ViewVC logotype

Contents of /code/trunk/ChangeLog

Parent Directory Parent Directory | Revision Log Revision Log

Revision 552 - (hide annotations) (download)
Wed Oct 13 10:15:41 2010 UTC (4 years, 5 months ago) by ph10
File size: 190635 byte(s)
Fix \s bug in character classes (always removing VT).

1 nigel 3 ChangeLog for PCRE
2     ------------------
4 ph10 550 Version 8.11 10-Oct-2010
5     ------------------------
7     1. (*THEN) was not working properly if there were untried alternatives prior
8     to it in the current branch. For example, in ((a|b)(*THEN)(*F)|c..) it
9     backtracked to try for "b" instead of moving to the next alternative branch
10     at the same level (in this case, to look for "c"). The Perl documentation
11     is clear that when (*THEN) is backtracked onto, it goes to the "next
12     alternative in the innermost enclosing group".
13 ph10 551
14     2. (*COMMIT) was not overriding (*THEN), as it does in Perl. In a pattern
15     such as (A(*COMMIT)B(*THEN)C|D) any failure after matching A should
16     result in overall failure. Similarly, (*COMMIT) now overrides (*PRUNE) and
17     (*SKIP), (*SKIP) overrides (*PRUNE) and (*THEN), and (*PRUNE) overrides
18     (*THEN).
19 ph10 552
20     3. If \s appeared in a character class, it removed the VT character from
21     the class, even if it had been included by some previous item, for example
22     in [\x00-\xff\s]. (This was a bug related to the fact that VT is not part
23     of \s, but is part of the POSIX "space" class.)
24 ph10 550
26 ph10 548 Version 8.10 25-Jun-2010
27 ph10 510 ------------------------
29 ph10 512 1. Added support for (*MARK:ARG) and for ARG additions to PRUNE, SKIP, and
30 ph10 510 THEN.
32 ph10 512 2. (*ACCEPT) was not working when inside an atomic group.
33 ph10 510
34 ph10 535 3. Inside a character class, \B is treated as a literal by default, but
35     faulted if PCRE_EXTRA is set. This mimics Perl's behaviour (the -w option
36 ph10 513 causes the error). The code is unchanged, but I tidied the documentation.
37 ph10 535
38     4. Inside a character class, PCRE always treated \R and \X as literals,
39 ph10 513 whereas Perl faults them if its -w option is set. I have changed PCRE so
40     that it faults them when PCRE_EXTRA is set.
41 ph10 535
42 ph10 514 5. Added support for \N, which always matches any character other than
43     newline. (It is the same as "." when PCRE_DOTALL is not set.)
44 ph10 535
45 ph10 515 6. When compiling pcregrep with newer versions of gcc which may have
46 ph10 535 FORTIFY_SOURCE set, several warnings "ignoring return value of 'fwrite',
47 ph10 515 declared with attribute warn_unused_result" were given. Just casting the
48 ph10 535 result to (void) does not stop the warnings; a more elaborate fudge is
49     needed. I've used a macro to implement this.
50 ph10 512
51 ph10 535 7. Minor change to pcretest.c to avoid a compiler warning.
53 ph10 517 8. Added four artifical Unicode properties to help with an option to make
54 ph10 518 \s etc use properties (see next item). The new properties are: Xan
55     (alphanumeric), Xsp (Perl space), Xps (POSIX space), and Xwd (word).
56 ph10 535
57 ph10 518 9. Added PCRE_UCP to make \b, \d, \s, \w, and certain POSIX character classes
58 ph10 535 use Unicode properties. (*UCP) at the start of a pattern can be used to set
59 ph10 518 this option. Modified pcretest to add /W to test this facility. Added
60     REG_UCP to make it available via the POSIX interface.
61 ph10 513
62 ph10 535 10. Added --line-buffered to pcregrep.
64     11. In UTF-8 mode, if a pattern that was compiled with PCRE_CASELESS was
65     studied, and the match started with a letter with a code point greater than
66 ph10 520 127 whose first byte was different to the first byte of the other case of
67 ph10 545 the letter, the other case of this starting letter was not recognized
68 ph10 536 (#976).
69 ph10 535
70 ph10 536 12. If a pattern that was studied started with a repeated Unicode property
71 ph10 523 test, for example, \p{Nd}+, there was the theoretical possibility of
72 ph10 535 setting up an incorrect bitmap of starting bytes, but fortunately it could
73     not have actually happened in practice until change 8 above was made (it
74 ph10 523 added property types that matched character-matching opcodes).
75 ph10 535
76 ph10 536 13. pcre_study() now recognizes \h, \v, and \R when constructing a bit map of
77 ph10 535 possible starting bytes for non-anchored patterns.
79 ph10 536 14. Extended the "auto-possessify" feature of pcre_compile(). It now recognizes
80 ph10 535 \R, and also a number of cases that involve Unicode properties, both
81 ph10 532 explicit and implicit when PCRE_UCP is set.
82 ph10 529
83 ph10 536 15. If a repeated Unicode property match (e.g. \p{Lu}*) was used with non-UTF-8
84 ph10 535 input, it could crash or give wrong results if characters with values
85     greater than 0xc0 were present in the subject string. (Detail: it assumed
86 ph10 529 UTF-8 input when processing these items.)
87 ph10 535
88 ph10 536 16. Added a lot of (int) casts to avoid compiler warnings in systems where
89     size_t is 64-bit (#991).
90 ph10 535
91 ph10 536 17. Added a check for running out of memory when PCRE is compiled with
92     --disable-stack-for-recursion (#990).
93 ph10 545
94 ph10 537 18. If the last data line in a file for pcretest does not have a newline on
95 ph10 545 the end, a newline was missing in the output.
97     19. The default pcre_chartables.c file recognizes only ASCII characters (values
98     less than 128) in its various bitmaps. However, there is a facility for
99     generating tables according to the current locale when PCRE is compiled. It
100     turns out that in some environments, 0x85 and 0xa0, which are Unicode space
101     characters, are recognized by isspace() and therefore were getting set in
102 ph10 541 these tables, and indeed these tables seem to approximate to ISO 8859. This
103     caused a problem in UTF-8 mode when pcre_study() was used to create a list
104     of bytes that can start a match. For \s, it was including 0x85 and 0xa0,
105     which of course cannot start UTF-8 characters. I have changed the code so
106     that only real ASCII characters (less than 128) and the correct starting
107 ph10 545 bytes for UTF-8 encodings are set for characters greater than 127 when in
108 ph10 541 UTF-8 mode. (When PCRE_UCP is set - see 9 above - the code is different
109     altogether.)
110 ph10 545
111 ph10 541 20. Added the /T option to pcretest so as to be able to run tests with non-
112     standard character tables, thus making it possible to include the tests
113 ph10 545 used for 19 above in the standard set of tests.
115     21. A pattern such as (?&t)(?#()(?(DEFINE)(?<t>a)) which has a forward
116     reference to a subpattern the other side of a comment that contains an
117     opening parenthesis caused either an internal compiling error, or a
118 ph10 544 reference to the wrong subpattern.
119 ph10 517
120 ph10 520
121 ph10 508 Version 8.02 19-Mar-2010
122 ph10 491 ------------------------
124     1. The Unicode data tables have been updated to Unicode 5.2.0.
126 ph10 492 2. Added the option --libs-cpp to pcre-config, but only when C++ support is
127     configured.
128 ph10 507
129 ph10 493 3. Updated the licensing terms in the pcregexp.pas file, as agreed with the
130 ph10 507 original author of that file, following a query about its status.
132     4. On systems that do not have stdint.h (e.g. Solaris), check for and include
133 ph10 494 inttypes.h instead. This fixes a bug that was introduced by change 8.01/8.
134 ph10 507
135 ph10 496 5. A pattern such as (?&t)*+(?(DEFINE)(?<t>.)) which has a possessive
136     quantifier applied to a forward-referencing subroutine call, could compile
137     incorrect code or give the error "internal error: previously-checked
138     referenced subpattern not found".
139 ph10 507
140     6. Both MS Visual Studio and Symbian OS have problems with initializing
141 ph10 497 variables to point to external functions. For these systems, therefore,
142     pcre_malloc etc. are now initialized to local functions that call the
143     relevant global functions.
144 ph10 507
145 ph10 498 7. There were two entries missing in the vectors called coptable and poptable
146     in pcre_dfa_exec.c. This could lead to memory accesses outsize the vectors.
147 ph10 507 I've fixed the data, and added a kludgy way of testing at compile time that
148     the lengths are correct (equal to the number of opcodes).
150     8. Following on from 7, I added a similar kludge to check the length of the
151     eint vector in pcreposix.c.
153     9. Error texts for pcre_compile() are held as one long string to avoid too
154     much relocation at load time. To find a text, the string is searched,
155 ph10 499 counting zeros. There was no check for running off the end of the string,
156     which could happen if a new error number was added without updating the
157 ph10 507 string.
159     10. \K gave a compile-time error if it appeared in a lookbehind assersion.
161 ph10 500 11. \K was not working if it appeared in an atomic group or in a group that
162 ph10 507 was called as a "subroutine", or in an assertion. Perl 5.11 documents that
163     \K is "not well defined" if used in an assertion. PCRE now accepts it if
164     the assertion is positive, but not if it is negative.
166 ph10 501 12. Change 11 fortuitously reduced the size of the stack frame used in the
167 ph10 507 "match()" function of pcre_exec.c by one pointer. Forthcoming
168     implementation of support for (*MARK) will need an extra pointer on the
169 ph10 501 stack; I have reserved it now, so that the stack frame size does not
170     decrease.
171 ph10 507
172     13. A pattern such as (?P<L1>(?P<L2>0)|(?P>L2)(?P>L1)) in which the only other
173     item in branch that calls a recursion is a subroutine call - as in the
174 ph10 503 second branch in the above example - was incorrectly given the compile-
175     time error "recursive call could loop indefinitely" because pcre_compile()
176 ph10 507 was not correctly checking the subroutine for matching a non-empty string.
178 ph10 505 14. The checks for overrunning compiling workspace could trigger after an
179 ph10 507 overrun had occurred. This is a "should never occur" error, but it can be
180 ph10 505 triggered by pathological patterns such as hundreds of nested parentheses.
181 ph10 507 The checks now trigger 100 bytes before the end of the workspace.
182 ph10 491
183 ph10 507 15. Fix typo in configure.ac: "srtoq" should be "strtoq".
184 ph10 492
185 ph10 507
186 ph10 489 Version 8.01 19-Jan-2010
187     ------------------------
188 ph10 471
189     1. If a pattern contained a conditional subpattern with only one branch (in
190 ph10 489 particular, this includes all (*DEFINE) patterns), a call to pcre_study()
191 ph10 473 computed the wrong minimum data length (which is of course zero for such
192 ph10 489 subpatterns). This could cause incorrect "no match" results.
193 ph10 473
194     2. For patterns such as (?i)a(?-i)b|c where an option setting at the start of
195 ph10 472 the pattern is reset in the first branch, pcre_compile() failed with
196     "internal error: code overflow at offset...". This happened only when
197 ph10 473 the reset was to the original external option setting. (An optimization
198     abstracts leading options settings into an external setting, which was the
199 ph10 472 cause of this.)
200 ph10 471
201 ph10 473 3. A pattern such as ^(?!a(*SKIP)b) where a negative assertion contained one
202     of the verbs SKIP, PRUNE, or COMMIT, did not work correctly. When the
203     assertion pattern did not match (meaning that the assertion was true), it
204     was incorrectly treated as false if the SKIP had been reached during the
205     matching. This also applied to assertions used as conditions.
206 ph10 471
207 ph10 473 4. If an item that is not supported by pcre_dfa_exec() was encountered in an
208     assertion subpattern, including such a pattern used as a condition,
209     unpredictable results occurred, instead of the error return
211 ph10 478
212 ph10 474 5. The C++ GlobalReplace function was not working like Perl for the special
213     situation when an empty string is matched. It now does the fancy magic
214 ph10 478 stuff that is necessary.
216     6. In pcre_internal.h, obsolete includes to setjmp.h and stdarg.h have been
217 ph10 475 removed. (These were left over from very, very early versions of PCRE.)
218 ph10 478
219 ph10 475 7. Some cosmetic changes to the code to make life easier when compiling it
220     as part of something else:
221 ph10 478
222     (a) Change DEBUG to PCRE_DEBUG.
224     (b) In pcre_compile(), rename the member of the "branch_chain" structure
225     called "current" as "current_branch", to prevent a collision with the
226 ph10 475 Linux macro when compiled as a kernel module.
227 ph10 478
228     (c) In pcre_study(), rename the function set_bit() as set_table_bit(), to
229     prevent a collision with the Linux macro when compiled as a kernel
230 ph10 475 module.
231 ph10 478
232 ph10 475 8. In pcre_compile() there are some checks for integer overflows that used to
233     cast potentially large values to (double). This has been changed to that
234 ph10 478 when building, a check for int64_t is made, and if it is found, it is used
235     instead, thus avoiding the use of floating point arithmetic. (There is no
236     other use of FP in PCRE.) If int64_t is not found, the fallback is to
237     double.
239     9. Added two casts to avoid signed/unsigned warnings from VS Studio Express
240 ph10 476 2005 (difference between two addresses compared to an unsigned value).
241 ph10 473
242 ph10 478 10. Change the standard AC_CHECK_LIB test for libbz2 in configure.ac to a
243     custom one, because of the following reported problem in Windows:
244 ph10 473
245 ph10 478 - libbz2 uses the Pascal calling convention (WINAPI) for the functions
246 ph10 489 under Win32.
247 ph10 478 - The standard autoconf AC_CHECK_LIB fails to include "bzlib.h",
248 ph10 489 therefore missing the function definition.
249 ph10 478 - The compiler thus generates a "C" signature for the test function.
250     - The linker fails to find the "C" function.
251     - PCRE fails to configure if asked to do so against libbz2.
252 ph10 486
253 ph10 479 11. When running libtoolize from libtool-2.2.6b as part of autogen.sh, these
254     messages were output:
255 ph10 478
256 ph10 479 Consider adding `AC_CONFIG_MACRO_DIR([m4])' to configure.ac and
257     rerunning libtoolize, to keep the correct libtool macros in-tree.
258     Consider adding `-I m4' to ACLOCAL_AMFLAGS in Makefile.am.
259 ph10 486
260 ph10 479 I have done both of these things.
261 ph10 486
262 ph10 480 12. Although pcre_dfa_exec() does not use nearly as much stack as pcre_exec()
263 ph10 486 most of the time, it *can* run out if it is given a pattern that contains a
264     runaway infinite recursion. I updated the discussion in the pcrestack man
265     page.
267     13. Now that we have gone to the x.xx style of version numbers, the minor
268     version may start with zero. Using 08 or 09 is a bad idea because users
269     might check the value of PCRE_MINOR in their code, and 08 or 09 may be
270     interpreted as invalid octal numbers. I've updated the previous comment in
271     configure.ac, and also added a check that gives an error if 08 or 09 are
272 ph10 481 used.
273 ph10 486
274 ph10 482 14. Change 8.00/11 was not quite complete: code had been accidentally omitted,
275 ph10 483 causing partial matching to fail when the end of the subject matched \W
276     in a UTF-8 pattern where \W was quantified with a minimum of 3.
277 ph10 486
278 ph10 483 15. There were some discrepancies between the declarations in pcre_internal.h
279 ph10 486 of _pcre_is_newline(), _pcre_was_newline(), and _pcre_valid_utf8() and
280 ph10 483 their definitions. The declarations used "const uschar *" and the
281     definitions used USPTR. Even though USPTR is normally defined as "const
282     unsigned char *" (and uschar is typedeffed as "unsigned char"), it was
283     reported that: "This difference in casting confuses some C++ compilers, for
284     example, SunCC recognizes above declarations as different functions and
285 ph10 486 generates broken code for hbpcre." I have changed the declarations to use
286 ph10 483 USPTR.
287 ph10 486
288     16. GNU libtool is named differently on some systems. The autogen.sh script now
289 ph10 484 tries several variants such as glibtoolize (MacOSX) and libtoolize1x
290     (FreeBSD).
291 ph10 478
292 ph10 486 17. Applied Craig's patch that fixes an HP aCC compile error in pcre 8.00
293     (strtoXX undefined when compiling pcrecpp.cc). The patch contains this
294     comment: "Figure out how to create a longlong from a string: strtoll and
295     equivalent. It's not enough to call AC_CHECK_FUNCS: hpux has a strtoll, for
296     instance, but it only takes 2 args instead of 3!"
297 ph10 479
298 ph10 488 18. A subtle bug concerned with back references has been fixed by a change of
299     specification, with a corresponding code fix. A pattern such as
300     ^(xa|=?\1a)+$ which contains a back reference inside the group to which it
301     refers, was giving matches when it shouldn't. For example, xa=xaaa would
302     match that pattern. Interestingly, Perl (at least up to 5.11.3) has the
303     same bug. Such groups have to be quantified to be useful, or contained
304     inside another quantified group. (If there's no repetition, the reference
305     can never match.) The problem arises because, having left the group and
306     moved on to the rest of the pattern, a later failure that backtracks into
307     the group uses the captured value from the final iteration of the group
308     rather than the correct earlier one. I have fixed this in PCRE by forcing
309     any group that contains a reference to itself to be an atomic group; that
310     is, there cannot be any backtracking into it once it has completed. This is
311     similar to recursive and subroutine calls.
312 ph10 486
313 ph10 488
314 ph10 469 Version 8.00 19-Oct-09
315 ph10 418 ----------------------
317     1. The table for translating pcre_compile() error codes into POSIX error codes
318 ph10 461 was out-of-date, and there was no check on the pcre_compile() error code
319     being within the table. This could lead to an OK return being given in
320 ph10 418 error.
321 ph10 461
322     2. Changed the call to open a subject file in pcregrep from fopen(pathname,
323     "r") to fopen(pathname, "rb"), which fixed a problem with some of the tests
324     in a Windows environment.
326 ph10 420 3. The pcregrep --count option prints the count for each file even when it is
327     zero, as does GNU grep. However, pcregrep was also printing all files when
328     --files-with-matches was added. Now, when both options are given, it prints
329     counts only for those files that have at least one match. (GNU grep just
330 ph10 461 prints the file name in this circumstance, but including the count seems
331     more useful - otherwise, why use --count?) Also ensured that the
332 ph10 420 combination -clh just lists non-zero counts, with no names.
333 ph10 461
334     4. The long form of the pcregrep -F option was incorrectly implemented as
335     --fixed_strings instead of --fixed-strings. This is an incompatible change,
336     but it seems right to fix it, and I didn't think it was worth preserving
337     the old behaviour.
339     5. The command line items --regex=pattern and --regexp=pattern were not
340 ph10 422 recognized by pcregrep, which required --regex pattern or --regexp pattern
341 ph10 461 (with a space rather than an '='). The man page documented the '=' forms,
342 ph10 422 which are compatible with GNU grep; these now work.
343 ph10 461
344     6. No libpcreposix.pc file was created for pkg-config; there was just
345 ph10 423 libpcre.pc and libpcrecpp.pc. The omission has been rectified.
346 ph10 461
347 ph10 425 7. Added #ifndef SUPPORT_UCP into the pcre_ucd.c module, to reduce its size
348 ph10 461 when UCP support is not needed, by modifying the Python script that
349 ph10 425 generates it from Unicode data files. This should not matter if the module
350     is correctly used as a library, but I received one complaint about 50K of
351     unwanted data. My guess is that the person linked everything into his
352     program rather than using a library. Anyway, it does no harm.
353 ph10 461
354 ph10 426 8. A pattern such as /\x{123}{2,2}+/8 was incorrectly compiled; the trigger
355 ph10 461 was a minimum greater than 1 for a wide character in a possessive
356 ph10 431 repetition. The same bug could also affect patterns like /(\x{ff}{0,2})*/8
357     which had an unlimited repeat of a nested, fixed maximum repeat of a wide
358     character. Chaos in the form of incorrect output or a compiling loop could
359     result.
360 ph10 461
361 ph10 426 9. The restrictions on what a pattern can contain when partial matching is
362 ph10 461 requested for pcre_exec() have been removed. All patterns can now be
363 ph10 426 partially matched by this function. In addition, if there are at least two
364 ph10 435 slots in the offset vector, the offset of the earliest inspected character
365     for the match and the offset of the end of the subject are set in them when
366 ph10 461 PCRE_ERROR_PARTIAL is returned.
368 ph10 428 10. Partial matching has been split into two forms: PCRE_PARTIAL_SOFT, which is
369     synonymous with PCRE_PARTIAL, for backwards compatibility, and
370     PCRE_PARTIAL_HARD, which causes a partial match to supersede a full match,
371 ph10 462 and may be more useful for multi-segment matching.
372 ph10 461
373     11. Partial matching with pcre_exec() is now more intuitive. A partial match
374     used to be given if ever the end of the subject was reached; now it is
375     given only if matching could not proceed because another character was
376     needed. This makes a difference in some odd cases such as Z(*FAIL) with the
377     string "Z", which now yields "no match" instead of "partial match". In the
378     case of pcre_dfa_exec(), "no match" is given if every matching path for the
379     final character ended with (*FAIL).
381 ph10 428 12. Restarting a match using pcre_dfa_exec() after a partial match did not work
382 ph10 461 if the pattern had a "must contain" character that was already found in the
383 ph10 428 earlier partial match, unless partial matching was again requested. For
384     example, with the pattern /dog.(body)?/, the "must contain" character is
385     "g". If the first part-match was for the string "dog", restarting with
386 ph10 435 "sbody" failed. This bug has been fixed.
387 ph10 461
388     13. The string returned by pcre_dfa_exec() after a partial match has been
389     changed so that it starts at the first inspected character rather than the
390     first character of the match. This makes a difference only if the pattern
391     starts with a lookbehind assertion or \b or \B (\K is not supported by
392     pcre_dfa_exec()). It's an incompatible change, but it makes the two
393 ph10 435 matching functions compatible, and I think it's the right thing to do.
394 ph10 461
395 ph10 435 14. Added a pcredemo man page, created automatically from the pcredemo.c file,
396 ph10 461 so that the demonstration program is easily available in environments where
397     PCRE has not been installed from source.
399 ph10 435 15. Arranged to add -DPCRE_STATIC to cflags in libpcre.pc, libpcreposix.cp,
400 ph10 430 libpcrecpp.pc and pcre-config when PCRE is not compiled as a shared
401     library.
402 ph10 461
403 ph10 435 16. Added REG_UNGREEDY to the pcreposix interface, at the request of a user.
404 ph10 432 It maps to PCRE_UNGREEDY. It is not, of course, POSIX-compatible, but it
405 ph10 461 is not the first non-POSIX option to be added. Clearly some people find
406 ph10 433 these options useful.
407 ph10 461
408     17. If a caller to the POSIX matching function regexec() passes a non-zero
409 ph10 438 value for nmatch with a NULL value for pmatch, the value of
410 ph10 461 nmatch is forced to zero.
412 ph10 437 18. RunGrepTest did not have a test for the availability of the -u option of
413 ph10 461 the diff command, as RunTest does. It now checks in the same way as
414 ph10 437 RunTest, and also checks for the -b option.
415 ph10 461
416 ph10 438 19. If an odd number of negated classes containing just a single character
417     interposed, within parentheses, between a forward reference to a named
418 ph10 461 subpattern and the definition of the subpattern, compilation crashed with
419     an internal error, complaining that it could not find the referenced
420 ph10 438 subpattern. An example of a crashing pattern is /(?&A)(([^m])(?<A>))/.
421 ph10 461 [The bug was that it was starting one character too far in when skipping
422     over the character class, thus treating the ] as data rather than
423     terminating the class. This meant it could skip too much.]
425 ph10 442 20. Added PCRE_NOTEMPTY_ATSTART in order to be able to correctly implement the
426 ph10 461 /g option in pcretest when the pattern contains \K, which makes it possible
427 ph10 442 to have an empty string match not at the start, even when the pattern is
428 ph10 461 anchored. Updated pcretest and pcredemo to use this option.
430 ph10 446 21. If the maximum number of capturing subpatterns in a recursion was greater
431 ph10 461 than the maximum at the outer level, the higher number was returned, but
432     with unset values at the outer level. The correct (outer level) value is
433 ph10 446 now given.
434 ph10 461
435 ph10 447 22. If (*ACCEPT) appeared inside capturing parentheses, previous releases of
436     PCRE did not set those parentheses (unlike Perl). I have now found a way to
437     make it do so. The string so far is captured, making this feature
438     compatible with Perl.
439 ph10 461
440     23. The tests have been re-organized, adding tests 11 and 12, to make it
441 ph10 448 possible to check the Perl 5.10 features against Perl 5.10.
442 ph10 461
443 ph10 454 24. Perl 5.10 allows subroutine calls in lookbehinds, as long as the subroutine
444 ph10 461 pattern matches a fixed length string. PCRE did not allow this; now it
445     does. Neither allows recursion.
447     25. I finally figured out how to implement a request to provide the minimum
448     length of subject string that was needed in order to match a given pattern.
449     (It was back references and recursion that I had previously got hung up
450     on.) This code has now been added to pcre_study(); it finds a lower bound
451 ph10 455 to the length of subject needed. It is not necessarily the greatest lower
452     bound, but using it to avoid searching strings that are too short does give
453     some useful speed-ups. The value is available to calling programs via
454     pcre_fullinfo().
455 ph10 461
456 ph10 455 26. While implementing 25, I discovered to my embarrassment that pcretest had
457     not been passing the result of pcre_study() to pcre_dfa_exec(), so the
458     study optimizations had never been tested with that matching function.
459     Oops. What is worse, even when it was passed study data, there was a bug in
460     pcre_dfa_exec() that meant it never actually used it. Double oops. There
461     were also very few tests of studied patterns with pcre_dfa_exec().
462 ph10 461
463 ph10 457 27. If (?| is used to create subpatterns with duplicate numbers, they are now
464     allowed to have the same name, even if PCRE_DUPNAMES is not set. However,
465     on the other side of the coin, they are no longer allowed to have different
466     names, because these cannot be distinguished in PCRE, and this has caused
467     confusion. (This is a difference from Perl.)
468 ph10 461
469     28. When duplicate subpattern names are present (necessarily with different
470     numbers, as required by 27 above), and a test is made by name in a
471     conditional pattern, either for a subpattern having been matched, or for
472     recursion in such a pattern, all the associated numbered subpatterns are
473 ph10 459 tested, and the overall condition is true if the condition is true for any
474     one of them. This is the way Perl works, and is also more like the way
475     testing by number works.
476 ph10 418
477 ph10 461
478 ph10 412 Version 7.9 11-Apr-09
479 ph10 376 ---------------------
481 ph10 392 1. When building with support for bzlib/zlib (pcregrep) and/or readline
482 ph10 376 (pcretest), all targets were linked against these libraries. This included
483     libpcre, libpcreposix, and libpcrecpp, even though they do not use these
484     libraries. This caused unwanted dependencies to be created. This problem
485 ph10 392 has been fixed, and now only pcregrep is linked with bzlib/zlib and only
486 ph10 376 pcretest is linked with readline.
487 ph10 392
488 ph10 376 2. The "typedef int BOOL" in pcre_internal.h that was included inside the
489     "#ifndef FALSE" condition by an earlier change (probably 7.8/18) has been
490     moved outside it again, because FALSE and TRUE are already defined in AIX,
491     but BOOL is not.
492 ph10 392
493     3. The pcre_config() function was treating the PCRE_MATCH_LIMIT and
494 ph10 409 PCRE_MATCH_LIMIT_RECURSION values as ints, when they should be long ints.
495 ph10 392
496 ph10 376 4. The pcregrep documentation said spaces were inserted as well as colons (or
497     hyphens) following file names and line numbers when outputting matching
498 ph10 392 lines. This is not true; no spaces are inserted. I have also clarified the
499 ph10 376 wording for the --colour (or --color) option.
500 ph10 392
501 ph10 378 5. In pcregrep, when --colour was used with -o, the list of matching strings
502     was not coloured; this is different to GNU grep, so I have changed it to be
503     the same.
504 ph10 392
505 ph10 378 6. When --colo(u)r was used in pcregrep, only the first matching substring in
506 ph10 392 each matching line was coloured. Now it goes on to look for further matches
507     of any of the test patterns, which is the same behaviour as GNU grep.
509     7. A pattern that could match an empty string could cause pcregrep to loop; it
510     doesn't make sense to accept an empty string match in pcregrep, so I have
511 ph10 379 locked it out (using PCRE's PCRE_NOTEMPTY option). By experiment, this
512     seems to be how GNU grep behaves.
513 ph10 392
514 ph10 380 8. The pattern (?(?=.*b)b|^) was incorrectly compiled as "match must be at
515 ph10 392 start or after a newline", because the conditional assertion was not being
516     correctly handled. The rule now is that both the assertion and what follows
517     in the first alternative must satisfy the test.
519 ph10 405 9. If auto-callout was enabled in a pattern with a conditional group whose
520 ph10 399 condition was an assertion, PCRE could crash during matching, both with
521     pcre_exec() and pcre_dfa_exec().
522 ph10 392
523     10. The PCRE_DOLLAR_ENDONLY option was not working when pcre_dfa_exec() was
524     used for matching.
526     11. Unicode property support in character classes was not working for
527 ph10 385 characters (bytes) greater than 127 when not in UTF-8 mode.
528 ph10 388
529 ph10 392 12. Added the -M command line option to pcretest.
531 ph10 388 14. Added the non-standard REG_NOTEMPTY option to the POSIX interface.
532 ph10 389
533     15. Added the PCRE_NO_START_OPTIMIZE match-time option.
534 ph10 390
535 ph10 392 16. Added comments and documentation about mis-use of no_arg in the C++
536 ph10 390 wrapper.
537 ph10 392
538 ph10 391 17. Implemented support for UTF-8 encoding in EBCDIC environments, a patch
539 ph10 392 from Martin Jerabek that uses macro names for all relevant character and
540 ph10 391 string constants.
541 ph10 405
542 ph10 393 18. Added to pcre_internal.h two configuration checks: (a) If both EBCDIC and
543     SUPPORT_UTF8 are set, give an error; (b) If SUPPORT_UCP is set without
544 ph10 405 SUPPORT_UTF8, define SUPPORT_UTF8. The "configure" script handles both of
545     these, but not everybody uses configure.
547     19. A conditional group that had only one branch was not being correctly
548     recognized as an item that could match an empty string. This meant that an
549     enclosing group might also not be so recognized, causing infinite looping
550     (and probably a segfault) for patterns such as ^"((?(?=[a])[^"])|b)*"$
551     with the subject "ab", where knowledge that the repeated group can match
552 ph10 396 nothing is needed in order to break the loop.
553 ph10 405
554 ph10 397 20. If a pattern that was compiled with callouts was matched using pcre_dfa_
555 ph10 405 exec(), but without supplying a callout function, matching went wrong.
557     21. If PCRE_ERROR_MATCHLIMIT occurred during a recursion, there was a memory
558     leak if the size of the offset vector was greater than 30. When the vector
559     is smaller, the saved offsets during recursion go onto a local stack
560     vector, but for larger vectors malloc() is used. It was failing to free
561     when the recursion yielded PCRE_ERROR_MATCH_LIMIT (or any other "abnormal"
562 ph10 400 error, in fact).
563 ph10 405
564 ph10 402 22. There was a missing #ifdef SUPPORT_UTF8 round one of the variables in the
565 ph10 405 heapframe that is used only when UTF-8 support is enabled. This caused no
566     problem, but was untidy.
567 ph10 376
568 ph10 405 23. Steven Van Ingelgem's patch to CMakeLists.txt to change the name
569     CMAKE_BINARY_DIR to PROJECT_BINARY_DIR so that it works when PCRE is
570     included within another project.
571 ph10 411
572 ph10 407 24. Steven Van Ingelgem's patches to add more options to the CMake support,
573     slightly modified by me:
574 ph10 411
575     (a) PCRE_BUILD_TESTS can be set OFF not to build the tests, including
576     not building pcregrep.
578 ph10 407 (b) PCRE_BUILD_PCREGREP can be see OFF not to build pcregrep, but only
579 ph10 411 if PCRE_BUILD_TESTS is also set OFF, because the tests use pcregrep.
581 ph10 408 25. Forward references, both numeric and by name, in patterns that made use of
582     duplicate group numbers, could behave incorrectly or give incorrect errors,
583     because when scanning forward to find the reference group, PCRE was not
584     taking into account the duplicate group numbers. A pattern such as
585     ^X(?3)(a)(?|(b)|(q))(Y) is an example.
586 ph10 411
587     26. Changed a few more instances of "const unsigned char *" to USPTR, making
588     the feature of a custom pointer more persuasive (as requested by a user).
589 ph10 416
590 ph10 411 27. Wrapped the definitions of fileno and isatty for Windows, which appear in
591 ph10 416 pcretest.c, inside #ifndefs, because it seems they are sometimes already
592     pre-defined.
593 ph10 392
594 ph10 416 28. Added support for (*UTF8) at the start of a pattern.
596 ph10 413 29. Arrange for flags added by the "release type" setting in CMake to be shown
597     in the configuration summary.
598 ph10 405
599 ph10 411
600 ph10 374 Version 7.8 05-Sep-08
601 ph10 349 ---------------------
603     1. Replaced UCP searching code with optimized version as implemented for Ad
604     Muncher (http://www.admuncher.com/) by Peter Kankowski. This uses a two-
605 ph10 351 stage table and inline lookup instead of a function, giving speed ups of 2
606     to 5 times on some simple patterns that I tested. Permission was given to
607     distribute the MultiStage2.py script that generates the tables (it's not in
608     the tarball, but is in the Subversion repository).
609 ph10 349
610 ph10 351 2. Updated the Unicode datatables to Unicode 5.1.0. This adds yet more
611     scripts.
612 ph10 358
613 ph10 353 3. Change 12 for 7.7 introduced a bug in pcre_study() when a pattern contained
614     a group with a zero qualifier. The result of the study could be incorrect,
615 ph10 358 or the function might crash, depending on the pattern.
617     4. Caseless matching was not working for non-ASCII characters in back
618 ph10 354 references. For example, /(\x{de})\1/8i was not matching \x{de}\x{fe}.
619 ph10 358 It now works when Unicode Property Support is available.
621 ph10 355 5. In pcretest, an escape such as \x{de} in the data was always generating
622     a UTF-8 string, even in non-UTF-8 mode. Now it generates a single byte in
623     non-UTF-8 mode. If the value is greater than 255, it gives a warning about
624 ph10 358 truncation.
625 ph10 349
626 ph10 358 6. Minor bugfix in pcrecpp.cc (change "" == ... to NULL == ...).
628 ph10 357 7. Added two (int) casts to pcregrep when printing the difference of two
629     pointers, in case they are 64-bit values.
630 ph10 371
631     8. Added comments about Mac OS X stack usage to the pcrestack man page and to
632     test 2 if it fails.
634 ph10 359 9. Added PCRE_CALL_CONVENTION just before the names of all exported functions,
635 ph10 371 and a #define of that name to empty if it is not externally set. This is to
636     allow users of MSVC to set it if necessary.
638     10. The PCRE_EXP_DEFN macro which precedes exported functions was missing from
639 ph10 359 the convenience functions in the pcre_get.c source file.
640 ph10 371
641 ph10 360 11. An option change at the start of a pattern that had top-level alternatives
642     could cause overwriting and/or a crash. This command provoked a crash in
643 ph10 371 some environments:
645     printf "/(?i)[\xc3\xa9\xc3\xbd]|[\xc3\xa9\xc3\xbdA]/8\n" | pcretest
647 ph10 360 This potential security problem was recorded as CVE-2008-2371.
648 ph10 371
649 ph10 361 12. For a pattern where the match had to start at the beginning or immediately
650     after a newline (e.g /.*anything/ without the DOTALL flag), pcre_exec() and
651 ph10 371 pcre_dfa_exec() could read past the end of the passed subject if there was
652 ph10 361 no match. To help with detecting such bugs (e.g. with valgrind), I modified
653     pcretest so that it places the subject at the end of its malloc-ed buffer.
654 ph10 371
655 ph10 364 13. The change to pcretest in 12 above threw up a couple more cases when pcre_
656 ph10 371 exec() might read past the end of the data buffer in UTF-8 mode.
658 ph10 364 14. A similar bug to 7.3/2 existed when the PCRE_FIRSTLINE option was set and
659 ph10 371 the data contained the byte 0x85 as part of a UTF-8 character within its
660     first line. This applied both to normal and DFA matching.
662 ph10 366 15. Lazy qualifiers were not working in some cases in UTF-8 mode. For example,
663 ph10 371 /^[^d]*?$/8 failed to match "abc".
664 ph10 351
665 ph10 371 16. Added a missing copyright notice to pcrecpp_internal.h.
667     17. Make it more clear in the documentation that values returned from
668 ph10 368 pcre_exec() in ovector are byte offsets, not character counts.
669 ph10 357
670 ph10 371 18. Tidied a few places to stop certain compilers from issuing warnings.
671 ph10 368
672 ph10 373 19. Updated the Virtual Pascal + BCC files to compile the latest v7.7, as
673     supplied by Stefan Weber. I made a further small update for 7.8 because
674 ph10 374 there is a change of source arrangements: the pcre_searchfuncs.c module is
675     replaced by pcre_ucd.c.
676 ph10 371
677 ph10 373
678 ph10 347 Version 7.7 07-May-08
679 ph10 321 ---------------------
681     1. Applied Craig's patch to sort out a long long problem: "If we can't convert
682 ph10 345 a string to a long long, pretend we don't even have a long long." This is
683 ph10 321 done by checking for the strtoq, strtoll, and _strtoi64 functions.
684 ph10 345
685 ph10 322 2. Applied Craig's patch to pcrecpp.cc to restore ABI compatibility with
686     pre-7.6 versions, which defined a global no_arg variable instead of putting
687 ph10 332 it in the RE class. (See also #8 below.)
688 ph10 345
689     3. Remove a line of dead code, identified by coverity and reported by Nuno
690     Lopes.
692 ph10 324 4. Fixed two related pcregrep bugs involving -r with --include or --exclude:
693 ph10 321
694 ph10 324 (1) The include/exclude patterns were being applied to the whole pathnames
695 ph10 345 of files, instead of just to the final components.
697 ph10 324 (2) If there was more than one level of directory, the subdirectories were
698     skipped unless they satisfied the include/exclude conditions. This is
699     inconsistent with GNU grep (and could even be seen as contrary to the
700     pcregrep specification - which I improved to make it absolutely clear).
701     The action now is always to scan all levels of directory, and just
702     apply the include/exclude patterns to regular files.
703 ph10 345
704 ph10 325 5. Added the --include_dir and --exclude_dir patterns to pcregrep, and used
705 ph10 345 --exclude_dir in the tests to avoid scanning .svn directories.
707 ph10 326 6. Applied Craig's patch to the QuoteMeta function so that it escapes the
708 ph10 345 NUL character as backslash + 0 rather than backslash + NUL, because PCRE
709 ph10 326 doesn't support NULs in patterns.
710 ph10 345
711     7. Added some missing "const"s to declarations of static tables in
712     pcre_compile.c and pcre_dfa_exec.c.
714 ph10 328 8. Applied Craig's patch to pcrecpp.cc to fix a problem in OS X that was
715 ph10 329 caused by fix #2 above. (Subsequently also a second patch to fix the
716 ph10 345 first patch. And a third patch - this was a messy problem.)
717 ph10 321
718 ph10 345 9. Applied Craig's patch to remove the use of push_back().
720 ph10 332 10. Applied Alan Lehotsky's patch to add REG_STARTEND support to the POSIX
721     matching function regexec().
722 ph10 345
723     11. Added support for the Oniguruma syntax \g<name>, \g<n>, \g'name', \g'n',
724     which, however, unlike Perl's \g{...}, are subroutine calls, not back
725 ph10 333 references. PCRE supports relative numbers with this syntax (I don't think
726     Oniguruma does).
727 ph10 345
728     12. Previously, a group with a zero repeat such as (...){0} was completely
729 ph10 335 omitted from the compiled regex. However, this means that if the group
730     was called as a subroutine from elsewhere in the pattern, things went wrong
731 ph10 345 (an internal error was given). Such groups are now left in the compiled
732     pattern, with a new opcode that causes them to be skipped at execution
733 ph10 335 time.
734 ph10 345
735 ph10 341 13. Added the PCRE_JAVASCRIPT_COMPAT option. This makes the following changes
736     to the way PCRE behaves:
737 ph10 345
738 ph10 336 (a) A lone ] character is dis-allowed (Perl treats it as data).
739 ph10 345
740     (b) A back reference to an unmatched subpattern matches an empty string
741 ph10 336 (Perl fails the current match path).
742 ph10 345
743 ph10 341 (c) A data ] in a character class must be notated as \] because if the
744 ph10 345 first data character in a class is ], it defines an empty class. (In
745 ph10 341 Perl it is not possible to have an empty class.) The empty class []
746     never matches; it forces failure and is equivalent to (*FAIL) or (?!).
747     The negative empty class [^] matches any one character, independently
748     of the DOTALL setting.
749 ph10 345
750     14. A pattern such as /(?2)[]a()b](abc)/ which had a forward reference to a
751 ph10 340 non-existent subpattern following a character class starting with ']' and
752     containing () gave an internal compiling error instead of "reference to
753     non-existent subpattern". Fortunately, when the pattern did exist, the
754 ph10 345 compiled code was correct. (When scanning forwards to check for the
755 ph10 340 existencd of the subpattern, it was treating the data ']' as terminating
756 ph10 345 the class, so got the count wrong. When actually compiling, the reference
757 ph10 340 was subsequently set up correctly.)
758 ph10 345
759 ph10 341 15. The "always fail" assertion (?!) is optimzed to (*FAIL) by pcre_compile;
760 ph10 345 it was being rejected as not supported by pcre_dfa_exec(), even though
761     other assertions are supported. I have made pcre_dfa_exec() support
762     (*FAIL).
764 ph10 342 16. The implementation of 13c above involved the invention of a new opcode,
765 ph10 345 OP_ALLANY, which is like OP_ANY but doesn't check the /s flag. Since /s
766     cannot be changed at match time, I realized I could make a small
767     improvement to matching performance by compiling OP_ALLANY instead of
768     OP_ANY for "." when DOTALL was set, and then removing the runtime tests
769     on the OP_ANY path.
771     17. Compiling pcretest on Windows with readline support failed without the
772 ph10 343 following two fixes: (1) Make the unistd.h include conditional on
773 ph10 345 HAVE_UNISTD_H; (2) #define isatty and fileno as _isatty and _fileno.
775 ph10 344 18. Changed CMakeLists.txt and cmake/FindReadline.cmake to arrange for the
776 ph10 345 ncurses library to be included for pcretest when ReadLine support is
777     requested, but also to allow for it to be overridden. This patch came from
778     Daniel Bergström.
779 ph10 347
780 ph10 346 19. There was a typo in the file ucpinternal.h where f0_rangeflag was defined
781 ph10 347 as 0x00f00000 instead of 0x00800000. Luckily, this would not have caused
782     any errors with the current Unicode tables. Thanks to Peter Kankowski for
783     spotting this.
784 ph10 324
785 ph10 332
786 ph10 319 Version 7.6 28-Jan-08
787 ph10 300 ---------------------
789 ph10 302 1. A character class containing a very large number of characters with
790     codepoints greater than 255 (in UTF-8 mode, of course) caused a buffer
791     overflow.
792 ph10 309
793     2. Patch to cut out the "long long" test in pcrecpp_unittest when
794     HAVE_LONG_LONG is not defined.
796 ph10 303 3. Applied Christian Ehrlicher's patch to update the CMake build files to
797 ph10 304 bring them up to date and include new features. This patch includes:
798 ph10 309
799 ph10 304 - Fixed PH's badly added libz and libbz2 support.
800     - Fixed a problem with static linking.
801 ph10 312 - Added pcredemo. [But later removed - see 7 below.]
802 ph10 304 - Fixed dftables problem and added an option.
803     - Added a number of HAVE_XXX tests, including HAVE_WINDOWS_H and
805     - Added readline support for pcretest.
806 ph10 309 - Added an listing of the option settings after cmake has run.
808 ph10 314 4. A user submitted a patch to Makefile that makes it easy to create
809     "pcre.dll" under mingw when using Configure/Make. I added stuff to
810     Makefile.am that cause it to include this special target, without
811     affecting anything else. Note that the same mingw target plus all
812     the other distribution libraries and programs are now supported
813     when configuring with CMake (see 6 below) instead of with
814     Configure/Make.
815 ph10 309
816 ph10 308 5. Applied Craig's patch that moves no_arg into the RE class in the C++ code.
817     This is an attempt to solve the reported problem "pcrecpp::no_arg is not
818 ph10 309 exported in the Windows port". It has not yet been confirmed that the patch
819     solves the problem, but it does no harm.
820 ph10 313
821 ph10 311 6. Applied Sheri's patch to CMakeLists.txt to add NON_STANDARD_LIB_PREFIX and
822 ph10 319 NON_STANDARD_LIB_SUFFIX for dll names built with mingw when configured
823 ph10 314 with CMake, and also correct the comment about stack recursion.
824 ph10 313
825 ph10 312 7. Remove the automatic building of pcredemo from the ./configure system and
826 ph10 313 from CMakeLists.txt. The whole idea of pcredemo.c is that it is an example
827     of a program that users should build themselves after PCRE is installed, so
828     building it automatically is not really right. What is more, it gave
829 ph10 312 trouble in some build environments.
830 ph10 300
831 ph10 319 8. Further tidies to CMakeLists.txt from Sheri and Christian.
832 ph10 308
833 ph10 319
834 ph10 298 Version 7.5 10-Jan-08
835 ph10 263 ---------------------
837     1. Applied a patch from Craig: "This patch makes it possible to 'ignore'
838     values in parens when parsing an RE using the C++ wrapper."
839 ph10 286
840 ph10 264 2. Negative specials like \S did not work in character classes in UTF-8 mode.
841     Characters greater than 255 were excluded from the class instead of being
842     included.
843 ph10 286
844     3. The same bug as (2) above applied to negated POSIX classes such as
845 ph10 265 [:^space:].
846 ph10 286
847 ph10 267 4. PCRECPP_STATIC was referenced in pcrecpp_internal.h, but nowhere was it
848     defined or documented. It seems to have been a typo for PCRE_STATIC, so
849 ph10 286 I have changed it.
851 ph10 268 5. The construct (?&) was not diagnosed as a syntax error (it referenced the
852 ph10 286 first named subpattern) and a construct such as (?&a) would reference the
853     first named subpattern whose name started with "a" (in other words, the
854 ph10 272 length check was missing). Both these problems are fixed. "Subpattern name
855     expected" is now given for (?&) (a zero-length name), and this patch also
856     makes it give the same error for \k'' (previously it complained that that
857     was a reference to a non-existent subpattern).
858 ph10 286
859 ph10 269 6. The erroneous patterns (?+-a) and (?-+a) give different error messages;
860 ph10 286 this is right because (?- can be followed by option settings as well as by
861 ph10 269 digits. I have, however, made the messages clearer.
862 ph10 286
863 ph10 270 7. Patterns such as (?(1)a|b) (a pattern that contains fewer subpatterns
864     than the number used in the conditional) now cause a compile-time error.
865     This is actually not compatible with Perl, which accepts such patterns, but
866     treats the conditional as always being FALSE (as PCRE used to), but it
867     seems to me that giving a diagnostic is better.
868 ph10 286
869 ph10 274 8. Change "alphameric" to the more common word "alphanumeric" in comments
870 ph10 275 and messages.
871 ph10 286
872     9. Fix two occurrences of "backslash" in comments that should have been
873     "backspace".
875     10. Remove two redundant lines of code that can never be obeyed (their function
876     was moved elsewhere).
878     11. The program that makes PCRE's Unicode character property table had a bug
879     which caused it to generate incorrect table entries for sequences of
880 ph10 277 characters that have the same character type, but are in different scripts.
881 ph10 286 It amalgamated them into a single range, with the script of the first of
882 ph10 277 them. In other words, some characters were in the wrong script. There were
883     thirteen such cases, affecting characters in the following ranges:
884 ph10 286
885 ph10 277 U+002b0 - U+002c1
886     U+0060c - U+0060d
887 ph10 286 U+0061e - U+00612
888 ph10 277 U+0064b - U+0065e
889     U+0074d - U+0076d
890     U+01800 - U+01805
891     U+01d00 - U+01d77
892     U+01d9b - U+01dbf
893     U+0200b - U+0200f
894     U+030fc - U+030fe
895     U+03260 - U+0327f
896     U+0fb46 - U+0fbb1
897     U+10450 - U+1049d
898 ph10 286
899 ph10 279 12. The -o option (show only the matching part of a line) for pcregrep was not
900     compatible with GNU grep in that, if there was more than one match in a
901     line, it showed only the first of them. It now behaves in the same way as
902     GNU grep.
903 ph10 286
904     13. If the -o and -v options were combined for pcregrep, it printed a blank
905     line for every non-matching line. GNU grep prints nothing, and pcregrep now
906     does the same. The return code can be used to tell if there were any
907     non-matching lines.
909 ph10 289 14. Added --file-offsets and --line-offsets to pcregrep.
911     15. The pattern (?=something)(?R) was not being diagnosed as a potentially
912 ph10 282 infinitely looping recursion. The bug was that positive lookaheads were not
913 ph10 286 being skipped when checking for a possible empty match (negative lookaheads
914     and both kinds of lookbehind were skipped).
916 ph10 289 16. Fixed two typos in the Windows-only code in pcregrep.c, and moved the
917 ph10 284 inclusion of <windows.h> to before rather than after the definition of
918     INVALID_FILE_ATTRIBUTES (patch from David Byron).
919 ph10 263
920 ph10 289 17. Specifying a possessive quantifier with a specific limit for a Unicode
921 ph10 286 character property caused pcre_compile() to compile bad code, which led at
922     runtime to PCRE_ERROR_INTERNAL (-14). Examples of patterns that caused this
923     are: /\p{Zl}{2,3}+/8 and /\p{Cc}{2}+/8. It was the possessive "+" that
924     caused the error; without that there was no problem.
925 ph10 263
926 ph10 289 18. Added --enable-pcregrep-libz and --enable-pcregrep-libbz2.
927 ph10 286
928 ph10 289 19. Added --enable-pcretest-libreadline.
929 ph10 286
930 ph10 289 20. In pcrecpp.cc, the variable 'count' was incremented twice in
931 ph10 288 RE::GlobalReplace(). As a result, the number of replacements returned was
932 ph10 298 double what it should be. I removed one of the increments, but Craig sent a
933     later patch that removed the other one (the right fix) and added unit tests
934     that check the return values (which was not done before).
935 ph10 292
936 ph10 291 21. Several CMake things:
937 ph10 287
938 ph10 291 (1) Arranged that, when cmake is used on Unix, the libraries end up with
939 ph10 292 the names libpcre and libpcreposix, not just pcre and pcreposix.
940 ph10 288
941 ph10 292 (2) The above change means that pcretest and pcregrep are now correctly
942     linked with the newly-built libraries, not previously installed ones.
943 ph10 291
945 ph10 298
946     22. In UTF-8 mode, with newline set to "any", a pattern such as .*a.*=.b.*
947     crashed when matching a string such as a\x{2029}b (note that \x{2029} is a
948     UTF-8 newline character). The key issue is that the pattern starts .*;
949     this means that the match must be either at the beginning, or after a
950     newline. The bug was in the code for advancing after a failed match and
951     checking that the new position followed a newline. It was not taking
952 ph10 294 account of UTF-8 characters correctly.
953 ph10 298
954     23. PCRE was behaving differently from Perl in the way it recognized POSIX
955     character classes. PCRE was not treating the sequence [:...:] as a
956     character class unless the ... were all letters. Perl, however, seems to
957     allow any characters between [: and :], though of course it rejects as
958     unknown any "names" that contain non-letters, because all the known class
959     names consist only of letters. Thus, Perl gives an error for [[:1234:]],
960     for example, whereas PCRE did not - it did not recognize a POSIX character
961     class. This seemed a bit dangerous, so the code has been changed to be
962     closer to Perl. The behaviour is not identical to Perl, because PCRE will
963     diagnose an unknown class for, for example, [[:l\ower:]] where Perl will
964     treat it as [[:lower:]]. However, PCRE does now give "unknown" errors where
965     Perl does, and where it didn't before.
967 ph10 296 24. Rewrite so as to remove the single use of %n from pcregrep because in some
968     Windows environments %n is disabled by default.
969 ph10 292
971 ph10 260 Version 7.4 21-Sep-07
972 ph10 230 ---------------------
974     1. Change 7.3/28 was implemented for classes by looking at the bitmap. This
975 ph10 231 means that a class such as [\s] counted as "explicit reference to CR or
976     LF". That isn't really right - the whole point of the change was to try to
977     help when there was an actual mention of one of the two characters. So now
978     the change happens only if \r or \n (or a literal CR or LF) character is
979 ph10 230 encountered.
980 ph10 231
981     2. The 32-bit options word was also used for 6 internal flags, but the numbers
982     of both had grown to the point where there were only 3 bits left.
983     Fortunately, there was spare space in the data structure, and so I have
984     moved the internal flags into a new 16-bit field to free up more option
985 ph10 230 bits.
986 ph10 231
987     3. The appearance of (?J) at the start of a pattern set the DUPNAMES option,
988     but did not set the internal JCHANGED flag - either of these is enough to
989     control the way the "get" function works - but the PCRE_INFO_JCHANGED
990     facility is supposed to tell if (?J) was ever used, so now (?J) at the
991 ph10 230 start sets both bits.
993 ph10 231 4. Added options (at build time, compile time, exec time) to change \R from
994     matching any Unicode line ending sequence to just matching CR, LF, or CRLF.
995 ph10 230
996 ph10 243 5. doc/pcresyntax.html was missing from the distribution.
998     6. Put back the definition of PCRE_ERROR_NULLWSLIMIT, for backward
999 ph10 233 compatibility, even though it is no longer used.
1000 ph10 243
1001     7. Added macro for snprintf to pcrecpp_unittest.cc and also for strtoll and
1002 ph10 254 strtoull to pcrecpp.cc to select the available functions in WIN32 when the
1003 ph10 259 windows.h file is present (where different names are used). [This was
1004 ph10 257 reversed later after testing - see 16 below.]
1005 ph10 243
1006     8. Changed all #include <config.h> to #include "config.h". There were also
1007 ph10 236 some further <pcre.h> cases that I changed to "pcre.h".
1008 ph10 243
1009 ph10 239 9. When pcregrep was used with the --colour option, it missed the line ending
1010 ph10 243 sequence off the lines that it output.
1012     10. It was pointed out to me that arrays of string pointers cause lots of
1013     relocations when a shared library is dynamically loaded. A technique of
1014     using a single long string with a table of offsets can drastically reduce
1015     these. I have refactored PCRE in four places to do this. The result is
1016 ph10 240 dramatic:
1017 ph10 243
1018 ph10 240 Originally: 290
1019     After changing UCP table: 187
1020 ph10 243 After changing error message table: 43
1021 ph10 240 After changing table of "verbs" 36
1022     After changing table of Posix names 22
1023 ph10 243
1024 ph10 240 Thanks to the folks working on Gregex for glib for this insight.
1025 ph10 247
1026 ph10 244 11. --disable-stack-for-recursion caused compiling to fail unless -enable-
1027 ph10 247 unicode-properties was also set.
1029 ph10 248 12. Updated the tests so that they work when \R is defaulted to ANYCRLF.
1030 ph10 231
1031 ph10 253 13. Added checks for ANY and ANYCRLF to pcrecpp.cc where it previously
1032     checked only for CRLF.
1033 ph10 233
1034 ph10 259 14. Added casts to pcretest.c to avoid compiler warnings.
1036 ph10 256 15. Added Craig's patch to various pcrecpp modules to avoid compiler warnings.
1037 ph10 253
1038 ph10 257 16. Added Craig's patch to remove the WINDOWS_H tests, that were not working,
1039 ph10 259 and instead check for _strtoi64 explicitly, and avoid the use of snprintf()
1040     entirely. This removes changes made in 7 above.
1041 ph10 256
1042 ph10 261 17. The CMake files have been updated, and there is now more information about
1043     building with CMake in the NON-UNIX-USE document.
1044 ph10 257
1045 ph10 261
1046 ph10 228 Version 7.3 28-Aug-07
1047 ph10 157 ---------------------
1049 ph10 189 1. In the rejigging of the build system that eventually resulted in 7.1, the
1050     line "#include <pcre.h>" was included in pcre_internal.h. The use of angle
1051     brackets there is not right, since it causes compilers to look for an
1052     installed pcre.h, not the version that is in the source that is being
1053     compiled (which of course may be different). I have changed it back to:
1054 ph10 197
1055 ph10 189 #include "pcre.h"
1056 ph10 197
1057     I have a vague recollection that the change was concerned with compiling in
1058     different directories, but in the new build system, that is taken care of
1059     by the VPATH setting the Makefile.
1061 ph10 190 2. The pattern .*$ when run in not-DOTALL UTF-8 mode with newline=any failed
1062     when the subject happened to end in the byte 0x85 (e.g. if the last
1063     character was \x{1ec5}). *Character* 0x85 is one of the "any" newline
1064     characters but of course it shouldn't be taken as a newline when it is part
1065     of another character. The bug was that, for an unlimited repeat of . in
1066     not-DOTALL UTF-8 mode, PCRE was advancing by bytes rather than by
1067     characters when looking for a newline.
1068 ph10 197
1069     3. A small performance improvement in the DOTALL UTF-8 mode .* case.
1071     4. Debugging: adjusted the names of opcodes for different kinds of parentheses
1072     in debug output.
1074 ph10 201 5. Arrange to use "%I64d" instead of "%lld" and "%I64u" instead of "%llu" for
1075     long printing in the pcrecpp unittest when running under MinGW.
1076 ph10 189
1077 ph10 197 6. ESC_K was left out of the EBCDIC table.
1078 ph10 189
1079 ph10 197 7. Change 7.0/38 introduced a new limit on the number of nested non-capturing
1080     parentheses; I made it 1000, which seemed large enough. Unfortunately, the
1081     limit also applies to "virtual nesting" when a pattern is recursive, and in
1082     this case 1000 isn't so big. I have been able to remove this limit at the
1083     expense of backing off one optimization in certain circumstances. Normally,
1084     when pcre_exec() would call its internal match() function recursively and
1085     immediately return the result unconditionally, it uses a "tail recursion"
1086     feature to save stack. However, when a subpattern that can match an empty
1087     string has an unlimited repetition quantifier, it no longer makes this
1088     optimization. That gives it a stack frame in which to save the data for
1089     checking that an empty string has been matched. Previously this was taken
1090     from the 1000-entry workspace that had been reserved. So now there is no
1091     explicit limit, but more stack is used.
1092 ph10 189
1093 ph10 200 8. Applied Daniel's patches to solve problems with the import/export magic
1094     syntax that is required for Windows, and which was going wrong for the
1095     pcreposix and pcrecpp parts of the library. These were overlooked when this
1096     problem was solved for the main library.
1097 ph10 197
1098 ph10 202 9. There were some crude static tests to avoid integer overflow when computing
1099     the size of patterns that contain repeated groups with explicit upper
1100     limits. As the maximum quantifier is 65535, the maximum group length was
1101     set at 30,000 so that the product of these two numbers did not overflow a
1102     32-bit integer. However, it turns out that people want to use groups that
1103     are longer than 30,000 bytes (though not repeat them that many times).
1104     Change 7.0/17 (the refactoring of the way the pattern size is computed) has
1105     made it possible to implement the integer overflow checks in a much more
1106     dynamic way, which I have now done. The artificial limitation on group
1107     length has been removed - we now have only the limit on the total length of
1108     the compiled pattern, which depends on the LINK_SIZE setting.
1109 ph10 208
1110     10. Fixed a bug in the documentation for get/copy named substring when
1111     duplicate names are permitted. If none of the named substrings are set, the
1112     functions return PCRE_ERROR_NOSUBSTRING (7); the doc said they returned an
1113     empty string.
1115     11. Because Perl interprets \Q...\E at a high level, and ignores orphan \E
1116     instances, patterns such as [\Q\E] or [\E] or even [^\E] cause an error,
1117     because the ] is interpreted as the first data character and the
1118     terminating ] is not found. PCRE has been made compatible with Perl in this
1119     regard. Previously, it interpreted [\Q\E] as an empty class, and [\E] could
1120     cause memory overwriting.
1122 ph10 206 10. Like Perl, PCRE automatically breaks an unlimited repeat after an empty
1123     string has been matched (to stop an infinite loop). It was not recognizing
1124 ph10 208 a conditional subpattern that could match an empty string if that
1125 ph10 206 subpattern was within another subpattern. For example, it looped when
1126 ph10 208 trying to match (((?(1)X|))*) but it was OK with ((?(1)X|)*) where the
1127 ph10 206 condition was not nested. This bug has been fixed.
1128 ph10 208
1129 ph10 207 12. A pattern like \X?\d or \P{L}?\d in non-UTF-8 mode could cause a backtrack
1130     past the start of the subject in the presence of bytes with the top bit
1131     set, for example "\x8aBCD".
1132 ph10 211
1133 ph10 210 13. Added Perl 5.10 experimental backtracking controls (*FAIL), (*F), (*PRUNE),
1134     (*SKIP), (*THEN), (*COMMIT), and (*ACCEPT).
1135 ph10 200
1136 ph10 211 14. Optimized (?!) to (*FAIL).
1137 ph10 202
1138 ph10 212 15. Updated the test for a valid UTF-8 string to conform to the later RFC 3629.
1139     This restricts code points to be within the range 0 to 0x10FFFF, excluding
1140     the "low surrogate" sequence 0xD800 to 0xDFFF. Previously, PCRE allowed the
1141     full range 0 to 0x7FFFFFFF, as defined by RFC 2279. Internally, it still
1142     does: it's just the validity check that is more restrictive.
1143 ph10 220
1144     16. Inserted checks for integer overflows during escape sequence (backslash)
1145     processing, and also fixed erroneous offset values for syntax errors during
1146     backslash processing.
1148 ph10 214 17. Fixed another case of looking too far back in non-UTF-8 mode (cf 12 above)
1149 ph10 220 for patterns like [\PPP\x8a]{1,}\x80 with the subject "A\x80".
1151 ph10 215 18. An unterminated class in a pattern like (?1)\c[ with a "forward reference"
1152     caused an overrun.
1153 ph10 220
1154     19. A pattern like (?:[\PPa*]*){8,} which had an "extended class" (one with
1155     something other than just ASCII characters) inside a group that had an
1156     unlimited repeat caused a loop at compile time (while checking to see
1157     whether the group could match an empty string).
1159 ph10 217 20. Debugging a pattern containing \p or \P could cause a crash. For example,
1160     [\P{Any}] did so. (Error in the code for printing property names.)
1161 ph10 210
1162 ph10 220 21. An orphan \E inside a character class could cause a crash.
1164     22. A repeated capturing bracket such as (A)? could cause a wild memory
1165 ph10 218 reference during compilation.
1166 ph10 220
1167     23. There are several functions in pcre_compile() that scan along a compiled
1168     expression for various reasons (e.g. to see if it's fixed length for look
1169 ph10 218 behind). There were bugs in these functions when a repeated \p or \P was
1170 ph10 220 present in the pattern. These operators have additional parameters compared
1171     with \d, etc, and these were not being taken into account when moving along
1172 ph10 218 the compiled data. Specifically:
1173 ph10 220
1174     (a) A item such as \p{Yi}{3} in a lookbehind was not treated as fixed
1175     length.
1177     (b) An item such as \pL+ within a repeated group could cause crashes or
1178 ph10 218 loops.
1179 ph10 220
1180 ph10 218 (c) A pattern such as \p{Yi}+(\P{Yi}+)(?1) could give an incorrect
1181 ph10 220 "reference to non-existent subpattern" error.
1183 ph10 221 (d) A pattern like (\P{Yi}{2}\277)? could loop at compile time.
1185 ph10 219 24. A repeated \S or \W in UTF-8 mode could give wrong answers when multibyte
1186 ph10 220 characters were involved (for example /\S{2}/8g with "A\x{a3}BC").
1187 ph10 211
1188 ph10 222 25. Using pcregrep in multiline, inverted mode (-Mv) caused it to loop.
1189 ph10 218
1190 ph10 227 26. Patterns such as [\P{Yi}A] which include \p or \P and just one other
1191 ph10 223 character were causing crashes (broken optimization).
1192 ph10 227
1193 ph10 224 27. Patterns such as (\P{Yi}*\277)* (group with possible zero repeat containing
1194 ph10 227 \p or \P) caused a compile-time loop.
1196 ph10 226 28. More problems have arisen in unanchored patterns when CRLF is a valid line
1197     break. For example, the unstudied pattern [\r\n]A does not match the string
1198     "\r\nA" because change 7.0/46 below moves the current point on by two
1199     characters after failing to match at the start. However, the pattern \nA
1200 ph10 227 *does* match, because it doesn't start till \n, and if [\r\n]A is studied,
1201     the same is true. There doesn't seem any very clean way out of this, but
1202     what I have chosen to do makes the common cases work: PCRE now takes note
1203     of whether there can be an explicit match for \r or \n anywhere in the
1204     pattern, and if so, 7.0/46 no longer applies. As part of this change,
1205     there's a new PCRE_INFO_HASCRORLF option for finding out whether a compiled
1206     pattern has explicit CR or LF references.
1207 ph10 222
1208 ph10 227 29. Added (*CR) etc for changing newline setting at start of pattern.
1209 ph10 223
1210 ph10 227
1211 ph10 189 Version 7.2 19-Jun-07
1212     ---------------------
1214 ph10 157 1. If the fr_FR locale cannot be found for test 3, try the "french" locale,
1215     which is apparently normally available under Windows.
1216 ph10 159
1217     2. Re-jig the pcregrep tests with different newline settings in an attempt
1218     to make them independent of the local environment's newline setting.
1219 ph10 157
1220 ph10 160 3. Add code to configure.ac to remove -g from the CFLAGS default settings.
1221 ph10 165
1222 ph10 161 4. Some of the "internals" tests were previously cut out when the link size
1223     was not 2, because the output contained actual offsets. The recent new
1224 ph10 165 "Z" feature of pcretest means that these can be cut out, making the tests
1225     usable with all link sizes.
1227 ph10 164 5. Implemented Stan Switzer's goto replacement for longjmp() when not using
1228     stack recursion. This gives a massive performance boost under BSD, but just
1229 ph10 165 a small improvement under Linux. However, it saves one field in the frame
1230 ph10 164 in all cases.
1231 ph10 172
1232 ph10 166 6. Added more features from the forthcoming Perl 5.10:
1233 ph10 172
1234 ph10 166 (a) (?-n) (where n is a string of digits) is a relative subroutine or
1235     recursion call. It refers to the nth most recently opened parentheses.
1236 ph10 172
1237 ph10 166 (b) (?+n) is also a relative subroutine call; it refers to the nth next
1238 ph10 172 to be opened parentheses.
1240     (c) Conditions that refer to capturing parentheses can be specified
1241 ph10 167 relatively, for example, (?(-2)... or (?(+3)...
1242 ph10 172
1243 ph10 168 (d) \K resets the start of the current match so that everything before
1244 ph10 172 is not part of it.
1246 ph10 171 (e) \k{name} is synonymous with \k<name> and \k'name' (.NET compatible).
1247 ph10 172
1248 ph10 171 (f) \g{name} is another synonym - part of Perl 5.10's unification of
1249 ph10 172 reference syntax.
1250 ph10 182
1251 ph10 175 (g) (?| introduces a group in which the numbering of parentheses in each
1252 ph10 182 alternative starts with the same number.
1253 ph10 172
1254 ph10 182 (h) \h, \H, \v, and \V match horizontal and vertical whitespace.
1256 ph10 172 7. Added two new calls to pcre_fullinfo(): PCRE_INFO_OKPARTIAL and
1259     8. A pattern such as (.*(.)?)* caused pcre_exec() to fail by either not
1260     terminating or by crashing. Diagnosed by Viktor Griph; it was in the code
1261 ph10 170 for detecting groups that can match an empty string.
1262 ph10 159
1263 ph10 172 9. A pattern with a very large number of alternatives (more than several
1264     hundred) was running out of internal workspace during the pre-compile
1265     phase, where pcre_compile() figures out how much memory will be needed. A
1266     bit of new cunning has reduced the workspace needed for groups with
1267     alternatives. The 1000-alternative test pattern now uses 12 bytes of
1268     workspace instead of running out of the 4096 that are available.
1269 ph10 182
1270 ph10 176 10. Inserted some missing (unsigned int) casts to get rid of compiler warnings.
1271 ph10 172
1272 ph10 179 11. Applied patch from Google to remove an optimization that didn't quite work.
1273     The report of the bug said:
1274 ph10 182
1275 ph10 179 pcrecpp::RE("a*").FullMatch("aaa") matches, while
1276     pcrecpp::RE("a*?").FullMatch("aaa") does not, and
1277 ph10 182 pcrecpp::RE("a*?\\z").FullMatch("aaa") does again.
1278 ph10 185
1279 ph10 184 12. If \p or \P was used in non-UTF-8 mode on a character greater than 127
1280 ph10 185 it matched the wrong number of bytes.
1281 ph10 172
1282 ph10 179
1283 ph10 155 Version 7.1 24-Apr-07
1284 ph10 98 ---------------------
1286 ph10 111 1. Applied Bob Rossi and Daniel G's patches to convert the build system to one
1287 ph10 122 that is more "standard", making use of automake and other Autotools. There
1288 ph10 99 is some re-arrangement of the files and adjustment of comments consequent
1289     on this.
1290 ph10 111
1291     2. Part of the patch fixed a problem with the pcregrep tests. The test of -r
1292     for recursive directory scanning broke on some systems because the files
1293     are not scanned in any specific order and on different systems the order
1294     was different. A call to "sort" has been inserted into RunGrepTest for the
1295     approprate test as a short-term fix. In the longer term there may be an
1296 ph10 100 alternative.
1297 ph10 111
1298 ph10 100 3. I had an email from Eric Raymond about problems translating some of PCRE's
1299 ph10 111 man pages to HTML (despite the fact that I distribute HTML pages, some
1300     people do their own conversions for various reasons). The problems
1301     concerned the use of low-level troff macros .br and .in. I have therefore
1302     removed all such uses from the man pages (some were redundant, some could
1303 ph10 113 be replaced by .nf/.fi pairs). The 132html script that I use to generate
1304     HTML has been updated to handle .nf/.fi and to complain if it encounters
1305     .br or .in.
1306 ph10 111
1307 ph10 100 4. Updated comments in configure.ac that get placed in config.h.in and also
1308 ph10 123 arranged for config.h to be included in the distribution, with the name
1309 ph10 111 config.h.generic, for the benefit of those who have to compile without
1310     Autotools (compare pcre.h, which is now distributed as pcre.h.generic).
1312     5. Updated the support (such as it is) for Virtual Pascal, thanks to Stefan
1313     Weber: (1) pcre_internal.h was missing some function renames; (2) updated
1314 ph10 127 makevp.bat for the current PCRE, using the additional files
1315 ph10 135 makevp_c.txt, makevp_l.txt, and pcregexp.pas.
1316 ph10 111
1317     6. A Windows user reported a minor discrepancy with test 2, which turned out
1318     to be caused by a trailing space on an input line that had got lost in his
1319 ph10 102 copy. The trailing space was an accident, so I've just removed it.
1320 ph10 111
1321 ph10 104 7. Add -Wl,-R... flags in pcre-config.in for *BSD* systems, as I'm told
1322 ph10 111 that is needed.
1324 ph10 105 8. Mark ucp_table (in ucptable.h) and ucp_gentype (in pcre_ucp_searchfuncs.c)
1325     as "const" (a) because they are and (b) because it helps the PHP
1326     maintainers who have recently made a script to detect big data structures
1327 ph10 111 in the php code that should be moved to the .rodata section. I remembered
1328     to update Builducptable as well, so it won't revert if ucptable.h is ever
1329 ph10 105 re-created.
1330 ph10 111
1331     9. Added some extra #ifdef SUPPORT_UTF8 conditionals into pcretest.c,
1332 ph10 107 pcre_printint.src, pcre_compile.c, pcre_study.c, and pcre_tables.c, in
1333 ph10 111 order to be able to cut out the UTF-8 tables in the latter when UTF-8
1334     support is not required. This saves 1.5-2K of code, which is important in
1335 ph10 107 some applications.
1336 ph10 111
1337 ph10 110 Later: more #ifdefs are needed in pcre_ord2utf8.c and pcre_valid_utf8.c
1338 ph10 111 so as not to refer to the tables, even though these functions will never be
1339     called when UTF-8 support is disabled. Otherwise there are problems with a
1340     shared library.
1342 ph10 118 10. Fixed two bugs in the emulated memmove() function in pcre_internal.h:
1344     (a) It was defining its arguments as char * instead of void *.
1346 ph10 123 (b) It was assuming that all moves were upwards in memory; this was true
1347     a long time ago when I wrote it, but is no longer the case.
1349 ph10 118 The emulated memove() is provided for those environments that have neither
1350 ph10 123 memmove() nor bcopy(). I didn't think anyone used it these days, but that
1351 ph10 118 is clearly not the case, as these two bugs were recently reported.
1352 ph10 123
1353 ph10 111 11. The script PrepareRelease is now distributed: it calls 132html, CleanTxt,
1354 ph10 123 and Detrail to create the HTML documentation, the .txt form of the man
1355     pages, and it removes trailing spaces from listed files. It also creates
1356     pcre.h.generic and config.h.generic from pcre.h and config.h. In the latter
1357     case, it wraps all the #defines with #ifndefs. This script should be run
1358 ph10 111 before "make dist".
1359 ph10 123
1360 ph10 115 12. Fixed two fairly obscure bugs concerned with quantified caseless matching
1361     with Unicode property support.
1362 ph10 123
1363     (a) For a maximizing quantifier, if the two different cases of the
1364     character were of different lengths in their UTF-8 codings (there are
1365     some cases like this - I found 11), and the matching function had to
1366 ph10 115 back up over a mixture of the two cases, it incorrectly assumed they
1367     were both the same length.
1368 ph10 123
1369     (b) When PCRE was configured to use the heap rather than the stack for
1370     recursion during matching, it was not correctly preserving the data for
1371     the other case of a UTF-8 character when checking ahead for a match
1372     while processing a minimizing repeat. If the check also involved
1373 ph10 115 matching a wide character, but failed, corruption could cause an
1374     erroneous result when trying to check for a repeat of the original
1375     character.
1376 ph10 123
1377 ph10 116 13. Some tidying changes to the testing mechanism:
1378 ph10 98
1379 ph10 116 (a) The RunTest script now detects the internal link size and whether there
1380     is UTF-8 and UCP support by running ./pcretest -C instead of relying on
1381 ph10 123 values substituted by "configure". (The RunGrepTest script already did
1382     this for UTF-8.) The configure.ac script no longer substitutes the
1383     relevant variables.
1385 ph10 116 (b) The debugging options /B and /D in pcretest show the compiled bytecode
1386     with length and offset values. This means that the output is different
1387     for different internal link sizes. Test 2 is skipped for link sizes
1388     other than 2 because of this, bypassing the problem. Unfortunately,
1389     there was also a test in test 3 (the locale tests) that used /B and
1390 ph10 123 failed for link sizes other than 2. Rather than cut the whole test out,
1391     I have added a new /Z option to pcretest that replaces the length and
1392     offset values with spaces. This is now used to make test 3 independent
1393 ph10 122 of link size. (Test 2 will be tidied up later.)
1394 ph10 123
1395     14. If erroroffset was passed as NULL to pcre_compile, it provoked a
1396 ph10 122 segmentation fault instead of returning the appropriate error message.
1397 ph10 134
1398 ph10 131 15. In multiline mode when the newline sequence was set to "any", the pattern
1399 ph10 134 ^$ would give a match between the \r and \n of a subject such as "A\r\nB".
1400     This doesn't seem right; it now treats the CRLF combination as the line
1401     ending, and so does not match in that case. It's only a pattern such as ^$
1402     that would hit this one: something like ^ABC$ would have failed after \r
1403     and then tried again after \r\n.
1405 ph10 131 16. Changed the comparison command for RunGrepTest from "diff -u" to "diff -ub"
1406 ph10 134 in an attempt to make files that differ only in their line terminators
1407     compare equal. This works on Linux.
1408 ph10 142
1409 ph10 141 17. Under certain error circumstances pcregrep might try to free random memory
1410     as it exited. This is now fixed, thanks to valgrind.
1411 ph10 142
1412 ph10 141 19. In pcretest, if the pattern /(?m)^$/g<any> was matched against the string
1413 ph10 142 "abc\r\n\r\n", it found an unwanted second match after the second \r. This
1414     was because its rules for how to advance for /g after matching an empty
1415 ph10 143 string at the end of a line did not allow for this case. They now check for
1416     it specially.
1417 ph10 150
1418     20. pcretest is supposed to handle patterns and data of any length, by
1419     extending its buffers when necessary. It was getting this wrong when the
1420 ph10 147 buffer for a data line had to be extended.
1421 ph10 150
1422 ph10 149 21. Added PCRE_NEWLINE_ANYCRLF which is like ANY, but matches only CR, LF, or
1423 ph10 150 CRLF as a newline sequence.
1424 ph10 152
1425 ph10 151 22. Code for handling Unicode properties in pcre_dfa_exec() wasn't being cut
1426 ph10 152 out by #ifdef SUPPORT_UCP. This did no harm, as it could never be used, but
1427     I have nevertheless tidied it up.
1428 ph10 111
1429 ph10 152 23. Added some casts to kill warnings from HP-UX ia64 compiler.
1430 ph10 123
1431 ph10 153 24. Added a man page for pcre-config.
1432 ph10 152
1433 ph10 153
1434 nigel 93 Version 7.0 19-Dec-06
1435     ---------------------
1437     1. Fixed a signed/unsigned compiler warning in pcre_compile.c, shown up by
1438     moving to gcc 4.1.1.
1440     2. The -S option for pcretest uses setrlimit(); I had omitted to #include
1441     sys/time.h, which is documented as needed for this function. It doesn't
1442     seem to matter on Linux, but it showed up on some releases of OS X.
1444     3. It seems that there are systems where bytes whose values are greater than
1445     127 match isprint() in the "C" locale. The "C" locale should be the
1446     default when a C program starts up. In most systems, only ASCII printing
1447     characters match isprint(). This difference caused the output from pcretest
1448     to vary, making some of the tests fail. I have changed pcretest so that:
1450     (a) When it is outputting text in the compiled version of a pattern, bytes
1451     other than 32-126 are always shown as hex escapes.
1453     (b) When it is outputting text that is a matched part of a subject string,
1454     it does the same, unless a different locale has been set for the match
1455     (using the /L modifier). In this case, it uses isprint() to decide.
1457     4. Fixed a major bug that caused incorrect computation of the amount of memory
1458     required for a compiled pattern when options that changed within the
1459     pattern affected the logic of the preliminary scan that determines the
1460     length. The relevant options are -x, and -i in UTF-8 mode. The result was
1461     that the computed length was too small. The symptoms of this bug were
1462     either the PCRE error "internal error: code overflow" from pcre_compile(),
1463     or a glibc crash with a message such as "pcretest: free(): invalid next
1464     size (fast)". Examples of patterns that provoked this bug (shown in
1465     pcretest format) are:
1467     /(?-x: )/x
1468     /(?x)(?-x: \s*#\s*)/
1469     /((?i)[\x{c0}])/8
1470     /(?i:[\x{c0}])/8
1472     HOWEVER: Change 17 below makes this fix obsolete as the memory computation
1473     is now done differently.
1475     5. Applied patches from Google to: (a) add a QuoteMeta function to the C++
1476     wrapper classes; (b) implement a new function in the C++ scanner that is
1477     more efficient than the old way of doing things because it avoids levels of
1478     recursion in the regex matching; (c) add a paragraph to the documentation
1479     for the FullMatch() function.
1481     6. The escape sequence \n was being treated as whatever was defined as
1482     "newline". Not only was this contrary to the documentation, which states
1483     that \n is character 10 (hex 0A), but it also went horribly wrong when
1484     "newline" was defined as CRLF. This has been fixed.
1486     7. In pcre_dfa_exec.c the value of an unsigned integer (the variable called c)
1487     was being set to -1 for the "end of line" case (supposedly a value that no
1488     character can have). Though this value is never used (the check for end of
1489     line is "zero bytes in current character"), it caused compiler complaints.
1490     I've changed it to 0xffffffff.
1492     8. In pcre_version.c, the version string was being built by a sequence of
1493     C macros that, in the event of PCRE_PRERELEASE being defined as an empty
1494     string (as it is for production releases) called a macro with an empty
1495     argument. The C standard says the result of this is undefined. The gcc
1496     compiler treats it as an empty string (which was what was wanted) but it is
1497     reported that Visual C gives an error. The source has been hacked around to
1498     avoid this problem.
1500     9. On the advice of a Windows user, included <io.h> and <fcntl.h> in Windows
1501     builds of pcretest, and changed the call to _setmode() to use _O_BINARY
1502     instead of 0x8000. Made all the #ifdefs test both _WIN32 and WIN32 (not all
1503     of them did).
1505     10. Originally, pcretest opened its input and output without "b"; then I was
1506     told that "b" was needed in some environments, so it was added for release
1507     5.0 to both the input and output. (It makes no difference on Unix-like
1508     systems.) Later I was told that it is wrong for the input on Windows. I've
1509     now abstracted the modes into two macros, to make it easier to fiddle with
1510     them, and removed "b" from the input mode under Windows.
1512     11. Added pkgconfig support for the C++ wrapper library, libpcrecpp.
1514     12. Added -help and --help to pcretest as an official way of being reminded
1515     of the options.
1517     13. Removed some redundant semicolons after macro calls in pcrecpparg.h.in
1518     and pcrecpp.cc because they annoy compilers at high warning levels.
1520     14. A bit of tidying/refactoring in pcre_exec.c in the main bumpalong loop.
1522     15. Fixed an occurrence of == in configure.ac that should have been = (shell
1523     scripts are not C programs :-) and which was not noticed because it works
1524     on Linux.
1526     16. pcretest is supposed to handle any length of pattern and data line (as one
1527     line or as a continued sequence of lines) by extending its input buffer if
1528     necessary. This feature was broken for very long pattern lines, leading to
1529     a string of junk being passed to pcre_compile() if the pattern was longer
1530     than about 50K.
1532     17. I have done a major re-factoring of the way pcre_compile() computes the
1533     amount of memory needed for a compiled pattern. Previously, there was code
1534     that made a preliminary scan of the pattern in order to do this. That was
1535     OK when PCRE was new, but as the facilities have expanded, it has become
1536     harder and harder to keep it in step with the real compile phase, and there
1537     have been a number of bugs (see for example, 4 above). I have now found a
1538     cunning way of running the real compile function in a "fake" mode that
1539     enables it to compute how much memory it would need, while actually only
1540     ever using a few hundred bytes of working memory and without too many
1541     tests of the mode. This should make future maintenance and development
1542     easier. A side effect of this work is that the limit of 200 on the nesting
1543     depth of parentheses has been removed (though this was never a serious
1544     limitation, I suspect). However, there is a downside: pcre_compile() now
1545     runs more slowly than before (30% or more, depending on the pattern). I
1546     hope this isn't a big issue. There is no effect on runtime performance.
1548     18. Fixed a minor bug in pcretest: if a pattern line was not terminated by a
1549     newline (only possible for the last line of a file) and it was a
1550     pattern that set a locale (followed by /Lsomething), pcretest crashed.
1552     19. Added additional timing features to pcretest. (1) The -tm option now times
1553     matching only, not compiling. (2) Both -t and -tm can be followed, as a
1554     separate command line item, by a number that specifies the number of
1555     repeats to use when timing. The default is 50000; this gives better
1556     precision, but takes uncomfortably long for very large patterns.
1558     20. Extended pcre_study() to be more clever in cases where a branch of a
1559     subpattern has no definite first character. For example, (a*|b*)[cd] would
1560     previously give no result from pcre_study(). Now it recognizes that the
1561     first character must be a, b, c, or d.
1563     21. There was an incorrect error "recursive call could loop indefinitely" if
1564     a subpattern (or the entire pattern) that was being tested for matching an
1565     empty string contained only one non-empty item after a nested subpattern.
1566     For example, the pattern (?>\x{100}*)\d(?R) provoked this error
1567     incorrectly, because the \d was being skipped in the check.
1569     22. The pcretest program now has a new pattern option /B and a command line
1570     option -b, which is equivalent to adding /B to every pattern. This causes
1571     it to show the compiled bytecode, without the additional information that
1572     -d shows. The effect of -d is now the same as -b with -i (and similarly, /D
1573     is the same as /B/I).
1575     23. A new optimization is now able automatically to treat some sequences such
1576     as a*b as a*+b. More specifically, if something simple (such as a character
1577     or a simple class like \d) has an unlimited quantifier, and is followed by
1578     something that cannot possibly match the quantified thing, the quantifier
1579     is automatically "possessified".
1581     24. A recursive reference to a subpattern whose number was greater than 39
1582     went wrong under certain circumstances in UTF-8 mode. This bug could also
1583     have affected the operation of pcre_study().
1585     25. Realized that a little bit of performance could be had by replacing
1586     (c & 0xc0) == 0xc0 with c >= 0xc0 when processing UTF-8 characters.
1588     26. Timing data from pcretest is now shown to 4 decimal places instead of 3.
1590     27. Possessive quantifiers such as a++ were previously implemented by turning
1591     them into atomic groups such as ($>a+). Now they have their own opcodes,
1592     which improves performance. This includes the automatically created ones
1593     from 23 above.
1595     28. A pattern such as (?=(\w+))\1: which simulates an atomic group using a
1596     lookahead was broken if it was not anchored. PCRE was mistakenly expecting
1597     the first matched character to be a colon. This applied both to named and
1598     numbered groups.
1600     29. The ucpinternal.h header file was missing its idempotency #ifdef.
1602     30. I was sent a "project" file called libpcre.a.dev which I understand makes
1603     building PCRE on Windows easier, so I have included it in the distribution.
1605     31. There is now a check in pcretest against a ridiculously large number being
1606     returned by pcre_exec() or pcre_dfa_exec(). If this happens in a /g or /G
1607     loop, the loop is abandoned.
1609     32. Forward references to subpatterns in conditions such as (?(2)...) where
1610     subpattern 2 is defined later cause pcre_compile() to search forwards in
1611     the pattern for the relevant set of parentheses. This search went wrong
1612     when there were unescaped parentheses in a character class, parentheses
1613     escaped with \Q...\E, or parentheses in a #-comment in /x mode.
1615     33. "Subroutine" calls and backreferences were previously restricted to
1616     referencing subpatterns earlier in the regex. This restriction has now
1617     been removed.
1619     34. Added a number of extra features that are going to be in Perl 5.10. On the
1620     whole, these are just syntactic alternatives for features that PCRE had
1621     previously implemented using the Python syntax or my own invention. The
1622     other formats are all retained for compatibility.
1624     (a) Named groups can now be defined as (?<name>...) or (?'name'...) as well
1625     as (?P<name>...). The new forms, as well as being in Perl 5.10, are
1626     also .NET compatible.
1628     (b) A recursion or subroutine call to a named group can now be defined as
1629     (?&name) as well as (?P>name).
1631     (c) A backreference to a named group can now be defined as \k<name> or
1632     \k'name' as well as (?P=name). The new forms, as well as being in Perl
1633     5.10, are also .NET compatible.
1635     (d) A conditional reference to a named group can now use the syntax
1636     (?(<name>) or (?('name') as well as (?(name).
1638     (e) A "conditional group" of the form (?(DEFINE)...) can be used to define
1639     groups (named and numbered) that are never evaluated inline, but can be
1640     called as "subroutines" from elsewhere. In effect, the DEFINE condition
1641     is always false. There may be only one alternative in such a group.
1643     (f) A test for recursion can be given as (?(R1).. or (?(R&name)... as well
1644     as the simple (?(R). The condition is true only if the most recent
1645     recursion is that of the given number or name. It does not search out
1646     through the entire recursion stack.
1648     (g) The escape \gN or \g{N} has been added, where N is a positive or
1649     negative number, specifying an absolute or relative reference.
1651     35. Tidied to get rid of some further signed/unsigned compiler warnings and
1652     some "unreachable code" warnings.
1654     36. Updated the Unicode property tables to Unicode version 5.0.0. Amongst other
1655     things, this adds five new scripts.
1657     37. Perl ignores orphaned \E escapes completely. PCRE now does the same.
1658     There were also incompatibilities regarding the handling of \Q..\E inside
1659     character classes, for example with patterns like [\Qa\E-\Qz\E] where the
1660     hyphen was adjacent to \Q or \E. I hope I've cleared all this up now.
1662     38. Like Perl, PCRE detects when an indefinitely repeated parenthesized group
1663     matches an empty string, and forcibly breaks the loop. There were bugs in
1664     this code in non-simple cases. For a pattern such as ^(a()*)* matched
1665     against aaaa the result was just "a" rather than "aaaa", for example. Two
1666     separate and independent bugs (that affected different cases) have been
1667     fixed.
1669     39. Refactored the code to abolish the use of different opcodes for small
1670     capturing bracket numbers. This is a tidy that I avoided doing when I
1671     removed the limit on the number of capturing brackets for 3.5 back in 2001.
1672     The new approach is not only tidier, it makes it possible to reduce the
1673     memory needed to fix the previous bug (38).
1675     40. Implemented PCRE_NEWLINE_ANY to recognize any of the Unicode newline
1676     sequences (http://unicode.org/unicode/reports/tr18/) as "newline" when
1677     processing dot, circumflex, or dollar metacharacters, or #-comments in /x
1678     mode.
1680     41. Add \R to match any Unicode newline sequence, as suggested in the Unicode
1681     report.
1683     42. Applied patch, originally from Ari Pollak, modified by Google, to allow
1684     copy construction and assignment in the C++ wrapper.
1686     43. Updated pcregrep to support "--newline=any". In the process, I fixed a
1687     couple of bugs that could have given wrong results in the "--newline=crlf"
1688     case.
1690     44. Added a number of casts and did some reorganization of signed/unsigned int
1691     variables following suggestions from Dair Grant. Also renamed the variable
1692     "this" as "item" because it is a C++ keyword.
1694     45. Arranged for dftables to add
1696     #include "pcre_internal.h"
1698     to pcre_chartables.c because without it, gcc 4.x may remove the array
1699     definition from the final binary if PCRE is built into a static library and
1700     dead code stripping is activated.
1702     46. For an unanchored pattern, if a match attempt fails at the start of a
1703     newline sequence, and the newline setting is CRLF or ANY, and the next two
1704     characters are CRLF, advance by two characters instead of one.
1707 nigel 91 Version 6.7 04-Jul-06
1708     ---------------------
1710     1. In order to handle tests when input lines are enormously long, pcretest has
1711     been re-factored so that it automatically extends its buffers when
1712     necessary. The code is crude, but this _is_ just a test program. The
1713     default size has been increased from 32K to 50K.
1715     2. The code in pcre_study() was using the value of the re argument before
1716     testing it for NULL. (Of course, in any sensible call of the function, it
1717     won't be NULL.)
1719     3. The memmove() emulation function in pcre_internal.h, which is used on
1720     systems that lack both memmove() and bcopy() - that is, hardly ever -
1721     was missing a "static" storage class specifier.
1723     4. When UTF-8 mode was not set, PCRE looped when compiling certain patterns
1724     containing an extended class (one that cannot be represented by a bitmap
1725     because it contains high-valued characters or Unicode property items, e.g.
1726     [\pZ]). Almost always one would set UTF-8 mode when processing such a
1727     pattern, but PCRE should not loop if you do not (it no longer does).
1728     [Detail: two cases were found: (a) a repeated subpattern containing an
1729     extended class; (b) a recursive reference to a subpattern that followed a
1730     previous extended class. It wasn't skipping over the extended class
1731     correctly when UTF-8 mode was not set.]
1733     5. A negated single-character class was not being recognized as fixed-length
1734     in lookbehind assertions such as (?<=[^f]), leading to an incorrect
1735     compile error "lookbehind assertion is not fixed length".
1737     6. The RunPerlTest auxiliary script was showing an unexpected difference
1738     between PCRE and Perl for UTF-8 tests. It turns out that it is hard to
1739     write a Perl script that can interpret lines of an input file either as
1740     byte characters or as UTF-8, which is what "perltest" was being required to
1741     do for the non-UTF-8 and UTF-8 tests, respectively. Essentially what you
1742     can't do is switch easily at run time between having the "use utf8;" pragma
1743     or not. In the end, I fudged it by using the RunPerlTest script to insert
1744     "use utf8;" explicitly for the UTF-8 tests.
1746     7. In multiline (/m) mode, PCRE was matching ^ after a terminating newline at
1747     the end of the subject string, contrary to the documentation and to what
1748     Perl does. This was true of both matching functions. Now it matches only at
1749     the start of the subject and immediately after *internal* newlines.
1751     8. A call of pcre_fullinfo() from pcretest to get the option bits was passing
1752     a pointer to an int instead of a pointer to an unsigned long int. This
1753     caused problems on 64-bit systems.
1755     9. Applied a patch from the folks at Google to pcrecpp.cc, to fix "another
1756     instance of the 'standard' template library not being so standard".
1758     10. There was no check on the number of named subpatterns nor the maximum
1759     length of a subpattern name. The product of these values is used to compute
1760     the size of the memory block for a compiled pattern. By supplying a very
1761     long subpattern name and a large number of named subpatterns, the size
1762     computation could be caused to overflow. This is now prevented by limiting
1763     the length of names to 32 characters, and the number of named subpatterns
1764     to 10,000.
1766     11. Subpatterns that are repeated with specific counts have to be replicated in
1767     the compiled pattern. The size of memory for this was computed from the
1768     length of the subpattern and the repeat count. The latter is limited to
1769     65535, but there was no limit on the former, meaning that integer overflow
1770     could in principle occur. The compiled length of a repeated subpattern is
1771     now limited to 30,000 bytes in order to prevent this.
1773     12. Added the optional facility to have named substrings with the same name.
1775     13. Added the ability to use a named substring as a condition, using the
1776     Python syntax: (?(name)yes|no). This overloads (?(R)... and names that
1777     are numbers (not recommended). Forward references are permitted.
1779     14. Added forward references in named backreferences (if you see what I mean).
1781     15. In UTF-8 mode, with the PCRE_DOTALL option set, a quantified dot in the
1782     pattern could run off the end of the subject. For example, the pattern
1783     "(?s)(.{1,5})"8 did this with the subject "ab".
1785     16. If PCRE_DOTALL or PCRE_MULTILINE were set, pcre_dfa_exec() behaved as if
1786     PCRE_CASELESS was set when matching characters that were quantified with ?
1787     or *.
1789     17. A character class other than a single negated character that had a minimum
1790     but no maximum quantifier - for example [ab]{6,} - was not handled
1791     correctly by pce_dfa_exec(). It would match only one character.
1793     18. A valid (though odd) pattern that looked like a POSIX character
1794     class but used an invalid character after [ (for example [[,abc,]]) caused
1795     pcre_compile() to give the error "Failed: internal error: code overflow" or
1796     in some cases to crash with a glibc free() error. This could even happen if
1797     the pattern terminated after [[ but there just happened to be a sequence of
1798     letters, a binary zero, and a closing ] in the memory that followed.
1800     19. Perl's treatment of octal escapes in the range \400 to \777 has changed
1801     over the years. Originally (before any Unicode support), just the bottom 8
1802     bits were taken. Thus, for example, \500 really meant \100. Nowadays the
1803     output from "man perlunicode" includes this:
1805     The regular expression compiler produces polymorphic opcodes. That
1806     is, the pattern adapts to the data and automatically switches to
1807     the Unicode character scheme when presented with Unicode data--or
1808     instead uses a traditional byte scheme when presented with byte
1809     data.
1811     Sadly, a wide octal escape does not cause a switch, and in a string with
1812     no other multibyte characters, these octal escapes are treated as before.
1813     Thus, in Perl, the pattern /\500/ actually matches \100 but the pattern
1814     /\500|\x{1ff}/ matches \500 or \777 because the whole thing is treated as a
1815     Unicode string.
1817     I have not perpetrated such confusion in PCRE. Up till now, it took just
1818     the bottom 8 bits, as in old Perl. I have now made octal escapes with
1819     values greater than \377 illegal in non-UTF-8 mode. In UTF-8 mode they
1820     translate to the appropriate multibyte character.
1822     29. Applied some refactoring to reduce the number of warnings from Microsoft
1823     and Borland compilers. This has included removing the fudge introduced
1824     seven years ago for the OS/2 compiler (see 2.02/2 below) because it caused
1825     a warning about an unused variable.
1827     21. PCRE has not included VT (character 0x0b) in the set of whitespace
1828     characters since release 4.0, because Perl (from release 5.004) does not.
1829     [Or at least, is documented not to: some releases seem to be in conflict
1830     with the documentation.] However, when a pattern was studied with
1831     pcre_study() and all its branches started with \s, PCRE still included VT
1832     as a possible starting character. Of course, this did no harm; it just
1833     caused an unnecessary match attempt.
1835     22. Removed a now-redundant internal flag bit that recorded the fact that case
1836     dependency changed within the pattern. This was once needed for "required
1837     byte" processing, but is no longer used. This recovers a now-scarce options
1838     bit. Also moved the least significant internal flag bit to the most-
1839     significant bit of the word, which was not previously used (hangover from
1840     the days when it was an int rather than a uint) to free up another bit for
1841     the future.
1843     23. Added support for CRLF line endings as well as CR and LF. As well as the
1844     default being selectable at build time, it can now be changed at runtime
1845     via the PCRE_NEWLINE_xxx flags. There are now options for pcregrep to
1846     specify that it is scanning data with non-default line endings.
1848     24. Changed the definition of CXXLINK to make it agree with the definition of
1849     LINK in the Makefile, by replacing LDFLAGS to CXXFLAGS.
1851     25. Applied Ian Taylor's patches to avoid using another stack frame for tail
1852     recursions. This makes a big different to stack usage for some patterns.
1854     26. If a subpattern containing a named recursion or subroutine reference such
1855     as (?P>B) was quantified, for example (xxx(?P>B)){3}, the calculation of
1856     the space required for the compiled pattern went wrong and gave too small a
1857     value. Depending on the environment, this could lead to "Failed: internal
1858     error: code overflow at offset 49" or "glibc detected double free or
1859     corruption" errors.
1861     27. Applied patches from Google (a) to support the new newline modes and (b) to
1862     advance over multibyte UTF-8 characters in GlobalReplace.
1864     28. Change free() to pcre_free() in pcredemo.c. Apparently this makes a
1865     difference for some implementation of PCRE in some Windows version.
1867     29. Added some extra testing facilities to pcretest:
1869     \q<number> in a data line sets the "match limit" value
1870     \Q<number> in a data line sets the "match recursion limt" value
1871     -S <number> sets the stack size, where <number> is in megabytes
1873     The -S option isn't available for Windows.
1876 nigel 89 Version 6.6 06-Feb-06
1877     ---------------------
1879     1. Change 16(a) for 6.5 broke things, because PCRE_DATA_SCOPE was not defined
1880     in pcreposix.h. I have copied the definition from pcre.h.
1882     2. Change 25 for 6.5 broke compilation in a build directory out-of-tree
1883     because pcre.h is no longer a built file.
1885     3. Added Jeff Friedl's additional debugging patches to pcregrep. These are
1886     not normally included in the compiled code.
1889 nigel 87 Version 6.5 01-Feb-06
1890     ---------------------
1892     1. When using the partial match feature with pcre_dfa_exec(), it was not
1893     anchoring the second and subsequent partial matches at the new starting
1894     point. This could lead to incorrect results. For example, with the pattern
1895     /1234/, partially matching against "123" and then "a4" gave a match.
1897     2. Changes to pcregrep:
1899     (a) All non-match returns from pcre_exec() were being treated as failures
1900     to match the line. Now, unless the error is PCRE_ERROR_NOMATCH, an
1901     error message is output. Some extra information is given for the
1903     probably the only errors that are likely to be caused by users (by
1904     specifying a regex that has nested indefinite repeats, for instance).
1905     If there are more than 20 of these errors, pcregrep is abandoned.
1907     (b) A binary zero was treated as data while matching, but terminated the
1908     output line if it was written out. This has been fixed: binary zeroes
1909     are now no different to any other data bytes.
1911     (c) Whichever of the LC_ALL or LC_CTYPE environment variables is set is
1912     used to set a locale for matching. The --locale=xxxx long option has
1913     been added (no short equivalent) to specify a locale explicitly on the
1914     pcregrep command, overriding the environment variables.
1916     (d) When -B was used with -n, some line numbers in the output were one less
1917     than they should have been.
1919     (e) Added the -o (--only-matching) option.
1921     (f) If -A or -C was used with -c (count only), some lines of context were
1922     accidentally printed for the final match.
1924     (g) Added the -H (--with-filename) option.
1926     (h) The combination of options -rh failed to suppress file names for files
1927     that were found from directory arguments.
1929     (i) Added the -D (--devices) and -d (--directories) options.
1931     (j) Added the -F (--fixed-strings) option.
1933     (k) Allow "-" to be used as a file name for -f as well as for a data file.
1935     (l) Added the --colo(u)r option.
1937     (m) Added Jeffrey Friedl's -S testing option, but within #ifdefs so that it
1938     is not present by default.
1940     3. A nasty bug was discovered in the handling of recursive patterns, that is,
1941     items such as (?R) or (?1), when the recursion could match a number of
1942     alternatives. If it matched one of the alternatives, but subsequently,
1943     outside the recursion, there was a failure, the code tried to back up into
1944     the recursion. However, because of the way PCRE is implemented, this is not
1945     possible, and the result was an incorrect result from the match.
1947     In order to prevent this happening, the specification of recursion has
1948     been changed so that all such subpatterns are automatically treated as
1949     atomic groups. Thus, for example, (?R) is treated as if it were (?>(?R)).
1951     4. I had overlooked the fact that, in some locales, there are characters for
1952     which isalpha() is true but neither isupper() nor islower() are true. In
1953     the fr_FR locale, for instance, the \xAA and \xBA characters (ordmasculine
1954     and ordfeminine) are like this. This affected the treatment of \w and \W
1955     when they appeared in character classes, but not when they appeared outside
1956     a character class. The bit map for "word" characters is now created
1957     separately from the results of isalnum() instead of just taking it from the
1958     upper, lower, and digit maps. (Plus the underscore character, of course.)
1960     5. The above bug also affected the handling of POSIX character classes such as
1961     [[:alpha:]] and [[:alnum:]]. These do not have their own bit maps in PCRE's
1962     permanent tables. Instead, the bit maps for such a class were previously
1963     created as the appropriate unions of the upper, lower, and digit bitmaps.
1964     Now they are created by subtraction from the [[:word:]] class, which has
1965     its own bitmap.
1967     6. The [[:blank:]] character class matches horizontal, but not vertical space.
1968     It is created by subtracting the vertical space characters (\x09, \x0a,
1969     \x0b, \x0c) from the [[:space:]] bitmap. Previously, however, the
1970     subtraction was done in the overall bitmap for a character class, meaning
1971     that a class such as [\x0c[:blank:]] was incorrect because \x0c would not
1972     be recognized. This bug has been fixed.
1974     7. Patches from the folks at Google:
1976     (a) pcrecpp.cc: "to handle a corner case that may or may not happen in
1977     real life, but is still worth protecting against".
1979     (b) pcrecpp.cc: "corrects a bug when negative radixes are used with
1980     regular expressions".
1982     (c) pcre_scanner.cc: avoid use of std::count() because not all systems
1983     have it.
1985     (d) Split off pcrecpparg.h from pcrecpp.h and had the former built by
1986     "configure" and the latter not, in order to fix a problem somebody had
1987     with compiling the Arg class on HP-UX.
1989     (e) Improve the error-handling of the C++ wrapper a little bit.
1991     (f) New tests for checking recursion limiting.
1993     8. The pcre_memmove() function, which is used only if the environment does not
1994     have a standard memmove() function (and is therefore rarely compiled),
1995     contained two bugs: (a) use of int instead of size_t, and (b) it was not
1996     returning a result (though PCRE never actually uses the result).
1998     9. In the POSIX regexec() interface, if nmatch is specified as a ridiculously
1999     large number - greater than INT_MAX/(3*sizeof(int)) - REG_ESPACE is
2000     returned instead of calling malloc() with an overflowing number that would
2001     most likely cause subsequent chaos.
2003     10. The debugging option of pcretest was not showing the NO_AUTO_CAPTURE flag.
2005     11. The POSIX flag REG_NOSUB is now supported. When a pattern that was compiled
2006     with this option is matched, the nmatch and pmatch options of regexec() are
2007     ignored.
2009     12. Added REG_UTF8 to the POSIX interface. This is not defined by POSIX, but is
2010     provided in case anyone wants to the the POSIX interface with UTF-8
2011     strings.
2013     13. Added CXXLDFLAGS to the Makefile parameters to provide settings only on the
2014     C++ linking (needed for some HP-UX environments).
2016     14. Avoid compiler warnings in get_ucpname() when compiled without UCP support
2017     (unused parameter) and in the pcre_printint() function (omitted "default"
2018     switch label when the default is to do nothing).
2020     15. Added some code to make it possible, when PCRE is compiled as a C++
2021     library, to replace subject pointers for pcre_exec() with a smart pointer
2022     class, thus making it possible to process discontinuous strings.
2024     16. The two macros PCRE_EXPORT and PCRE_DATA_SCOPE are confusing, and perform
2025     much the same function. They were added by different people who were trying
2026     to make PCRE easy to compile on non-Unix systems. It has been suggested
2027     that PCRE_EXPORT be abolished now that there is more automatic apparatus
2028     for compiling on Windows systems. I have therefore replaced it with
2029     PCRE_DATA_SCOPE. This is set automatically for Windows; if not set it
2030     defaults to "extern" for C or "extern C" for C++, which works fine on
2031     Unix-like systems. It is now possible to override the value of PCRE_DATA_
2032     SCOPE with something explicit in config.h. In addition:
2034     (a) pcreposix.h still had just "extern" instead of either of these macros;
2035     I have replaced it with PCRE_DATA_SCOPE.
2037     (b) Functions such as _pcre_xclass(), which are internal to the library,
2038     but external in the C sense, all had PCRE_EXPORT in their definitions.
2039     This is apparently wrong for the Windows case, so I have removed it.
2040     (It makes no difference on Unix-like systems.)
2042     17. Added a new limit, MATCH_LIMIT_RECURSION, which limits the depth of nesting
2043     of recursive calls to match(). This is different to MATCH_LIMIT because
2044     that limits the total number of calls to match(), not all of which increase
2045     the depth of recursion. Limiting the recursion depth limits the amount of
2046     stack (or heap if NO_RECURSE is set) that is used. The default can be set
2047     when PCRE is compiled, and changed at run time. A patch from Google adds
2048     this functionality to the C++ interface.
2050     18. Changes to the handling of Unicode character properties:
2052     (a) Updated the table to Unicode 4.1.0.
2054     (b) Recognize characters that are not in the table as "Cn" (undefined).
2056     (c) I revised the way the table is implemented to a much improved format
2057     which includes recognition of ranges. It now supports the ranges that
2058     are defined in UnicodeData.txt, and it also amalgamates other
2059     characters into ranges. This has reduced the number of entries in the
2060     table from around 16,000 to around 3,000, thus reducing its size
2061     considerably. I realized I did not need to use a tree structure after
2062     all - a binary chop search is just as efficient. Having reduced the
2063     number of entries, I extended their size from 6 bytes to 8 bytes to
2064     allow for more data.
2066     (d) Added support for Unicode script names via properties such as \p{Han}.
2068     19. In UTF-8 mode, a backslash followed by a non-Ascii character was not
2069     matching that character.
2071     20. When matching a repeated Unicode property with a minimum greater than zero,
2072     (for example \pL{2,}), PCRE could look past the end of the subject if it
2073     reached it while seeking the minimum number of characters. This could
2074     happen only if some of the characters were more than one byte long, because
2075     there is a check for at least the minimum number of bytes.
2077     21. Refactored the implementation of \p and \P so as to be more general, to
2078     allow for more different types of property in future. This has changed the
2079     compiled form incompatibly. Anybody with saved compiled patterns that use
2080     \p or \P will have to recompile them.
2082     22. Added "Any" and "L&" to the supported property types.
2084     23. Recognize \x{...} as a code point specifier, even when not in UTF-8 mode,
2085     but give a compile time error if the value is greater than 0xff.
2087     24. The man pages for pcrepartial, pcreprecompile, and pcre_compile2 were
2088     accidentally not being installed or uninstalled.
2090     25. The pcre.h file was built from pcre.h.in, but the only changes that were
2091     made were to insert the current release number. This seemed silly, because
2092     it made things harder for people building PCRE on systems that don't run
2093     "configure". I have turned pcre.h into a distributed file, no longer built
2094     by "configure", with the version identification directly included. There is
2095     no longer a pcre.h.in file.
2097     However, this change necessitated a change to the pcre-config script as
2098     well. It is built from pcre-config.in, and one of the substitutions was the
2099     release number. I have updated configure.ac so that ./configure now finds
2100     the release number by grepping pcre.h.
2102     26. Added the ability to run the tests under valgrind.
2105 nigel 85 Version 6.4 05-Sep-05
2106     ---------------------
2108     1. Change 6.0/10/(l) to pcregrep introduced a bug that caused separator lines
2109     "--" to be printed when multiple files were scanned, even when none of the
2110     -A, -B, or -C options were used. This is not compatible with Gnu grep, so I
2111     consider it to be a bug, and have restored the previous behaviour.
2113     2. A couple of code tidies to get rid of compiler warnings.
2115     3. The pcretest program used to cheat by referring to symbols in the library
2116     whose names begin with _pcre_. These are internal symbols that are not
2117     really supposed to be visible externally, and in some environments it is
2118     possible to suppress them. The cheating is now confined to including
2119     certain files from the library's source, which is a bit cleaner.
2121     4. Renamed pcre.in as pcre.h.in to go with pcrecpp.h.in; it also makes the
2122     file's purpose clearer.
2124     5. Reorganized pcre_ucp_findchar().
2127 nigel 83 Version 6.3 15-Aug-05
2128     ---------------------
2130     1. The file libpcre.pc.in did not have general read permission in the tarball.
2132     2. There were some problems when building without C++ support:
2134     (a) If C++ support was not built, "make install" and "make test" still
2135     tried to test it.
2137     (b) There were problems when the value of CXX was explicitly set. Some
2138     changes have been made to try to fix these, and ...
2140     (c) --disable-cpp can now be used to explicitly disable C++ support.
2142     (d) The use of @CPP_OBJ@ directly caused a blank line preceded by a
2143     backslash in a target when C++ was disabled. This confuses some
2144     versions of "make", apparently. Using an intermediate variable solves
2145     this. (Same for CPP_LOBJ.)
2147     3. $(LINK_FOR_BUILD) now includes $(CFLAGS_FOR_BUILD) and $(LINK)
2148     (non-Windows) now includes $(CFLAGS) because these flags are sometimes
2149     necessary on certain architectures.
2151     4. Added a setting of -export-symbols-regex to the link command to remove
2152     those symbols that are exported in the C sense, but actually are local
2153     within the library, and not documented. Their names all begin with
2154     "_pcre_". This is not a perfect job, because (a) we have to except some
2155     symbols that pcretest ("illegally") uses, and (b) the facility isn't always
2156     available (and never for static libraries). I have made a note to try to
2157     find a way round (a) in the future.
2160 nigel 81 Version 6.2 01-Aug-05
2161     ---------------------
2163     1. There was no test for integer overflow of quantifier values. A construction
2164     such as {1111111111111111} would give undefined results. What is worse, if
2165     a minimum quantifier for a parenthesized subpattern overflowed and became
2166     negative, the calculation of the memory size went wrong. This could have
2167     led to memory overwriting.
2169     2. Building PCRE using VPATH was broken. Hopefully it is now fixed.
2171     3. Added "b" to the 2nd argument of fopen() in dftables.c, for non-Unix-like
2172     operating environments where this matters.
2174     4. Applied Giuseppe Maxia's patch to add additional features for controlling
2175     PCRE options from within the C++ wrapper.
2177     5. Named capturing subpatterns were not being correctly counted when a pattern
2178     was compiled. This caused two problems: (a) If there were more than 100
2179     such subpatterns, the calculation of the memory needed for the whole
2180     compiled pattern went wrong, leading to an overflow error. (b) Numerical
2181     back references of the form \12, where the number was greater than 9, were
2182     not recognized as back references, even though there were sufficient
2183     previous subpatterns.
2185     6. Two minor patches to pcrecpp.cc in order to allow it to compile on older
2186     versions of gcc, e.g. 2.95.4.
2189 nigel 79 Version 6.1 21-Jun-05
2190     ---------------------
2192     1. There was one reference to the variable "posix" in pcretest.c that was not
2193     surrounded by "#if !defined NOPOSIX".
2195     2. Make it possible to compile pcretest without DFA support, UTF8 support, or
2196     the cross-check on the old pcre_info() function, for the benefit of the
2197     cut-down version of PCRE that is currently imported into Exim.
2199     3. A (silly) pattern starting with (?i)(?-i) caused an internal space
2200     allocation error. I've done the easy fix, which wastes 2 bytes for sensible
2201     patterns that start (?i) but I don't think that matters. The use of (?i) is
2202     just an example; this all applies to the other options as well.
2204     4. Since libtool seems to echo the compile commands it is issuing, the output
2205     from "make" can be reduced a bit by putting "@" in front of each libtool
2206     compile command.
2208     5. Patch from the folks at Google for configure.in to be a bit more thorough
2209     in checking for a suitable C++ installation before trying to compile the
2210     C++ stuff. This should fix a reported problem when a compiler was present,
2211     but no suitable headers.
2213     6. The man pages all had just "PCRE" as their title. I have changed them to
2214     be the relevant file name. I have also arranged that these names are
2215     retained in the file doc/pcre.txt, which is a concatenation in text format
2216     of all the man pages except the little individual ones for each function.
2218     7. The NON-UNIX-USE file had not been updated for the different set of source
2219     files that come with release 6. I also added a few comments about the C++
2220     wrapper.
2223 nigel 77 Version 6.0 07-Jun-05
2224     ---------------------
2226     1. Some minor internal re-organization to help with my DFA experiments.
2228     2. Some missing #ifdef SUPPORT_UCP conditionals in pcretest and printint that
2229     didn't matter for the library itself when fully configured, but did matter
2230     when compiling without UCP support, or within Exim, where the ucp files are
2231     not imported.
2233     3. Refactoring of the library code to split up the various functions into
2234     different source modules. The addition of the new DFA matching code (see
2235     below) to a single monolithic source would have made it really too
2236     unwieldy, quite apart from causing all the code to be include in a
2237     statically linked application, when only some functions are used. This is
2238     relevant even without the DFA addition now that patterns can be compiled in
2239     one application and matched in another.
2241     The downside of splitting up is that there have to be some external
2242     functions and data tables that are used internally in different modules of
2243     the library but which are not part of the API. These have all had their
2244     names changed to start with "_pcre_" so that they are unlikely to clash
2245     with other external names.
2247     4. Added an alternate matching function, pcre_dfa_exec(), which matches using
2248     a different (DFA) algorithm. Although it is slower than the original
2249     function, it does have some advantages for certain types of matching
2250     problem.
2252     5. Upgrades to pcretest in order to test the features of pcre_dfa_exec(),
2253     including restarting after a partial match.
2255     6. A patch for pcregrep that defines INVALID_FILE_ATTRIBUTES if it is not
2256     defined when compiling for Windows was sent to me. I have put it into the
2257     code, though I have no means of testing or verifying it.
2259     7. Added the pcre_refcount() auxiliary function.
2261     8. Added the PCRE_FIRSTLINE option. This constrains an unanchored pattern to
2262     match before or at the first newline in the subject string. In pcretest,
2263     the /f option on a pattern can be used to set this.
2265     9. A repeated \w when used in UTF-8 mode with characters greater than 256
2266     would behave wrongly. This has been present in PCRE since release 4.0.
2268     10. A number of changes to the pcregrep command:
2270     (a) Refactored how -x works; insert ^(...)$ instead of setting
2271     PCRE_ANCHORED and checking the length, in preparation for adding
2272     something similar for -w.
2274     (b) Added the -w (match as a word) option.
2276     (c) Refactored the way lines are read and buffered so as to have more
2277     than one at a time available.
2279     (d) Implemented a pcregrep test script.
2281     (e) Added the -M (multiline match) option. This allows patterns to match
2282     over several lines of the subject. The buffering ensures that at least
2283     8K, or the rest of the document (whichever is the shorter) is available
2284     for matching (and similarly the previous 8K for lookbehind assertions).
2286     (f) Changed the --help output so that it now says
2288     -w, --word-regex(p)
2290     instead of two lines, one with "regex" and the other with "regexp"
2291     because that confused at least one person since the short forms are the
2292     same. (This required a bit of code, as the output is generated
2293     automatically from a table. It wasn't just a text change.)
2295     (g) -- can be used to terminate pcregrep options if the next thing isn't an
2296     option but starts with a hyphen. Could be a pattern or a path name
2297     starting with a hyphen, for instance.
2299     (h) "-" can be given as a file name to represent stdin.
2301     (i) When file names are being printed, "(standard input)" is used for
2302     the standard input, for compatibility with GNU grep. Previously
2303     "<stdin>" was used.
2305     (j) The option --label=xxx can be used to supply a name to be used for
2306     stdin when file names are being printed. There is no short form.
2308     (k) Re-factored the options decoding logic because we are going to add
2309     two more options that take data. Such options can now be given in four
2310     different ways, e.g. "-fname", "-f name", "--file=name", "--file name".
2312     (l) Added the -A, -B, and -C options for requesting that lines of context
2313     around matches be printed.
2315     (m) Added the -L option to print the names of files that do not contain
2316     any matching lines, that is, the complement of -l.
2318     (n) The return code is 2 if any file cannot be opened, but pcregrep does
2319     continue to scan other files.
2321     (o) The -s option was incorrectly implemented. For compatibility with other
2322     greps, it now suppresses the error message for a non-existent or non-
2323     accessible file (but not the return code). There is a new option called
2324     -q that suppresses the output of matching lines, which was what -s was
2325     previously doing.
2327     (p) Added --include and --exclude options to specify files for inclusion
2328     and exclusion when recursing.
2330     11. The Makefile was not using the Autoconf-supported LDFLAGS macro properly.
2331     Hopefully, it now does.
2333     12. Missing cast in pcre_study().
2335     13. Added an "uninstall" target to the makefile.
2337     14. Replaced "extern" in the function prototypes in Makefile.in with
2338     "PCRE_DATA_SCOPE", which defaults to 'extern' or 'extern "C"' in the Unix
2339     world, but is set differently for Windows.
2341     15. Added a second compiling function called pcre_compile2(). The only
2342     difference is that it has an extra argument, which is a pointer to an
2343     integer error code. When there is a compile-time failure, this is set
2344     non-zero, in addition to the error test pointer being set to point to an
2345     error message. The new argument may be NULL if no error number is required
2346     (but then you may as well call pcre_compile(), which is now just a
2347     wrapper). This facility is provided because some applications need a
2348     numeric error indication, but it has also enabled me to tidy up the way
2349     compile-time errors are handled in the POSIX wrapper.
2351     16. Added VPATH=.libs to the makefile; this should help when building with one
2352     prefix path and installing with another. (Or so I'm told by someone who
2353     knows more about this stuff than I do.)
2355     17. Added a new option, REG_DOTALL, to the POSIX function regcomp(). This
2356     passes PCRE_DOTALL to the pcre_compile() function, making the "." character
2357     match everything, including newlines. This is not POSIX-compatible, but
2358     somebody wanted the feature. From pcretest it can be activated by using
2359     both the P and the s flags.
2361     18. AC_PROG_LIBTOOL appeared twice in Makefile.in. Removed one.
2363     19. libpcre.pc was being incorrectly installed as executable.
2365     20. A couple of places in pcretest check for end-of-line by looking for '\n';
2366     it now also looks for '\r' so that it will work unmodified on Windows.
2368     21. Added Google's contributed C++ wrapper to the distribution.
2370     22. Added some untidy missing memory free() calls in pcretest, to keep
2371     Electric Fence happy when testing.
2375 nigel 75 Version 5.0 13-Sep-04
2376     ---------------------
2378     1. Internal change: literal characters are no longer packed up into items
2379     containing multiple characters in a single byte-string. Each character
2380     is now matched using a separate opcode. However, there may be more than one
2381     byte in the character in UTF-8 mode.
2383     2. The pcre_callout_block structure has two new fields: pattern_position and
2384     next_item_length. These contain the offset in the pattern to the next match
2385     item, and its length, respectively.
2387     3. The PCRE_AUTO_CALLOUT option for pcre_compile() requests the automatic
2388     insertion of callouts before each pattern item. Added the /C option to
2389     pcretest to make use of this.
2391     4. On the advice of a Windows user, the lines
2393     #if defined(_WIN32) || defined(WIN32)
2394     _setmode( _fileno( stdout ), 0x8000 );
2395     #endif /* defined(_WIN32) || defined(WIN32) */
2397     have been added to the source of pcretest. This apparently does useful
2398     magic in relation to line terminators.
2400     5. Changed "r" and "w" in the calls to fopen() in pcretest to "rb" and "wb"
2401     for the benefit of those environments where the "b" makes a difference.
2403     6. The icc compiler has the same options as gcc, but "configure" doesn't seem
2404     to know about it. I have put a hack into configure.in that adds in code
2405     to set GCC=yes if CC=icc. This seems to end up at a point in the
2406     generated configure script that is early enough to affect the setting of
2407     compiler options, which is what is needed, but I have no means of testing
2408     whether it really works. (The user who reported this had patched the
2409     generated configure script, which of course I cannot do.)
2411     LATER: After change 22 below (new libtool files), the configure script
2412     seems to know about icc (and also ecc). Therefore, I have commented out
2413     this hack in configure.in.
2415     7. Added support for pkg-config (2 patches were sent in).
2417     8. Negated POSIX character classes that used a combination of internal tables
2418     were completely broken. These were [[:^alpha:]], [[:^alnum:]], and
2419     [[:^ascii]]. Typically, they would match almost any characters. The other
2420     POSIX classes were not broken in this way.
2422     9. Matching the pattern "\b.*?" against "ab cd", starting at offset 1, failed
2423     to find the match, as PCRE was deluded into thinking that the match had to
2424     start at the start point or following a newline. The same bug applied to
2425     patterns with negative forward assertions or any backward assertions
2426     preceding ".*" at the start, unless the pattern required a fixed first
2427     character. This was a failing pattern: "(?!.bcd).*". The bug is now fixed.
2429     10. In UTF-8 mode, when moving forwards in the subject after a failed match
2430     starting at the last subject character, bytes beyond the end of the subject
2431     string were read.
2433     11. Renamed the variable "class" as "classbits" to make life easier for C++
2434     users. (Previously there was a macro definition, but it apparently wasn't
2435     enough.)
2437     12. Added the new field "tables" to the extra data so that tables can be passed
2438     in at exec time, or the internal tables can be re-selected. This allows
2439     a compiled regex to be saved and re-used at a later time by a different
2440     program that might have everything at different addresses.
2442     13. Modified the pcre-config script so that, when run on Solaris, it shows a
2443     -R library as well as a -L library.
2445     14. The debugging options of pcretest (-d on the command line or D on a
2446     pattern) showed incorrect output for anything following an extended class
2447     that contained multibyte characters and which was followed by a quantifier.
2449     15. Added optional support for general category Unicode character properties
2450     via the \p, \P, and \X escapes. Unicode property support implies UTF-8
2451     support. It adds about 90K to the size of the library. The meanings of the
2452     inbuilt class escapes such as \d and \s have NOT been changed.
2454     16. Updated pcredemo.c to include calls to free() to release the memory for the
2455     compiled pattern.
2457     17. The generated file chartables.c was being created in the source directory
2458     instead of in the building directory. This caused the build to fail if the
2459     source directory was different from the building directory, and was
2460     read-only.
2462     18. Added some sample Win commands from Mark Tetrode into the NON-UNIX-USE
2463     file. No doubt somebody will tell me if they don't make sense... Also added
2464     Dan Mooney's comments about building on OpenVMS.
2466     19. Added support for partial matching via the PCRE_PARTIAL option for
2467     pcre_exec() and the \P data escape in pcretest.
2469     20. Extended pcretest with 3 new pattern features:
2471     (i) A pattern option of the form ">rest-of-line" causes pcretest to
2472     write the compiled pattern to the file whose name is "rest-of-line".
2473     This is a straight binary dump of the data, with the saved pointer to
2474     the character tables forced to be NULL. The study data, if any, is
2475     written too. After writing, pcretest reads a new pattern.
2477     (ii) If, instead of a pattern, "<rest-of-line" is given, pcretest reads a
2478     compiled pattern from the given file. There must not be any
2479     occurrences of "<" in the file name (pretty unlikely); if there are,
2480     pcretest will instead treat the initial "<" as a pattern delimiter.
2481     After reading in the pattern, pcretest goes on to read data lines as
2482     usual.
2484     (iii) The F pattern option causes pcretest to flip the bytes in the 32-bit
2485     and 16-bit fields in a compiled pattern, to simulate a pattern that
2486     was compiled on a host of opposite endianness.
2488     21. The pcre-exec() function can now cope with patterns that were compiled on
2489     hosts of opposite endianness, with this restriction:
2491     As for any compiled expression that is saved and used later, the tables
2492     pointer field cannot be preserved; the extra_data field in the arguments
2493     to pcre_exec() should be used to pass in a tables address if a value
2494     other than the default internal tables were used at compile time.
2496     22. Calling pcre_exec() with a negative value of the "ovecsize" parameter is
2497     now diagnosed as an error. Previously, most of the time, a negative number
2498     would have been treated as zero, but if in addition "ovector" was passed as
2499     NULL, a crash could occur.
2501     23. Updated the files ltmain.sh, config.sub, config.guess, and aclocal.m4 with
2502     new versions from the libtool 1.5 distribution (the last one is a copy of
2503     a file called libtool.m4). This seems to have fixed the need to patch
2504     "configure" to support Darwin 1.3 (which I used to do). However, I still
2505     had to patch ltmain.sh to ensure that ${SED} is set (it isn't on my
2506     workstation).
2508     24. Changed the PCRE licence to be the more standard "BSD" licence.
2511 nigel 73 Version 4.5 01-Dec-03
2512     ---------------------
2514     1. There has been some re-arrangement of the code for the match() function so
2515     that it can be compiled in a version that does not call itself recursively.
2516     Instead, it keeps those local variables that need separate instances for
2517     each "recursion" in a frame on the heap, and gets/frees frames whenever it
2518     needs to "recurse". Keeping track of where control must go is done by means
2519     of setjmp/longjmp. The whole thing is implemented by a set of macros that
2520     hide most of the details from the main code, and operates only if
2521     NO_RECURSE is defined while compiling pcre.c. If PCRE is built using the
2522     "configure" mechanism, "--disable-stack-for-recursion" turns on this way of
2523     operating.
2525     To make it easier for callers to provide specially tailored get/free
2526     functions for this usage, two new functions, pcre_stack_malloc, and
2527     pcre_stack_free, are used. They are always called in strict stacking order,
2528     and the size of block requested is always the same.
2530     The PCRE_CONFIG_STACKRECURSE info parameter can be used to find out whether
2531     PCRE has been compiled to use the stack or the heap for recursion. The
2532     -C option of pcretest uses this to show which version is compiled.
2534     A new data escape \S, is added to pcretest; it causes the amounts of store
2535     obtained and freed by both kinds of malloc/free at match time to be added
2536     to the output.
2538     2. Changed the locale test to use "fr_FR" instead of "fr" because that's
2539     what's available on my current Linux desktop machine.
2541     3. When matching a UTF-8 string, the test for a valid string at the start has
2542     been extended. If start_offset is not zero, PCRE now checks that it points
2543     to a byte that is the start of a UTF-8 character. If not, it returns
2544     PCRE_ERROR_BADUTF8_OFFSET (-11). Note: the whole string is still checked;
2545     this is necessary because there may be backward assertions in the pattern.
2546     When matching the same subject several times, it may save resources to use
2547     PCRE_NO_UTF8_CHECK on all but the first call if the string is long.
2549     4. The code for checking the validity of UTF-8 strings has been tightened so
2550     that it rejects (a) strings containing 0xfe or 0xff bytes and (b) strings
2551     containing "overlong sequences".
2553     5. Fixed a bug (appearing twice) that I could not find any way of exploiting!
2554     I had written "if ((digitab[*p++] && chtab_digit) == 0)" where the "&&"
2555     should have been "&", but it just so happened that all the cases this let
2556     through by mistake were picked up later in the function.
2558     6. I had used a variable called "isblank" - this is a C99 function, causing
2559     some compilers to warn. To avoid this, I renamed it (as "blankclass").
2561     7. Cosmetic: (a) only output another newline at the end of pcretest if it is
2562     prompting; (b) run "./pcretest /dev/null" at the start of the test script
2563     so the version is shown; (c) stop "make test" echoing "./RunTest".
2565     8. Added patches from David Burgess to enable PCRE to run on EBCDIC systems.
2567     9. The prototype for memmove() for systems that don't have it was using
2568     size_t, but the inclusion of the header that defines size_t was later. I've
2569     moved the #includes for the C headers earlier to avoid this.
2571     10. Added some adjustments to the code to make it easier to compiler on certain
2572     special systems:
2574     (a) Some "const" qualifiers were missing.
2575     (b) Added the macro EXPORT before all exported functions; by default this
2576     is defined to be empty.
2577     (c) Changed the dftables auxiliary program (that builds chartables.c) so
2578     that it reads its output file name as an argument instead of writing
2579     to the standard output and assuming this can be redirected.
2581     11. In UTF-8 mode, if a recursive reference (e.g. (?1)) followed a character
2582     class containing characters with values greater than 255, PCRE compilation
2583     went into a loop.
2585     12. A recursive reference to a subpattern that was within another subpattern
2586     that had a minimum quantifier of zero caused PCRE to crash. For example,
2587     (x(y(?2))z)? provoked this bug with a subject that got as far as the
2588     recursion. If the recursively-called subpattern itself had a zero repeat,
2589     that was OK.
2591     13. In pcretest, the buffer for reading a data line was set at 30K, but the
2592     buffer into which it was copied (for escape processing) was still set at
2593     1024, so long lines caused crashes.
2595     14. A pattern such as /[ab]{1,3}+/ failed to compile, giving the error
2596     "internal error: code overflow...". This applied to any character class
2597     that was followed by a possessive quantifier.
2599     15. Modified the Makefile to add libpcre.la as a prerequisite for
2600     libpcreposix.la because I was told this is needed for a parallel build to
2601     work.
2603     16. If a pattern that contained .* following optional items at the start was
2604     studied, the wrong optimizing data was generated, leading to matching
2605     errors. For example, studying /[ab]*.*c/ concluded, erroneously, that any
2606     matching string must start with a or b or c. The correct conclusion for
2607     this pattern is that a match can start with any character.
2610 nigel 71 Version 4.4 13-Aug-03
2611     ---------------------
2613     1. In UTF-8 mode, a character class containing characters with values between
2614     127 and 255 was not handled correctly if the compiled pattern was studied.
2615     In fixing this, I have also improved the studying algorithm for such
2616     classes (slightly).
2618     2. Three internal functions had redundant arguments passed to them. Removal
2619     might give a very teeny performance improvement.
2621     3. Documentation bug: the value of the capture_top field in a callout is *one
2622     more than* the number of the hightest numbered captured substring.
2624     4. The Makefile linked pcretest and pcregrep with -lpcre, which could result
2625     in incorrectly linking with a previously installed version. They now link
2626     explicitly with libpcre.la.
2628     5. configure.in no longer needs to recognize Cygwin specially.
2630     6. A problem in pcre.in for Windows platforms is fixed.
2632     7. If a pattern was successfully studied, and the -d (or /D) flag was given to
2633     pcretest, it used to include the size of the study block as part of its
2634     output. Unfortunately, the structure contains a field that has a different
2635     size on different hardware architectures. This meant that the tests that
2636     showed this size failed. As the block is currently always of a fixed size,
2637     this information isn't actually particularly useful in pcretest output, so
2638     I have just removed it.
2640     8. Three pre-processor statements accidentally did not start in column 1.
2641     Sadly, there are *still* compilers around that complain, even though
2642     standard C has not required this for well over a decade. Sigh.
2644     9. In pcretest, the code for checking callouts passed small integers in the
2645     callout_data field, which is a void * field. However, some picky compilers
2646     complained about the casts involved for this on 64-bit systems. Now
2647     pcretest passes the address of the small integer instead, which should get
2648     rid of the warnings.
2650     10. By default, when in UTF-8 mode, PCRE now checks for valid UTF-8 strings at
2651     both compile and run time, and gives an error if an invalid UTF-8 sequence
2652     is found. There is a option for disabling this check in cases where the
2653     string is known to be correct and/or the maximum performance is wanted.
2655     11. In response to a bug report, I changed one line in Makefile.in from
2657     -Wl,--out-implib,.libs/lib@WIN_PREFIX@pcreposix.dll.a \
2658     to
2659     -Wl,--out-implib,.libs/@WIN_PREFIX@libpcreposix.dll.a \
2661     to look similar to other lines, but I have no way of telling whether this
2662     is the right thing to do, as I do not use Windows. No doubt I'll get told
2663     if it's wrong...
2666 nigel 69 Version 4.3 21-May-03
2667     ---------------------
2669     1. Two instances of @WIN_PREFIX@ omitted from the Windows targets in the
2670     Makefile.
2672     2. Some refactoring to improve the quality of the code:
2674     (i) The utf8_table... variables are now declared "const".
2676     (ii) The code for \cx, which used the "case flipping" table to upper case
2677     lower case letters, now just substracts 32. This is ASCII-specific,
2678     but the whole concept of \cx is ASCII-specific, so it seems
2679     reasonable.
2681     (iii) PCRE was using its character types table to recognize decimal and
2682     hexadecimal digits in the pattern. This is silly, because it handles
2683     only 0-9, a-f, and A-F, but the character types table is locale-
2684     specific, which means strange things might happen. A private
2685     table is now used for this - though it costs 256 bytes, a table is
2686     much faster than multiple explicit tests. Of course, the standard
2687     character types table is still used for matching digits in subject
2688     strings against \d.
2690     (iv) Strictly, the identifier ESC_t is reserved by POSIX (all identifiers
2691     ending in _t are). So I've renamed it as ESC_tee.
2693     3. The first argument for regexec() in the POSIX wrapper should have been
2694     defined as "const".
2696     4. Changed pcretest to use malloc() for its buffers so that they can be
2697     Electric Fenced for debugging.
2699     5. There were several places in the code where, in UTF-8 mode, PCRE would try
2700     to read one or more bytes before the start of the subject string. Often this
2701     had no effect on PCRE's behaviour, but in some circumstances it could
2702     provoke a segmentation fault.
2704     6. A lookbehind at the start of a pattern in UTF-8 mode could also cause PCRE
2705     to try to read one or more bytes before the start of the subject string.
2707     7. A lookbehind in a pattern matched in non-UTF-8 mode on a PCRE compiled with
2708     UTF-8 support could misbehave in various ways if the subject string
2709     contained bytes with the 0x80 bit set and the 0x40 bit unset in a lookbehind
2710     area. (PCRE was not checking for the UTF-8 mode flag, and trying to move
2711     back over UTF-8 characters.)
2714 nigel 67 Version 4.2 14-Apr-03
2715     ---------------------
2717     1. Typo "#if SUPPORT_UTF8" instead of "#ifdef SUPPORT_UTF8" fixed.
2719     2. Changes to the building process, supplied by Ronald Landheer-Cieslak
2720     [ON_WINDOWS]: new variable, "#" on non-Windows platforms
2721     [NOT_ON_WINDOWS]: new variable, "#" on Windows platforms
2722     [WIN_PREFIX]: new variable, "cyg" for Cygwin
2723     * Makefile.in: use autoconf substitution for OBJEXT, EXEEXT, BUILD_OBJEXT
2724     and BUILD_EXEEXT
2725     Note: automatic setting of the BUILD variables is not yet working
2726     set CPPFLAGS and BUILD_CPPFLAGS (but don't use yet) - should be used at
2727     compile-time but not at link-time
2728     [LINK]: use for linking executables only
2729     make different versions for Windows and non-Windows
2730     [LINKLIB]: new variable, copy of UNIX-style LINK, used for linking
2731     libraries
2732     [LINK_FOR_BUILD]: new variable
2733     [OBJEXT]: use throughout
2734     [EXEEXT]: use throughout
2735     <winshared>: new target
2736     <wininstall>: new target
2737     <dftables.o>: use native compiler
2738     <dftables>: use native linker
2739     <install>: handle Windows platform correctly
2740     <clean>: ditto
2741     <check>: ditto
2742     copy DLL to top builddir before testing
2744     As part of these changes, -no-undefined was removed again. This was reported
2745     to give trouble on HP-UX 11.0, so getting rid of it seems like a good idea
2746     in any case.
2748     3. Some tidies to get rid of compiler warnings:
2750     . In the match_data structure, match_limit was an unsigned long int, whereas
2751     match_call_count was an int. I've made them both unsigned long ints.
2753     . In pcretest the fact that a const uschar * doesn't automatically cast to
2754     a void * provoked a warning.
2756     . Turning on some more compiler warnings threw up some "shadow" variables
2757     and a few more missing casts.
2759     4. If PCRE was complied with UTF-8 support, but called without the PCRE_UTF8
2760     option, a class that contained a single character with a value between 128
2761     and 255 (e.g. /[\xFF]/) caused PCRE to crash.
2763     5. If PCRE was compiled with UTF-8 support, but called without the PCRE_UTF8
2764     option, a class that contained several characters, but with at least one
2765     whose value was between 128 and 255 caused PCRE to crash.
2768 nigel 65 Version 4.1 12-Mar-03
2769     ---------------------
2770 nigel 63
2771 nigel 65 1. Compiling with gcc -pedantic found a couple of places where casts were
2772     needed, and a string in dftables.c that was longer than standard compilers are
2773     required to support.
2775     2. Compiling with Sun's compiler found a few more places where the code could
2776     be tidied up in order to avoid warnings.
2778     3. The variables for cross-compiling were called HOST_CC and HOST_CFLAGS; the
2779     first of these names is deprecated in the latest Autoconf in favour of the name
2780     CC_FOR_BUILD, because "host" is typically used to mean the system on which the
2781     compiled code will be run. I can't find a reference for HOST_CFLAGS, but by
2782     analogy I have changed it to CFLAGS_FOR_BUILD.
2784     4. Added -no-undefined to the linking command in the Makefile, because this is
2785     apparently helpful for Windows. To make it work, also added "-L. -lpcre" to the
2786     linking step for the pcreposix library.
2788     5. PCRE was failing to diagnose the case of two named groups with the same
2789     name.
2791     6. A problem with one of PCRE's optimizations was discovered. PCRE remembers a
2792     literal character that is needed in the subject for a match, and scans along to
2793     ensure that it is present before embarking on the full matching process. This
2794     saves time in cases of nested unlimited repeats that are never going to match.
2795     Problem: the scan can take a lot of time if the subject is very long (e.g.
2796     megabytes), thus penalizing straightforward matches. It is now done only if the
2797     amount of subject to be scanned is less than 1000 bytes.
2799     7. A lesser problem with the same optimization is that it was recording the
2800     first character of an anchored pattern as "needed", thus provoking a search
2801     right along the subject, even when the first match of the pattern was going to
2802     fail. The "needed" character is now not set for anchored patterns, unless it
2803     follows something in the pattern that is of non-fixed length. Thus, it still
2804     fulfils its original purpose of finding quick non-matches in cases of nested
2805     unlimited repeats, but isn't used for simple anchored patterns such as /^abc/.
2808     Version 4.0 17-Feb-03
2809     ---------------------
2811 nigel 63 1. If a comment in an extended regex that started immediately after a meta-item
2812     extended to the end of string, PCRE compiled incorrect data. This could lead to
2813     all kinds of weird effects. Example: /#/ was bad; /()#/ was bad; /a#/ was not.
2815     2. Moved to autoconf 2.53 and libtool 1.4.2.
2817     3. Perl 5.8 no longer needs "use utf8" for doing UTF-8 things. Consequently,
2818     the special perltest8 script is no longer needed - all the tests can be run
2819     from a single perltest script.
2821     4. From 5.004, Perl has not included the VT character (0x0b) in the set defined
2822     by \s. It has now been removed in PCRE. This means it isn't recognized as
2823     whitespace in /x regexes too, which is the same as Perl. Note that the POSIX
2824     class [:space:] *does* include VT, thereby creating a mess.
2826     5. Added the class [:blank:] (a GNU extension from Perl 5.8) to match only
2827     space and tab.
2829     6. Perl 5.005 was a long time ago. It's time to amalgamate the tests that use
2830     its new features into the main test script, reducing the number of scripts.
2832     7. Perl 5.8 has changed the meaning of patterns like /a(?i)b/. Earlier versions
2833     were backward compatible, and made the (?i) apply to the whole pattern, as if
2834     /i were given. Now it behaves more logically, and applies the option setting
2835     only to what follows. PCRE has been changed to follow suit. However, if it
2836     finds options settings right at the start of the pattern, it extracts them into
2837     the global options, as before. Thus, they show up in the info data.
2839     8. Added support for the \Q...\E escape sequence. Characters in between are
2840     treated as literals. This is slightly different from Perl in that $ and @ are
2841     also handled as literals inside the quotes. In Perl, they will cause variable
2842     interpolation. Note the following examples:
2844     Pattern PCRE matches Perl matches
2846     \Qabc$xyz\E abc$xyz abc followed by the contents of $xyz
2847     \Qabc\$xyz\E abc\$xyz abc\$xyz
2848     \Qabc\E\$\Qxyz\E abc$xyz abc$xyz
2850     For compatibility with Perl, \Q...\E sequences are recognized inside character
2851     classes as well as outside them.
2853     9. Re-organized 3 code statements in pcretest to avoid "overflow in
2854     floating-point constant arithmetic" warnings from a Microsoft compiler. Added a
2855     (size_t) cast to one statement in pcretest and one in pcreposix to avoid
2856     signed/unsigned warnings.
2858     10. SunOS4 doesn't have strtoul(). This was used only for unpicking the -o
2859     option for pcretest, so I've replaced it by a simple function that does just
2860     that job.
2862     11. pcregrep was ending with code 0 instead of 2 for the commands "pcregrep" or
2863     "pcregrep -".
2865     12. Added "possessive quantifiers" ?+, *+, ++, and {,}+ which come from Sun's
2866     Java package. This provides some syntactic sugar for simple cases of what my
2867     documentation calls "once-only subpatterns". A pattern such as x*+ is the same
2868     as (?>x*). In other words, if what is inside (?>...) is just a single repeated
2869     item, you can use this simplified notation. Note that only makes sense with
2870     greedy quantifiers. Consequently, the use of the possessive quantifier forces
2871     greediness, whatever the setting of the PCRE_UNGREEDY option.
2873     13. A change of greediness default within a pattern was not taking effect at
2874     the current level for patterns like /(b+(?U)a+)/. It did apply to parenthesized
2875     subpatterns that followed. Patterns like /b+(?U)a+/ worked because the option
2876     was abstracted outside.
2878     14. PCRE now supports the \G assertion. It is true when the current matching
2879     position is at the start point of the match. This differs from \A when the
2880     starting offset is non-zero. Used with the /g option of pcretest (or similar
2881     code), it works in the same way as it does for Perl's /g option. If all
2882     alternatives of a regex begin with \G, the expression is anchored to the start
2883     match position, and the "anchored" flag is set in the compiled expression.
2885     15. Some bugs concerning the handling of certain option changes within patterns
2886     have been fixed. These applied to options other than (?ims). For example,
2887     "a(?x: b c )d" did not match "XabcdY" but did match "Xa b c dY". It should have
2888     been the other way round. Some of this was related to change 7 above.
2890     16. PCRE now gives errors for /[.x.]/ and /[=x=]/ as unsupported POSIX
2891     features, as Perl does. Previously, PCRE gave the warnings only for /[[.x.]]/
2892     and /[[=x=]]/. PCRE now also gives an error for /[:name:]/ because it supports
2893     POSIX classes only within a class (e.g. /[[:alpha:]]/).
2895     17. Added support for Perl's \C escape. This matches one byte, even in UTF8
2896     mode. Unlike ".", it always matches newline, whatever the setting of
2897     PCRE_DOTALL. However, PCRE does not permit \C to appear in lookbehind
2898     assertions. Perl allows it, but it doesn't (in general) work because it can't
2899     calculate the length of the lookbehind. At least, that's the case for Perl
2900     5.8.0 - I've been told they are going to document that it doesn't work in
2901     future.
2903     18. Added an error diagnosis for escapes that PCRE does not support: these are
2904     \L, \l, \N, \P, \p, \U, \u, and \X.
2906     19. Although correctly diagnosing a missing ']' in a character class, PCRE was
2907     reading past the end of the pattern in cases such as /[abcd/.
2909     20. PCRE was getting more memory than necessary for patterns with classes that
2910     contained both POSIX named classes and other characters, e.g. /[[:space:]abc/.
2912     21. Added some code, conditional on #ifdef VPCOMPAT, to make life easier for
2913     compiling PCRE for use with Virtual Pascal.
2915     22. Small fix to the Makefile to make it work properly if the build is done
2916     outside the source tree.
2918     23. Added a new extension: a condition to go with recursion. If a conditional
2919     subpattern starts with (?(R) the "true" branch is used if recursion has
2920     happened, whereas the "false" branch is used only at the top level.
2922     24. When there was a very long string of literal characters (over 255 bytes
2923     without UTF support, over 250 bytes with UTF support), the computation of how
2924     much memory was required could be incorrect, leading to segfaults or other
2925     strange effects.
2927     25. PCRE was incorrectly assuming anchoring (either to start of subject or to
2928     start of line for a non-DOTALL pattern) when a pattern started with (.*) and
2929     there was a subsequent back reference to those brackets. This meant that, for
2930     example, /(.*)\d+\1/ failed to match "abc123bc". Unfortunately, it isn't
2931     possible to check for precisely this case. All we can do is abandon the
2932     optimization if .* occurs inside capturing brackets when there are any back
2933     references whatsoever. (See below for a better fix that came later.)
2935     26. The handling of the optimization for finding the first character of a
2936     non-anchored pattern, and for finding a character that is required later in the
2937     match were failing in some cases. This didn't break the matching; it just
2938     failed to optimize when it could. The way this is done has been re-implemented.
2940     27. Fixed typo in error message for invalid (?R item (it said "(?p").
2942     28. Added a new feature that provides some of the functionality that Perl
2943     provides with (?{...}). The facility is termed a "callout". The way it is done
2944     in PCRE is for the caller to provide an optional function, by setting
2945     pcre_callout to its entry point. Like pcre_malloc and pcre_free, this is a
2946     global variable. By default it is unset, which disables all calling out. To get
2947     the function called, the regex must include (?C) at appropriate points. This
2948     is, in fact, equivalent to (?C0), and any number <= 255 may be given with (?C).
2949     This provides a means of identifying different callout points. When PCRE
2950     reaches such a point in the regex, if pcre_callout has been set, the external
2951     function is called. It is provided with data in a structure called
2952     pcre_callout_block, which is defined in pcre.h. If the function returns 0,
2953     matching continues; if it returns a non-zero value, the match at the current
2954     point fails. However, backtracking will occur if possible. [This was changed
2955     later and other features added - see item 49 below.]
2957     29. pcretest is upgraded to test the callout functionality. It provides a
2958     callout function that displays information. By default, it shows the start of
2959     the match and the current position in the text. There are some new data escapes
2960     to vary what happens:
2962     \C+ in addition, show current contents of captured substrings
2963     \C- do not supply a callout function
2964     \C!n return 1 when callout number n is reached
2965     \C!n!m return 1 when callout number n is reached for the mth time
2967     30. If pcregrep was called with the -l option and just a single file name, it
2968     output "<stdin>" if a match was found, instead of the file name.
2970     31. Improve the efficiency of the POSIX API to PCRE. If the number of capturing
2971     slots is less than POSIX_MALLOC_THRESHOLD, use a block on the stack to pass to
2972     pcre_exec(). This saves a malloc/free per call. The default value of
2973     POSIX_MALLOC_THRESHOLD is 10; it can be changed by --with-posix-malloc-threshold
2974     when configuring.
2976     32. The default maximum size of a compiled pattern is 64K. There have been a
2977     few cases of people hitting this limit. The code now uses macros to handle the
2978     storing of links as offsets within the compiled pattern. It defaults to 2-byte
2979     links, but this can be changed to 3 or 4 bytes by --with-link-size when
2980     configuring. Tests 2 and 5 work only with 2-byte links because they output
2981     debugging information about compiled patterns.
2983     33. Internal code re-arrangements:
2985     (a) Moved the debugging function for printing out a compiled regex into
2986     its own source file (printint.c) and used #include to pull it into
2987     pcretest.c and, when DEBUG is defined, into pcre.c, instead of having two
2988     separate copies.
2990     (b) Defined the list of op-code names for debugging as a macro in
2991     internal.h so that it is next to the definition of the opcodes.
2993     (c) Defined a table of op-code lengths for simpler skipping along compiled
2994     code. This is again a macro in internal.h so that it is next to the
2995     definition of the opcodes.
2997     34. Added support for recursive calls to individual subpatterns, along the
2998     lines of Robin Houston's patch (but implemented somewhat differently).
3000     35. Further mods to the Makefile to help Win32. Also, added code to pcregrep to
3001     allow it to read and process whole directories in Win32. This code was
3002     contributed by Lionel Fourquaux; it has not been tested by me.
3004     36. Added support for named subpatterns. The Python syntax (?P<name>...) is
3005     used to name a group. Names consist of alphanumerics and underscores, and must
3006     be unique. Back references use the syntax (?P=name) and recursive calls use
3007     (?P>name) which is a PCRE extension to the Python extension. Groups still have
3008     numbers. The function pcre_fullinfo() can be used after compilation to extract
3009     a name/number map. There are three relevant calls:
3011     PCRE_INFO_NAMEENTRYSIZE yields the size of each entry in the map
3012     PCRE_INFO_NAMECOUNT yields the number of entries
3013     PCRE_INFO_NAMETABLE yields a pointer to the map.
3015     The map is a vector of fixed-size entries. The size of each entry depends on
3016     the length of the longest name used. The first two bytes of each entry are the
3017     group number, most significant byte first. There follows the corresponding
3018     name, zero terminated. The names are in alphabetical order.
3020     37. Make the maximum literal string in the compiled code 250 for the non-UTF-8
3021     case instead of 255. Making it the same both with and without UTF-8 support
3022     means that the same test output works with both.
3024     38. There was a case of malloc(0) in the POSIX testing code in pcretest. Avoid
3025     calling malloc() with a zero argument.
3027     39. Change 25 above had to resort to a heavy-handed test for the .* anchoring
3028     optimization. I've improved things by keeping a bitmap of backreferences with
3029     numbers 1-31 so that if .* occurs inside capturing brackets that are not in
3030     fact referenced, the optimization can be applied. It is unlikely that a
3031     relevant occurrence of .* (i.e. one which might indicate anchoring or forcing
3032     the match to follow \n) will appear inside brackets with a number greater than
3033     31, but if it does, any back reference > 31 suppresses the optimization.
3035     40. Added a new compile-time option PCRE_NO_AUTO_CAPTURE. This has the effect
3036     of disabling numbered capturing parentheses. Any opening parenthesis that is
3037     not followed by ? behaves as if it were followed by ?: but named parentheses
3038     can still be used for capturing (and they will acquire numbers in the usual
3039     way).
3041     41. Redesigned the return codes from the match() function into yes/no/error so
3042     that errors can be passed back from deep inside the nested calls. A malloc
3043     failure while inside a recursive subpattern call now causes the
3044     PCRE_ERROR_NOMEMORY return instead of quietly going wrong.
3046     42. It is now possible to set a limit on the number of times the match()
3047     function is called in a call to pcre_exec(). This facility makes it possible to
3048     limit the amount of recursion and backtracking, though not in a directly
3049     obvious way, because the match() function is used in a number of different
3050     circumstances. The count starts from zero for each position in the subject
3051     string (for non-anchored patterns). The default limit is, for compatibility, a
3052     large number, namely 10 000 000. You can change this in two ways:
3054     (a) When configuring PCRE before making, you can use --with-match-limit=n
3055     to set a default value for the compiled library.
3057     (b) For each call to pcre_exec(), you can pass a pcre_extra block in which
3058     a different value is set. See 45 below.
3060     If the limit is exceeded, pcre_exec() returns PCRE_ERROR_MATCHLIMIT.
3062     43. Added a new function pcre_config(int, void *) to enable run-time extraction
3063     of things that can be changed at compile time. The first argument specifies
3064     what is wanted and the second points to where the information is to be placed.
3065     The current list of available information is:
3069     The output is an integer that is set to one if UTF-8 support is available;
3070     otherwise it is set to zero.
3074     The output is an integer that it set to the value of the code that is used for
3075     newline. It is either LF (10) or CR (13).
3079     The output is an integer that contains the number of bytes used for internal
3080     linkage in compiled expressions. The value is 2, 3, or 4. See item 32 above.
3084     The output is an integer that contains the threshold above which the POSIX
3085     interface uses malloc() for output vectors. See item 31 above.
3089     The output is an unsigned integer that contains the default limit of the number
3090     of match() calls in a pcre_exec() execution. See 42 above.
3092     44. pcretest has been upgraded by the addition of the -C option. This causes it
3093     to extract all the available output from the new pcre_config() function, and to
3094     output it. The program then exits immediately.
3096     45. A need has arisen to pass over additional data with calls to pcre_exec() in
3097     order to support additional features. One way would have been to define
3098     pcre_exec2() (for example) with extra arguments, but this would not have been
3099     extensible, and would also have required all calls to the original function to
3100     be mapped to the new one. Instead, I have chosen to extend the mechanism that
3101     is used for passing in "extra" data from pcre_study().
3103     The pcre_extra structure is now exposed and defined in pcre.h. It currently
3104     contains the following fields:
3106     flags a bitmap indicating which of the following fields are set
3107     study_data opaque data from pcre_study()
3108     match_limit a way of specifying a limit on match() calls for a specific
3109     call to pcre_exec()
3110     callout_data data for callouts (see 49 below)
3112     The flag bits are also defined in pcre.h, and are
3118     The pcre_study() function now returns one of these new pcre_extra blocks, with
3119     the actual study data pointed to by the study_data field, and the
3120     PCRE_EXTRA_STUDY_DATA flag set. This can be passed directly to pcre_exec() as
3121     before. That is, this change is entirely upwards-compatible and requires no
3122     change to existing code.
3124     If you want to pass in additional data to pcre_exec(), you can either place it
3125     in a pcre_extra block provided by pcre_study(), or create your own pcre_extra
3126     block.
3128     46. pcretest has been extended to test the PCRE_EXTRA_MATCH_LIMIT feature. If a
3129     data string contains the escape sequence \M, pcretest calls pcre_exec() several
3130     times with different match limits, until it finds the minimum value needed for
3131     pcre_exec() to complete. The value is then output. This can be instructive; for
3132     most simple matches the number is quite small, but for pathological cases it
3133     gets very large very quickly.
3135     47. There's a new option for pcre_fullinfo() called PCRE_INFO_STUDYSIZE. It
3136     returns the size of the data block pointed to by the study_data field in a
3137     pcre_extra block, that is, the value that was passed as the argument to
3138     pcre_malloc() when PCRE was getting memory in which to place the information
3139     created by pcre_study(). The fourth argument should point to a size_t variable.
3140     pcretest has been extended so that this information is shown after a successful
3141     pcre_study() call when information about the compiled regex is being displayed.
3143     48. Cosmetic change to Makefile: there's no need to have / after $(DESTDIR)
3144     because what follows is always an absolute path. (Later: it turns out that this
3145     is more than cosmetic for MinGW, because it doesn't like empty path
3146     components.)
3148     49. Some changes have been made to the callout feature (see 28 above):
3150     (i) A callout function now has three choices for what it returns:
3152     0 => success, carry on matching
3153     > 0 => failure at this point, but backtrack if possible
3154     < 0 => serious error, return this value from pcre_exec()
3156     Negative values should normally be chosen from the set of PCRE_ERROR_xxx
3157     values. In particular, returning PCRE_ERROR_NOMATCH forces a standard
3158     "match failed" error. The error number PCRE_ERROR_CALLOUT is reserved for
3159     use by callout functions. It will never be used by PCRE itself.
3161     (ii) The pcre_extra structure (see 45 above) has a void * field called
3162     callout_data, with corresponding flag bit PCRE_EXTRA_CALLOUT_DATA. The
3163     pcre_callout_block structure has a field of the same name. The contents of
3164     the field passed in the pcre_extra structure are passed to the callout
3165     function in the corresponding field in the callout block. This makes it
3166     easier to use the same callout-containing regex from multiple threads. For
3167     testing, the pcretest program has a new data escape
3169     \C*n pass the number n (may be negative) as callout_data
3171     If the callout function in pcretest receives a non-zero value as
3172     callout_data, it returns that value.
3174     50. Makefile wasn't handling CFLAGS properly when compiling dftables. Also,
3175     there were some redundant $(CFLAGS) in commands that are now specified as
3176     $(LINK), which already includes $(CFLAGS).
3178     51. Extensions to UTF-8 support are listed below. These all apply when (a) PCRE
3179     has been compiled with UTF-8 support *and* pcre_compile() has been compiled
3180     with the PCRE_UTF8 flag. Patterns that are compiled without that flag assume
3181     one-byte characters throughout. Note that case-insensitive matching applies
3182     only to characters whose values are less than 256. PCRE doesn't support the
3183     notion of cases for higher-valued characters.
3185     (i) A character class whose characters are all within 0-255 is handled as
3186     a bit map, and the map is inverted for negative classes. Previously, a
3187     character > 255 always failed to match such a class; however it should
3188     match if the class was a negative one (e.g. [^ab]). This has been fixed.
3190     (ii) A negated character class with a single character < 255 is coded as
3191     "not this character" (OP_NOT). This wasn't working properly when the test
3192     character was multibyte, either singly or repeated.
3194     (iii) Repeats of multibyte characters are now handled correctly in UTF-8
3195     mode, for example: \x{100}{2,3}.
3197     (iv) The character escapes \b, \B, \d, \D, \s, \S, \w, and \W (either
3198     singly or repeated) now correctly test multibyte characters. However,
3199     PCRE doesn't recognize any characters with values greater than 255 as
3200     digits, spaces, or word characters. Such characters always match \D, \S,
3201     and \W, and never match \d, \s, or \w.
3203     (v) Classes may now contain characters and character ranges with values
3204     greater than 255. For example: [ab\x{100}-\x{400}].
3206     (vi) pcregrep now has a --utf-8 option (synonym -u) which makes it call
3207     PCRE in UTF-8 mode.
3209     52. The info request value PCRE_INFO_FIRSTCHAR has been renamed
3210     PCRE_INFO_FIRSTBYTE because it is a byte value. However, the old name is
3211     retained for backwards compatibility. (Note that LASTLITERAL is also a byte
3212     value.)
3214     53. The single man page has become too large. I have therefore split it up into
3215     a number of separate man pages. These also give rise to individual HTML pages;
3216     these are now put in a separate directory, and there is an index.html page that
3217     lists them all. Some hyperlinking between the pages has been installed.
3219     54. Added convenience functions for handling named capturing parentheses.
3221     55. Unknown escapes inside character classes (e.g. [\M]) and escapes that
3222     aren't interpreted therein (e.g. [\C]) are literals in Perl. This is now also
3223     true in PCRE, except when the PCRE_EXTENDED option is set, in which case they
3224     are faulted.
3226     56. Introduced HOST_CC and HOST_CFLAGS which can be set in the environment when
3227     calling configure. These values are used when compiling the dftables.c program
3228     which is run to generate the source of the default character tables. They
3229     default to the values of CC and CFLAGS. If you are cross-compiling PCRE,
3230     you will need to set these values.
3232     57. Updated the building process for Windows DLL, as provided by Fred Cox.
3235     Version 3.9 02-Jan-02
3236 nigel 61 ---------------------
3238     1. A bit of extraneous text had somehow crept into the pcregrep documentation.
3240     2. If --disable-static was given, the building process failed when trying to
3241     build pcretest and pcregrep. (For some reason it was using libtool to compile
3242     them, which is not right, as they aren't part of the library.)
3245 nigel 59 Version 3.8 18-Dec-01
3246     ---------------------
3248     1. The experimental UTF-8 code was completely screwed up. It was packing the
3249     bytes in the wrong order. How dumb can you get?
3252 nigel 57 Version 3.7 29-Oct-01
3253     ---------------------
3255     1. In updating pcretest to check change 1 of version 3.6, I screwed up.
3256     This caused pcretest, when used on the test data, to segfault. Unfortunately,
3257     this didn't happen under Solaris 8, where I normally test things.
3259 nigel 59 2. The Makefile had to be changed to make it work on BSD systems, where 'make'
3260     doesn't seem to recognize that ./xxx and xxx are the same file. (This entry
3261     isn't in ChangeLog distributed with 3.7 because I forgot when I hastily made
3262     this fix an hour or so after the initial 3.7 release.)
3263 nigel 57
3264 nigel 59
3265 nigel 55 Version 3.6 23-Oct-01
3266     ---------------------
3268     1. Crashed with /(sens|respons)e and \1ibility/ and "sense and sensibility" if
3269     offsets passed as NULL with zero offset count.
3271     2. The config.guess and config.sub files had not been updated when I moved to
3272     the latest autoconf.
3275 nigel 53 Version 3.5 15-Aug-01
3276     ---------------------
3277 nigel 5
3278 nigel 53 1. Added some missing #if !defined NOPOSIX conditionals in pcretest.c that
3279     had been forgotten.
3281     2. By using declared but undefined structures, we can avoid using "void"
3282     definitions in pcre.h while keeping the internal definitions of the structures
3283     private.
3285     3. The distribution is now built using autoconf 2.50 and libtool 1.4. From a
3286     user point of view, this means that both static and shared libraries are built
3287     by default, but this can be individually controlled. More of the work of
3288     handling this static/shared cases is now inside libtool instead of PCRE's make
3289     file.
3291     4. The pcretest utility is now installed along with pcregrep because it is
3292     useful for users (to test regexs) and by doing this, it automatically gets
3293     relinked by libtool. The documentation has been turned into a man page, so
3294     there are now .1, .txt, and .html versions in /doc.
3296     5. Upgrades to pcregrep:
3297     (i) Added long-form option names like gnu grep.
3298     (ii) Added --help to list all options with an explanatory phrase.
3299     (iii) Added -r, --recursive to recurse into sub-directories.
3300     (iv) Added -f, --file to read patterns from a file.
3302     6. pcre_exec() was referring to its "code" argument before testing that
3303     argument for NULL (and giving an error if it was NULL).
3305     7. Upgraded Makefile.in to allow for compiling in a different directory from
3306     the source directory.
3308     8. Tiny buglet in pcretest: when pcre_fullinfo() was called to retrieve the
3309     options bits, the pointer it was passed was to an int instead of to an unsigned
3310     long int. This mattered only on 64-bit systems.
3312     9. Fixed typo (3.4/1) in pcre.h again. Sigh. I had changed pcre.h (which is
3313     generated) instead of pcre.in, which it its source. Also made the same change
3314     in several of the .c files.
3316     10. A new release of gcc defines printf() as a macro, which broke pcretest
3317     because it had an ifdef in the middle of a string argument for printf(). Fixed
3318     by using separate calls to printf().
3320     11. Added --enable-newline-is-cr and --enable-newline-is-lf to the configure
3321     script, to force use of CR or LF instead of \n in the source. On non-Unix
3322     systems, the value can be set in config.h.
3324     12. The limit of 200 on non-capturing parentheses is a _nesting_ limit, not an
3325     absolute limit. Changed the text of the error message to make this clear, and
3326     likewise updated the man page.
3328     13. The limit of 99 on the number of capturing subpatterns has been removed.
3329     The new limit is 65535, which I hope will not be a "real" limit.
3332 nigel 51 Version 3.4 22-Aug-00
3333     ---------------------
3335     1. Fixed typo in pcre.h: unsigned const char * changed to const unsigned char *.
3337     2. Diagnose condition (?(0) as an error instead of crashing on matching.
3340 nigel 49 Version 3.3 01-Aug-00
3341     ---------------------
3343     1. If an octal character was given, but the value was greater than \377, it
3344     was not getting masked to the least significant bits, as documented. This could
3345     lead to crashes in some systems.
3347     2. Perl 5.6 (if not earlier versions) accepts classes like [a-\d] and treats
3348     the hyphen as a literal. PCRE used to give an error; it now behaves like Perl.
3350     3. Added the functions pcre_free_substring() and pcre_free_substring_list().
3351     These just pass their arguments on to (pcre_free)(), but they are provided
3352     because some uses of PCRE bind it to non-C systems that can call its functions,
3353     but cannot call free() or pcre_free() directly.
3355     4. Add "make test" as a synonym for "make check". Corrected some comments in
3356     the Makefile.
3358     5. Add $(DESTDIR)/ in front of all the paths in the "install" target in the
3359     Makefile.
3361     6. Changed the name of pgrep to pcregrep, because Solaris has introduced a
3362     command called pgrep for grepping around the active processes.
3364     7. Added the beginnings of support for UTF-8 character strings.
3366     8. Arranged for the Makefile to pass over the settings of CC, CFLAGS, and
3367     RANLIB to ./ltconfig so that they are used by libtool. I think these are all
3368     the relevant ones. (AR is not passed because ./ltconfig does its own figuring
3369     out for the ar command.)
3372 nigel 47 Version 3.2 12-May-00
3373     ---------------------
3375     This is purely a bug fixing release.
3377     1. If the pattern /((Z)+|A)*/ was matched agained ZABCDEFG it matched Z instead
3378     of ZA. This was just one example of several cases that could provoke this bug,
3379     which was introduced by change 9 of version 2.00. The code for breaking
3380     infinite loops after an iteration that matches an empty string was't working
3381     correctly.
3383     2. The pcretest program was not imitating Perl correctly for the pattern /a*/g
3384     when matched against abbab (for example). After matching an empty string, it
3385     wasn't forcing anchoring when setting PCRE_NOTEMPTY for the next attempt; this
3386     caused it to match further down the string than it should.
3388     3. The code contained an inclusion of sys/types.h. It isn't clear why this
3389     was there because it doesn't seem to be needed, and it causes trouble on some
3390     systems, as it is not a Standard C header. It has been removed.
3392     4. Made 4 silly changes to the source to avoid stupid compiler warnings that
3393     were reported on the Macintosh. The changes were from
3395     while ((c = *(++ptr)) != 0 && c != '\n');
3396     to
3397     while ((c = *(++ptr)) != 0 && c != '\n') ;
3399     Totally extraordinary, but if that's what it takes...
3401     5. PCRE is being used in one environment where neither memmove() nor bcopy() is
3402     available. Added HAVE_BCOPY and an autoconf test for it; if neither
3403     HAVE_MEMMOVE nor HAVE_BCOPY is set, use a built-in emulation function which
3404     assumes the way PCRE uses memmove() (always moving upwards).
3406     6. PCRE is being used in one environment where strchr() is not available. There
3407     was only one use in pcre.c, and writing it out to avoid strchr() probably gives
3408     faster code anyway.
3411 nigel 45 Version 3.1 09-Feb-00
3412     ---------------------
3414     The only change in this release is the fixing of some bugs in Makefile.in for
3415     the "install" target:
3417     (1) It was failing to install pcreposix.h.
3419     (2) It was overwriting the pcre.3 man page with the pcreposix.3 man page.
3422 nigel 43 Version 3.0 01-Feb-00
3423     ---------------------
3424 nigel 41
3425     1. Add support for the /+ modifier to perltest (to output $` like it does in
3426     pcretest).
3428     2. Add support for the /g modifier to perltest.
3430     3. Fix pcretest so that it behaves even more like Perl for /g when the pattern
3431     matches null strings.
3433     4. Fix perltest so that it doesn't do unwanted things when fed an empty
3434     pattern. Perl treats empty patterns specially - it reuses the most recent
3435     pattern, which is not what we want. Replace // by /(?#)/ in order to avoid this
3436     effect.
3438     5. The POSIX interface was broken in that it was just handing over the POSIX
3439     captured string vector to pcre_exec(), but (since release 2.00) PCRE has
3440     required a bigger vector, with some working space on the end. This means that
3441     the POSIX wrapper now has to get and free some memory, and copy the results.
3443 nigel 43 6. Added some simple autoconf support, placing the test data and the
3444     documentation in separate directories, re-organizing some of the
3445     information files, and making it build pcre-config (a GNU standard). Also added
3446     libtool support for building PCRE as a shared library, which is now the
3447     default.
3448 nigel 41
3449 nigel 43 7. Got rid of the leading zero in the definition of PCRE_MINOR because 08 and
3450     09 are not valid octal constants. Single digits will be used for minor values
3451     less than 10.
3453     8. Defined REG_EXTENDED and REG_NOSUB as zero in the POSIX header, so that
3454     existing programs that set these in the POSIX interface can use PCRE without
3455     modification.
3457     9. Added a new function, pcre_fullinfo() with an extensible interface. It can
3458     return all that pcre_info() returns, plus additional data. The pcre_info()
3459     function is retained for compatibility, but is considered to be obsolete.
3461     10. Added experimental recursion feature (?R) to handle one common case that
3462     Perl 5.6 will be able to do with (?p{...}).
3464     11. Added support for POSIX character classes like [:alpha:], which Perl is
3465     adopting.
3468 nigel 39 Version 2.08 31-Aug-99
3469     ----------------------
3471     1. When startoffset was not zero and the pattern began with ".*", PCRE was not
3472     trying to match at the startoffset position, but instead was moving forward to
3473     the next newline as if a previous match had failed.
3475     2. pcretest was not making use of PCRE_NOTEMPTY when repeating for /g and /G,
3476     and could get into a loop if a null string was matched other than at the start
3477     of the subject.
3479     3. Added definitions of PCRE_MAJOR and PCRE_MINOR to pcre.h so the version can
3480     be distinguished at compile time, and for completeness also added PCRE_DATE.
3482     5. Added Paul Sokolovsky's minor changes to make it easy to compile a Win32 DLL
3483     in GnuWin32 environments.
3486 nigel 37 Version 2.07 29-Jul-99
3487     ----------------------
3489     1. The documentation is now supplied in plain text form and HTML as well as in
3490     the form of man page sources.
3492     2. C++ compilers don't like assigning (void *) values to other pointer types.
3493     In particular this affects malloc(). Although there is no problem in Standard
3494     C, I've put in casts to keep C++ compilers happy.
3496     3. Typo on pcretest.c; a cast of (unsigned char *) in the POSIX regexec() call
3497     should be (const char *).
3499     4. If NOPOSIX is defined, pcretest.c compiles without POSIX support. This may
3500     be useful for non-Unix systems who don't want to bother with the POSIX stuff.
3501     However, I haven't made this a standard facility. The documentation doesn't
3502     mention it, and the Makefile doesn't support it.
3504     5. The Makefile now contains an "install" target, with editable destinations at
3505     the top of the file. The pcretest program is not installed.
3507     6. pgrep -V now gives the PCRE version number and date.
3509     7. Fixed bug: a zero repetition after a literal string (e.g. /abcde{0}/) was
3510     causing the entire string to be ignored, instead of just the last character.
3512     8. If a pattern like /"([^\\"]+|\\.)*"/ is applied in the normal way to a
3513     non-matching string, it can take a very, very long time, even for strings of
3514     quite modest length, because of the nested recursion. PCRE now does better in
3515     some of these cases. It does this by remembering the last required literal
3516     character in the pattern, and pre-searching the subject to ensure it is present
3517     before running the real match. In other words, it applies a heuristic to detect
3518     some types of certain failure quickly, and in the above example, if presented
3519     with a string that has no trailing " it gives "no match" very quickly.
3521     9. A new runtime option PCRE_NOTEMPTY causes null string matches to be ignored;
3522     other alternatives are tried instead.
3525 nigel 35 Version 2.06 09-Jun-99
3526     ----------------------
3528     1. Change pcretest's output for amount of store used to show just the code
3529     space, because the remainder (the data block) varies in size between 32-bit and
3530     64-bit systems.
3532     2. Added an extra argument to pcre_exec() to supply an offset in the subject to
3533     start matching at. This allows lookbehinds to work when searching for multiple
3534     occurrences in a string.
3536     3. Added additional options to pcretest for testing multiple occurrences:
3538     /+ outputs the rest of the string that follows a match
3539     /g loops for multiple occurrences, using the new startoffset argument
3540     /G loops for multiple occurrences by passing an incremented pointer
3542     4. PCRE wasn't doing the "first character" optimization for patterns starting
3543     with \b or \B, though it was doing it for other lookbehind assertions. That is,
3544     it wasn't noticing that a match for a pattern such as /\bxyz/ has to start with
3545     the letter 'x'. On long subject strings, this gives a significant speed-up.
3548 nigel 33 Version 2.05 21-Apr-99
3549     ----------------------
3551     1. Changed the type of magic_number from int to long int so that it works
3552     properly on 16-bit systems.
3554     2. Fixed a bug which caused patterns starting with .* not to work correctly
3555     when the subject string contained newline characters. PCRE was assuming
3556     anchoring for such patterns in all cases, which is not correct because .* will
3557     not pass a newline unless PCRE_DOTALL is set. It now assumes anchoring only if
3558     DOTALL is set at top level; otherwise it knows that patterns starting with .*
3559     must be retried after every newline in the subject.
3562 nigel 31 Version 2.04 18-Feb-99
3563     ----------------------
3565     1. For parenthesized subpatterns with repeats whose minimum was zero, the
3566     computation of the store needed to hold the pattern was incorrect (too large).
3567     If such patterns were nested a few deep, this could multiply and become a real
3568     problem.
3570     2. Added /M option to pcretest to show the memory requirement of a specific
3571     pattern. Made -m a synonym of -s (which does this globally) for compatibility.
3573     3. Subpatterns of the form (regex){n,m} (i.e. limited maximum) were being
3574     compiled in such a way that the backtracking after subsequent failure was
3575     pessimal. Something like (a){0,3} was compiled as (a)?(a)?(a)? instead of
3576     ((a)((a)(a)?)?)? with disastrous performance if the maximum was of any size.
3579 nigel 29 Version 2.03 02-Feb-99
3580     ----------------------
3582     1. Fixed typo and small mistake in man page.
3584 nigel 31 2. Added 4th condition (GPL supersedes if conflict) and created separate
3585     LICENCE file containing the conditions.
3586 nigel 29
3587     3. Updated pcretest so that patterns such as /abc\/def/ work like they do in
3588     Perl, that is the internal \ allows the delimiter to be included in the
3589     pattern. Locked out the use of \ as a delimiter. If \ immediately follows
3590     the final delimiter, add \ to the end of the pattern (to test the error).
3592     4. Added the convenience functions for extracting substrings after a successful
3593     match. Updated pcretest to make it able to test these functions.
3596 nigel 27 Version 2.02 14-Jan-99
3597     ----------------------
3599     1. Initialized the working variables associated with each extraction so that
3600     their saving and restoring doesn't refer to uninitialized store.
3602     2. Put dummy code into study.c in order to trick the optimizer of the IBM C
3603     compiler for OS/2 into generating correct code. Apparently IBM isn't going to
3604     fix the problem.
3606     3. Pcretest: the timing code wasn't using LOOPREPEAT for timing execution
3607     calls, and wasn't printing the correct value for compiling calls. Increased the
3608     default value of LOOPREPEAT, and the number of significant figures in the
3609     times.
3611     4. Changed "/bin/rm" in the Makefile to "-rm" so it works on Windows NT.
3613     5. Renamed "deftables" as "dftables" to get it down to 8 characters, to avoid
3614     a building problem on Windows NT with a FAT file system.
3617 nigel 25 Version 2.01 21-Oct-98
3618     ----------------------
3620     1. Changed the API for pcre_compile() to allow for the provision of a pointer
3621     to character tables built by pcre_maketables() in the current locale. If NULL
3622     is passed, the default tables are used.
3625 nigel 23 Version 2.00 24-Sep-98
3626 nigel 21 ----------------------
3628 nigel 23 1. Since the (>?) facility is in Perl 5.005, don't require PCRE_EXTRA to enable
3629     it any more.
3631     2. Allow quantification of (?>) groups, and make it work correctly.
3633     3. The first character computation wasn't working for (?>) groups.
3635     4. Correct the implementation of \Z (it is permitted to match on the \n at the
3636     end of the subject) and add 5.005's \z, which really does match only at the
3637     very end of the subject.
3639     5. Remove the \X "cut" facility; Perl doesn't have it, and (?> is neater.
3641     6. Remove the ability to specify CASELESS, MULTILINE, DOTALL, and
3642     DOLLAR_END_ONLY at runtime, to make it possible to implement the Perl 5.005
3643     localized options. All options to pcre_study() were also removed.
3645     7. Add other new features from 5.005:
3647     $(?<= positive lookbehind
3648     $(?<! negative lookbehind
3649     (?imsx-imsx) added the unsetting capability
3650     such a setting is global if at outer level; local otherwise
3651     (?imsx-imsx:) non-capturing groups with option setting
3652     (?(cond)re|re) conditional pattern matching
3654     A backreference to itself in a repeated group matches the previous
3655     captured string.
3657     8. General tidying up of studying (both automatic and via "study")
3658     consequential on the addition of new assertions.
3660     9. As in 5.005, unlimited repeated groups that could match an empty substring
3661     are no longer faulted at compile time. Instead, the loop is forcibly broken at
3662     runtime if any iteration does actually match an empty substring.
3664     10. Include the RunTest script in the distribution.
3666     11. Added tests from the Perl 5.005_02 distribution. This showed up a few
3667     discrepancies, some of which were old and were also with respect to 5.004. They
3668     have now been fixed.
3671     Version 1.09 28-Apr-98
3672     ----------------------
3674 nigel 21 1. A negated single character class followed by a quantifier with a minimum
3675     value of one (e.g. [^x]{1,6} ) was not compiled correctly. This could lead to
3676     program crashes, or just wrong answers. This did not apply to negated classes
3677     containing more than one character, or to minima other than one.
3680 nigel 19 Version 1.08 27-Mar-98
3681     ----------------------
3683     1. Add PCRE_UNGREEDY to invert the greediness of quantifiers.
3685     2. Add (?U) and (?X) to set PCRE_UNGREEDY and PCRE_EXTRA respectively. The
3686     latter must appear before anything that relies on it in the pattern.
3689 nigel 17 Version 1.07 16-Feb-98
3690     ----------------------
3692     1. A pattern such as /((a)*)*/ was not being diagnosed as in error (unlimited
3693     repeat of a potentially empty string).
3696 nigel 15 Version 1.06 23-Jan-98
3697     ----------------------
3699     1. Added Markus Oberhumer's little patches for C++.
3701     2. Literal strings longer than 255 characters were broken.
3704 nigel 13 Version 1.05 23-Dec-97
3705     ----------------------
3707     1. Negated character classes containing more than one character were failing if
3708     PCRE_CASELESS was set at run time.
3711 nigel 11 Version 1.04 19-Dec-97
3712     ----------------------
3714     1. Corrected the man page, where some "const" qualifiers had been omitted.
3716     2. Made debugging output print "{0,xxx}" instead of just "{,xxx}" to agree with
3717     input syntax.
3719     3. Fixed memory leak which occurred when a regex with back references was
3720     matched with an offsets vector that wasn't big enough. The temporary memory
3721     that is used in this case wasn't being freed if the match failed.
3723     4. Tidied pcretest to ensure it frees memory that it gets.
3725     5. Temporary memory was being obtained in the case where the passed offsets
3726     vector was exactly big enough.
3728     6. Corrected definition of offsetof() from change 5 below.
3730     7. I had screwed up change 6 below and broken the rules for the use of
3731     setjmp(). Now fixed.
3734 nigel 9 Version 1.03 18-Dec-97
3735     ----------------------
3737     1. A erroneous regex with a missing opening parenthesis was correctly
3738     diagnosed, but PCRE attempted to access brastack[-1], which could cause crashes
3739     on some systems.
3741     2. Replaced offsetof(real_pcre, code) by offsetof(real_pcre, code[0]) because
3742     it was reported that one broken compiler failed on the former because "code" is
3743     also an independent variable.
3745     3. The erroneous regex a[]b caused an array overrun reference.
3747     4. A regex ending with a one-character negative class (e.g. /[^k]$/) did not
3748     fail on data ending with that character. (It was going on too far, and checking
3749     the next character, typically a binary zero.) This was specific to the
3750     optimized code for single-character negative classes.
3752     5. Added a contributed patch from the TIN world which does the following:
3754     + Add an undef for memmove, in case the the system defines a macro for it.
3756     + Add a definition of offsetof(), in case there isn't one. (I don't know
3757     the reason behind this - offsetof() is part of the ANSI standard - but
3758     it does no harm).
3760     + Reduce the ifdef's in pcre.c using macro DPRINTF, thereby eliminating
3761     most of the places where whitespace preceded '#'. I have given up and
3762     allowed the remaining 2 cases to be at the margin.
3764     + Rename some variables in pcre to eliminate shadowing. This seems very
3765     pedantic, but does no harm, of course.
3767     6. Moved the call to setjmp() into its own function, to get rid of warnings
3768     from gcc -Wall, and avoided calling it at all unless PCRE_EXTRA is used.
3770     7. Constructs such as \d{8,} were compiling into the equivalent of
3771 nigel 11 \d{8}\d{0,65527} instead of \d{8}\d* which didn't make much difference to the
3772 nigel 9 outcome, but in this particular case used more store than had been allocated,
3773     which caused the bug to be discovered because it threw up an internal error.
3775     8. The debugging code in both pcre and pcretest for outputting the compiled
3776     form of a regex was going wrong in the case of back references followed by
3777     curly-bracketed repeats.
3780 nigel 7 Version 1.02 12-Dec-97
3781     ----------------------
3783     1. Typos in pcre.3 and comments in the source fixed.
3785     2. Applied a contributed patch to get rid of places where it used to remove
3786     'const' from variables, and fixed some signed/unsigned and uninitialized
3787     variable warnings.
3789     3. Added the "runtest" target to Makefile.
3791     4. Set default compiler flag to -O2 rather than just -O.
3794 nigel 5 Version 1.01 19-Nov-97
3795     ----------------------
3797     1. PCRE was failing to diagnose unlimited repeat of empty string for patterns
3798     like /([ab]*)*/, that is, for classes with more than one character in them.
3800     2. Likewise, it wasn't diagnosing patterns with "once-only" subpatterns, such
3801     as /((?>a*))*/ (a PCRE_EXTRA facility).
3804     Version 1.00 18-Nov-97
3805     ----------------------
3807     1. Added compile-time macros to support systems such as SunOS4 which don't have
3808     memmove() or strerror() but have other things that can be used instead.
3810     2. Arranged that "make clean" removes the executables.