.h two configuration checks: (a) If both EBCDIC and
500     SUPPORT_UTF8 are set, give an error; (b) If SUPPORT_UCP is set without
501 ph10 405 SUPPORT_UTF8, define SUPPORT_UTF8. The "configure" script handles both of
502     these, but not everybody uses configure.
503    
504     19. A conditional group that had only one branch was not being correctly
505     recognized as an item that could match an empty string. This meant that an
506     enclosing group might also not be so recognized, causing infinite looping
507     (and probably a segfault) for patterns such as ^"((?(?=[a])[^"])|b)*"$
508     with the subject "ab", where knowledge that the repeated group can match
509 ph10 396 nothing is needed in order to break the loop.
510 ph10 405
511 ph10 397 20. If a pattern that was compiled with callouts was matched using pcre_dfa_
512 ph10 405 exec(), but without supplying a callout function, matching went wrong.
513    
514     21. If PCRE_ERROR_MATCHLIMIT occurred during a recursion, there was a memory
515     leak if the size of the offset vector was greater than 30. When the vector
516     is smaller, the saved offsets during recursion go onto a local stack
517     vector, but for larger vectors malloc() is used. It was failing to free
518     when the recursion yielded PCRE_ERROR_MATCH_LIMIT (or any other "abnormal"
519 ph10 400 error, in fact).
520 ph10 405
521 ph10 402 22. There was a missing #ifdef SUPPORT_UTF8 round one of the variables in the
522 ph10 405 heapframe that is used only when UTF-8 support is enabled. This caused no
523     problem, but was untidy.
524 ph10 376
525 ph10 405 23. Steven Van Ingelgem's patch to CMakeLists.txt to change the name
526     CMAKE_BINARY_DIR to PROJECT_BINARY_DIR so that it works when PCRE is
527     included within another project.
528 ph10 411
529 ph10 407 24. Steven Van Ingelgem's patches to add more options to the CMake support,
530     slightly modified by me:
531 ph10 411
532     (a) PCRE_BUILD_TESTS can be set OFF not to build the tests, including
533     not building pcregrep.
534    
535 ph10 407 (b) PCRE_BUILD_PCREGREP can be see OFF not to build pcregrep, but only
536 ph10 411 if PCRE_BUILD_TESTS is also set OFF, because the tests use pcregrep.
537    
538 ph10 408 25. Forward references, both numeric and by name, in patterns that made use of
539     duplicate group numbers, could behave incorrectly or give incorrect errors,
540     because when scanning forward to find the reference group, PCRE was not
541     taking into account the duplicate group numbers. A pattern such as
542     ^X(?3)(a)(?|(b)|(q))(Y) is an example.
543 ph10 411
544     26. Changed a few more instances of "const unsigned char *" to USPTR, making
545     the feature of a custom pointer more persuasive (as requested by a user).
546 ph10 416
547 ph10 411 27. Wrapped the definitions of fileno and isatty for Windows, which appear in
548 ph10 416 pcretest.c, inside #ifndefs, because it seems they are sometimes already
549     pre-defined.
550 ph10 392
551 ph10 416 28. Added support for (*UTF8) at the start of a pattern.
552    
553 ph10 413 29. Arrange for flags added by the "release type" setting in CMake to be shown
554     in the configuration summary.
555 ph10 405
556 ph10 411
557 ph10 374 Version 7.8 05-Sep-08
558 ph10 349 ---------------------
559    
560     1. Replaced UCP searching code with optimized version as implemented for Ad
561     Muncher (http://www.admuncher.com/) by Peter Kankowski. This uses a two-
562 ph10 351 stage table and inline lookup instead of a function, giving speed ups of 2
563     to 5 times on some simple patterns that I tested. Permission was given to
564     distribute the MultiStage2.py script that generates the tables (it's not in
565     the tarball, but is in the Subversion repository).
566 ph10 349
567 ph10 351 2. Updated the Unicode datatables to Unicode 5.1.0. This adds yet more
568     scripts.
569 ph10 358
570 ph10 353 3. Change 12 for 7.7 introduced a bug in pcre_study() when a pattern contained
571     a group with a zero qualifier. The result of the study could be incorrect,
572 ph10 358 or the function might crash, depending on the pattern.
573    
574     4. Caseless matching was not working for non-ASCII characters in back
575 ph10 354 references. For example, /(\x{de})\1/8i was not matching \x{de}\x{fe}.
576 ph10 358 It now works when Unicode Property Support is available.
577    
578 ph10 355 5. In pcretest, an escape such as \x{de} in the data was always generating
579     a UTF-8 string, even in non-UTF-8 mode. Now it generates a single byte in
580     non-UTF-8 mode. If the value is greater than 255, it gives a warning about
581 ph10 358 truncation.
582 ph10 349
583 ph10 358 6. Minor bugfix in pcrecpp.cc (change "" == ... to NULL == ...).
584    
585 ph10 357 7. Added two (int) casts to pcregrep when printing the difference of two
586     pointers, in case they are 64-bit values.
587 ph10 371
588     8. Added comments about Mac OS X stack usage to the pcrestack man page and to
589     test 2 if it fails.
590    
591 ph10 359 9. Added PCRE_CALL_CONVENTION just before the names of all exported functions,
592 ph10 371 and a #define of that name to empty if it is not externally set. This is to
593     allow users of MSVC to set it if necessary.
594    
595     10. The PCRE_EXP_DEFN macro which precedes exported functions was missing from
596 ph10 359 the convenience functions in the pcre_get.c source file.
597 ph10 371
598 ph10 360 11. An option change at the start of a pattern that had top-level alternatives
599     could cause overwriting and/or a crash. This command provoked a crash in
600 ph10 371 some environments:
601    
602     printf "/(?i)[\xc3\xa9\xc3\xbd]|[\xc3\xa9\xc3\xbdA]/8\n" | pcretest
603    
604 ph10 360 This potential security problem was recorded as CVE-2008-2371.
605 ph10 371
606 ph10 361 12. For a pattern where the match had to start at the beginning or immediately
607     after a newline (e.g /.*anything/ without the DOTALL flag), pcre_exec() and
608 ph10 371 pcre_dfa_exec() could read past the end of the passed subject if there was
609 ph10 361 no match. To help with detecting such bugs (e.g. with valgrind), I modified
610     pcretest so that it places the subject at the end of its malloc-ed buffer.
611 ph10 371
612 ph10 364 13. The change to pcretest in 12 above threw up a couple more cases when pcre_
613 ph10 371 exec() might read past the end of the data buffer in UTF-8 mode.
614    
615 ph10 364 14. A similar bug to 7.3/2 existed when the PCRE_FIRSTLINE option was set and
616 ph10 371 the data contained the byte 0x85 as part of a UTF-8 character within its
617     first line. This applied both to normal and DFA matching.
618    
619 ph10 366 15. Lazy qualifiers were not working in some cases in UTF-8 mode. For example,
620 ph10 371 /^[^d]*?$/8 failed to match "abc".
621 ph10 351
622 ph10 371 16. Added a missing copyright notice to pcrecpp_internal.h.
623    
624     17. Make it more clear in the documentation that values returned from
625 ph10 368 pcre_exec() in ovector are byte offsets, not character counts.
626 ph10 357
627 ph10 371 18. Tidied a few places to stop certain compilers from issuing warnings.
628 ph10 368
629 ph10 373 19. Updated the Virtual Pascal + BCC files to compile the latest v7.7, as
630     supplied by Stefan Weber. I made a further small update for 7.8 because
631 ph10 374 there is a change of source arrangements: the pcre_searchfuncs.c module is
632     replaced by pcre_ucd.c.
633 ph10 371
634 ph10 373
635 ph10 347 Version 7.7 07-May-08
636 ph10 321 ---------------------
637    
638     1. Applied Craig's patch to sort out a long long problem: "If we can't convert
639 ph10 345 a string to a long long, pretend we don't even have a long long." This is
640 ph10 321 done by checking for the strtoq, strtoll, and _strtoi64 functions.
641 ph10 345
642 ph10 322 2. Applied Craig's patch to pcrecpp.cc to restore ABI compatibility with
643     pre-7.6 versions, which defined a global no_arg variable instead of putting
644 ph10 332 it in the RE class. (See also #8 below.)
645 ph10 345
646     3. Remove a line of dead code, identified by coverity and reported by Nuno
647     Lopes.
648    
649 ph10 324 4. Fixed two related pcregrep bugs involving -r with --include or --exclude:
650 ph10 321
651 ph10 324 (1) The include/exclude patterns were being applied to the whole pathnames
652 ph10 345 of files, instead of just to the final components.
653    
654 ph10 324 (2) If there was more than one level of directory, the subdirectories were
655     skipped unless they satisfied the include/exclude conditions. This is
656     inconsistent with GNU grep (and could even be seen as contrary to the
657     pcregrep specification - which I improved to make it absolutely clear).
658     The action now is always to scan all levels of directory, and just
659     apply the include/exclude patterns to regular files.
660 ph10 345
661 ph10 325 5. Added the --include_dir and --exclude_dir patterns to pcregrep, and used
662 ph10 345 --exclude_dir in the tests to avoid scanning .svn directories.
663    
664 ph10 326 6. Applied Craig's patch to the QuoteMeta function so that it escapes the
665 ph10 345 NUL character as backslash + 0 rather than backslash + NUL, because PCRE
666 ph10 326 doesn't support NULs in patterns.
667 ph10 345
668     7. Added some missing "const"s to declarations of static tables in
669     pcre_compile.c and pcre_dfa_exec.c.
670    
671 ph10 328 8. Applied Craig's patch to pcrecpp.cc to fix a problem in OS X that was
672 ph10 329 caused by fix #2 above. (Subsequently also a second patch to fix the
673 ph10 345 first patch. And a third patch - this was a messy problem.)
674 ph10 321
675 ph10 345 9. Applied Craig's patch to remove the use of push_back().
676    
677 ph10 332 10. Applied Alan Lehotsky's patch to add REG_STARTEND support to the POSIX
678     matching function regexec().
679 ph10 345
680     11. Added support for the Oniguruma syntax \g<name>, \g<n>, \g'name', \g'n',
681     which, however, unlike Perl's \g{...}, are subroutine calls, not back
682 ph10 333 references. PCRE supports relative numbers with this syntax (I don't think
683     Oniguruma does).
684 ph10 345
685     12. Previously, a group with a zero repeat such as (...){0} was completely
686 ph10 335 omitted from the compiled regex. However, this means that if the group
687     was called as a subroutine from elsewhere in the pattern, things went wrong
688 ph10 345 (an internal error was given). Such groups are now left in the compiled
689     pattern, with a new opcode that causes them to be skipped at execution
690 ph10 335 time.
691 ph10 345
692 ph10 341 13. Added the PCRE_JAVASCRIPT_COMPAT option. This makes the following changes
693     to the way PCRE behaves:
694 ph10 345
695 ph10 336 (a) A lone ] character is dis-allowed (Perl treats it as data).
696 ph10 345
697     (b) A back reference to an unmatched subpattern matches an empty string
698 ph10 336 (Perl fails the current match path).
699 ph10 345
700 ph10 341 (c) A data ] in a character class must be notated as \] because if the
701 ph10 345 first data character in a class is ], it defines an empty class. (In
702 ph10 341 Perl it is not possible to have an empty class.) The empty class []
703     never matches; it forces failure and is equivalent to (*FAIL) or (?!).
704     The negative empty class [^] matches any one character, independently
705     of the DOTALL setting.
706 ph10 345
707     14. A pattern such as /(?2)[]a()b](abc)/ which had a forward reference to a
708 ph10 340 non-existent subpattern following a character class starting with ']' and
709     containing () gave an internal compiling error instead of "reference to
710     non-existent subpattern". Fortunately, when the pattern did exist, the
711 ph10 345 compiled code was correct. (When scanning forwards to check for the
712 ph10 340 existencd of the subpattern, it was treating the data ']' as terminating
713 ph10 345 the class, so got the count wrong. When actually compiling, the reference
714 ph10 340 was subsequently set up correctly.)
715 ph10 345
716 ph10 341 15. The "always fail" assertion (?!) is optimzed to (*FAIL) by pcre_compile;
717 ph10 345 it was being rejected as not supported by pcre_dfa_exec(), even though
718     other assertions are supported. I have made pcre_dfa_exec() support
719     (*FAIL).
720    
721 ph10 342 16. The implementation of 13c above involved the invention of a new opcode,
722 ph10 345 OP_ALLANY, which is like OP_ANY but doesn't check the /s flag. Since /s
723     cannot be changed at match time, I realized I could make a small
724     improvement to matching performance by compiling OP_ALLANY instead of
725     OP_ANY for "." when DOTALL was set, and then removing the runtime tests
726     on the OP_ANY path.
727    
728     17. Compiling pcretest on Windows with readline support failed without the
729 ph10 343 following two fixes: (1) Make the unistd.h include conditional on
730 ph10 345 HAVE_UNISTD_H; (2) #define isatty and fileno as _isatty and _fileno.
731    
732 ph10 344 18. Changed CMakeLists.txt and cmake/FindReadline.cmake to arrange for the
733 ph10 345 ncurses library to be included for pcretest when ReadLine support is
734     requested, but also to allow for it to be overridden. This patch came from
735     Daniel Bergström.
736 ph10 347
737 ph10 346 19. There was a typo in the file ucpinternal.h where f0_rangeflag was defined
738 ph10 347 as 0x00f00000 instead of 0x00800000. Luckily, this would not have caused
739     any errors with the current Unicode tables. Thanks to Peter Kankowski for
740     spotting this.
741 ph10 324
742 ph10 332
743 ph10 319 Version 7.6 28-Jan-08
744 ph10 300 ---------------------
745    
746 ph10 302 1. A character class containing a very large number of characters with
747     codepoints greater than 255 (in UTF-8 mode, of course) caused a buffer
748     overflow.
749 ph10 309
750     2. Patch to cut out the "long long" test in pcrecpp_unittest when
751     HAVE_LONG_LONG is not defined.
752    
753 ph10 303 3. Applied Christian Ehrlicher's patch to update the CMake build files to
754 ph10 304 bring them up to date and include new features. This patch includes:
755 ph10 309
756 ph10 304 - Fixed PH's badly added libz and libbz2 support.
757     - Fixed a problem with static linking.
758 ph10 312 - Added pcredemo. [But later removed - see 7 below.]
759 ph10 304 - Fixed dftables problem and added an option.
760     - Added a number of HAVE_XXX tests, including HAVE_WINDOWS_H and
761     HAVE_LONG_LONG.
762     - Added readline support for pcretest.
763 ph10 309 - Added an listing of the option settings after cmake has run.
764    
765 ph10 314 4. A user submitted a patch to Makefile that makes it easy to create
766     "pcre.dll" under mingw when using Configure/Make. I added stuff to
767     Makefile.am that cause it to include this special target, without
768     affecting anything else. Note that the same mingw target plus all
769     the other distribution libraries and programs are now supported
770     when configuring with CMake (see 6 below) instead of with
771     Configure/Make.
772 ph10 309
773 ph10 308 5. Applied Craig's patch that moves no_arg into the RE class in the C++ code.
774     This is an attempt to solve the reported problem "pcrecpp::no_arg is not
775 ph10 309 exported in the Windows port". It has not yet been confirmed that the patch
776     solves the problem, but it does no harm.
777 ph10 313
778 ph10 311 6. Applied Sheri's patch to CMakeLists.txt to add NON_STANDARD_LIB_PREFIX and
779 ph10 319 NON_STANDARD_LIB_SUFFIX for dll names built with mingw when configured
780 ph10 314 with CMake, and also correct the comment about stack recursion.
781 ph10 313
782 ph10 312 7. Remove the automatic building of pcredemo from the ./configure system and
783 ph10 313 from CMakeLists.txt. The whole idea of pcredemo.c is that it is an example
784     of a program that users should build themselves after PCRE is installed, so
785     building it automatically is not really right. What is more, it gave
786 ph10 312 trouble in some build environments.
787 ph10 300
788 ph10 319 8. Further tidies to CMakeLists.txt from Sheri and Christian.
789 ph10 308
790 ph10 319
791 ph10 298 Version 7.5 10-Jan-08
792 ph10 263 ---------------------
793    
794     1. Applied a patch from Craig: "This patch makes it possible to 'ignore'
795     values in parens when parsing an RE using the C++ wrapper."
796 ph10 286
797 ph10 264 2. Negative specials like \S did not work in character classes in UTF-8 mode.
798     Characters greater than 255 were excluded from the class instead of being
799     included.
800 ph10 286
801     3. The same bug as (2) above applied to negated POSIX classes such as
802 ph10 265 [:^space:].
803 ph10 286
804 ph10 267 4. PCRECPP_STATIC was referenced in pcrecpp_internal.h, but nowhere was it
805     defined or documented. It seems to have been a typo for PCRE_STATIC, so
806 ph10 286 I have changed it.
807    
808 ph10 268 5. The construct (?&) was not diagnosed as a syntax error (it referenced the
809 ph10 286 first named subpattern) and a construct such as (?&a) would reference the
810     first named subpattern whose name started with "a" (in other words, the
811 ph10 272 length check was missing). Both these problems are fixed. "Subpattern name
812     expected" is now given for (?&) (a zero-length name), and this patch also
813     makes it give the same error for \k'' (previously it complained that that
814     was a reference to a non-existent subpattern).
815 ph10 286
816 ph10 269 6. The erroneous patterns (?+-a) and (?-+a) give different error messages;
817 ph10 286 this is right because (?- can be followed by option settings as well as by
818 ph10 269 digits. I have, however, made the messages clearer.
819 ph10 286
820 ph10 270 7. Patterns such as (?(1)a|b) (a pattern that contains fewer subpatterns
821     than the number used in the conditional) now cause a compile-time error.
822     This is actually not compatible with Perl, which accepts such patterns, but
823     treats the conditional as always being FALSE (as PCRE used to), but it
824     seems to me that giving a diagnostic is better.
825 ph10 286
826 ph10 274 8. Change "alphameric" to the more common word "alphanumeric" in comments
827 ph10 275 and messages.
828 ph10 286
829     9. Fix two occurrences of "backslash" in comments that should have been
830     "backspace".
831    
832     10. Remove two redundant lines of code that can never be obeyed (their function
833     was moved elsewhere).
834    
835     11. The program that makes PCRE's Unicode character property table had a bug
836     which caused it to generate incorrect table entries for sequences of
837 ph10 277 characters that have the same character type, but are in different scripts.
838 ph10 286 It amalgamated them into a single range, with the script of the first of
839 ph10 277 them. In other words, some characters were in the wrong script. There were
840     thirteen such cases, affecting characters in the following ranges:
841 ph10 286
842 ph10 277 U+002b0 - U+002c1
843     U+0060c - U+0060d
844 ph10 286 U+0061e - U+00612
845 ph10 277 U+0064b - U+0065e
846     U+0074d - U+0076d
847     U+01800 - U+01805
848     U+01d00 - U+01d77
849     U+01d9b - U+01dbf
850     U+0200b - U+0200f
851     U+030fc - U+030fe
852     U+03260 - U+0327f
853     U+0fb46 - U+0fbb1
854     U+10450 - U+1049d
855 ph10 286
856 ph10 279 12. The -o option (show only the matching part of a line) for pcregrep was not
857     compatible with GNU grep in that, if there was more than one match in a
858     line, it showed only the first of them. It now behaves in the same way as
859     GNU grep.
860 ph10 286
861     13. If the -o and -v options were combined for pcregrep, it printed a blank
862     line for every non-matching line. GNU grep prints nothing, and pcregrep now
863     does the same. The return code can be used to tell if there were any
864     non-matching lines.
865    
866 ph10 289 14. Added --file-offsets and --line-offsets to pcregrep.
867    
868     15. The pattern (?=something)(?R) was not being diagnosed as a potentially
869 ph10 282 infinitely looping recursion. The bug was that positive lookaheads were not
870 ph10 286 being skipped when checking for a possible empty match (negative lookaheads
871     and both kinds of lookbehind were skipped).
872    
873 ph10 289 16. Fixed two typos in the Windows-only code in pcregrep.c, and moved the
874 ph10 284 inclusion of <windows.h> to before rather than after the definition of
875     INVALID_FILE_ATTRIBUTES (patch from David Byron).
876 ph10 263
877 ph10 289 17. Specifying a possessive quantifier with a specific limit for a Unicode
878 ph10 286 character property caused pcre_compile() to compile bad code, which led at
879     runtime to PCRE_ERROR_INTERNAL (-14). Examples of patterns that caused this
880     are: /\p{Zl}{2,3}+/8 and /\p{Cc}{2}+/8. It was the possessive "+" that
881     caused the error; without that there was no problem.
882 ph10 263
883 ph10 289 18. Added --enable-pcregrep-libz and --enable-pcregrep-libbz2.
884 ph10 286
885 ph10 289 19. Added --enable-pcretest-libreadline.
886 ph10 286
887 ph10 289 20. In pcrecpp.cc, the variable 'count' was incremented twice in
888 ph10 288 RE::GlobalReplace(). As a result, the number of replacements returned was
889 ph10 298 double what it should be. I removed one of the increments, but Craig sent a
890     later patch that removed the other one (the right fix) and added unit tests
891     that check the return values (which was not done before).
892 ph10 292
893 ph10 291 21. Several CMake things:
894 ph10 287
895 ph10 291 (1) Arranged that, when cmake is used on Unix, the libraries end up with
896 ph10 292 the names libpcre and libpcreposix, not just pcre and pcreposix.
897 ph10 288
898 ph10 292 (2) The above change means that pcretest and pcregrep are now correctly
899     linked with the newly-built libraries, not previously installed ones.
900 ph10 291
901 ph10 292 (3) Added PCRE_SUPPORT_LIBREADLINE, PCRE_SUPPORT_LIBZ, PCRE_SUPPORT_LIBBZ2.
902 ph10 298
903     22. In UTF-8 mode, with newline set to "any", a pattern such as .*a.*=.b.*
904     crashed when matching a string such as a\x{2029}b (note that \x{2029} is a
905     UTF-8 newline character). The key issue is that the pattern starts .*;
906     this means that the match must be either at the beginning, or after a
907     newline. The bug was in the code for advancing after a failed match and
908     checking that the new position followed a newline. It was not taking
909 ph10 294 account of UTF-8 characters correctly.
910 ph10 298
911     23. PCRE was behaving differently from Perl in the way it recognized POSIX
912     character classes. PCRE was not treating the sequence [:...:] as a
913     character class unless the ... were all letters. Perl, however, seems to
914     allow any characters between [: and :], though of course it rejects as
915     unknown any "names" that contain non-letters, because all the known class
916     names consist only of letters. Thus, Perl gives an error for [[:1234:]],
917     for example, whereas PCRE did not - it did not recognize a POSIX character
918     class. This seemed a bit dangerous, so the code has been changed to be
919     closer to Perl. The behaviour is not identical to Perl, because PCRE will
920     diagnose an unknown class for, for example, [[:l\ower:]] where Perl will
921     treat it as [[:lower:]]. However, PCRE does now give "unknown" errors where
922     Perl does, and where it didn't before.
923    
924 ph10 296 24. Rewrite so as to remove the single use of %n from pcregrep because in some
925     Windows environments %n is disabled by default.
926 ph10 292
927    
928 ph10 260 Version 7.4 21-Sep-07
929 ph10 230 ---------------------
930    
931     1. Change 7.3/28 was implemented for classes by looking at the bitmap. This
932 ph10 231 means that a class such as [\s] counted as "explicit reference to CR or
933     LF". That isn't really right - the whole point of the change was to try to
934     help when there was an actual mention of one of the two characters. So now
935     the change happens only if \r or \n (or a literal CR or LF) character is
936 ph10 230 encountered.
937 ph10 231
938     2. The 32-bit options word was also used for 6 internal flags, but the numbers
939     of both had grown to the point where there were only 3 bits left.
940     Fortunately, there was spare space in the data structure, and so I have
941     moved the internal flags into a new 16-bit field to free up more option
942 ph10 230 bits.
943 ph10 231
944     3. The appearance of (?J) at the start of a pattern set the DUPNAMES option,
945     but did not set the internal JCHANGED flag - either of these is enough to
946     control the way the "get" function works - but the PCRE_INFO_JCHANGED
947     facility is supposed to tell if (?J) was ever used, so now (?J) at the
948 ph10 230 start sets both bits.
949    
950 ph10 231 4. Added options (at build time, compile time, exec time) to change \R from
951     matching any Unicode line ending sequence to just matching CR, LF, or CRLF.
952 ph10 230
953 ph10 243 5. doc/pcresyntax.html was missing from the distribution.
954    
955     6. Put back the definition of PCRE_ERROR_NULLWSLIMIT, for backward
956 ph10 233 compatibility, even though it is no longer used.
957 ph10 243
958     7. Added macro for snprintf to pcrecpp_unittest.cc and also for strtoll and
959 ph10 254 strtoull to pcrecpp.cc to select the available functions in WIN32 when the
960 ph10 259 windows.h file is present (where different names are used). [This was
961 ph10 257 reversed later after testing - see 16 below.]
962 ph10 243
963     8. Changed all #include <config.h> to #include "config.h". There were also
964 ph10 236 some further <pcre.h> cases that I changed to "pcre.h".
965 ph10 243
966 ph10 239 9. When pcregrep was used with the --colour option, it missed the line ending
967 ph10 243 sequence off the lines that it output.
968    
969     10. It was pointed out to me that arrays of string pointers cause lots of
970     relocations when a shared library is dynamically loaded. A technique of
971     using a single long string with a table of offsets can drastically reduce
972     these. I have refactored PCRE in four places to do this. The result is
973 ph10 240 dramatic:
974 ph10 243
975 ph10 240 Originally: 290
976     After changing UCP table: 187
977 ph10 243 After changing error message table: 43
978 ph10 240 After changing table of "verbs" 36
979     After changing table of Posix names 22
980 ph10 243
981 ph10 240 Thanks to the folks working on Gregex for glib for this insight.
982 ph10 247
983 ph10 244 11. --disable-stack-for-recursion caused compiling to fail unless -enable-
984 ph10 247 unicode-properties was also set.
985    
986 ph10 248 12. Updated the tests so that they work when \R is defaulted to ANYCRLF.
987 ph10 231
988 ph10 253 13. Added checks for ANY and ANYCRLF to pcrecpp.cc where it previously
989     checked only for CRLF.
990 ph10 233
991 ph10 259 14. Added casts to pcretest.c to avoid compiler warnings.
992    
993 ph10 256 15. Added Craig's patch to various pcrecpp modules to avoid compiler warnings.
994 ph10 253
995 ph10 257 16. Added Craig's patch to remove the WINDOWS_H tests, that were not working,
996 ph10 259 and instead check for _strtoi64 explicitly, and avoid the use of snprintf()
997     entirely. This removes changes made in 7 above.
998 ph10 256
999 ph10 261 17. The CMake files have been updated, and there is now more information about
1000     building with CMake in the NON-UNIX-USE document.
1001 ph10 257
1002 ph10 261
1003 ph10 228 Version 7.3 28-Aug-07
1004 ph10 157 ---------------------
1005    
1006 ph10 189 1. In the rejigging of the build system that eventually resulted in 7.1, the
1007     line "#include <pcre.h>" was included in pcre_internal.h. The use of angle
1008     brackets there is not right, since it causes compilers to look for an
1009     installed pcre.h, not the version that is in the source that is being
1010     compiled (which of course may be different). I have changed it back to:
1011 ph10 197
1012 ph10 189 #include "pcre.h"
1013 ph10 197
1014     I have a vague recollection that the change was concerned with compiling in
1015     different directories, but in the new build system, that is taken care of
1016     by the VPATH setting the Makefile.
1017    
1018 ph10 190 2. The pattern .*$ when run in not-DOTALL UTF-8 mode with newline=any failed
1019     when the subject happened to end in the byte 0x85 (e.g. if the last
1020     character was \x{1ec5}). *Character* 0x85 is one of the "any" newline
1021     characters but of course it shouldn't be taken as a newline when it is part
1022     of another character. The bug was that, for an unlimited repeat of . in
1023     not-DOTALL UTF-8 mode, PCRE was advancing by bytes rather than by
1024     characters when looking for a newline.
1025 ph10 197
1026     3. A small performance improvement in the DOTALL UTF-8 mode .* case.
1027    
1028     4. Debugging: adjusted the names of opcodes for different kinds of parentheses
1029     in debug output.
1030    
1031 ph10 201 5. Arrange to use "%I64d" instead of "%lld" and "%I64u" instead of "%llu" for
1032     long printing in the pcrecpp unittest when running under MinGW.
1033 ph10 189
1034 ph10 197 6. ESC_K was left out of the EBCDIC table.
1035 ph10 189
1036 ph10 197 7. Change 7.0/38 introduced a new limit on the number of nested non-capturing
1037     parentheses; I made it 1000, which seemed large enough. Unfortunately, the
1038     limit also applies to "virtual nesting" when a pattern is recursive, and in
1039     this case 1000 isn't so big. I have been able to remove this limit at the
1040     expense of backing off one optimization in certain circumstances. Normally,
1041     when pcre_exec() would call its internal match() function recursively and
1042     immediately return the result unconditionally, it uses a "tail recursion"
1043     feature to save stack. However, when a subpattern that can match an empty
1044     string has an unlimited repetition quantifier, it no longer makes this
1045     optimization. That gives it a stack frame in which to save the data for
1046     checking that an empty string has been matched. Previously this was taken
1047     from the 1000-entry workspace that had been reserved. So now there is no
1048     explicit limit, but more stack is used.
1049 ph10 189
1050 ph10 200 8. Applied Daniel's patches to solve problems with the import/export magic
1051     syntax that is required for Windows, and which was going wrong for the
1052     pcreposix and pcrecpp parts of the library. These were overlooked when this
1053     problem was solved for the main library.
1054 ph10 197
1055 ph10 202 9. There were some crude static tests to avoid integer overflow when computing
1056     the size of patterns that contain repeated groups with explicit upper
1057     limits. As the maximum quantifier is 65535, the maximum group length was
1058     set at 30,000 so that the product of these two numbers did not overflow a
1059     32-bit integer. However, it turns out that people want to use groups that
1060     are longer than 30,000 bytes (though not repeat them that many times).
1061     Change 7.0/17 (the refactoring of the way the pattern size is computed) has
1062     made it possible to implement the integer overflow checks in a much more
1063     dynamic way, which I have now done. The artificial limitation on group
1064     length has been removed - we now have only the limit on the total length of
1065     the compiled pattern, which depends on the LINK_SIZE setting.
1066 ph10 208
1067     10. Fixed a bug in the documentation for get/copy named substring when
1068     duplicate names are permitted. If none of the named substrings are set, the
1069     functions return PCRE_ERROR_NOSUBSTRING (7); the doc said they returned an
1070     empty string.
1071    
1072     11. Because Perl interprets \Q...\E at a high level, and ignores orphan \E
1073     instances, patterns such as [\Q\E] or [\E] or even [^\E] cause an error,
1074     because the ] is interpreted as the first data character and the
1075     terminating ] is not found. PCRE has been made compatible with Perl in this
1076     regard. Previously, it interpreted [\Q\E] as an empty class, and [\E] could
1077     cause memory overwriting.
1078    
1079 ph10 206 10. Like Perl, PCRE automatically breaks an unlimited repeat after an empty
1080     string has been matched (to stop an infinite loop). It was not recognizing
1081 ph10 208 a conditional subpattern that could match an empty string if that
1082 ph10 206 subpattern was within another subpattern. For example, it looped when
1083 ph10 208 trying to match (((?(1)X|))*) but it was OK with ((?(1)X|)*) where the
1084 ph10 206 condition was not nested. This bug has been fixed.
1085 ph10 208
1086 ph10 207 12. A pattern like \X?\d or \P{L}?\d in non-UTF-8 mode could cause a backtrack
1087     past the start of the subject in the presence of bytes with the top bit
1088     set, for example "\x8aBCD".
1089 ph10 211
1090 ph10 210 13. Added Perl 5.10 experimental backtracking controls (*FAIL), (*F), (*PRUNE),
1091     (*SKIP), (*THEN), (*COMMIT), and (*ACCEPT).
1092 ph10 200
1093 ph10 211 14. Optimized (?!) to (*FAIL).
1094 ph10 202
1095 ph10 212 15. Updated the test for a valid UTF-8 string to conform to the later RFC 3629.
1096     This restricts code points to be within the range 0 to 0x10FFFF, excluding
1097     the "low surrogate" sequence 0xD800 to 0xDFFF. Previously, PCRE allowed the
1098     full range 0 to 0x7FFFFFFF, as defined by RFC 2279. Internally, it still
1099     does: it's just the validity check that is more restrictive.
1100 ph10 220
1101     16. Inserted checks for integer overflows during escape sequence (backslash)
1102     processing, and also fixed erroneous offset values for syntax errors during
1103     backslash processing.
1104    
1105 ph10 214 17. Fixed another case of looking too far back in non-UTF-8 mode (cf 12 above)
1106 ph10 220 for patterns like [\PPP\x8a]{1,}\x80 with the subject "A\x80".
1107    
1108 ph10 215 18. An unterminated class in a pattern like (?1)\c[ with a "forward reference"
1109     caused an overrun.
1110 ph10 220
1111     19. A pattern like (?:[\PPa*]*){8,} which had an "extended class" (one with
1112     something other than just ASCII characters) inside a group that had an
1113     unlimited repeat caused a loop at compile time (while checking to see
1114     whether the group could match an empty string).
1115    
1116 ph10 217 20. Debugging a pattern containing \p or \P could cause a crash. For example,
1117     [\P{Any}] did so. (Error in the code for printing property names.)
1118 ph10 210
1119 ph10 220 21. An orphan \E inside a character class could cause a crash.
1120    
1121     22. A repeated capturing bracket such as (A)? could cause a wild memory
1122 ph10 218 reference during compilation.
1123 ph10 220
1124     23. There are several functions in pcre_compile() that scan along a compiled
1125     expression for various reasons (e.g. to see if it's fixed length for look
1126 ph10 218 behind). There were bugs in these functions when a repeated \p or \P was
1127 ph10 220 present in the pattern. These operators have additional parameters compared
1128     with \d, etc, and these were not being taken into account when moving along
1129 ph10 218 the compiled data. Specifically:
1130 ph10 220
1131     (a) A item such as \p{Yi}{3} in a lookbehind was not treated as fixed
1132     length.
1133    
1134     (b) An item such as \pL+ within a repeated group could cause crashes or
1135 ph10 218 loops.
1136 ph10 220
1137 ph10 218 (c) A pattern such as \p{Yi}+(\P{Yi}+)(?1) could give an incorrect
1138 ph10 220 "reference to non-existent subpattern" error.
1139    
1140 ph10 221 (d) A pattern like (\P{Yi}{2}\277)? could loop at compile time.
1141    
1142 ph10 219 24. A repeated \S or \W in UTF-8 mode could give wrong answers when multibyte
1143 ph10 220 characters were involved (for example /\S{2}/8g with "A\x{a3}BC").
1144 ph10 211
1145 ph10 222 25. Using pcregrep in multiline, inverted mode (-Mv) caused it to loop.
1146 ph10 218
1147 ph10 227 26. Patterns such as [\P{Yi}A] which include \p or \P and just one other
1148 ph10 223 character were causing crashes (broken optimization).
1149 ph10 227
1150 ph10 224 27. Patterns such as (\P{Yi}*\277)* (group with possible zero repeat containing
1151 ph10 227 \p or \P) caused a compile-time loop.
1152    
1153 ph10 226 28. More problems have arisen in unanchored patterns when CRLF is a valid line
1154     break. For example, the unstudied pattern [\r\n]A does not match the string
1155     "\r\nA" because change 7.0/46 below moves the current point on by two
1156     characters after failing to match at the start. However, the pattern \nA
1157 ph10 227 *does* match, because it doesn't start till \n, and if [\r\n]A is studied,
1158     the same is true. There doesn't seem any very clean way out of this, but
1159     what I have chosen to do makes the common cases work: PCRE now takes note
1160     of whether there can be an explicit match for \r or \n anywhere in the
1161     pattern, and if so, 7.0/46 no longer applies. As part of this change,
1162     there's a new PCRE_INFO_HASCRORLF option for finding out whether a compiled
1163     pattern has explicit CR or LF references.
1164 ph10 222
1165 ph10 227 29. Added (*CR) etc for changing newline setting at start of pattern.
1166 ph10 223
1167 ph10 227
1168 ph10 189 Version 7.2 19-Jun-07
1169     ---------------------
1170    
1171 ph10 157 1. If the fr_FR locale cannot be found for test 3, try the "french" locale,
1172     which is apparently normally available under Windows.
1173 ph10 159
1174     2. Re-jig the pcregrep tests with different newline settings in an attempt
1175     to make them independent of the local environment's newline setting.
1176 ph10 157
1177 ph10 160 3. Add code to configure.ac to remove -g from the CFLAGS default settings.
1178 ph10 165
1179 ph10 161 4. Some of the "internals" tests were previously cut out when the link size
1180     was not 2, because the output contained actual offsets. The recent new
1181 ph10 165 "Z" feature of pcretest means that these can be cut out, making the tests
1182     usable with all link sizes.
1183    
1184 ph10 164 5. Implemented Stan Switzer's goto replacement for longjmp() when not using
1185     stack recursion. This gives a massive performance boost under BSD, but just
1186 ph10 165 a small improvement under Linux. However, it saves one field in the frame
1187 ph10 164 in all cases.
1188 ph10 172
1189 ph10 166 6. Added more features from the forthcoming Perl 5.10:
1190 ph10 172
1191 ph10 166 (a) (?-n) (where n is a string of digits) is a relative subroutine or
1192     recursion call. It refers to the nth most recently opened parentheses.
1193 ph10 172
1194 ph10 166 (b) (?+n) is also a relative subroutine call; it refers to the nth next
1195 ph10 172 to be opened parentheses.
1196    
1197     (c) Conditions that refer to capturing parentheses can be specified
1198 ph10 167 relatively, for example, (?(-2)... or (?(+3)...
1199 ph10 172
1200 ph10 168 (d) \K resets the start of the current match so that everything before
1201 ph10 172 is not part of it.
1202    
1203 ph10 171 (e) \k{name} is synonymous with \k<name> and \k'name' (.NET compatible).
1204 ph10 172
1205 ph10 171 (f) \g{name} is another synonym - part of Perl 5.10's unification of
1206 ph10 172 reference syntax.
1207 ph10 182
1208 ph10 175 (g) (?| introduces a group in which the numbering of parentheses in each
1209 ph10 182 alternative starts with the same number.
1210 ph10 172
1211 ph10 182 (h) \h, \H, \v, and \V match horizontal and vertical whitespace.
1212    
1213 ph10 172 7. Added two new calls to pcre_fullinfo(): PCRE_INFO_OKPARTIAL and
1214     PCRE_INFO_JCHANGED.
1215    
1216     8. A pattern such as (.*(.)?)* caused pcre_exec() to fail by either not
1217     terminating or by crashing. Diagnosed by Viktor Griph; it was in the code
1218 ph10 170 for detecting groups that can match an empty string.
1219 ph10 159
1220 ph10 172 9. A pattern with a very large number of alternatives (more than several
1221     hundred) was running out of internal workspace during the pre-compile
1222     phase, where pcre_compile() figures out how much memory will be needed. A
1223     bit of new cunning has reduced the workspace needed for groups with
1224     alternatives. The 1000-alternative test pattern now uses 12 bytes of
1225     workspace instead of running out of the 4096 that are available.
1226 ph10 182
1227 ph10 176 10. Inserted some missing (unsigned int) casts to get rid of compiler warnings.
1228 ph10 172
1229 ph10 179 11. Applied patch from Google to remove an optimization that didn't quite work.
1230     The report of the bug said:
1231 ph10 182
1232 ph10 179 pcrecpp::RE("a*").FullMatch("aaa") matches, while
1233     pcrecpp::RE("a*?").FullMatch("aaa") does not, and
1234 ph10 182 pcrecpp::RE("a*?\\z").FullMatch("aaa") does again.
1235 ph10 185
1236 ph10 184 12. If \p or \P was used in non-UTF-8 mode on a character greater than 127
1237 ph10 185 it matched the wrong number of bytes.
1238 ph10 172
1239 ph10 179
1240 ph10 155 Version 7.1 24-Apr-07
1241 ph10 98 ---------------------
1242    
1243 ph10 111 1. Applied Bob Rossi and Daniel G's patches to convert the build system to one
1244 ph10 122 that is more "standard", making use of automake and other Autotools. There
1245 ph10 99 is some re-arrangement of the files and adjustment of comments consequent
1246     on this.
1247 ph10 111
1248     2. Part of the patch fixed a problem with the pcregrep tests. The test of -r
1249     for recursive directory scanning broke on some systems because the files
1250     are not scanned in any specific order and on different systems the order
1251     was different. A call to "sort" has been inserted into RunGrepTest for the
1252     approprate test as a short-term fix. In the longer term there may be an
1253 ph10 100 alternative.
1254 ph10 111
1255 ph10 100 3. I had an email from Eric Raymond about problems translating some of PCRE's
1256 ph10 111 man pages to HTML (despite the fact that I distribute HTML pages, some
1257     people do their own conversions for various reasons). The problems
1258     concerned the use of low-level troff macros .br and .in. I have therefore
1259     removed all such uses from the man pages (some were redundant, some could
1260 ph10 113 be replaced by .nf/.fi pairs). The 132html script that I use to generate
1261     HTML has been updated to handle .nf/.fi and to complain if it encounters
1262     .br or .in.
1263 ph10 111
1264 ph10 100 4. Updated comments in configure.ac that get placed in config.h.in and also
1265 ph10 123 arranged for config.h to be included in the distribution, with the name
1266 ph10 111 config.h.generic, for the benefit of those who have to compile without
1267     Autotools (compare pcre.h, which is now distributed as pcre.h.generic).
1268    
1269     5. Updated the support (such as it is) for Virtual Pascal, thanks to Stefan
1270     Weber: (1) pcre_internal.h was missing some function renames; (2) updated
1271 ph10 127 makevp.bat for the current PCRE, using the additional files
1272 ph10 135 makevp_c.txt, makevp_l.txt, and pcregexp.pas.
1273 ph10 111
1274     6. A Windows user reported a minor discrepancy with test 2, which turned out
1275     to be caused by a trailing space on an input line that had got lost in his
1276 ph10 102 copy. The trailing space was an accident, so I've just removed it.
1277 ph10 111
1278 ph10 104 7. Add -Wl,-R... flags in pcre-config.in for *BSD* systems, as I'm told
1279 ph10 111 that is needed.
1280    
1281 ph10 105 8. Mark ucp_table (in ucptable.h) and ucp_gentype (in pcre_ucp_searchfuncs.c)
1282     as "const" (a) because they are and (b) because it helps the PHP
1283     maintainers who have recently made a script to detect big data structures
1284 ph10 111 in the php code that should be moved to the .rodata section. I remembered
1285     to update Builducptable as well, so it won't revert if ucptable.h is ever
1286 ph10 105 re-created.
1287 ph10 111
1288     9. Added some extra #ifdef SUPPORT_UTF8 conditionals into pcretest.c,
1289 ph10 107 pcre_printint.src, pcre_compile.c, pcre_study.c, and pcre_tables.c, in
1290 ph10 111 order to be able to cut out the UTF-8 tables in the latter when UTF-8
1291     support is not required. This saves 1.5-2K of code, which is important in
1292 ph10 107 some applications.
1293 ph10 111
1294 ph10 110 Later: more #ifdefs are needed in pcre_ord2utf8.c and pcre_valid_utf8.c
1295 ph10 111 so as not to refer to the tables, even though these functions will never be
1296     called when UTF-8 support is disabled. Otherwise there are problems with a
1297     shared library.
1298    
1299 ph10 118 10. Fixed two bugs in the emulated memmove() function in pcre_internal.h:
1300    
1301     (a) It was defining its arguments as char * instead of void *.
1302    
1303 ph10 123 (b) It was assuming that all moves were upwards in memory; this was true
1304     a long time ago when I wrote it, but is no longer the case.
1305    
1306 ph10 118 The emulated memove() is provided for those environments that have neither
1307 ph10 123 memmove() nor bcopy(). I didn't think anyone used it these days, but that
1308 ph10 118 is clearly not the case, as these two bugs were recently reported.
1309 ph10 123
1310 ph10 111 11. The script PrepareRelease is now distributed: it calls 132html, CleanTxt,
1311 ph10 123 and Detrail to create the HTML documentation, the .txt form of the man
1312     pages, and it removes trailing spaces from listed files. It also creates
1313     pcre.h.generic and config.h.generic from pcre.h and config.h. In the latter
1314     case, it wraps all the #defines with #ifndefs. This script should be run
1315 ph10 111 before "make dist".
1316 ph10 123
1317 ph10 115 12. Fixed two fairly obscure bugs concerned with quantified caseless matching
1318     with Unicode property support.
1319 ph10 123
1320     (a) For a maximizing quantifier, if the two different cases of the
1321     character were of different lengths in their UTF-8 codings (there are
1322     some cases like this - I found 11), and the matching function had to
1323 ph10 115 back up over a mixture of the two cases, it incorrectly assumed they
1324     were both the same length.
1325 ph10 123
1326     (b) When PCRE was configured to use the heap rather than the stack for
1327     recursion during matching, it was not correctly preserving the data for
1328     the other case of a UTF-8 character when checking ahead for a match
1329     while processing a minimizing repeat. If the check also involved
1330 ph10 115 matching a wide character, but failed, corruption could cause an
1331     erroneous result when trying to check for a repeat of the original
1332     character.
1333 ph10 123
1334 ph10 116 13. Some tidying changes to the testing mechanism:
1335 ph10 98
1336 ph10 116 (a) The RunTest script now detects the internal link size and whether there
1337     is UTF-8 and UCP support by running ./pcretest -C instead of relying on
1338 ph10 123 values substituted by "configure". (The RunGrepTest script already did
1339     this for UTF-8.) The configure.ac script no longer substitutes the
1340     relevant variables.
1341    
1342 ph10 116 (b) The debugging options /B and /D in pcretest show the compiled bytecode
1343     with length and offset values. This means that the output is different
1344     for different internal link sizes. Test 2 is skipped for link sizes
1345     other than 2 because of this, bypassing the problem. Unfortunately,
1346     there was also a test in test 3 (the locale tests) that used /B and
1347 ph10 123 failed for link sizes other than 2. Rather than cut the whole test out,
1348     I have added a new /Z option to pcretest that replaces the length and
1349     offset values with spaces. This is now used to make test 3 independent
1350 ph10 122 of link size. (Test 2 will be tidied up later.)
1351 ph10 123
1352     14. If erroroffset was passed as NULL to pcre_compile, it provoked a
1353 ph10 122 segmentation fault instead of returning the appropriate error message.
1354 ph10 134
1355 ph10 131 15. In multiline mode when the newline sequence was set to "any", the pattern
1356 ph10 134 ^$ would give a match between the \r and \n of a subject such as "A\r\nB".
1357     This doesn't seem right; it now treats the CRLF combination as the line
1358     ending, and so does not match in that case. It's only a pattern such as ^$
1359     that would hit this one: something like ^ABC$ would have failed after \r
1360     and then tried again after \r\n.
1361    
1362 ph10 131 16. Changed the comparison command for RunGrepTest from "diff -u" to "diff -ub"
1363 ph10 134 in an attempt to make files that differ only in their line terminators
1364     compare equal. This works on Linux.
1365 ph10 142
1366 ph10 141 17. Under certain error circumstances pcregrep might try to free random memory
1367     as it exited. This is now fixed, thanks to valgrind.
1368 ph10 142
1369 ph10 141 19. In pcretest, if the pattern /(?m)^$/g<any> was matched against the string
1370 ph10 142 "abc\r\n\r\n", it found an unwanted second match after the second \r. This
1371     was because its rules for how to advance for /g after matching an empty
1372 ph10 143 string at the end of a line did not allow for this case. They now check for
1373     it specially.
1374 ph10 150
1375     20. pcretest is supposed to handle patterns and data of any length, by
1376     extending its buffers when necessary. It was getting this wrong when the
1377 ph10 147 buffer for a data line had to be extended.
1378 ph10 150
1379 ph10 149 21. Added PCRE_NEWLINE_ANYCRLF which is like ANY, but matches only CR, LF, or
1380 ph10 150 CRLF as a newline sequence.
1381 ph10 152
1382 ph10 151 22. Code for handling Unicode properties in pcre_dfa_exec() wasn't being cut
1383 ph10 152 out by #ifdef SUPPORT_UCP. This did no harm, as it could never be used, but
1384     I have nevertheless tidied it up.
1385 ph10 111
1386 ph10 152 23. Added some casts to kill warnings from HP-UX ia64 compiler.
1387 ph10 123
1388 ph10 153 24. Added a man page for pcre-config.
1389 ph10 152
1390 ph10 153
1391 nigel 93 Version 7.0 19-Dec-06
1392     ---------------------
1393    
1394     1. Fixed a signed/unsigned compiler warning in pcre_compile.c, shown up by
1395     moving to gcc 4.1.1.
1396    
1397     2. The -S option for pcretest uses setrlimit(); I had omitted to #include
1398     sys/time.h, which is documented as needed for this function. It doesn't
1399     seem to matter on Linux, but it showed up on some releases of OS X.
1400    
1401     3. It seems that there are systems where bytes whose values are greater than
1402     127 match isprint() in the "C" locale. The "C" locale should be the
1403     default when a C program starts up. In most systems, only ASCII printing
1404     characters match isprint(). This difference caused the output from pcretest
1405     to vary, making some of the tests fail. I have changed pcretest so that:
1406    
1407     (a) When it is outputting text in the compiled version of a pattern, bytes
1408     other than 32-126 are always shown as hex escapes.
1409    
1410     (b) When it is outputting text that is a matched part of a subject string,
1411     it does the same, unless a different locale has been set for the match
1412     (using the /L modifier). In this case, it uses isprint() to decide.
1413    
1414     4. Fixed a major bug that caused incorrect computation of the amount of memory
1415     required for a compiled pattern when options that changed within the
1416     pattern affected the logic of the preliminary scan that determines the
1417     length. The relevant options are -x, and -i in UTF-8 mode. The result was
1418     that the computed length was too small. The symptoms of this bug were
1419     either the PCRE error "internal error: code overflow" from pcre_compile(),
1420     or a glibc crash with a message such as "pcretest: free(): invalid next
1421     size (fast)". Examples of patterns that provoked this bug (shown in
1422     pcretest format) are:
1423    
1424     /(?-x: )/x
1425     /(?x)(?-x: \s*#\s*)/
1426     /((?i)[\x{c0}])/8
1427     /(?i:[\x{c0}])/8
1428    
1429     HOWEVER: Change 17 below makes this fix obsolete as the memory computation
1430     is now done differently.
1431    
1432     5. Applied patches from Google to: (a) add a QuoteMeta function to the C++
1433     wrapper classes; (b) implement a new function in the C++ scanner that is
1434     more efficient than the old way of doing things because it avoids levels of
1435     recursion in the regex matching; (c) add a paragraph to the documentation
1436     for the FullMatch() function.
1437    
1438     6. The escape sequence \n was being treated as whatever was defined as
1439     "newline". Not only was this contrary to the documentation, which states
1440     that \n is character 10 (hex 0A), but it also went horribly wrong when
1441     "newline" was defined as CRLF. This has been fixed.
1442    
1443     7. In pcre_dfa_exec.c the value of an unsigned integer (the variable called c)
1444     was being set to -1 for the "end of line" case (supposedly a value that no
1445     character can have). Though this value is never used (the check for end of
1446     line is "zero bytes in current character"), it caused compiler complaints.
1447     I've changed it to 0xffffffff.
1448    
1449     8. In pcre_version.c, the version string was being built by a sequence of
1450     C macros that, in the event of PCRE_PRERELEASE being defined as an empty
1451     string (as it is for production releases) called a macro with an empty
1452     argument. The C standard says the result of this is undefined. The gcc
1453     compiler treats it as an empty string (which was what was wanted) but it is
1454     reported that Visual C gives an error. The source has been hacked around to
1455     avoid this problem.
1456    
1457     9. On the advice of a Windows user, included <io.h> and <fcntl.h> in Windows
1458     builds of pcretest, and changed the call to _setmode() to use _O_BINARY
1459     instead of 0x8000. Made all the #ifdefs test both _WIN32 and WIN32 (not all
1460     of them did).
1461    
1462     10. Originally, pcretest opened its input and output without "b"; then I was
1463     told that "b" was needed in some environments, so it was added for release
1464     5.0 to both the input and output. (It makes no difference on Unix-like
1465     systems.) Later I was told that it is wrong for the input on Windows. I've
1466     now abstracted the modes into two macros, to make it easier to fiddle with
1467     them, and removed "b" from the input mode under Windows.
1468    
1469     11. Added pkgconfig support for the C++ wrapper library, libpcrecpp.
1470    
1471     12. Added -help and --help to pcretest as an official way of being reminded
1472     of the options.
1473    
1474     13. Removed some redundant semicolons after macro calls in pcrecpparg.h.in
1475     and pcrecpp.cc because they annoy compilers at high warning levels.
1476    
1477     14. A bit of tidying/refactoring in pcre_exec.c in the main bumpalong loop.
1478    
1479     15. Fixed an occurrence of == in configure.ac that should have been = (shell
1480     scripts are not C programs :-) and which was not noticed because it works
1481     on Linux.
1482    
1483     16. pcretest is supposed to handle any length of pattern and data line (as one
1484     line or as a continued sequence of lines) by extending its input buffer if
1485     necessary. This feature was broken for very long pattern lines, leading to
1486     a string of junk being passed to pcre_compile() if the pattern was longer
1487     than about 50K.
1488    
1489     17. I have done a major re-factoring of the way pcre_compile() computes the
1490     amount of memory needed for a compiled pattern. Previously, there was code
1491     that made a preliminary scan of the pattern in order to do this. That was
1492     OK when PCRE was new, but as the facilities have expanded, it has become
1493     harder and harder to keep it in step with the real compile phase, and there
1494     have been a number of bugs (see for example, 4 above). I have now found a
1495     cunning way of running the real compile function in a "fake" mode that
1496     enables it to compute how much memory it would need, while actually only
1497     ever using a few hundred bytes of working memory and without too many
1498     tests of the mode. This should make future maintenance and development
1499     easier. A side effect of this work is that the limit of 200 on the nesting
1500     depth of parentheses has been removed (though this was never a serious
1501     limitation, I suspect). However, there is a downside: pcre_compile() now
1502     runs more slowly than before (30% or more, depending on the pattern). I
1503     hope this isn't a big issue. There is no effect on runtime performance.
1504    
1505     18. Fixed a minor bug in pcretest: if a pattern line was not terminated by a
1506     newline (only possible for the last line of a file) and it was a
1507     pattern that set a locale (followed by /Lsomething), pcretest crashed.
1508    
1509     19. Added additional timing features to pcretest. (1) The -tm option now times
1510     matching only, not compiling. (2) Both -t and -tm can be followed, as a
1511     separate command line item, by a number that specifies the number of
1512     repeats to use when timing. The default is 50000; this gives better
1513     precision, but takes uncomfortably long for very large patterns.
1514    
1515     20. Extended pcre_study() to be more clever in cases where a branch of a
1516     subpattern has no definite first character. For example, (a*|b*)[cd] would
1517     previously give no result from pcre_study(). Now it recognizes that the
1518     first character must be a, b, c, or d.
1519    
1520     21. There was an incorrect error "recursive call could loop indefinitely" if
1521     a subpattern (or the entire pattern) that was being tested for matching an
1522     empty string contained only one non-empty item after a nested subpattern.
1523     For example, the pattern (?>\x{100}*)\d(?R) provoked this error
1524     incorrectly, because the \d was being skipped in the check.
1525    
1526     22. The pcretest program now has a new pattern option /B and a command line
1527     option -b, which is equivalent to adding /B to every pattern. This causes
1528     it to show the compiled bytecode, without the additional information that
1529     -d shows. The effect of -d is now the same as -b with -i (and similarly, /D
1530     is the same as /B/I).
1531    
1532     23. A new optimization is now able automatically to treat some sequences such
1533     as a*b as a*+b. More specifically, if something simple (such as a character
1534     or a simple class like \d) has an unlimited quantifier, and is followed by
1535     something that cannot possibly match the quantified thing, the quantifier
1536     is automatically "possessified".
1537    
1538     24. A recursive reference to a subpattern whose number was greater than 39
1539     went wrong under certain circumstances in UTF-8 mode. This bug could also
1540     have affected the operation of pcre_study().
1541    
1542     25. Realized that a little bit of performance could be had by replacing
1543     (c & 0xc0) == 0xc0 with c >= 0xc0 when processing UTF-8 characters.
1544    
1545     26. Timing data from pcretest is now shown to 4 decimal places instead of 3.
1546    
1547     27. Possessive quantifiers such as a++ were previously implemented by turning
1548     them into atomic groups such as ($>a+). Now they have their own opcodes,
1549     which improves performance. This includes the automatically created ones
1550     from 23 above.
1551    
1552     28. A pattern such as (?=(\w+))\1: which simulates an atomic group using a
1553     lookahead was broken if it was not anchored. PCRE was mistakenly expecting
1554     the first matched character to be a colon. This applied both to named and
1555     numbered groups.
1556    
1557     29. The ucpinternal.h header file was missing its idempotency #ifdef.
1558    
1559     30. I was sent a "project" file called libpcre.a.dev which I understand makes
1560     building PCRE on Windows easier, so I have included it in the distribution.
1561    
1562     31. There is now a check in pcretest against a ridiculously large number being
1563     returned by pcre_exec() or pcre_dfa_exec(). If this happens in a /g or /G
1564     loop, the loop is abandoned.
1565    
1566     32. Forward references to subpatterns in conditions such as (?(2)...) where
1567     subpattern 2 is defined later cause pcre_compile() to search forwards in
1568     the pattern for the relevant set of parentheses. This search went wrong
1569     when there were unescaped parentheses in a character class, parentheses
1570     escaped with \Q...\E, or parentheses in a #-comment in /x mode.
1571    
1572     33. "Subroutine" calls and backreferences were previously restricted to
1573     referencing subpatterns earlier in the regex. This restriction has now
1574     been removed.
1575    
1576     34. Added a number of extra features that are going to be in Perl 5.10. On the
1577     whole, these are just syntactic alternatives for features that PCRE had
1578     previously implemented using the Python syntax or my own invention. The
1579     other formats are all retained for compatibility.
1580    
1581     (a) Named groups can now be defined as (?<name>...) or (?'name'...) as well
1582     as (?P<name>...). The new forms, as well as being in Perl 5.10, are
1583     also .NET compatible.
1584    
1585     (b) A recursion or subroutine call to a named group can now be defined as
1586     (?&name) as well as (?P>name).
1587    
1588     (c) A backreference to a named group can now be defined as \k<name> or
1589     \k'name' as well as (?P=name). The new forms, as well as being in Perl
1590     5.10, are also .NET compatible.
1591    
1592     (d) A conditional reference to a named group can now use the syntax
1593     (?(<name>) or (?('name') as well as (?(name).
1594    
1595     (e) A "conditional group" of the form (?(DEFINE)...) can be used to define
1596     groups (named and numbered) that are never evaluated inline, but can be
1597     called as "subroutines" from elsewhere. In effect, the DEFINE condition
1598     is always false. There may be only one alternative in such a group.
1599    
1600     (f) A test for recursion can be given as (?(R1).. or (?(R&name)... as well
1601     as the simple (?(R). The condition is true only if the most recent
1602     recursion is that of the given number or name. It does not search out
1603     through the entire recursion stack.
1604    
1605     (g) The escape \gN or \g{N} has been added, where N is a positive or
1606     negative number, specifying an absolute or relative reference.
1607    
1608     35. Tidied to get rid of some further signed/unsigned compiler warnings and
1609     some "unreachable code" warnings.
1610    
1611     36. Updated the Unicode property tables to Unicode version 5.0.0. Amongst other
1612     things, this adds five new scripts.
1613    
1614     37. Perl ignores orphaned \E escapes completely. PCRE now does the same.
1615     There were also incompatibilities regarding the handling of \Q..\E inside
1616     character classes, for example with patterns like [\Qa\E-\Qz\E] where the
1617     hyphen was adjacent to \Q or \E. I hope I've cleared all this up now.
1618    
1619     38. Like Perl, PCRE detects when an indefinitely repeated parenthesized group
1620     matches an empty string, and forcibly breaks the loop. There were bugs in
1621     this code in non-simple cases. For a pattern such as ^(a()*)* matched
1622     against aaaa the result was just "a" rather than "aaaa", for example. Two
1623     separate and independent bugs (that affected different cases) have been
1624     fixed.
1625    
1626     39. Refactored the code to abolish the use of different opcodes for small
1627     capturing bracket numbers. This is a tidy that I avoided doing when I
1628     removed the limit on the number of capturing brackets for 3.5 back in 2001.
1629     The new approach is not only tidier, it makes it possible to reduce the
1630     memory needed to fix the previous bug (38).
1631    
1632     40. Implemented PCRE_NEWLINE_ANY to recognize any of the Unicode newline
1633     sequences (http://unicode.org/unicode/reports/tr18/) as "newline" when
1634     processing dot, circumflex, or dollar metacharacters, or #-comments in /x
1635     mode.
1636    
1637     41. Add \R to match any Unicode newline sequence, as suggested in the Unicode
1638     report.
1639    
1640     42. Applied patch, originally from Ari Pollak, modified by Google, to allow
1641     copy construction and assignment in the C++ wrapper.
1642    
1643     43. Updated pcregrep to support "--newline=any". In the process, I fixed a
1644     couple of bugs that could have given wrong results in the "--newline=crlf"
1645     case.
1646    
1647     44. Added a number of casts and did some reorganization of signed/unsigned int
1648     variables following suggestions from Dair Grant. Also renamed the variable
1649     "this" as "item" because it is a C++ keyword.
1650    
1651     45. Arranged for dftables to add
1652    
1653     #include "pcre_internal.h"
1654    
1655     to pcre_chartables.c because without it, gcc 4.x may remove the array
1656     definition from the final binary if PCRE is built into a static library and
1657     dead code stripping is activated.
1658    
1659     46. For an unanchored pattern, if a match attempt fails at the start of a
1660     newline sequence, and the newline setting is CRLF or ANY, and the next two
1661     characters are CRLF, advance by two characters instead of one.
1662    
1663    
1664 nigel 91 Version 6.7 04-Jul-06
1665     ---------------------
1666    
1667     1. In order to handle tests when input lines are enormously long, pcretest has
1668     been re-factored so that it automatically extends its buffers when
1669     necessary. The code is crude, but this _is_ just a test program. The
1670     default size has been increased from 32K to 50K.
1671    
1672     2. The code in pcre_study() was using the value of the re argument before
1673     testing it for NULL. (Of course, in any sensible call of the function, it
1674     won't be NULL.)
1675    
1676     3. The memmove() emulation function in pcre_internal.h, which is used on
1677     systems that lack both memmove() and bcopy() - that is, hardly ever -
1678     was missing a "static" storage class specifier.
1679    
1680     4. When UTF-8 mode was not set, PCRE looped when compiling certain patterns
1681     containing an extended class (one that cannot be represented by a bitmap
1682     because it contains high-valued characters or Unicode property items, e.g.
1683     [\pZ]). Almost always one would set UTF-8 mode when processing such a
1684     pattern, but PCRE should not loop if you do not (it no longer does).
1685     [Detail: two cases were found: (a) a repeated subpattern containing an
1686     extended class; (b) a recursive reference to a subpattern that followed a
1687     previous extended class. It wasn't skipping over the extended class
1688     correctly when UTF-8 mode was not set.]
1689    
1690     5. A negated single-character class was not being recognized as fixed-length
1691     in lookbehind assertions such as (?<=[^f]), leading to an incorrect
1692     compile error "lookbehind assertion is not fixed length".
1693    
1694     6. The RunPerlTest auxiliary script was showing an unexpected difference
1695     between PCRE and Perl for UTF-8 tests. It turns out that it is hard to
1696     write a Perl script that can interpret lines of an input file either as
1697     byte characters or as UTF-8, which is what "perltest" was being required to
1698     do for the non-UTF-8 and UTF-8 tests, respectively. Essentially what you
1699     can't do is switch easily at run time between having the "use utf8;" pragma
1700     or not. In the end, I fudged it by using the RunPerlTest script to insert
1701     "use utf8;" explicitly for the UTF-8 tests.
1702    
1703     7. In multiline (/m) mode, PCRE was matching ^ after a terminating newline at
1704     the end of the subject string, contrary to the documentation and to what
1705     Perl does. This was true of both matching functions. Now it matches only at
1706     the start of the subject and immediately after *internal* newlines.
1707    
1708     8. A call of pcre_fullinfo() from pcretest to get the option bits was passing
1709     a pointer to an int instead of a pointer to an unsigned long int. This
1710     caused problems on 64-bit systems.
1711    
1712     9. Applied a patch from the folks at Google to pcrecpp.cc, to fix "another
1713     instance of the 'standard' template library not being so standard".
1714    
1715     10. There was no check on the number of named subpatterns nor the maximum
1716     length of a subpattern name. The product of these values is used to compute
1717     the size of the memory block for a compiled pattern. By supplying a very
1718     long subpattern name and a large number of named subpatterns, the size
1719     computation could be caused to overflow. This is now prevented by limiting
1720     the length of names to 32 characters, and the number of named subpatterns
1721     to 10,000.
1722    
1723     11. Subpatterns that are repeated with specific counts have to be replicated in
1724     the compiled pattern. The size of memory for this was computed from the
1725     length of the subpattern and the repeat count. The latter is limited to
1726     65535, but there was no limit on the former, meaning that integer overflow
1727     could in principle occur. The compiled length of a repeated subpattern is
1728     now limited to 30,000 bytes in order to prevent this.
1729    
1730     12. Added the optional facility to have named substrings with the same name.
1731    
1732     13. Added the ability to use a named substring as a condition, using the
1733     Python syntax: (?(name)yes|no). This overloads (?(R)... and names that
1734     are numbers (not recommended). Forward references are permitted.
1735    
1736     14. Added forward references in named backreferences (if you see what I mean).
1737    
1738     15. In UTF-8 mode, with the PCRE_DOTALL option set, a quantified dot in the
1739     pattern could run off the end of the subject. For example, the pattern
1740     "(?s)(.{1,5})"8 did this with the subject "ab".
1741    
1742     16. If PCRE_DOTALL or PCRE_MULTILINE were set, pcre_dfa_exec() behaved as if
1743     PCRE_CASELESS was set when matching characters that were quantified with ?
1744     or *.
1745    
1746     17. A character class other than a single negated character that had a minimum
1747     but no maximum quantifier - for example [ab]{6,} - was not handled
1748     correctly by pce_dfa_exec(). It would match only one character.
1749    
1750     18. A valid (though odd) pattern that looked like a POSIX character
1751     class but used an invalid character after [ (for example [[,abc,]]) caused
1752     pcre_compile() to give the error "Failed: internal error: code overflow" or
1753     in some cases to crash with a glibc free() error. This could even happen if
1754     the pattern terminated after [[ but there just happened to be a sequence of
1755     letters, a binary zero, and a closing ] in the memory that followed.
1756    
1757     19. Perl's treatment of octal escapes in the range \400 to \777 has changed
1758     over the years. Originally (before any Unicode support), just the bottom 8
1759     bits were taken. Thus, for example, \500 really meant \100. Nowadays the
1760     output from "man perlunicode" includes this:
1761    
1762     The regular expression compiler produces polymorphic opcodes. That
1763     is, the pattern adapts to the data and automatically switches to
1764     the Unicode character scheme when presented with Unicode data--or
1765     instead uses a traditional byte scheme when presented with byte
1766     data.
1767    
1768     Sadly, a wide octal escape does not cause a switch, and in a string with
1769     no other multibyte characters, these octal escapes are treated as before.
1770     Thus, in Perl, the pattern /\500/ actually matches \100 but the pattern
1771     /\500|\x{1ff}/ matches \500 or \777 because the whole thing is treated as a
1772     Unicode string.
1773    
1774     I have not perpetrated such confusion in PCRE. Up till now, it took just
1775     the bottom 8 bits, as in old Perl. I have now made octal escapes with
1776     values greater than \377 illegal in non-UTF-8 mode. In UTF-8 mode they
1777     translate to the appropriate multibyte character.
1778    
1779     29. Applied some refactoring to reduce the number of warnings from Microsoft
1780     and Borland compilers. This has included removing the fudge introduced
1781     seven years ago for the OS/2 compiler (see 2.02/2 below) because it caused
1782     a warning about an unused variable.
1783    
1784     21. PCRE has not included VT (character 0x0b) in the set of whitespace
1785     characters since release 4.0, because Perl (from release 5.004) does not.
1786     [Or at least, is documented not to: some releases seem to be in conflict
1787     with the documentation.] However, when a pattern was studied with
1788     pcre_study() and all its branches started with \s, PCRE still included VT
1789     as a possible starting character. Of course, this did no harm; it just
1790     caused an unnecessary match attempt.
1791    
1792     22. Removed a now-redundant internal flag bit that recorded the fact that case
1793     dependency changed within the pattern. This was once needed for "required
1794     byte" processing, but is no longer used. This recovers a now-scarce options
1795     bit. Also moved the least significant internal flag bit to the most-
1796     significant bit of the word, which was not previously used (hangover from
1797     the days when it was an int rather than a uint) to free up another bit for
1798     the future.
1799    
1800     23. Added support for CRLF line endings as well as CR and LF. As well as the
1801     default being selectable at build time, it can now be changed at runtime
1802     via the PCRE_NEWLINE_xxx flags. There are now options for pcregrep to
1803     specify that it is scanning data with non-default line endings.
1804    
1805     24. Changed the definition of CXXLINK to make it agree with the definition of
1806     LINK in the Makefile, by replacing LDFLAGS to CXXFLAGS.
1807    
1808     25. Applied Ian Taylor's patches to avoid using another stack frame for tail
1809     recursions. This makes a big different to stack usage for some patterns.
1810    
1811     26. If a subpattern containing a named recursion or subroutine reference such
1812     as (?P>B) was quantified, for example (xxx(?P>B)){3}, the calculation of
1813     the space required for the compiled pattern went wrong and gave too small a
1814     value. Depending on the environment, this could lead to "Failed: internal
1815     error: code overflow at offset 49" or "glibc detected double free or
1816     corruption" errors.
1817    
1818     27. Applied patches from Google (a) to support the new newline modes and (b) to
1819     advance over multibyte UTF-8 characters in GlobalReplace.
1820    
1821     28. Change free() to pcre_free() in pcredemo.c. Apparently this makes a
1822     difference for some implementation of PCRE in some Windows version.
1823    
1824     29. Added some extra testing facilities to pcretest:
1825    
1826     \q<number> in a data line sets the "match limit" value
1827     \Q<number> in a data line sets the "match recursion limt" value
1828     -S <number> sets the stack size, where <number> is in megabytes
1829    
1830     The -S option isn't available for Windows.
1831    
1832    
1833 nigel 89 Version 6.6 06-Feb-06
1834     ---------------------
1835    
1836     1. Change 16(a) for 6.5 broke things, because PCRE_DATA_SCOPE was not defined
1837     in pcreposix.h. I have copied the definition from pcre.h.
1838    
1839     2. Change 25 for 6.5 broke compilation in a build directory out-of-tree
1840     because pcre.h is no longer a built file.
1841    
1842     3. Added Jeff Friedl's additional debugging patches to pcregrep. These are
1843     not normally included in the compiled code.
1844    
1845    
1846 nigel 87 Version 6.5 01-Feb-06
1847     ---------------------
1848    
1849     1. When using the partial match feature with pcre_dfa_exec(), it was not
1850     anchoring the second and subsequent partial matches at the new starting
1851     point. This could lead to incorrect results. For example, with the pattern
1852     /1234/, partially matching against "123" and then "a4" gave a match.
1853    
1854     2. Changes to pcregrep:
1855    
1856     (a) All non-match returns from pcre_exec() were being treated as failures
1857     to match the line. Now, unless the error is PCRE_ERROR_NOMATCH, an
1858     error message is output. Some extra information is given for the
1859     PCRE_ERROR_MATCHLIMIT and PCRE_ERROR_RECURSIONLIMIT errors, which are
1860     probably the only errors that are likely to be caused by users (by
1861     specifying a regex that has nested indefinite repeats, for instance).
1862     If there are more than 20 of these errors, pcregrep is abandoned.
1863    
1864     (b) A binary zero was treated as data while matching, but terminated the
1865     output line if it was written out. This has been fixed: binary zeroes
1866     are now no different to any other data bytes.
1867    
1868     (c) Whichever of the LC_ALL or LC_CTYPE environment variables is set is
1869     used to set a locale for matching. The --locale=xxxx long option has
1870     been added (no short equivalent) to specify a locale explicitly on the
1871     pcregrep command, overriding the environment variables.
1872    
1873     (d) When -B was used with -n, some line numbers in the output were one less
1874     than they should have been.
1875    
1876     (e) Added the -o (--only-matching) option.
1877    
1878     (f) If -A or -C was used with -c (count only), some lines of context were
1879     accidentally printed for the final match.
1880    
1881     (g) Added the -H (--with-filename) option.
1882    
1883     (h) The combination of options -rh failed to suppress file names for files
1884     that were found from directory arguments.
1885    
1886     (i) Added the -D (--devices) and -d (--directories) options.
1887    
1888     (j) Added the -F (--fixed-strings) option.
1889    
1890     (k) Allow "-" to be used as a file name for -f as well as for a data file.
1891    
1892     (l) Added the --colo(u)r option.
1893    
1894     (m) Added Jeffrey Friedl's -S testing option, but within #ifdefs so that it
1895     is not present by default.
1896    
1897     3. A nasty bug was discovered in the handling of recursive patterns, that is,
1898     items such as (?R) or (?1), when the recursion could match a number of
1899     alternatives. If it matched one of the alternatives, but subsequently,
1900     outside the recursion, there was a failure, the code tried to back up into
1901     the recursion. However, because of the way PCRE is implemented, this is not
1902     possible, and the result was an incorrect result from the match.
1903    
1904     In order to prevent this happening, the specification of recursion has
1905     been changed so that all such subpatterns are automatically treated as
1906     atomic groups. Thus, for example, (?R) is treated as if it were (?>(?R)).
1907    
1908     4. I had overlooked the fact that, in some locales, there are characters for
1909     which isalpha() is true but neither isupper() nor islower() are true. In
1910     the fr_FR locale, for instance, the \xAA and \xBA characters (ordmasculine
1911     and ordfeminine) are like this. This affected the treatment of \w and \W
1912     when they appeared in character classes, but not when they appeared outside
1913     a character class. The bit map for "word" characters is now created
1914     separately from the results of isalnum() instead of just taking it from the
1915     upper, lower, and digit maps. (Plus the underscore character, of course.)
1916    
1917     5. The above bug also affected the handling of POSIX character classes such as
1918     [[:alpha:]] and [[:alnum:]]. These do not have their own bit maps in PCRE's
1919     permanent tables. Instead, the bit maps for such a class were previously
1920     created as the appropriate unions of the upper, lower, and digit bitmaps.
1921     Now they are created by subtraction from the [[:word:]] class, which has
1922     its own bitmap.
1923    
1924     6. The [[:blank:]] character class matches horizontal, but not vertical space.
1925     It is created by subtracting the vertical space characters (\x09, \x0a,
1926     \x0b, \x0c) from the [[:space:]] bitmap. Previously, however, the
1927     subtraction was done in the overall bitmap for a character class, meaning
1928     that a class such as [\x0c[:blank:]] was incorrect because \x0c would not
1929     be recognized. This bug has been fixed.
1930    
1931     7. Patches from the folks at Google:
1932    
1933     (a) pcrecpp.cc: "to handle a corner case that may or may not happen in
1934     real life, but is still worth protecting against".
1935    
1936     (b) pcrecpp.cc: "corrects a bug when negative radixes are used with
1937     regular expressions".
1938    
1939     (c) pcre_scanner.cc: avoid use of std::count() because not all systems
1940     have it.
1941    
1942     (d) Split off pcrecpparg.h from pcrecpp.h and had the former built by
1943     "configure" and the latter not, in order to fix a problem somebody had
1944     with compiling the Arg class on HP-UX.
1945    
1946     (e) Improve the error-handling of the C++ wrapper a little bit.
1947    
1948     (f) New tests for checking recursion limiting.
1949    
1950     8. The pcre_memmove() function, which is used only if the environment does not
1951     have a standard memmove() function (and is therefore rarely compiled),
1952     contained two bugs: (a) use of int instead of size_t, and (b) it was not
1953     returning a result (though PCRE never actually uses the result).
1954    
1955     9. In the POSIX regexec() interface, if nmatch is specified as a ridiculously
1956     large number - greater than INT_MAX/(3*sizeof(int)) - REG_ESPACE is
1957     returned instead of calling malloc() with an overflowing number that would
1958     most likely cause subsequent chaos.
1959    
1960     10. The debugging option of pcretest was not showing the NO_AUTO_CAPTURE flag.
1961    
1962     11. The POSIX flag REG_NOSUB is now supported. When a pattern that was compiled
1963     with this option is matched, the nmatch and pmatch options of regexec() are
1964     ignored.
1965    
1966     12. Added REG_UTF8 to the POSIX interface. This is not defined by POSIX, but is
1967     provided in case anyone wants to the the POSIX interface with UTF-8
1968     strings.
1969    
1970     13. Added CXXLDFLAGS to the Makefile parameters to provide settings only on the
1971     C++ linking (needed for some HP-UX environments).
1972    
1973     14. Avoid compiler warnings in get_ucpname() when compiled without UCP support
1974     (unused parameter) and in the pcre_printint() function (omitted "default"
1975     switch label when the default is to do nothing).
1976    
1977     15. Added some code to make it possible, when PCRE is compiled as a C++
1978     library, to replace subject pointers for pcre_exec() with a smart pointer
1979     class, thus making it possible to process discontinuous strings.
1980    
1981     16. The two macros PCRE_EXPORT and PCRE_DATA_SCOPE are confusing, and perform
1982     much the same function. They were added by different people who were trying
1983     to make PCRE easy to compile on non-Unix systems. It has been suggested
1984     that PCRE_EXPORT be abolished now that there is more automatic apparatus
1985     for compiling on Windows systems. I have therefore replaced it with
1986     PCRE_DATA_SCOPE. This is set automatically for Windows; if not set it
1987     defaults to "extern" for C or "extern C" for C++, which works fine on
1988     Unix-like systems. It is now possible to override the value of PCRE_DATA_
1989     SCOPE with something explicit in config.h. In addition:
1990    
1991     (a) pcreposix.h still had just "extern" instead of either of these macros;
1992     I have replaced it with PCRE_DATA_SCOPE.
1993    
1994     (b) Functions such as _pcre_xclass(), which are internal to the library,
1995     but external in the C sense, all had PCRE_EXPORT in their definitions.
1996     This is apparently wrong for the Windows case, so I have removed it.
1997     (It makes no difference on Unix-like systems.)
1998    
1999     17. Added a new limit, MATCH_LIMIT_RECURSION, which limits the depth of nesting
2000     of recursive calls to match(). This is different to MATCH_LIMIT because
2001     that limits the total number of calls to match(), not all of which increase
2002     the depth of recursion. Limiting the recursion depth limits the amount of
2003     stack (or heap if NO_RECURSE is set) that is used. The default can be set
2004     when PCRE is compiled, and changed at run time. A patch from Google adds
2005     this functionality to the C++ interface.
2006    
2007     18. Changes to the handling of Unicode character properties:
2008    
2009     (a) Updated the table to Unicode 4.1.0.
2010    
2011     (b) Recognize characters that are not in the table as "Cn" (undefined).
2012    
2013     (c) I revised the way the table is implemented to a much improved format
2014     which includes recognition of ranges. It now supports the ranges that
2015     are defined in UnicodeData.txt, and it also amalgamates other
2016     characters into ranges. This has reduced the number of entries in the
2017     table from around 16,000 to around 3,000, thus reducing its size
2018     considerably. I realized I did not need to use a tree structure after
2019     all - a binary chop search is just as efficient. Having reduced the
2020     number of entries, I extended their size from 6 bytes to 8 bytes to
2021     allow for more data.
2022    
2023     (d) Added support for Unicode script names via properties such as \p{Han}.
2024    
2025     19. In UTF-8 mode, a backslash followed by a non-Ascii character was not
2026     matching that character.
2027    
2028     20. When matching a repeated Unicode property with a minimum greater than zero,
2029     (for example \pL{2,}), PCRE could look past the end of the subject if it
2030     reached it while seeking the minimum number of characters. This could
2031     happen only if some of the characters were more than one byte long, because
2032     there is a check for at least the minimum number of bytes.
2033    
2034     21. Refactored the implementation of \p and \P so as to be more general, to
2035     allow for more different types of property in future. This has changed the
2036     compiled form incompatibly. Anybody with saved compiled patterns that use
2037     \p or \P will have to recompile them.
2038    
2039     22. Added "Any" and "L&" to the supported property types.
2040    
2041     23. Recognize \x{...} as a code point specifier, even when not in UTF-8 mode,
2042     but give a compile time error if the value is greater than 0xff.
2043    
2044     24. The man pages for pcrepartial, pcreprecompile, and pcre_compile2 were
2045     accidentally not being installed or uninstalled.
2046    
2047     25. The pcre.h file was built from pcre.h.in, but the only changes that were
2048     made were to insert the current release number. This seemed silly, because
2049     it made things harder for people building PCRE on systems that don't run
2050     "configure". I have turned pcre.h into a distributed file, no longer built
2051     by "configure", with the version identification directly included. There is
2052     no longer a pcre.h.in file.
2053    
2054     However, this change necessitated a change to the pcre-config script as
2055     well. It is built from pcre-config.in, and one of the substitutions was the
2056     release number. I have updated configure.ac so that ./configure now finds
2057     the release number by grepping pcre.h.
2058    
2059     26. Added the ability to run the tests under valgrind.
2060    
2061    
2062 nigel 85 Version 6.4 05-Sep-05
2063     ---------------------
2064    
2065     1. Change 6.0/10/(l) to pcregrep introduced a bug that caused separator lines
2066     "--" to be printed when multiple files were scanned, even when none of the
2067     -A, -B, or -C options were used. This is not compatible with Gnu grep, so I
2068     consider it to be a bug, and have restored the previous behaviour.
2069    
2070     2. A couple of code tidies to get rid of compiler warnings.
2071    
2072     3. The pcretest program used to cheat by referring to symbols in the library
2073     whose names begin with _pcre_. These are internal symbols that are not
2074     really supposed to be visible externally, and in some environments it is
2075     possible to suppress them. The cheating is now confined to including
2076     certain files from the library's source, which is a bit cleaner.
2077    
2078     4. Renamed pcre.in as pcre.h.in to go with pcrecpp.h.in; it also makes the
2079     file's purpose clearer.
2080    
2081     5. Reorganized pcre_ucp_findchar().
2082    
2083    
2084 nigel 83 Version 6.3 15-Aug-05
2085     ---------------------
2086    
2087     1. The file libpcre.pc.in did not have general read permission in the tarball.
2088    
2089     2. There were some problems when building without C++ support:
2090    
2091     (a) If C++ support was not built, "make install" and "make test" still
2092     tried to test it.
2093    
2094     (b) There were problems when the value of CXX was explicitly set. Some
2095     changes have been made to try to fix these, and ...
2096    
2097     (c) --disable-cpp can now be used to explicitly disable C++ support.
2098    
2099     (d) The use of @CPP_OBJ@ directly caused a blank line preceded by a
2100     backslash in a target when C++ was disabled. This confuses some
2101     versions of "make", apparently. Using an intermediate variable solves
2102     this. (Same for CPP_LOBJ.)
2103    
2104     3. $(LINK_FOR_BUILD) now includes $(CFLAGS_FOR_BUILD) and $(LINK)
2105     (non-Windows) now includes $(CFLAGS) because these flags are sometimes
2106     necessary on certain architectures.
2107    
2108     4. Added a setting of -export-symbols-regex to the link command to remove
2109     those symbols that are exported in the C sense, but actually are local
2110     within the library, and not documented. Their names all begin with
2111     "_pcre_". This is not a perfect job, because (a) we have to except some
2112     symbols that pcretest ("illegally") uses, and (b) the facility isn't always
2113     available (and never for static libraries). I have made a note to try to
2114     find a way round (a) in the future.
2115    
2116    
2117 nigel 81 Version 6.2 01-Aug-05
2118     ---------------------
2119    
2120     1. There was no test for integer overflow of quantifier values. A construction
2121     such as {1111111111111111} would give undefined results. What is worse, if
2122     a minimum quantifier for a parenthesized subpattern overflowed and became
2123     negative, the calculation of the memory size went wrong. This could have
2124     led to memory overwriting.
2125    
2126     2. Building PCRE using VPATH was broken. Hopefully it is now fixed.
2127    
2128     3. Added "b" to the 2nd argument of fopen() in dftables.c, for non-Unix-like
2129     operating environments where this matters.
2130    
2131     4. Applied Giuseppe Maxia's patch to add additional features for controlling
2132     PCRE options from within the C++ wrapper.
2133    
2134     5. Named capturing subpatterns were not being correctly counted when a pattern
2135     was compiled. This caused two problems: (a) If there were more than 100
2136     such subpatterns, the calculation of the memory needed for the whole
2137     compiled pattern went wrong, leading to an overflow error. (b) Numerical
2138     back references of the form \12, where the number was greater than 9, were
2139     not recognized as back references, even though there were sufficient
2140     previous subpatterns.
2141    
2142     6. Two minor patches to pcrecpp.cc in order to allow it to compile on older
2143     versions of gcc, e.g. 2.95.4.
2144    
2145    
2146 nigel 79 Version 6.1 21-Jun-05
2147     ---------------------
2148    
2149     1. There was one reference to the variable "posix" in pcretest.c that was not
2150     surrounded by "#if !defined NOPOSIX".
2151    
2152     2. Make it possible to compile pcretest without DFA support, UTF8 support, or
2153     the cross-check on the old pcre_info() function, for the benefit of the
2154     cut-down version of PCRE that is currently imported into Exim.
2155    
2156     3. A (silly) pattern starting with (?i)(?-i) caused an internal space
2157     allocation error. I've done the easy fix, which wastes 2 bytes for sensible
2158     patterns that start (?i) but I don't think that matters. The use of (?i) is
2159     just an example; this all applies to the other options as well.
2160    
2161     4. Since libtool seems to echo the compile commands it is issuing, the output
2162     from "make" can be reduced a bit by putting "@" in front of each libtool
2163     compile command.
2164    
2165     5. Patch from the folks at Google for configure.in to be a bit more thorough
2166     in checking for a suitable C++ installation before trying to compile the
2167     C++ stuff. This should fix a reported problem when a compiler was present,
2168     but no suitable headers.
2169    
2170     6. The man pages all had just "PCRE" as their title. I have changed them to
2171     be the relevant file name. I have also arranged that these names are
2172     retained in the file doc/pcre.txt, which is a concatenation in text format
2173     of all the man pages except the little individual ones for each function.
2174    
2175     7. The NON-UNIX-USE file had not been updated for the different set of source
2176     files that come with release 6. I also added a few comments about the C++
2177     wrapper.
2178    
2179    
2180 nigel 77 Version 6.0 07-Jun-05
2181     ---------------------
2182    
2183     1. Some minor internal re-organization to help with my DFA experiments.
2184    
2185     2. Some missing #ifdef SUPPORT_UCP conditionals in pcretest and printint that
2186     didn't matter for the library itself when fully configured, but did matter
2187     when compiling without UCP support, or within Exim, where the ucp files are
2188     not imported.
2189    
2190     3. Refactoring of the library code to split up the various functions into
2191     different source modules. The addition of the new DFA matching code (see
2192     below) to a single monolithic source would have made it really too
2193     unwieldy, quite apart from causing all the code to be include in a
2194     statically linked application, when only some functions are used. This is
2195     relevant even without the DFA addition now that patterns can be compiled in
2196     one application and matched in another.
2197    
2198     The downside of splitting up is that there have to be some external
2199     functions and data tables that are used internally in different modules of
2200     the library but which are not part of the API. These have all had their
2201     names changed to start with "_pcre_" so that they are unlikely to clash
2202     with other external names.
2203    
2204     4. Added an alternate matching function, pcre_dfa_exec(), which matches using
2205     a different (DFA) algorithm. Although it is slower than the original
2206     function, it does have some advantages for certain types of matching
2207     problem.
2208    
2209     5. Upgrades to pcretest in order to test the features of pcre_dfa_exec(),
2210     including restarting after a partial match.
2211    
2212     6. A patch for pcregrep that defines INVALID_FILE_ATTRIBUTES if it is not
2213     defined when compiling for Windows was sent to me. I have put it into the
2214     code, though I have no means of testing or verifying it.
2215    
2216     7. Added the pcre_refcount() auxiliary function.
2217    
2218     8. Added the PCRE_FIRSTLINE option. This constrains an unanchored pattern to
2219     match before or at the first newline in the subject string. In pcretest,
2220     the /f option on a pattern can be used to set this.
2221    
2222     9. A repeated \w when used in UTF-8 mode with characters greater than 256
2223     would behave wrongly. This has been present in PCRE since release 4.0.
2224    
2225     10. A number of changes to the pcregrep command:
2226    
2227     (a) Refactored how -x works; insert ^(...)$ instead of setting
2228     PCRE_ANCHORED and checking the length, in preparation for adding
2229     something similar for -w.
2230    
2231     (b) Added the -w (match as a word) option.
2232    
2233     (c) Refactored the way lines are read and buffered so as to have more
2234     than one at a time available.
2235    
2236     (d) Implemented a pcregrep test script.
2237    
2238     (e) Added the -M (multiline match) option. This allows patterns to match
2239     over several lines of the subject. The buffering ensures that at least
2240     8K, or the rest of the document (whichever is the shorter) is available
2241     for matching (and similarly the previous 8K for lookbehind assertions).
2242    
2243     (f) Changed the --help output so that it now says
2244    
2245     -w, --word-regex(p)
2246    
2247     instead of two lines, one with "regex" and the other with "regexp"
2248     because that confused at least one person since the short forms are the
2249     same. (This required a bit of code, as the output is generated
2250     automatically from a table. It wasn't just a text change.)
2251    
2252     (g) -- can be used to terminate pcregrep options if the next thing isn't an
2253     option but starts with a hyphen. Could be a pattern or a path name
2254     starting with a hyphen, for instance.
2255    
2256     (h) "-" can be given as a file name to represent stdin.
2257    
2258     (i) When file names are being printed, "(standard input)" is used for
2259     the standard input, for compatibility with GNU grep. Previously
2260     "<stdin>" was used.
2261    
2262     (j) The option --label=xxx can be used to supply a name to be used for
2263     stdin when file names are being printed. There is no short form.
2264    
2265     (k) Re-factored the options decoding logic because we are going to add
2266     two more options that take data. Such options can now be given in four
2267     different ways, e.g. "-fname", "-f name", "--file=name", "--file name".
2268    
2269     (l) Added the -A, -B, and -C options for requesting that lines of context
2270     around matches be printed.
2271    
2272     (m) Added the -L option to print the names of files that do not contain
2273     any matching lines, that is, the complement of -l.
2274    
2275     (n) The return code is 2 if any file cannot be opened, but pcregrep does
2276     continue to scan other files.
2277    
2278     (o) The -s option was incorrectly implemented. For compatibility with other
2279     greps, it now suppresses the error message for a non-existent or non-
2280     accessible file (but not the return code). There is a new option called
2281     -q that suppresses the output of matching lines, which was what -s was
2282     previously doing.
2283    
2284     (p) Added --include and --exclude options to specify files for inclusion
2285     and exclusion when recursing.
2286    
2287     11. The Makefile was not using the Autoconf-supported LDFLAGS macro properly.
2288     Hopefully, it now does.
2289    
2290     12. Missing cast in pcre_study().
2291    
2292     13. Added an "uninstall" target to the makefile.
2293    
2294     14. Replaced "extern" in the function prototypes in Makefile.in with
2295     "PCRE_DATA_SCOPE", which defaults to 'extern' or 'extern "C"' in the Unix
2296     world, but is set differently for Windows.
2297    
2298     15. Added a second compiling function called pcre_compile2(). The only
2299     difference is that it has an extra argument, which is a pointer to an
2300     integer error code. When there is a compile-time failure, this is set
2301     non-zero, in addition to the error test pointer being set to point to an
2302     error message. The new argument may be NULL if no error number is required
2303     (but then you may as well call pcre_compile(), which is now just a
2304     wrapper). This facility is provided because some applications need a
2305     numeric error indication, but it has also enabled me to tidy up the way
2306     compile-time errors are handled in the POSIX wrapper.
2307    
2308     16. Added VPATH=.libs to the makefile; this should help when building with one
2309     prefix path and installing with another. (Or so I'm told by someone who
2310     knows more about this stuff than I do.)
2311    
2312     17. Added a new option, REG_DOTALL, to the POSIX function regcomp(). This
2313     passes PCRE_DOTALL to the pcre_compile() function, making the "." character
2314     match everything, including newlines. This is not POSIX-compatible, but
2315     somebody wanted the feature. From pcretest it can be activated by using
2316     both the P and the s flags.
2317    
2318     18. AC_PROG_LIBTOOL appeared twice in Makefile.in. Removed one.
2319    
2320     19. libpcre.pc was being incorrectly installed as executable.
2321    
2322     20. A couple of places in pcretest check for end-of-line by looking for '\n';
2323     it now also looks for '\r' so that it will work unmodified on Windows.
2324    
2325     21. Added Google's contributed C++ wrapper to the distribution.
2326    
2327     22. Added some untidy missing memory free() calls in pcretest, to keep
2328     Electric Fence happy when testing.
2329    
2330    
2331    
2332 nigel 75 Version 5.0 13-Sep-04
2333     ---------------------
2334    
2335     1. Internal change: literal characters are no longer packed up into items
2336     containing multiple characters in a single byte-string. Each character
2337     is now matched using a separate opcode. However, there may be more than one
2338     byte in the character in UTF-8 mode.
2339    
2340     2. The pcre_callout_block structure has two new fields: pattern_position and
2341     next_item_length. These contain the offset in the pattern to the next match
2342     item, and its length, respectively.
2343    
2344     3. The PCRE_AUTO_CALLOUT option for pcre_compile() requests the automatic
2345     insertion of callouts before each pattern item. Added the /C option to
2346     pcretest to make use of this.
2347    
2348     4. On the advice of a Windows user, the lines
2349    
2350     #if defined(_WIN32) || defined(WIN32)
2351     _setmode( _fileno( stdout ), 0x8000 );
2352     #endif /* defined(_WIN32) || defined(WIN32) */
2353    
2354     have been added to the source of pcretest. This apparently does useful
2355     magic in relation to line terminators.
2356    
2357     5. Changed "r" and "w" in the calls to fopen() in pcretest to "rb" and "wb"
2358     for the benefit of those environments where the "b" makes a difference.
2359    
2360     6. The icc compiler has the same options as gcc, but "configure" doesn't seem
2361     to know about it. I have put a hack into configure.in that adds in code
2362     to set GCC=yes if CC=icc. This seems to end up at a point in the
2363     generated configure script that is early enough to affect the setting of
2364     compiler options, which is what is needed, but I have no means of testing
2365     whether it really works. (The user who reported this had patched the
2366     generated configure script, which of course I cannot do.)
2367    
2368     LATER: After change 22 below (new libtool files), the configure script
2369     seems to know about icc (and also ecc). Therefore, I have commented out
2370     this hack in configure.in.
2371    
2372     7. Added support for pkg-config (2 patches were sent in).
2373    
2374     8. Negated POSIX character classes that used a combination of internal tables
2375     were completely broken. These were [[:^alpha:]], [[:^alnum:]], and
2376     [[:^ascii]]. Typically, they would match almost any characters. The other
2377     POSIX classes were not broken in this way.
2378    
2379     9. Matching the pattern "\b.*?" against "ab cd", starting at offset 1, failed
2380     to find the match, as PCRE was deluded into thinking that the match had to
2381     start at the start point or following a newline. The same bug applied to
2382     patterns with negative forward assertions or any backward assertions
2383     preceding ".*" at the start, unless the pattern required a fixed first
2384     character. This was a failing pattern: "(?!.bcd).*". The bug is now fixed.
2385    
2386     10. In UTF-8 mode, when moving forwards in the subject after a failed match
2387     starting at the last subject character, bytes beyond the end of the subject
2388     string were read.
2389    
2390     11. Renamed the variable "class" as "classbits" to make life easier for C++
2391     users. (Previously there was a macro definition, but it apparently wasn't
2392     enough.)
2393    
2394     12. Added the new field "tables" to the extra data so that tables can be passed
2395     in at exec time, or the internal tables can be re-selected. This allows
2396     a compiled regex to be saved and re-used at a later time by a different
2397     program that might have everything at different addresses.
2398    
2399     13. Modified the pcre-config script so that, when run on Solaris, it shows a
2400     -R library as well as a -L library.
2401    
2402     14. The debugging options of pcretest (-d on the command line or D on a
2403     pattern) showed incorrect output for anything following an extended class
2404     that contained multibyte characters and which was followed by a quantifier.
2405    
2406     15. Added optional support for general category Unicode character properties
2407     via the \p, \P, and \X escapes. Unicode property support implies UTF-8
2408     support. It adds about 90K to the size of the library. The meanings of the
2409     inbuilt class escapes such as \d and \s have NOT been changed.
2410    
2411     16. Updated pcredemo.c to include calls to free() to release the memory for the
2412     compiled pattern.
2413    
2414     17. The generated file chartables.c was being created in the source directory
2415     instead of in the building directory. This caused the build to fail if the
2416     source directory was different from the building directory, and was
2417     read-only.
2418    
2419     18. Added some sample Win commands from Mark Tetrode into the NON-UNIX-USE
2420     file. No doubt somebody will tell me if they don't make sense... Also added
2421     Dan Mooney's comments about building on OpenVMS.
2422    
2423     19. Added support for partial matching via the PCRE_PARTIAL option for
2424     pcre_exec() and the \P data escape in pcretest.
2425    
2426     20. Extended pcretest with 3 new pattern features:
2427    
2428     (i) A pattern option of the form ">rest-of-line" causes pcretest to
2429     write the compiled pattern to the file whose name is "rest-of-line".
2430     This is a straight binary dump of the data, with the saved pointer to
2431     the character tables forced to be NULL. The study data, if any, is
2432     written too. After writing, pcretest reads a new pattern.
2433    
2434     (ii) If, instead of a pattern, "<rest-of-line" is given, pcretest reads a
2435     compiled pattern from the given file. There must not be any
2436     occurrences of "<" in the file name (pretty unlikely); if there are,
2437     pcretest will instead treat the initial "<" as a pattern delimiter.
2438     After reading in the pattern, pcretest goes on to read data lines as
2439     usual.
2440    
2441     (iii) The F pattern option causes pcretest to flip the bytes in the 32-bit
2442     and 16-bit fields in a compiled pattern, to simulate a pattern that
2443     was compiled on a host of opposite endianness.
2444    
2445     21. The pcre-exec() function can now cope with patterns that were compiled on
2446     hosts of opposite endianness, with this restriction:
2447    
2448     As for any compiled expression that is saved and used later, the tables
2449     pointer field cannot be preserved; the extra_data field in the arguments
2450     to pcre_exec() should be used to pass in a tables address if a value
2451     other than the default internal tables were used at compile time.
2452    
2453     22. Calling pcre_exec() with a negative value of the "ovecsize" parameter is
2454     now diagnosed as an error. Previously, most of the time, a negative number
2455     would have been treated as zero, but if in addition "ovector" was passed as
2456     NULL, a crash could occur.
2457    
2458     23. Updated the files ltmain.sh, config.sub, config.guess, and aclocal.m4 with
2459     new versions from the libtool 1.5 distribution (the last one is a copy of
2460     a file called libtool.m4). This seems to have fixed the need to patch
2461     "configure" to support Darwin 1.3 (which I used to do). However, I still
2462     had to patch ltmain.sh to ensure that ${SED} is set (it isn't on my
2463     workstation).
2464    
2465     24. Changed the PCRE licence to be the more standard "BSD" licence.
2466    
2467    
2468 nigel 73 Version 4.5 01-Dec-03
2469     ---------------------
2470    
2471     1. There has been some re-arrangement of the code for the match() function so
2472     that it can be compiled in a version that does not call itself recursively.
2473     Instead, it keeps those local variables that need separate instances for
2474     each "recursion" in a frame on the heap, and gets/frees frames whenever it
2475     needs to "recurse". Keeping track of where control must go is done by means
2476     of setjmp/longjmp. The whole thing is implemented by a set of macros that
2477     hide most of the details from the main code, and operates only if
2478     NO_RECURSE is defined while compiling pcre.c. If PCRE is built using the
2479     "configure" mechanism, "--disable-stack-for-recursion" turns on this way of
2480     operating.
2481    
2482     To make it easier for callers to provide specially tailored get/free
2483     functions for this usage, two new functions, pcre_stack_malloc, and
2484     pcre_stack_free, are used. They are always called in strict stacking order,
2485     and the size of block requested is always the same.
2486    
2487     The PCRE_CONFIG_STACKRECURSE info parameter can be used to find out whether
2488     PCRE has been compiled to use the stack or the heap for recursion. The
2489     -C option of pcretest uses this to show which version is compiled.
2490    
2491     A new data escape \S, is added to pcretest; it causes the amounts of store
2492     obtained and freed by both kinds of malloc/free at match time to be added
2493     to the output.
2494    
2495     2. Changed the locale test to use "fr_FR" instead of "fr" because that's
2496     what's available on my current Linux desktop machine.
2497    
2498     3. When matching a UTF-8 string, the test for a valid string at the start has
2499     been extended. If start_offset is not zero, PCRE now checks that it points
2500     to a byte that is the start of a UTF-8 character. If not, it returns
2501     PCRE_ERROR_BADUTF8_OFFSET (-11). Note: the whole string is still checked;
2502     this is necessary because there may be backward assertions in the pattern.
2503     When matching the same subject several times, it may save resources to use
2504     PCRE_NO_UTF8_CHECK on all but the first call if the string is long.
2505    
2506     4. The code for checking the validity of UTF-8 strings has been tightened so
2507     that it rejects (a) strings containing 0xfe or 0xff bytes and (b) strings
2508     containing "overlong sequences".
2509    
2510     5. Fixed a bug (appearing twice) that I could not find any way of exploiting!
2511     I had written "if ((digitab[*p++] && chtab_digit) == 0)" where the "&&"
2512     should have been "&", but it just so happened that all the cases this let
2513     through by mistake were picked up later in the function.
2514    
2515     6. I had used a variable called "isblank" - this is a C99 function, causing
2516     some compilers to warn. To avoid this, I renamed it (as "blankclass").
2517    
2518     7. Cosmetic: (a) only output another newline at the end of pcretest if it is
2519     prompting; (b) run "./pcretest /dev/null" at the start of the test script
2520     so the version is shown; (c) stop "make test" echoing "./RunTest".
2521    
2522     8. Added patches from David Burgess to enable PCRE to run on EBCDIC systems.
2523    
2524     9. The prototype for memmove() for systems that don't have it was using
2525     size_t, but the inclusion of the header that defines size_t was later. I've
2526     moved the #includes for the C headers earlier to avoid this.
2527    
2528     10. Added some adjustments to the code to make it easier to compiler on certain
2529     special systems:
2530    
2531     (a) Some "const" qualifiers were missing.
2532     (b) Added the macro EXPORT before all exported functions; by default this
2533     is defined to be empty.
2534     (c) Changed the dftables auxiliary program (that builds chartables.c) so
2535     that it reads its output file name as an argument instead of writing
2536     to the standard output and assuming this can be redirected.
2537    
2538     11. In UTF-8 mode, if a recursive reference (e.g. (?1)) followed a character
2539     class containing characters with values greater than 255, PCRE compilation
2540     went into a loop.
2541    
2542     12. A recursive reference to a subpattern that was within another subpattern
2543     that had a minimum quantifier of zero caused PCRE to crash. For example,
2544     (x(y(?2))z)? provoked this bug with a subject that got as far as the
2545     recursion. If the recursively-called subpattern itself had a zero repeat,
2546     that was OK.
2547    
2548     13. In pcretest, the buffer for reading a data line was set at 30K, but the
2549     buffer into which it was copied (for escape processing) was still set at
2550     1024, so long lines caused crashes.
2551    
2552     14. A pattern such as /[ab]{1,3}+/ failed to compile, giving the error
2553     "internal error: code overflow...". This applied to any character class
2554     that was followed by a possessive quantifier.
2555    
2556     15. Modified the Makefile to add libpcre.la as a prerequisite for
2557     libpcreposix.la because I was told this is needed for a parallel build to
2558     work.
2559    
2560     16. If a pattern that contained .* following optional items at the start was
2561     studied, the wrong optimizing data was generated, leading to matching
2562     errors. For example, studying /[ab]*.*c/ concluded, erroneously, that any
2563     matching string must start with a or b or c. The correct conclusion for
2564     this pattern is that a match can start with any character.
2565    
2566    
2567 nigel 71 Version 4.4 13-Aug-03
2568     ---------------------
2569    
2570     1. In UTF-8 mode, a character class containing characters with values between
2571     127 and 255 was not handled correctly if the compiled pattern was studied.
2572     In fixing this, I have also improved the studying algorithm for such
2573     classes (slightly).
2574    
2575     2. Three internal functions had redundant arguments passed to them. Removal
2576     might give a very teeny performance improvement.
2577    
2578     3. Documentation bug: the value of the capture_top field in a callout is *one
2579     more than* the number of the hightest numbered captured substring.
2580    
2581     4. The Makefile linked pcretest and pcregrep with -lpcre, which could result
2582     in incorrectly linking with a previously installed version. They now link
2583     explicitly with libpcre.la.
2584    
2585     5. configure.in no longer needs to recognize Cygwin specially.
2586    
2587     6. A problem in pcre.in for Windows platforms is fixed.
2588    
2589     7. If a pattern was successfully studied, and the -d (or /D) flag was given to
2590     pcretest, it used to include the size of the study block as part of its
2591     output. Unfortunately, the structure contains a field that has a different
2592     size on different hardware architectures. This meant that the tests that
2593     showed this size failed. As the block is currently always of a fixed size,
2594     this information isn't actually particularly useful in pcretest output, so
2595     I have just removed it.
2596    
2597     8. Three pre-processor statements accidentally did not start in column 1.
2598     Sadly, there are *still* compilers around that complain, even though
2599     standard C has not required this for well over a decade. Sigh.
2600    
2601     9. In pcretest, the code for checking callouts passed small integers in the
2602     callout_data field, which is a void * field. However, some picky compilers
2603     complained about the casts involved for this on 64-bit systems. Now
2604     pcretest passes the address of the small integer instead, which should get
2605     rid of the warnings.
2606    
2607     10. By default, when in UTF-8 mode, PCRE now checks for valid UTF-8 strings at
2608     both compile and run time, and gives an error if an invalid UTF-8 sequence
2609     is found. There is a option for disabling this check in cases where the
2610     string is known to be correct and/or the maximum performance is wanted.
2611    
2612     11. In response to a bug report, I changed one line in Makefile.in from
2613    
2614     -Wl,--out-implib,.libs/lib@WIN_PREFIX@pcreposix.dll.a \
2615     to
2616     -Wl,--out-implib,.libs/@WIN_PREFIX@libpcreposix.dll.a \
2617    
2618     to look similar to other lines, but I have no way of telling whether this
2619     is the right thing to do, as I do not use Windows. No doubt I'll get told
2620     if it's wrong...
2621    
2622    
2623 nigel 69 Version 4.3 21-May-03
2624     ---------------------
2625    
2626     1. Two instances of @WIN_PREFIX@ omitted from the Windows targets in the
2627     Makefile.
2628    
2629     2. Some refactoring to improve the quality of the code:
2630    
2631     (i) The utf8_table... variables are now declared "const".
2632    
2633     (ii) The code for \cx, which used the "case flipping" table to upper case
2634     lower case letters, now just substracts 32. This is ASCII-specific,
2635     but the whole concept of \cx is ASCII-specific, so it seems
2636     reasonable.
2637    
2638     (iii) PCRE was using its character types table to recognize decimal and
2639     hexadecimal digits in the pattern. This is silly, because it handles
2640     only 0-9, a-f, and A-F, but the character types table is locale-
2641     specific, which means strange things might happen. A private
2642     table is now used for this - though it costs 256 bytes, a table is
2643     much faster than multiple explicit tests. Of course, the standard
2644     character types table is still used for matching digits in subject
2645     strings against \d.
2646    
2647     (iv) Strictly, the identifier ESC_t is reserved by POSIX (all identifiers
2648     ending in _t are). So I've renamed it as ESC_tee.
2649    
2650     3. The first argument for regexec() in the POSIX wrapper should have been
2651     defined as "const".
2652    
2653     4. Changed pcretest to use malloc() for its buffers so that they can be
2654     Electric Fenced for debugging.
2655    
2656     5. There were several places in the code where, in UTF-8 mode, PCRE would try
2657     to read one or more bytes before the start of the subject string. Often this
2658     had no effect on PCRE's behaviour, but in some circumstances it could
2659     provoke a segmentation fault.
2660    
2661     6. A lookbehind at the start of a pattern in UTF-8 mode could also cause PCRE
2662     to try to read one or more bytes before the start of the subject string.
2663    
2664     7. A lookbehind in a pattern matched in non-UTF-8 mode on a PCRE compiled with
2665     UTF-8 support could misbehave in various ways if the subject string
2666     contained bytes with the 0x80 bit set and the 0x40 bit unset in a lookbehind
2667     area. (PCRE was not checking for the UTF-8 mode flag, and trying to move
2668     back over UTF-8 characters.)
2669    
2670    
2671 nigel 67 Version 4.2 14-Apr-03
2672     ---------------------
2673    
2674     1. Typo "#if SUPPORT_UTF8" instead of "#ifdef SUPPORT_UTF8" fixed.
2675    
2676     2. Changes to the building process, supplied by Ronald Landheer-Cieslak
2677     [ON_WINDOWS]: new variable, "#" on non-Windows platforms
2678     [NOT_ON_WINDOWS]: new variable, "#" on Windows platforms
2679     [WIN_PREFIX]: new variable, "cyg" for Cygwin
2680     * Makefile.in: use autoconf substitution for OBJEXT, EXEEXT, BUILD_OBJEXT
2681     and BUILD_EXEEXT
2682     Note: automatic setting of the BUILD variables is not yet working
2683     set CPPFLAGS and BUILD_CPPFLAGS (but don't use yet) - should be used at
2684     compile-time but not at link-time
2685     [LINK]: use for linking executables only
2686     make different versions for Windows and non-Windows
2687     [LINKLIB]: new variable, copy of UNIX-style LINK, used for linking
2688     libraries
2689     [LINK_FOR_BUILD]: new variable
2690     [OBJEXT]: use throughout
2691     [EXEEXT]: use throughout
2692     <winshared>: new target
2693     <wininstall>: new target
2694     <dftables.o>: use native compiler
2695     <dftables>: use native linker
2696     <install>: handle Windows platform correctly
2697     <clean>: ditto
2698     <check>: ditto
2699     copy DLL to top builddir before testing
2700    
2701     As part of these changes, -no-undefined was removed again. This was reported
2702     to give trouble on HP-UX 11.0, so getting rid of it seems like a good idea
2703     in any case.
2704    
2705     3. Some tidies to get rid of compiler warnings:
2706    
2707     . In the match_data structure, match_limit was an unsigned long int, whereas
2708     match_call_count was an int. I've made them both unsigned long ints.
2709    
2710     . In pcretest the fact that a const uschar * doesn't automatically cast to
2711     a void * provoked a warning.
2712    
2713     . Turning on some more compiler warnings threw up some "shadow" variables
2714     and a few more missing casts.
2715    
2716     4. If PCRE was complied with UTF-8 support, but called without the PCRE_UTF8
2717     option, a class that contained a single character with a value between 128
2718     and 255 (e.g. /[\xFF]/) caused PCRE to crash.
2719    
2720     5. If PCRE was compiled with UTF-8 support, but called without the PCRE_UTF8
2721     option, a class that contained several characters, but with at least one
2722     whose value was between 128 and 255 caused PCRE to crash.
2723    
2724    
2725 nigel 65 Version 4.1 12-Mar-03
2726     ---------------------
2727 nigel 63
2728 nigel 65 1. Compiling with gcc -pedantic found a couple of places where casts were
2729     needed, and a string in dftables.c that was longer than standard compilers are
2730     required to support.
2731    
2732     2. Compiling with Sun's compiler found a few more places where the code could
2733     be tidied up in order to avoid warnings.
2734    
2735     3. The variables for cross-compiling were called HOST_CC and HOST_CFLAGS; the
2736     first of these names is deprecated in the latest Autoconf in favour of the name
2737     CC_FOR_BUILD, because "host" is typically used to mean the system on which the
2738     compiled code will be run. I can't find a reference for HOST_CFLAGS, but by
2739     analogy I have changed it to CFLAGS_FOR_BUILD.
2740    
2741     4. Added -no-undefined to the linking command in the Makefile, because this is
2742     apparently helpful for Windows. To make it work, also added "-L. -lpcre" to the
2743     linking step for the pcreposix library.
2744    
2745     5. PCRE was failing to diagnose the case of two named groups with the same
2746     name.
2747    
2748     6. A problem with one of PCRE's optimizations was discovered. PCRE remembers a
2749     literal character that is needed in the subject for a match, and scans along to
2750     ensure that it is present before embarking on the full matching process. This
2751     saves time in cases of nested unlimited repeats that are never going to match.
2752     Problem: the scan can take a lot of time if the subject is very long (e.g.
2753     megabytes), thus penalizing straightforward matches. It is now done only if the
2754     amount of subject to be scanned is less than 1000 bytes.
2755    
2756     7. A lesser problem with the same optimization is that it was recording the
2757     first character of an anchored pattern as "needed", thus provoking a search
2758     right along the subject, even when the first match of the pattern was going to
2759     fail. The "needed" character is now not set for anchored patterns, unless it
2760     follows something in the pattern that is of non-fixed length. Thus, it still
2761     fulfils its original purpose of finding quick non-matches in cases of nested
2762     unlimited repeats, but isn't used for simple anchored patterns such as /^abc/.
2763    
2764    
2765     Version 4.0 17-Feb-03
2766     ---------------------
2767    
2768 nigel 63 1. If a comment in an extended regex that started immediately after a meta-item
2769     extended to the end of string, PCRE compiled incorrect data. This could lead to
2770     all kinds of weird effects. Example: /#/ was bad; /()#/ was bad; /a#/ was not.
2771    
2772     2. Moved to autoconf 2.53 and libtool 1.4.2.
2773    
2774     3. Perl 5.8 no longer needs "use utf8" for doing UTF-8 things. Consequently,
2775     the special perltest8 script is no longer needed - all the tests can be run
2776     from a single perltest script.
2777    
2778     4. From 5.004, Perl has not included the VT character (0x0b) in the set defined
2779     by \s. It has now been removed in PCRE. This means it isn't recognized as
2780     whitespace in /x regexes too, which is the same as Perl. Note that the POSIX
2781     class [:space:] *does* include VT, thereby creating a mess.
2782    
2783     5. Added the class [:blank:] (a GNU extension from Perl 5.8) to match only
2784     space and tab.
2785    
2786     6. Perl 5.005 was a long time ago. It's time to amalgamate the tests that use
2787     its new features into the main test script, reducing the number of scripts.
2788    
2789     7. Perl 5.8 has changed the meaning of patterns like /a(?i)b/. Earlier versions
2790     were backward compatible, and made the (?i) apply to the whole pattern, as if
2791     /i were given. Now it behaves more logically, and applies the option setting
2792     only to what follows. PCRE has been changed to follow suit. However, if it
2793     finds options settings right at the start of the pattern, it extracts them into
2794     the global options, as before. Thus, they show up in the info data.
2795    
2796     8. Added support for the \Q...\E escape sequence. Characters in between are
2797     treated as literals. This is slightly different from Perl in that $ and @ are
2798     also handled as literals inside the quotes. In Perl, they will cause variable
2799     interpolation. Note the following examples:
2800    
2801     Pattern PCRE matches Perl matches
2802    
2803     \Qabc$xyz\E abc$xyz abc followed by the contents of $xyz
2804     \Qabc\$xyz\E abc\$xyz abc\$xyz
2805     \Qabc\E\$\Qxyz\E abc$xyz abc$xyz
2806    
2807     For compatibility with Perl, \Q...\E sequences are recognized inside character
2808     classes as well as outside them.
2809    
2810     9. Re-organized 3 code statements in pcretest to avoid "overflow in
2811     floating-point constant arithmetic" warnings from a Microsoft compiler. Added a
2812     (size_t) cast to one statement in pcretest and one in pcreposix to avoid
2813     signed/unsigned warnings.
2814    
2815     10. SunOS4 doesn't have strtoul(). This was used only for unpicking the -o
2816     option for pcretest, so I've replaced it by a simple function that does just
2817     that job.
2818    
2819     11. pcregrep was ending with code 0 instead of 2 for the commands "pcregrep" or
2820     "pcregrep -".
2821    
2822     12. Added "possessive quantifiers" ?+, *+, ++, and {,}+ which come from Sun's
2823     Java package. This provides some syntactic sugar for simple cases of what my
2824     documentation calls "once-only subpatterns". A pattern such as x*+ is the same
2825     as (?>x*). In other words, if what is inside (?>...) is just a single repeated
2826     item, you can use this simplified notation. Note that only makes sense with
2827     greedy quantifiers. Consequently, the use of the possessive quantifier forces
2828     greediness, whatever the setting of the PCRE_UNGREEDY option.
2829    
2830     13. A change of greediness default within a pattern was not taking effect at
2831     the current level for patterns like /(b+(?U)a+)/. It did apply to parenthesized
2832     subpatterns that followed. Patterns like /b+(?U)a+/ worked because the option
2833     was abstracted outside.
2834    
2835     14. PCRE now supports the \G assertion. It is true when the current matching
2836     position is at the start point of the match. This differs from \A when the
2837     starting offset is non-zero. Used with the /g option of pcretest (or similar
2838     code), it works in the same way as it does for Perl's /g option. If all
2839     alternatives of a regex begin with \G, the expression is anchored to the start
2840     match position, and the "anchored" flag is set in the compiled expression.
2841    
2842     15. Some bugs concerning the handling of certain option changes within patterns
2843     have been fixed. These applied to options other than (?ims). For example,
2844     "a(?x: b c )d" did not match "XabcdY" but did match "Xa b c dY". It should have
2845     been the other way round. Some of this was related to change 7 above.
2846    
2847     16. PCRE now gives errors for /[.x.]/ and /[=x=]/ as unsupported POSIX
2848     features, as Perl does. Previously, PCRE gave the warnings only for /[[.x.]]/
2849     and /[[=x=]]/. PCRE now also gives an error for /[:name:]/ because it supports
2850     POSIX classes only within a class (e.g. /[[:alpha:]]/).
2851    
2852     17. Added support for Perl's \C escape. This matches one byte, even in UTF8
2853     mode. Unlike ".", it always matches newline, whatever the setting of
2854     PCRE_DOTALL. However, PCRE does not permit \C to appear in lookbehind
2855     assertions. Perl allows it, but it doesn't (in general) work because it can't
2856     calculate the length of the lookbehind. At least, that's the case for Perl
2857     5.8.0 - I've been told they are going to document that it doesn't work in
2858     future.
2859    
2860     18. Added an error diagnosis for escapes that PCRE does not support: these are
2861     \L, \l, \N, \P, \p, \U, \u, and \X.
2862    
2863     19. Although correctly diagnosing a missing ']' in a character class, PCRE was
2864     reading past the end of the pattern in cases such as /[abcd/.
2865    
2866     20. PCRE was getting more memory than necessary for patterns with classes that
2867     contained both POSIX named classes and other characters, e.g. /[[:space:]abc/.
2868    
2869     21. Added some code, conditional on #ifdef VPCOMPAT, to make life easier for
2870     compiling PCRE for use with Virtual Pascal.
2871    
2872     22. Small fix to the Makefile to make it work properly if the build is done
2873     outside the source tree.
2874    
2875     23. Added a new extension: a condition to go with recursion. If a conditional
2876     subpattern starts with (?(R) the "true" branch is used if recursion has
2877     happened, whereas the "false" branch is used only at the top level.
2878    
2879     24. When there was a very long string of literal characters (over 255 bytes
2880     without UTF support, over 250 bytes with UTF support), the computation of how
2881     much memory was required could be incorrect, leading to segfaults or other
2882     strange effects.
2883    
2884     25. PCRE was incorrectly assuming anchoring (either to start of subject or to
2885     start of line for a non-DOTALL pattern) when a pattern started with (.*) and
2886     there was a subsequent back reference to those brackets. This meant that, for
2887     example, /(.*)\d+\1/ failed to match "abc123bc". Unfortunately, it isn't
2888     possible to check for precisely this case. All we can do is abandon the
2889     optimization if .* occurs inside capturing brackets when there are any back
2890     references whatsoever. (See below for a better fix that came later.)
2891    
2892     26. The handling of the optimization for finding the first character of a
2893     non-anchored pattern, and for finding a character that is required later in the
2894     match were failing in some cases. This didn't break the matching; it just
2895     failed to optimize when it could. The way this is done has been re-implemented.
2896    
2897     27. Fixed typo in error message for invalid (?R item (it said "(?p").
2898    
2899     28. Added a new feature that provides some of the functionality that Perl
2900     provides with (?{...}). The facility is termed a "callout". The way it is done
2901     in PCRE is for the caller to provide an optional function, by setting
2902     pcre_callout to its entry point. Like pcre_malloc and pcre_free, this is a
2903     global variable. By default it is unset, which disables all calling out. To get
2904     the function called, the regex must include (?C) at appropriate points. This
2905     is, in fact, equivalent to (?C0), and any number <= 255 may be given with (?C).
2906     This provides a means of identifying different callout points. When PCRE
2907     reaches such a point in the regex, if pcre_callout has been set, the external
2908     function is called. It is provided with data in a structure called
2909     pcre_callout_block, which is defined in pcre.h. If the function returns 0,
2910     matching continues; if it returns a non-zero value, the match at the current
2911     point fails. However, backtracking will occur if possible. [This was changed
2912     later and other features added - see item 49 below.]
2913    
2914     29. pcretest is upgraded to test the callout functionality. It provides a
2915     callout function that displays information. By default, it shows the start of
2916     the match and the current position in the text. There are some new data escapes
2917     to vary what happens:
2918    
2919     \C+ in addition, show current contents of captured substrings
2920     \C- do not supply a callout function
2921     \C!n return 1 when callout number n is reached
2922     \C!n!m return 1 when callout number n is reached for the mth time
2923    
2924     30. If pcregrep was called with the -l option and just a single file name, it
2925     output "<stdin>" if a match was found, instead of the file name.
2926    
2927     31. Improve the efficiency of the POSIX API to PCRE. If the number of capturing
2928     slots is less than POSIX_MALLOC_THRESHOLD, use a block on the stack to pass to
2929     pcre_exec(). This saves a malloc/free per call. The default value of
2930     POSIX_MALLOC_THRESHOLD is 10; it can be changed by --with-posix-malloc-threshold
2931     when configuring.
2932    
2933     32. The default maximum size of a compiled pattern is 64K. There have been a
2934     few cases of people hitting this limit. The code now uses macros to handle the
2935     storing of links as offsets within the compiled pattern. It defaults to 2-byte
2936     links, but this can be changed to 3 or 4 bytes by --with-link-size when
2937     configuring. Tests 2 and 5 work only with 2-byte links because they output
2938     debugging information about compiled patterns.
2939    
2940     33. Internal code re-arrangements:
2941    
2942     (a) Moved the debugging function for printing out a compiled regex into
2943     its own source file (printint.c) and used #include to pull it into
2944     pcretest.c and, when DEBUG is defined, into pcre.c, instead of having two
2945     separate copies.
2946    
2947     (b) Defined the list of op-code names for debugging as a macro in
2948     internal.h so that it is next to the definition of the opcodes.
2949    
2950     (c) Defined a table of op-code lengths for simpler skipping along compiled
2951     code. This is again a macro in internal.h so that it is next to the
2952     definition of the opcodes.
2953    
2954     34. Added support for recursive calls to individual subpatterns, along the
2955     lines of Robin Houston's patch (but implemented somewhat differently).
2956    
2957     35. Further mods to the Makefile to help Win32. Also, added code to pcregrep to
2958     allow it to read and process whole directories in Win32. This code was
2959     contributed by Lionel Fourquaux; it has not been tested by me.
2960    
2961     36. Added support for named subpatterns. The Python syntax (?P<name>...) is
2962     used to name a group. Names consist of alphanumerics and underscores, and must
2963     be unique. Back references use the syntax (?P=name) and recursive calls use
2964     (?P>name) which is a PCRE extension to the Python extension. Groups still have
2965     numbers. The function pcre_fullinfo() can be used after compilation to extract
2966     a name/number map. There are three relevant calls:
2967    
2968     PCRE_INFO_NAMEENTRYSIZE yields the size of each entry in the map
2969     PCRE_INFO_NAMECOUNT yields the number of entries
2970     PCRE_INFO_NAMETABLE yields a pointer to the map.
2971    
2972     The map is a vector of fixed-size entries. The size of each entry depends on
2973     the length of the longest name used. The first two bytes of each entry are the
2974     group number, most significant byte first. There follows the corresponding
2975     name, zero terminated. The names are in alphabetical order.
2976    
2977     37. Make the maximum literal string in the compiled code 250 for the non-UTF-8
2978     case instead of 255. Making it the same both with and without UTF-8 support
2979     means that the same test output works with both.
2980    
2981     38. There was a case of malloc(0) in the POSIX testing code in pcretest. Avoid
2982     calling malloc() with a zero argument.
2983    
2984     39. Change 25 above had to resort to a heavy-handed test for the .* anchoring
2985     optimization. I've improved things by keeping a bitmap of backreferences with
2986     numbers 1-31 so that if .* occurs inside capturing brackets that are not in
2987     fact referenced, the optimization can be applied. It is unlikely that a
2988     relevant occurrence of .* (i.e. one which might indicate anchoring or forcing
2989     the match to follow \n) will appear inside brackets with a number greater than
2990     31, but if it does, any back reference > 31 suppresses the optimization.
2991    
2992     40. Added a new compile-time option PCRE_NO_AUTO_CAPTURE. This has the effect
2993     of disabling numbered capturing parentheses. Any opening parenthesis that is
2994     not followed by ? behaves as if it were followed by ?: but named parentheses
2995     can still be used for capturing (and they will acquire numbers in the usual
2996     way).
2997    
2998     41. Redesigned the return codes from the match() function into yes/no/error so
2999     that errors can be passed back from deep inside the nested calls. A malloc
3000     failure while inside a recursive subpattern call now causes the
3001     PCRE_ERROR_NOMEMORY return instead of quietly going wrong.
3002    
3003     42. It is now possible to set a limit on the number of times the match()
3004     function is called in a call to pcre_exec(). This facility makes it possible to
3005     limit the amount of recursion and backtracking, though not in a directly
3006     obvious way, because the match() function is used in a number of different
3007     circumstances. The count starts from zero for each position in the subject
3008     string (for non-anchored patterns). The default limit is, for compatibility, a
3009     large number, namely 10 000 000. You can change this in two ways:
3010    
3011     (a) When configuring PCRE before making, you can use --with-match-limit=n
3012     to set a default value for the compiled library.
3013    
3014     (b) For each call to pcre_exec(), you can pass a pcre_extra block in which
3015     a different value is set. See 45 below.
3016    
3017     If the limit is exceeded, pcre_exec() returns PCRE_ERROR_MATCHLIMIT.
3018    
3019     43. Added a new function pcre_config(int, void *) to enable run-time extraction
3020     of things that can be changed at compile time. The first argument specifies
3021     what is wanted and the second points to where the information is to be placed.
3022     The current list of available information is:
3023    
3024     PCRE_CONFIG_UTF8
3025    
3026     The output is an integer that is set to one if UTF-8 support is available;
3027     otherwise it is set to zero.
3028    
3029     PCRE_CONFIG_NEWLINE
3030    
3031     The output is an integer that it set to the value of the code that is used for
3032     newline. It is either LF (10) or CR (13).
3033    
3034     PCRE_CONFIG_LINK_SIZE
3035    
3036     The output is an integer that contains the number of bytes used for internal
3037     linkage in compiled expressions. The value is 2, 3, or 4. See item 32 above.
3038    
3039     PCRE_CONFIG_POSIX_MALLOC_THRESHOLD
3040    
3041     The output is an integer that contains the threshold above which the POSIX
3042     interface uses malloc() for output vectors. See item 31 above.
3043    
3044     PCRE_CONFIG_MATCH_LIMIT
3045    
3046     The output is an unsigned integer that contains the default limit of the number
3047     of match() calls in a pcre_exec() execution. See 42 above.
3048    
3049     44. pcretest has been upgraded by the addition of the -C option. This causes it
3050     to extract all the available output from the new pcre_config() function, and to
3051     output it. The program then exits immediately.
3052    
3053     45. A need has arisen to pass over additional data with calls to pcre_exec() in
3054     order to support additional features. One way would have been to define
3055     pcre_exec2() (for example) with extra arguments, but this would not have been
3056     extensible, and would also have required all calls to the original function to
3057     be mapped to the new one. Instead, I have chosen to extend the mechanism that
3058     is used for passing in "extra" data from pcre_study().
3059    
3060     The pcre_extra structure is now exposed and defined in pcre.h. It currently
3061     contains the following fields:
3062    
3063     flags a bitmap indicating which of the following fields are set
3064     study_data opaque data from pcre_study()
3065     match_limit a way of specifying a limit on match() calls for a specific
3066     call to pcre_exec()
3067     callout_data data for callouts (see 49 below)
3068    
3069     The flag bits are also defined in pcre.h, and are
3070    
3071     PCRE_EXTRA_STUDY_DATA
3072     PCRE_EXTRA_MATCH_LIMIT
3073     PCRE_EXTRA_CALLOUT_DATA
3074    
3075     The pcre_study() function now returns one of these new pcre_extra blocks, with
3076     the actual study data pointed to by the study_data field, and the
3077     PCRE_EXTRA_STUDY_DATA flag set. This can be passed directly to pcre_exec() as
3078     before. That is, this change is entirely upwards-compatible and requires no
3079     change to existing code.
3080    
3081     If you want to pass in additional data to pcre_exec(), you can either place it
3082     in a pcre_extra block provided by pcre_study(), or create your own pcre_extra
3083     block.
3084    
3085     46. pcretest has been extended to test the PCRE_EXTRA_MATCH_LIMIT feature. If a
3086     data string contains the escape sequence \M, pcretest calls pcre_exec() several
3087     times with different match limits, until it finds the minimum value needed for
3088     pcre_exec() to complete. The value is then output. This can be instructive; for
3089     most simple matches the number is quite small, but for pathological cases it
3090     gets very large very quickly.
3091    
3092     47. There's a new option for pcre_fullinfo() called PCRE_INFO_STUDYSIZE. It
3093     returns the size of the data block pointed to by the study_data field in a
3094     pcre_extra block, that is, the value that was passed as the argument to
3095     pcre_malloc() when PCRE was getting memory in which to place the information
3096     created by pcre_study(). The fourth argument should point to a size_t variable.
3097     pcretest has been extended so that this information is shown after a successful
3098     pcre_study() call when information about the compiled regex is being displayed.
3099    
3100     48. Cosmetic change to Makefile: there's no need to have / after $(DESTDIR)
3101     because what follows is always an absolute path. (Later: it turns out that this
3102     is more than cosmetic for MinGW, because it doesn't like empty path
3103     components.)
3104    
3105     49. Some changes have been made to the callout feature (see 28 above):
3106    
3107     (i) A callout function now has three choices for what it returns:
3108    
3109     0 => success, carry on matching
3110     > 0 => failure at this point, but backtrack if possible
3111     < 0 => serious error, return this value from pcre_exec()
3112    
3113     Negative values should normally be chosen from the set of PCRE_ERROR_xxx
3114     values. In particular, returning PCRE_ERROR_NOMATCH forces a standard
3115     "match failed" error. The error number PCRE_ERROR_CALLOUT is reserved for
3116     use by callout functions. It will never be used by PCRE itself.
3117    
3118     (ii) The pcre_extra structure (see 45 above) has a void * field called
3119     callout_data, with corresponding flag bit PCRE_EXTRA_CALLOUT_DATA. The
3120     pcre_callout_block structure has a field of the same name. The contents of
3121     the field passed in the pcre_extra structure are passed to the callout
3122     function in the corresponding field in the callout block. This makes it
3123     easier to use the same callout-containing regex from multiple threads. For
3124     testing, the pcretest program has a new data escape
3125    
3126     \C*n pass the number n (may be negative) as callout_data
3127    
3128     If the callout function in pcretest receives a non-zero value as
3129     callout_data, it returns that value.
3130    
3131     50. Makefile wasn't handling CFLAGS properly when compiling dftables. Also,
3132     there were some redundant $(CFLAGS) in commands that are now specified as
3133     $(LINK), which already includes $(CFLAGS).
3134    
3135     51. Extensions to UTF-8 support are listed below. These all apply when (a) PCRE
3136     has been compiled with UTF-8 support *and* pcre_compile() has been compiled
3137     with the PCRE_UTF8 flag. Patterns that are compiled without that flag assume
3138     one-byte characters throughout. Note that case-insensitive matching applies
3139     only to characters whose values are less than 256. PCRE doesn't support the
3140     notion of cases for higher-valued characters.
3141    
3142     (i) A character class whose characters are all within 0-255 is handled as
3143     a bit map, and the map is inverted for negative classes. Previously, a
3144     character > 255 always failed to match such a class; however it should
3145     match if the class was a negative one (e.g. [^ab]). This has been fixed.
3146    
3147     (ii) A negated character class with a single character < 255 is coded as
3148     "not this character" (OP_NOT). This wasn't working properly when the test
3149     character was multibyte, either singly or repeated.
3150    
3151     (iii) Repeats of multibyte characters are now handled correctly in UTF-8
3152     mode, for example: \x{100}{2,3}.
3153    
3154     (iv) The character escapes \b, \B, \d, \D, \s, \S, \w, and \W (either
3155     singly or repeated) now correctly test multibyte characters. However,
3156     PCRE doesn't recognize any characters with values greater than 255 as
3157     digits, spaces, or word characters. Such characters always match \D, \S,
3158     and \W, and never match \d, \s, or \w.
3159    
3160     (v) Classes may now contain characters and character ranges with values
3161     greater than 255. For example: [ab\x{100}-\x{400}].
3162    
3163     (vi) pcregrep now has a --utf-8 option (synonym -u) which makes it call
3164     PCRE in UTF-8 mode.
3165    
3166     52. The info request value PCRE_INFO_FIRSTCHAR has been renamed
3167     PCRE_INFO_FIRSTBYTE because it is a byte value. However, the old name is
3168     retained for backwards compatibility. (Note that LASTLITERAL is also a byte
3169     value.)
3170    
3171     53. The single man page has become too large. I have therefore split it up into
3172     a number of separate man pages. These also give rise to individual HTML pages;
3173     these are now put in a separate directory, and there is an index.html page that
3174     lists them all. Some hyperlinking between the pages has been installed.
3175    
3176     54. Added convenience functions for handling named capturing parentheses.
3177    
3178     55. Unknown escapes inside character classes (e.g. [\M]) and escapes that
3179     aren't interpreted therein (e.g. [\C]) are literals in Perl. This is now also
3180     true in PCRE, except when the PCRE_EXTENDED option is set, in which case they
3181     are faulted.
3182    
3183     56. Introduced HOST_CC and HOST_CFLAGS which can be set in the environment when
3184     calling configure. These values are used when compiling the dftables.c program
3185     which is run to generate the source of the default character tables. They
3186     default to the values of CC and CFLAGS. If you are cross-compiling PCRE,
3187     you will need to set these values.
3188    
3189     57. Updated the building process for Windows DLL, as provided by Fred Cox.
3190    
3191    
3192     Version 3.9 02-Jan-02
3193 nigel 61 ---------------------
3194    
3195     1. A bit of extraneous text had somehow crept into the pcregrep documentation.
3196    
3197     2. If --disable-static was given, the building process failed when trying to
3198     build pcretest and pcregrep. (For some reason it was using libtool to compile
3199     them, which is not right, as they aren't part of the library.)
3200    
3201    
3202 nigel 59 Version 3.8 18-Dec-01
3203     ---------------------
3204    
3205     1. The experimental UTF-8 code was completely screwed up. It was packing the
3206     bytes in the wrong order. How dumb can you get?
3207    
3208    
3209 nigel 57 Version 3.7 29-Oct-01
3210     ---------------------
3211    
3212     1. In updating pcretest to check change 1 of version 3.6, I screwed up.
3213     This caused pcretest, when used on the test data, to segfault. Unfortunately,
3214     this didn't happen under Solaris 8, where I normally test things.
3215    
3216 nigel 59 2. The Makefile had to be changed to make it work on BSD systems, where 'make'
3217     doesn't seem to recognize that ./xxx and xxx are the same file. (This entry
3218     isn't in ChangeLog distributed with 3.7 because I forgot when I hastily made
3219     this fix an hour or so after the initial 3.7 release.)
3220 nigel 57
3221 nigel 59
3222 nigel 55 Version 3.6 23-Oct-01
3223     ---------------------
3224    
3225     1. Crashed with /(sens|respons)e and \1ibility/ and "sense and sensibility" if
3226     offsets passed as NULL with zero offset count.
3227    
3228     2. The config.guess and config.sub files had not been updated when I moved to
3229     the latest autoconf.
3230    
3231    
3232 nigel 53 Version 3.5 15-Aug-01
3233     ---------------------
3234 nigel 5
3235 nigel 53 1. Added some missing #if !defined NOPOSIX conditionals in pcretest.c that
3236     had been forgotten.
3237    
3238     2. By using declared but undefined structures, we can avoid using "void"
3239     definitions in pcre.h while keeping the internal definitions of the structures
3240     private.
3241    
3242     3. The distribution is now built using autoconf 2.50 and libtool 1.4. From a
3243     user point of view, this means that both static and shared libraries are built
3244     by default, but this can be individually controlled. More of the work of
3245     handling this static/shared cases is now inside libtool instead of PCRE's make
3246     file.
3247    
3248     4. The pcretest utility is now installed along with pcregrep because it is
3249     useful for users (to test regexs) and by doing this, it automatically gets
3250     relinked by libtool. The documentation has been turned into a man page, so
3251     there are now .1, .txt, and .html versions in /doc.
3252    
3253     5. Upgrades to pcregrep:
3254     (i) Added long-form option names like gnu grep.
3255     (ii) Added --help to list all options with an explanatory phrase.
3256     (iii) Added -r, --recursive to recurse into sub-directories.
3257     (iv) Added -f, --file to read patterns from a file.
3258    
3259     6. pcre_exec() was referring to its "code" argument before testing that
3260     argument for NULL (and giving an error if it was NULL).
3261    
3262     7. Upgraded Makefile.in to allow for compiling in a different directory from
3263     the source directory.
3264    
3265     8. Tiny buglet in pcretest: when pcre_fullinfo() was called to retrieve the
3266     options bits, the pointer it was passed was to an int instead of to an unsigned
3267     long int. This mattered only on 64-bit systems.
3268    
3269     9. Fixed typo (3.4/1) in pcre.h again. Sigh. I had changed pcre.h (which is
3270     generated) instead of pcre.in, which it its source. Also made the same change
3271     in several of the .c files.
3272    
3273     10. A new release of gcc defines printf() as a macro, which broke pcretest
3274     because it had an ifdef in the middle of a string argument for printf(). Fixed
3275     by using separate calls to printf().
3276    
3277     11. Added --enable-newline-is-cr and --enable-newline-is-lf to the configure
3278     script, to force use of CR or LF instead of \n in the source. On non-Unix
3279     systems, the value can be set in config.h.
3280    
3281     12. The limit of 200 on non-capturing parentheses is a _nesting_ limit, not an
3282     absolute limit. Changed the text of the error message to make this clear, and
3283     likewise updated the man page.
3284    
3285     13. The limit of 99 on the number of capturing subpatterns has been removed.
3286     The new limit is 65535, which I hope will not be a "real" limit.
3287    
3288    
3289 nigel 51 Version 3.4 22-Aug-00
3290     ---------------------
3291    
3292     1. Fixed typo in pcre.h: unsigned const char * changed to const unsigned char *.
3293    
3294     2. Diagnose condition (?(0) as an error instead of crashing on matching.
3295    
3296    
3297 nigel 49 Version 3.3 01-Aug-00
3298     ---------------------
3299    
3300     1. If an octal character was given, but the value was greater than \377, it
3301     was not getting masked to the least significant bits, as documented. This could
3302     lead to crashes in some systems.
3303    
3304     2. Perl 5.6 (if not earlier versions) accepts classes like [a-\d] and treats
3305     the hyphen as a literal. PCRE used to give an error; it now behaves like Perl.
3306    
3307     3. Added the functions pcre_free_substring() and pcre_free_substring_list().
3308     These just pass their arguments on to (pcre_free)(), but they are provided
3309     because some uses of PCRE bind it to non-C systems that can call its functions,
3310     but cannot call free() or pcre_free() directly.
3311    
3312     4. Add "make test" as a synonym for "make check". Corrected some comments in
3313     the Makefile.
3314    
3315     5. Add $(DESTDIR)/ in front of all the paths in the "install" target in the
3316     Makefile.
3317    
3318     6. Changed the name of pgrep to pcregrep, because Solaris has introduced a
3319     command called pgrep for grepping around the active processes.
3320    
3321     7. Added the beginnings of support for UTF-8 character strings.
3322    
3323     8. Arranged for the Makefile to pass over the settings of CC, CFLAGS, and
3324     RANLIB to ./ltconfig so that they are used by libtool. I think these are all
3325     the relevant ones. (AR is not passed because ./ltconfig does its own figuring
3326     out for the ar command.)
3327    
3328    
3329 nigel 47 Version 3.2 12-May-00
3330     ---------------------
3331    
3332     This is purely a bug fixing release.
3333    
3334     1. If the pattern /((Z)+|A)*/ was matched agained ZABCDEFG it matched Z instead
3335     of ZA. This was just one example of several cases that could provoke this bug,
3336     which was introduced by change 9 of version 2.00. The code for breaking
3337     infinite loops after an iteration that matches an empty string was't working
3338     correctly.
3339    
3340     2. The pcretest program was not imitating Perl correctly for the pattern /a*/g
3341     when matched against abbab (for example). After matching an empty string, it
3342     wasn't forcing anchoring when setting PCRE_NOTEMPTY for the next attempt; this
3343     caused it to match further down the string than it should.
3344    
3345     3. The code contained an inclusion of sys/types.h. It isn't clear why this
3346     was there because it doesn't seem to be needed, and it causes trouble on some
3347     systems, as it is not a Standard C header. It has been removed.
3348    
3349     4. Made 4 silly changes to the source to avoid stupid compiler warnings that
3350     were reported on the Macintosh. The changes were from
3351    
3352     while ((c = *(++ptr)) != 0 && c != '\n');
3353     to
3354     while ((c = *(++ptr)) != 0 && c != '\n') ;
3355    
3356     Totally extraordinary, but if that's what it takes...
3357    
3358     5. PCRE is being used in one environment where neither memmove() nor bcopy() is
3359     available. Added HAVE_BCOPY and an autoconf test for it; if neither
3360     HAVE_MEMMOVE nor HAVE_BCOPY is set, use a built-in emulation function which
3361     assumes the way PCRE uses memmove() (always moving upwards).
3362    
3363     6. PCRE is being used in one environment where strchr() is not available. There
3364     was only one use in pcre.c, and writing it out to avoid strchr() probably gives
3365     faster code anyway.
3366    
3367    
3368 nigel 45 Version 3.1 09-Feb-00
3369     ---------------------
3370    
3371     The only change in this release is the fixing of some bugs in Makefile.in for
3372     the "install" target:
3373    
3374     (1) It was failing to install pcreposix.h.
3375    
3376     (2) It was overwriting the pcre.3 man page with the pcreposix.3 man page.
3377    
3378    
3379 nigel 43 Version 3.0 01-Feb-00
3380     ---------------------
3381 nigel 41
3382     1. Add support for the /+ modifier to perltest (to output $` like it does in
3383     pcretest).
3384    
3385     2. Add support for the /g modifier to perltest.
3386    
3387     3. Fix pcretest so that it behaves even more like Perl for /g when the pattern
3388     matches null strings.
3389    
3390     4. Fix perltest so that it doesn't do unwanted things when fed an empty
3391     pattern. Perl treats empty patterns specially - it reuses the most recent
3392     pattern, which is not what we want. Replace // by /(?#)/ in order to avoid this
3393     effect.
3394    
3395     5. The POSIX interface was broken in that it was just handing over the POSIX
3396     captured string vector to pcre_exec(), but (since release 2.00) PCRE has
3397     required a bigger vector, with some working space on the end. This means that
3398     the POSIX wrapper now has to get and free some memory, and copy the results.
3399    
3400 nigel 43 6. Added some simple autoconf support, placing the test data and the
3401     documentation in separate directories, re-organizing some of the
3402     information files, and making it build pcre-config (a GNU standard). Also added
3403     libtool support for building PCRE as a shared library, which is now the
3404     default.
3405 nigel 41
3406 nigel 43 7. Got rid of the leading zero in the definition of PCRE_MINOR because 08 and
3407     09 are not valid octal constants. Single digits will be used for minor values
3408     less than 10.
3409    
3410     8. Defined REG_EXTENDED and REG_NOSUB as zero in the POSIX header, so that
3411     existing programs that set these in the POSIX interface can use PCRE without
3412     modification.
3413    
3414     9. Added a new function, pcre_fullinfo() with an extensible interface. It can
3415     return all that pcre_info() returns, plus additional data. The pcre_info()
3416     function is retained for compatibility, but is considered to be obsolete.
3417    
3418     10. Added experimental recursion feature (?R) to handle one common case that
3419     Perl 5.6 will be able to do with (?p{...}).
3420    
3421     11. Added support for POSIX character classes like [:alpha:], which Perl is
3422     adopting.
3423    
3424    
3425 nigel 39 Version 2.08 31-Aug-99
3426     ----------------------
3427    
3428     1. When startoffset was not zero and the pattern began with ".*", PCRE was not
3429     trying to match at the startoffset position, but instead was moving forward to
3430     the next newline as if a previous match had failed.
3431    
3432     2. pcretest was not making use of PCRE_NOTEMPTY when repeating for /g and /G,
3433     and could get into a loop if a null string was matched other than at the start
3434     of the subject.
3435    
3436     3. Added definitions of PCRE_MAJOR and PCRE_MINOR to pcre.h so the version can
3437     be distinguished at compile time, and for completeness also added PCRE_DATE.
3438    
3439     5. Added Paul Sokolovsky's minor changes to make it easy to compile a Win32 DLL
3440     in GnuWin32 environments.
3441    
3442    
3443 nigel 37 Version 2.07 29-Jul-99
3444     ----------------------
3445    
3446     1. The documentation is now supplied in plain text form and HTML as well as in
3447     the form of man page sources.
3448    
3449     2. C++ compilers don't like assigning (void *) values to other pointer types.
3450     In particular this affects malloc(). Although there is no problem in Standard
3451     C, I've put in casts to keep C++ compilers happy.
3452    
3453     3. Typo on pcretest.c; a cast of (unsigned char *) in the POSIX regexec() call
3454     should be (const char *).
3455    
3456     4. If NOPOSIX is defined, pcretest.c compiles without POSIX support. This may
3457     be useful for non-Unix systems who don't want to bother with the POSIX stuff.
3458     However, I haven't made this a standard facility. The documentation doesn't
3459     mention it, and the Makefile doesn't support it.
3460    
3461     5. The Makefile now contains an "install" target, with editable destinations at
3462     the top of the file. The pcretest program is not installed.
3463    
3464     6. pgrep -V now gives the PCRE version number and date.
3465    
3466     7. Fixed bug: a zero repetition after a literal string (e.g. /abcde{0}/) was
3467     causing the entire string to be ignored, instead of just the last character.
3468    
3469     8. If a pattern like /"([^\\"]+|\\.)*"/ is applied in the normal way to a
3470     non-matching string, it can take a very, very long time, even for strings of
3471     quite modest length, because of the nested recursion. PCRE now does better in
3472     some of these cases. It does this by remembering the last required literal
3473     character in the pattern, and pre-searching the subject to ensure it is present
3474     before running the real match. In other words, it applies a heuristic to detect
3475     some types of certain failure quickly, and in the above example, if presented
3476     with a string that has no trailing " it gives "no match" very quickly.
3477    
3478     9. A new runtime option PCRE_NOTEMPTY causes null string matches to be ignored;
3479     other alternatives are tried instead.
3480    
3481    
3482 nigel 35 Version 2.06 09-Jun-99
3483     ----------------------
3484    
3485     1. Change pcretest's output for amount of store used to show just the code
3486     space, because the remainder (the data block) varies in size between 32-bit and
3487     64-bit systems.
3488    
3489     2. Added an extra argument to pcre_exec() to supply an offset in the subject to
3490     start matching at. This allows lookbehinds to work when searching for multiple
3491     occurrences in a string.
3492    
3493     3. Added additional options to pcretest for testing multiple occurrences:
3494    
3495     /+ outputs the rest of the string that follows a match
3496     /g loops for multiple occurrences, using the new startoffset argument
3497     /G loops for multiple occurrences by passing an incremented pointer
3498    
3499     4. PCRE wasn't doing the "first character" optimization for patterns starting
3500     with \b or \B, though it was doing it for other lookbehind assertions. That is,
3501     it wasn't noticing that a match for a pattern such as /\bxyz/ has to start with
3502     the letter 'x'. On long subject strings, this gives a significant speed-up.
3503    
3504    
3505 nigel 33 Version 2.05 21-Apr-99
3506     ----------------------
3507    
3508     1. Changed the type of magic_number from int to long int so that it works
3509     properly on 16-bit systems.
3510    
3511     2. Fixed a bug which caused patterns starting with .* not to work correctly
3512     when the subject string contained newline characters. PCRE was assuming
3513     anchoring for such patterns in all cases, which is not correct because .* will
3514     not pass a newline unless PCRE_DOTALL is set. It now assumes anchoring only if
3515     DOTALL is set at top level; otherwise it knows that patterns starting with .*
3516     must be retried after every newline in the subject.
3517    
3518    
3519 nigel 31 Version 2.04 18-Feb-99
3520     ----------------------
3521    
3522     1. For parenthesized subpatterns with repeats whose minimum was zero, the
3523     computation of the store needed to hold the pattern was incorrect (too large).
3524     If such patterns were nested a few deep, this could multiply and become a real
3525     problem.
3526    
3527     2. Added /M option to pcretest to show the memory requirement of a specific
3528     pattern. Made -m a synonym of -s (which does this globally) for compatibility.
3529    
3530     3. Subpatterns of the form (regex){n,m} (i.e. limited maximum) were being
3531     compiled in such a way that the backtracking after subsequent failure was
3532     pessimal. Something like (a){0,3} was compiled as (a)?(a)?(a)? instead of
3533     ((a)((a)(a)?)?)? with disastrous performance if the maximum was of any size.
3534    
3535    
3536 nigel 29 Version 2.03 02-Feb-99
3537     ----------------------
3538    
3539     1. Fixed typo and small mistake in man page.
3540    
3541 nigel 31 2. Added 4th condition (GPL supersedes if conflict) and created separate
3542     LICENCE file containing the conditions.
3543 nigel 29
3544     3. Updated pcretest so that patterns such as /abc\/def/ work like they do in
3545     Perl, that is the internal \ allows the delimiter to be included in the
3546     pattern. Locked out the use of \ as a delimiter. If \ immediately follows
3547     the final delimiter, add \ to the end of the pattern (to test the error).
3548    
3549     4. Added the convenience functions for extracting substrings after a successful
3550     match. Updated pcretest to make it able to test these functions.
3551    
3552    
3553 nigel 27 Version 2.02 14-Jan-99
3554     ----------------------
3555    
3556     1. Initialized the working variables associated with each extraction so that
3557     their saving and restoring doesn't refer to uninitialized store.
3558    
3559     2. Put dummy code into study.c in order to trick the optimizer of the IBM C
3560     compiler for OS/2 into generating correct code. Apparently IBM isn't going to
3561     fix the problem.
3562    
3563     3. Pcretest: the timing code wasn't using LOOPREPEAT for timing execution
3564     calls, and wasn't printing the correct value for compiling calls. Increased the
3565     default value of LOOPREPEAT, and the number of significant figures in the
3566     times.
3567    
3568     4. Changed "/bin/rm" in the Makefile to "-rm" so it works on Windows NT.
3569    
3570     5. Renamed "deftables" as "dftables" to get it down to 8 characters, to avoid
3571     a building problem on Windows NT with a FAT file system.
3572    
3573    
3574 nigel 25 Version 2.01 21-Oct-98
3575     ----------------------
3576    
3577     1. Changed the API for pcre_compile() to allow for the provision of a pointer
3578     to character tables built by pcre_maketables() in the current locale. If NULL
3579     is passed, the default tables are used.
3580    
3581    
3582 nigel 23 Version 2.00 24-Sep-98
3583 nigel 21 ----------------------
3584    
3585 nigel 23 1. Since the (>?) facility is in Perl 5.005, don't require PCRE_EXTRA to enable
3586     it any more.
3587    
3588     2. Allow quantification of (?>) groups, and make it work correctly.
3589    
3590     3. The first character computation wasn't working for (?>) groups.
3591    
3592     4. Correct the implementation of \Z (it is permitted to match on the \n at the
3593     end of the subject) and add 5.005's \z, which really does match only at the
3594     very end of the subject.
3595    
3596     5. Remove the \X "cut" facility; Perl doesn't have it, and (?> is neater.
3597    
3598     6. Remove the ability to specify CASELESS, MULTILINE, DOTALL, and
3599     DOLLAR_END_ONLY at runtime, to make it possible to implement the Perl 5.005
3600     localized options. All options to pcre_study() were also removed.
3601    
3602     7. Add other new features from 5.005:
3603    
3604     $(?<= positive lookbehind
3605     $(?<! negative lookbehind
3606     (?imsx-imsx) added the unsetting capability
3607     such a setting is global if at outer level; local otherwise
3608     (?imsx-imsx:) non-capturing groups with option setting
3609     (?(cond)re|re) conditional pattern matching
3610    
3611     A backreference to itself in a repeated group matches the previous
3612     captured string.
3613    
3614     8. General tidying up of studying (both automatic and via "study")
3615     consequential on the addition of new assertions.
3616    
3617     9. As in 5.005, unlimited repeated groups that could match an empty substring
3618     are no longer faulted at compile time. Instead, the loop is forcibly broken at
3619     runtime if any iteration does actually match an empty substring.
3620    
3621     10. Include the RunTest script in the distribution.
3622    
3623     11. Added tests from the Perl 5.005_02 distribution. This showed up a few
3624     discrepancies, some of which were old and were also with respect to 5.004. They
3625     have now been fixed.
3626    
3627    
3628     Version 1.09 28-Apr-98
3629     ----------------------
3630    
3631 nigel 21 1. A negated single character class followed by a quantifier with a minimum
3632     value of one (e.g. [^x]{1,6} ) was not compiled correctly. This could lead to
3633     program crashes, or just wrong answers. This did not apply to negated classes
3634     containing more than one character, or to minima other than one.
3635    
3636    
3637 nigel 19 Version 1.08 27-Mar-98
3638     ----------------------
3639    
3640     1. Add PCRE_UNGREEDY to invert the greediness of quantifiers.
3641    
3642     2. Add (?U) and (?X) to set PCRE_UNGREEDY and PCRE_EXTRA respectively. The
3643     latter must appear before anything that relies on it in the pattern.
3644    
3645    
3646 nigel 17 Version 1.07 16-Feb-98
3647     ----------------------
3648    
3649     1. A pattern such as /((a)*)*/ was not being diagnosed as in error (unlimited
3650     repeat of a potentially empty string).
3651    
3652    
3653 nigel 15 Version 1.06 23-Jan-98
3654     ----------------------
3655    
3656     1. Added Markus Oberhumer's little patches for C++.
3657    
3658     2. Literal strings longer than 255 characters were broken.
3659    
3660    
3661 nigel 13 Version 1.05 23-Dec-97
3662     ----------------------
3663    
3664     1. Negated character classes containing more than one character were failing if
3665     PCRE_CASELESS was set at run time.
3666    
3667    
3668 nigel 11 Version 1.04 19-Dec-97
3669     ----------------------
3670    
3671     1. Corrected the man page, where some "const" qualifiers had been omitted.
3672    
3673     2. Made debugging output print "{0,xxx}" instead of just "{,xxx}" to agree with
3674     input syntax.
3675    
3676     3. Fixed memory leak which occurred when a regex with back references was
3677     matched with an offsets vector that wasn't big enough. The temporary memory
3678     that is used in this case wasn't being freed if the match failed.
3679    
3680     4. Tidied pcretest to ensure it frees memory that it gets.
3681    
3682     5. Temporary memory was being obtained in the case where the passed offsets
3683     vector was exactly big enough.
3684    
3685     6. Corrected definition of offsetof() from change 5 below.
3686    
3687     7. I had screwed up change 6 below and broken the rules for the use of
3688     setjmp(). Now fixed.
3689    
3690    
3691 nigel 9 Version 1.03 18-Dec-97
3692     ----------------------
3693    
3694     1. A erroneous regex with a missing opening parenthesis was correctly
3695     diagnosed, but PCRE attempted to access brastack[-1], which could cause crashes
3696     on some systems.
3697    
3698     2. Replaced offsetof(real_pcre, code) by offsetof(real_pcre, code[0]) because
3699     it was reported that one broken compiler failed on the former because "code" is
3700     also an independent variable.
3701    
3702     3. The erroneous regex a[]b caused an array overrun reference.
3703    
3704     4. A regex ending with a one-character negative class (e.g. /[^k]$/) did not
3705     fail on data ending with that character. (It was going on too far, and checking
3706     the next character, typically a binary zero.) This was specific to the
3707     optimized code for single-character negative classes.
3708    
3709     5. Added a contributed patch from the TIN world which does the following:
3710    
3711     + Add an unde