ViewVC logotype

Contents of /code/trunk/ChangeLog

Parent Directory Parent Directory | Revision Log Revision Log

Revision 500 - (hide annotations) (download)
Sat Mar 6 19:00:29 2010 UTC (5 years, 1 month ago) by ph10
File size: 183722 byte(s)
Fix bugs with \K in atomic groups, subroutines, and assertions.

1 nigel 3 ChangeLog for PCRE
2     ------------------
4 ph10 491 Version 8.02 01-Mar-2010
5     ------------------------
7     1. The Unicode data tables have been updated to Unicode 5.2.0.
9 ph10 492 2. Added the option --libs-cpp to pcre-config, but only when C++ support is
10     configured.
11 ph10 493
12     3. Updated the licensing terms in the pcregexp.pas file, as agreed with the
13     original author of that file, following a query about its status.
14 ph10 494
15     4. On systems that do not have stdint.h (e.g. Solaris), check for and include
16     inttypes.h instead. This fixes a bug that was introduced by change 8.01/8.
17 ph10 496
18     5. A pattern such as (?&t)*+(?(DEFINE)(?<t>.)) which has a possessive
19     quantifier applied to a forward-referencing subroutine call, could compile
20     incorrect code or give the error "internal error: previously-checked
21     referenced subpattern not found".
22 ph10 497
23     6. Both MS Visual Studio and Symbian OS have problems with initializing
24     variables to point to external functions. For these systems, therefore,
25     pcre_malloc etc. are now initialized to local functions that call the
26     relevant global functions.
27 ph10 498
28     7. There were two entries missing in the vectors called coptable and poptable
29     in pcre_dfa_exec.c. This could lead to memory accesses outsize the vectors.
30     I've fixed the data, and added a kludgy way of testing at compile time that
31     the lengths are correct (equal to the number of opcodes).
32 ph10 499
33     8. Following on from 7, I added a similar kludge to check the length of the
34     eint vector in pcreposix.c.
36     9. Error texts for pcre_compile() are held as one long string to avoid too
37     much relocation at load time. To find a text, the string is searched,
38     counting zeros. There was no check for running off the end of the string,
39     which could happen if a new error number was added without updating the
40     string.
41 ph10 500
42     10. \K gave a compile-time error if it appeared in a lookbehind assersion.
44     11. \K was not working if it appeared in an atomic group or in a group that
45     was called as a "subroutine", or in an assertion. Perl 5.11 documents that
46     \K is "not well defined" if used in an assertion. PCRE now accepts it if
47     the assertion is positive, but not if it is negative.
48 ph10 491
49 ph10 492
50 ph10 489 Version 8.01 19-Jan-2010
51     ------------------------
52 ph10 471
53     1. If a pattern contained a conditional subpattern with only one branch (in
54 ph10 489 particular, this includes all (*DEFINE) patterns), a call to pcre_study()
55 ph10 473 computed the wrong minimum data length (which is of course zero for such
56 ph10 489 subpatterns). This could cause incorrect "no match" results.
57 ph10 473
58     2. For patterns such as (?i)a(?-i)b|c where an option setting at the start of
59 ph10 472 the pattern is reset in the first branch, pcre_compile() failed with
60     "internal error: code overflow at offset...". This happened only when
61 ph10 473 the reset was to the original external option setting. (An optimization
62     abstracts leading options settings into an external setting, which was the
63 ph10 472 cause of this.)
64 ph10 471
65 ph10 473 3. A pattern such as ^(?!a(*SKIP)b) where a negative assertion contained one
66     of the verbs SKIP, PRUNE, or COMMIT, did not work correctly. When the
67     assertion pattern did not match (meaning that the assertion was true), it
68     was incorrectly treated as false if the SKIP had been reached during the
69     matching. This also applied to assertions used as conditions.
70 ph10 471
71 ph10 473 4. If an item that is not supported by pcre_dfa_exec() was encountered in an
72     assertion subpattern, including such a pattern used as a condition,
73     unpredictable results occurred, instead of the error return
75 ph10 478
76 ph10 474 5. The C++ GlobalReplace function was not working like Perl for the special
77     situation when an empty string is matched. It now does the fancy magic
78 ph10 478 stuff that is necessary.
80     6. In pcre_internal.h, obsolete includes to setjmp.h and stdarg.h have been
81 ph10 475 removed. (These were left over from very, very early versions of PCRE.)
82 ph10 478
83 ph10 475 7. Some cosmetic changes to the code to make life easier when compiling it
84     as part of something else:
85 ph10 478
86     (a) Change DEBUG to PCRE_DEBUG.
88     (b) In pcre_compile(), rename the member of the "branch_chain" structure
89     called "current" as "current_branch", to prevent a collision with the
90 ph10 475 Linux macro when compiled as a kernel module.
91 ph10 478
92     (c) In pcre_study(), rename the function set_bit() as set_table_bit(), to
93     prevent a collision with the Linux macro when compiled as a kernel
94 ph10 475 module.
95 ph10 478
96 ph10 475 8. In pcre_compile() there are some checks for integer overflows that used to
97     cast potentially large values to (double). This has been changed to that
98 ph10 478 when building, a check for int64_t is made, and if it is found, it is used
99     instead, thus avoiding the use of floating point arithmetic. (There is no
100     other use of FP in PCRE.) If int64_t is not found, the fallback is to
101     double.
103     9. Added two casts to avoid signed/unsigned warnings from VS Studio Express
104 ph10 476 2005 (difference between two addresses compared to an unsigned value).
105 ph10 473
106 ph10 478 10. Change the standard AC_CHECK_LIB test for libbz2 in configure.ac to a
107     custom one, because of the following reported problem in Windows:
108 ph10 473
109 ph10 478 - libbz2 uses the Pascal calling convention (WINAPI) for the functions
110 ph10 489 under Win32.
111 ph10 478 - The standard autoconf AC_CHECK_LIB fails to include "bzlib.h",
112 ph10 489 therefore missing the function definition.
113 ph10 478 - The compiler thus generates a "C" signature for the test function.
114     - The linker fails to find the "C" function.
115     - PCRE fails to configure if asked to do so against libbz2.
116 ph10 486
117 ph10 479 11. When running libtoolize from libtool-2.2.6b as part of autogen.sh, these
118     messages were output:
119 ph10 478
120 ph10 479 Consider adding `AC_CONFIG_MACRO_DIR([m4])' to configure.ac and
121     rerunning libtoolize, to keep the correct libtool macros in-tree.
122     Consider adding `-I m4' to ACLOCAL_AMFLAGS in Makefile.am.
123 ph10 486
124 ph10 479 I have done both of these things.
125 ph10 486
126 ph10 480 12. Although pcre_dfa_exec() does not use nearly as much stack as pcre_exec()
127 ph10 486 most of the time, it *can* run out if it is given a pattern that contains a
128     runaway infinite recursion. I updated the discussion in the pcrestack man
129     page.
131     13. Now that we have gone to the x.xx style of version numbers, the minor
132     version may start with zero. Using 08 or 09 is a bad idea because users
133     might check the value of PCRE_MINOR in their code, and 08 or 09 may be
134     interpreted as invalid octal numbers. I've updated the previous comment in
135     configure.ac, and also added a check that gives an error if 08 or 09 are
136 ph10 481 used.
137 ph10 486
138 ph10 482 14. Change 8.00/11 was not quite complete: code had been accidentally omitted,
139 ph10 483 causing partial matching to fail when the end of the subject matched \W
140     in a UTF-8 pattern where \W was quantified with a minimum of 3.
141 ph10 486
142 ph10 483 15. There were some discrepancies between the declarations in pcre_internal.h
143 ph10 486 of _pcre_is_newline(), _pcre_was_newline(), and _pcre_valid_utf8() and
144 ph10 483 their definitions. The declarations used "const uschar *" and the
145     definitions used USPTR. Even though USPTR is normally defined as "const
146     unsigned char *" (and uschar is typedeffed as "unsigned char"), it was
147     reported that: "This difference in casting confuses some C++ compilers, for
148     example, SunCC recognizes above declarations as different functions and
149 ph10 486 generates broken code for hbpcre." I have changed the declarations to use
150 ph10 483 USPTR.
151 ph10 486
152     16. GNU libtool is named differently on some systems. The autogen.sh script now
153 ph10 484 tries several variants such as glibtoolize (MacOSX) and libtoolize1x
154     (FreeBSD).
155 ph10 478
156 ph10 486 17. Applied Craig's patch that fixes an HP aCC compile error in pcre 8.00
157     (strtoXX undefined when compiling pcrecpp.cc). The patch contains this
158     comment: "Figure out how to create a longlong from a string: strtoll and
159     equivalent. It's not enough to call AC_CHECK_FUNCS: hpux has a strtoll, for
160     instance, but it only takes 2 args instead of 3!"
161 ph10 479
162 ph10 488 18. A subtle bug concerned with back references has been fixed by a change of
163     specification, with a corresponding code fix. A pattern such as
164     ^(xa|=?\1a)+$ which contains a back reference inside the group to which it
165     refers, was giving matches when it shouldn't. For example, xa=xaaa would
166     match that pattern. Interestingly, Perl (at least up to 5.11.3) has the
167     same bug. Such groups have to be quantified to be useful, or contained
168     inside another quantified group. (If there's no repetition, the reference
169     can never match.) The problem arises because, having left the group and
170     moved on to the rest of the pattern, a later failure that backtracks into
171     the group uses the captured value from the final iteration of the group
172     rather than the correct earlier one. I have fixed this in PCRE by forcing
173     any group that contains a reference to itself to be an atomic group; that
174     is, there cannot be any backtracking into it once it has completed. This is
175     similar to recursive and subroutine calls.
176 ph10 486
177 ph10 488
178 ph10 469 Version 8.00 19-Oct-09
179 ph10 418 ----------------------
181     1. The table for translating pcre_compile() error codes into POSIX error codes
182 ph10 461 was out-of-date, and there was no check on the pcre_compile() error code
183     being within the table. This could lead to an OK return being given in
184 ph10 418 error.
185 ph10 461
186     2. Changed the call to open a subject file in pcregrep from fopen(pathname,
187     "r") to fopen(pathname, "rb"), which fixed a problem with some of the tests
188     in a Windows environment.
190 ph10 420 3. The pcregrep --count option prints the count for each file even when it is
191     zero, as does GNU grep. However, pcregrep was also printing all files when
192     --files-with-matches was added. Now, when both options are given, it prints
193     counts only for those files that have at least one match. (GNU grep just
194 ph10 461 prints the file name in this circumstance, but including the count seems
195     more useful - otherwise, why use --count?) Also ensured that the
196 ph10 420 combination -clh just lists non-zero counts, with no names.
197 ph10 461
198     4. The long form of the pcregrep -F option was incorrectly implemented as
199     --fixed_strings instead of --fixed-strings. This is an incompatible change,
200     but it seems right to fix it, and I didn't think it was worth preserving
201     the old behaviour.
203     5. The command line items --regex=pattern and --regexp=pattern were not
204 ph10 422 recognized by pcregrep, which required --regex pattern or --regexp pattern
205 ph10 461 (with a space rather than an '='). The man page documented the '=' forms,
206 ph10 422 which are compatible with GNU grep; these now work.
207 ph10 461
208     6. No libpcreposix.pc file was created for pkg-config; there was just
209 ph10 423 libpcre.pc and libpcrecpp.pc. The omission has been rectified.
210 ph10 461
211 ph10 425 7. Added #ifndef SUPPORT_UCP into the pcre_ucd.c module, to reduce its size
212 ph10 461 when UCP support is not needed, by modifying the Python script that
213 ph10 425 generates it from Unicode data files. This should not matter if the module
214     is correctly used as a library, but I received one complaint about 50K of
215     unwanted data. My guess is that the person linked everything into his
216     program rather than using a library. Anyway, it does no harm.
217 ph10 461
218 ph10 426 8. A pattern such as /\x{123}{2,2}+/8 was incorrectly compiled; the trigger
219 ph10 461 was a minimum greater than 1 for a wide character in a possessive
220 ph10 431 repetition. The same bug could also affect patterns like /(\x{ff}{0,2})*/8
221     which had an unlimited repeat of a nested, fixed maximum repeat of a wide
222     character. Chaos in the form of incorrect output or a compiling loop could
223     result.
224 ph10 461
225 ph10 426 9. The restrictions on what a pattern can contain when partial matching is
226 ph10 461 requested for pcre_exec() have been removed. All patterns can now be
227 ph10 426 partially matched by this function. In addition, if there are at least two
228 ph10 435 slots in the offset vector, the offset of the earliest inspected character
229     for the match and the offset of the end of the subject are set in them when
230 ph10 461 PCRE_ERROR_PARTIAL is returned.
232 ph10 428 10. Partial matching has been split into two forms: PCRE_PARTIAL_SOFT, which is
233     synonymous with PCRE_PARTIAL, for backwards compatibility, and
234     PCRE_PARTIAL_HARD, which causes a partial match to supersede a full match,
235 ph10 462 and may be more useful for multi-segment matching.
236 ph10 461
237     11. Partial matching with pcre_exec() is now more intuitive. A partial match
238     used to be given if ever the end of the subject was reached; now it is
239     given only if matching could not proceed because another character was
240     needed. This makes a difference in some odd cases such as Z(*FAIL) with the
241     string "Z", which now yields "no match" instead of "partial match". In the
242     case of pcre_dfa_exec(), "no match" is given if every matching path for the
243     final character ended with (*FAIL).
245 ph10 428 12. Restarting a match using pcre_dfa_exec() after a partial match did not work
246 ph10 461 if the pattern had a "must contain" character that was already found in the
247 ph10 428 earlier partial match, unless partial matching was again requested. For
248     example, with the pattern /dog.(body)?/, the "must contain" character is
249     "g". If the first part-match was for the string "dog", restarting with
250 ph10 435 "sbody" failed. This bug has been fixed.
251 ph10 461
252     13. The string returned by pcre_dfa_exec() after a partial match has been
253     changed so that it starts at the first inspected character rather than the
254     first character of the match. This makes a difference only if the pattern
255     starts with a lookbehind assertion or \b or \B (\K is not supported by
256     pcre_dfa_exec()). It's an incompatible change, but it makes the two
257 ph10 435 matching functions compatible, and I think it's the right thing to do.
258 ph10 461
259 ph10 435 14. Added a pcredemo man page, created automatically from the pcredemo.c file,
260 ph10 461 so that the demonstration program is easily available in environments where
261     PCRE has not been installed from source.
263 ph10 435 15. Arranged to add -DPCRE_STATIC to cflags in libpcre.pc, libpcreposix.cp,
264 ph10 430 libpcrecpp.pc and pcre-config when PCRE is not compiled as a shared
265     library.
266 ph10 461
267 ph10 435 16. Added REG_UNGREEDY to the pcreposix interface, at the request of a user.
268 ph10 432 It maps to PCRE_UNGREEDY. It is not, of course, POSIX-compatible, but it
269 ph10 461 is not the first non-POSIX option to be added. Clearly some people find
270 ph10 433 these options useful.
271 ph10 461
272     17. If a caller to the POSIX matching function regexec() passes a non-zero
273 ph10 438 value for nmatch with a NULL value for pmatch, the value of
274 ph10 461 nmatch is forced to zero.
276 ph10 437 18. RunGrepTest did not have a test for the availability of the -u option of
277 ph10 461 the diff command, as RunTest does. It now checks in the same way as
278 ph10 437 RunTest, and also checks for the -b option.
279 ph10 461
280 ph10 438 19. If an odd number of negated classes containing just a single character
281     interposed, within parentheses, between a forward reference to a named
282 ph10 461 subpattern and the definition of the subpattern, compilation crashed with
283     an internal error, complaining that it could not find the referenced
284 ph10 438 subpattern. An example of a crashing pattern is /(?&A)(([^m])(?<A>))/.
285 ph10 461 [The bug was that it was starting one character too far in when skipping
286     over the character class, thus treating the ] as data rather than
287     terminating the class. This meant it could skip too much.]
289 ph10 442 20. Added PCRE_NOTEMPTY_ATSTART in order to be able to correctly implement the
290 ph10 461 /g option in pcretest when the pattern contains \K, which makes it possible
291 ph10 442 to have an empty string match not at the start, even when the pattern is
292 ph10 461 anchored. Updated pcretest and pcredemo to use this option.
294 ph10 446 21. If the maximum number of capturing subpatterns in a recursion was greater
295 ph10 461 than the maximum at the outer level, the higher number was returned, but
296     with unset values at the outer level. The correct (outer level) value is
297 ph10 446 now given.
298 ph10 461
299 ph10 447 22. If (*ACCEPT) appeared inside capturing parentheses, previous releases of
300     PCRE did not set those parentheses (unlike Perl). I have now found a way to
301     make it do so. The string so far is captured, making this feature
302     compatible with Perl.
303 ph10 461
304     23. The tests have been re-organized, adding tests 11 and 12, to make it
305 ph10 448 possible to check the Perl 5.10 features against Perl 5.10.
306 ph10 461
307 ph10 454 24. Perl 5.10 allows subroutine calls in lookbehinds, as long as the subroutine
308 ph10 461 pattern matches a fixed length string. PCRE did not allow this; now it
309     does. Neither allows recursion.
311     25. I finally figured out how to implement a request to provide the minimum
312     length of subject string that was needed in order to match a given pattern.
313     (It was back references and recursion that I had previously got hung up
314     on.) This code has now been added to pcre_study(); it finds a lower bound
315 ph10 455 to the length of subject needed. It is not necessarily the greatest lower
316     bound, but using it to avoid searching strings that are too short does give
317     some useful speed-ups. The value is available to calling programs via
318     pcre_fullinfo().
319 ph10 461
320 ph10 455 26. While implementing 25, I discovered to my embarrassment that pcretest had
321     not been passing the result of pcre_study() to pcre_dfa_exec(), so the
322     study optimizations had never been tested with that matching function.
323     Oops. What is worse, even when it was passed study data, there was a bug in
324     pcre_dfa_exec() that meant it never actually used it. Double oops. There
325     were also very few tests of studied patterns with pcre_dfa_exec().
326 ph10 461
327 ph10 457 27. If (?| is used to create subpatterns with duplicate numbers, they are now
328     allowed to have the same name, even if PCRE_DUPNAMES is not set. However,
329     on the other side of the coin, they are no longer allowed to have different
330     names, because these cannot be distinguished in PCRE, and this has caused
331     confusion. (This is a difference from Perl.)
332 ph10 461
333     28. When duplicate subpattern names are present (necessarily with different
334     numbers, as required by 27 above), and a test is made by name in a
335     conditional pattern, either for a subpattern having been matched, or for
336     recursion in such a pattern, all the associated numbered subpatterns are
337 ph10 459 tested, and the overall condition is true if the condition is true for any
338     one of them. This is the way Perl works, and is also more like the way
339     testing by number works.
340 ph10 418
341 ph10 461
342 ph10 412 Version 7.9 11-Apr-09
343 ph10 376 ---------------------
345 ph10 392 1. When building with support for bzlib/zlib (pcregrep) and/or readline
346 ph10 376 (pcretest), all targets were linked against these libraries. This included
347     libpcre, libpcreposix, and libpcrecpp, even though they do not use these
348     libraries. This caused unwanted dependencies to be created. This problem
349 ph10 392 has been fixed, and now only pcregrep is linked with bzlib/zlib and only
350 ph10 376 pcretest is linked with readline.
351 ph10 392
352 ph10 376 2. The "typedef int BOOL" in pcre_internal.h that was included inside the
353     "#ifndef FALSE" condition by an earlier change (probably 7.8/18) has been
354     moved outside it again, because FALSE and TRUE are already defined in AIX,
355     but BOOL is not.
356 ph10 392
357     3. The pcre_config() function was treating the PCRE_MATCH_LIMIT and
358 ph10 409 PCRE_MATCH_LIMIT_RECURSION values as ints, when they should be long ints.
359 ph10 392
360 ph10 376 4. The pcregrep documentation said spaces were inserted as well as colons (or
361     hyphens) following file names and line numbers when outputting matching
362 ph10 392 lines. This is not true; no spaces are inserted. I have also clarified the
363 ph10 376 wording for the --colour (or --color) option.
364 ph10 392
365 ph10 378 5. In pcregrep, when --colour was used with -o, the list of matching strings
366     was not coloured; this is different to GNU grep, so I have changed it to be
367     the same.
368 ph10 392
369 ph10 378 6. When --colo(u)r was used in pcregrep, only the first matching substring in
370 ph10 392 each matching line was coloured. Now it goes on to look for further matches
371     of any of the test patterns, which is the same behaviour as GNU grep.
373     7. A pattern that could match an empty string could cause pcregrep to loop; it
374     doesn't make sense to accept an empty string match in pcregrep, so I have
375 ph10 379 locked it out (using PCRE's PCRE_NOTEMPTY option). By experiment, this
376     seems to be how GNU grep behaves.
377 ph10 392
378 ph10 380 8. The pattern (?(?=.*b)b|^) was incorrectly compiled as "match must be at
379 ph10 392 start or after a newline", because the conditional assertion was not being
380     correctly handled. The rule now is that both the assertion and what follows
381     in the first alternative must satisfy the test.
383 ph10 405 9. If auto-callout was enabled in a pattern with a conditional group whose
384 ph10 399 condition was an assertion, PCRE could crash during matching, both with
385     pcre_exec() and pcre_dfa_exec().
386 ph10 392
387     10. The PCRE_DOLLAR_ENDONLY option was not working when pcre_dfa_exec() was
388     used for matching.
390     11. Unicode property support in character classes was not working for
391 ph10 385 characters (bytes) greater than 127 when not in UTF-8 mode.
392 ph10 388
393 ph10 392 12. Added the -M command line option to pcretest.
395 ph10 388 14. Added the non-standard REG_NOTEMPTY option to the POSIX interface.
396 ph10 389
397     15. Added the PCRE_NO_START_OPTIMIZE match-time option.
398 ph10 390
399 ph10 392 16. Added comments and documentation about mis-use of no_arg in the C++
400 ph10 390 wrapper.
401 ph10 392
402 ph10 391 17. Implemented support for UTF-8 encoding in EBCDIC environments, a patch
403 ph10 392 from Martin Jerabek that uses macro names for all relevant character and
404 ph10 391 string constants.
405 ph10 405
406 ph10 393 18. Added to pcre_internal.h two configuration checks: (a) If both EBCDIC and
407     SUPPORT_UTF8 are set, give an error; (b) If SUPPORT_UCP is set without
408 ph10 405 SUPPORT_UTF8, define SUPPORT_UTF8. The "configure" script handles both of
409     these, but not everybody uses configure.
411     19. A conditional group that had only one branch was not being correctly
412     recognized as an item that could match an empty string. This meant that an
413     enclosing group might also not be so recognized, causing infinite looping
414     (and probably a segfault) for patterns such as ^"((?(?=[a])[^"])|b)*"$
415     with the subject "ab", where knowledge that the repeated group can match
416 ph10 396 nothing is needed in order to break the loop.
417 ph10 405
418 ph10 397 20. If a pattern that was compiled with callouts was matched using pcre_dfa_
419 ph10 405 exec(), but without supplying a callout function, matching went wrong.
421     21. If PCRE_ERROR_MATCHLIMIT occurred during a recursion, there was a memory
422     leak if the size of the offset vector was greater than 30. When the vector
423     is smaller, the saved offsets during recursion go onto a local stack
424     vector, but for larger vectors malloc() is used. It was failing to free
425     when the recursion yielded PCRE_ERROR_MATCH_LIMIT (or any other "abnormal"
426 ph10 400 error, in fact).
427 ph10 405
428 ph10 402 22. There was a missing #ifdef SUPPORT_UTF8 round one of the variables in the
429 ph10 405 heapframe that is used only when UTF-8 support is enabled. This caused no
430     problem, but was untidy.
431 ph10 376
432 ph10 405 23. Steven Van Ingelgem's patch to CMakeLists.txt to change the name
433     CMAKE_BINARY_DIR to PROJECT_BINARY_DIR so that it works when PCRE is
434     included within another project.
435 ph10 411
436 ph10 407 24. Steven Van Ingelgem's patches to add more options to the CMake support,
437     slightly modified by me:
438 ph10 411
439     (a) PCRE_BUILD_TESTS can be set OFF not to build the tests, including
440     not building pcregrep.
442 ph10 407 (b) PCRE_BUILD_PCREGREP can be see OFF not to build pcregrep, but only
443 ph10 411 if PCRE_BUILD_TESTS is also set OFF, because the tests use pcregrep.
445 ph10 408 25. Forward references, both numeric and by name, in patterns that made use of
446     duplicate group numbers, could behave incorrectly or give incorrect errors,
447     because when scanning forward to find the reference group, PCRE was not
448     taking into account the duplicate group numbers. A pattern such as
449     ^X(?3)(a)(?|(b)|(q))(Y) is an example.
450 ph10 411
451     26. Changed a few more instances of "const unsigned char *" to USPTR, making
452     the feature of a custom pointer more persuasive (as requested by a user).
453 ph10 416
454 ph10 411 27. Wrapped the definitions of fileno and isatty for Windows, which appear in
455 ph10 416 pcretest.c, inside #ifndefs, because it seems they are sometimes already
456     pre-defined.
457 ph10 392
458 ph10 416 28. Added support for (*UTF8) at the start of a pattern.
460 ph10 413 29. Arrange for flags added by the "release type" setting in CMake to be shown
461     in the configuration summary.
462 ph10 405
463 ph10 411
464 ph10 374 Version 7.8 05-Sep-08
465 ph10 349 ---------------------
467     1. Replaced UCP searching code with optimized version as implemented for Ad
468     Muncher (http://www.admuncher.com/) by Peter Kankowski. This uses a two-
469 ph10 351 stage table and inline lookup instead of a function, giving speed ups of 2
470     to 5 times on some simple patterns that I tested. Permission was given to
471     distribute the MultiStage2.py script that generates the tables (it's not in
472     the tarball, but is in the Subversion repository).
473 ph10 349
474 ph10 351 2. Updated the Unicode datatables to Unicode 5.1.0. This adds yet more
475     scripts.
476 ph10 358
477 ph10 353 3. Change 12 for 7.7 introduced a bug in pcre_study() when a pattern contained
478     a group with a zero qualifier. The result of the study could be incorrect,
479 ph10 358 or the function might crash, depending on the pattern.
481     4. Caseless matching was not working for non-ASCII characters in back
482 ph10 354 references. For example, /(\x{de})\1/8i was not matching \x{de}\x{fe}.
483 ph10 358 It now works when Unicode Property Support is available.
485 ph10 355 5. In pcretest, an escape such as \x{de} in the data was always generating
486     a UTF-8 string, even in non-UTF-8 mode. Now it generates a single byte in
487     non-UTF-8 mode. If the value is greater than 255, it gives a warning about
488 ph10 358 truncation.
489 ph10 349
490 ph10 358 6. Minor bugfix in pcrecpp.cc (change "" == ... to NULL == ...).
492 ph10 357 7. Added two (int) casts to pcregrep when printing the difference of two
493     pointers, in case they are 64-bit values.
494 ph10 371
495     8. Added comments about Mac OS X stack usage to the pcrestack man page and to
496     test 2 if it fails.
498 ph10 359 9. Added PCRE_CALL_CONVENTION just before the names of all exported functions,
499 ph10 371 and a #define of that name to empty if it is not externally set. This is to
500     allow users of MSVC to set it if necessary.
502     10. The PCRE_EXP_DEFN macro which precedes exported functions was missing from
503 ph10 359 the convenience functions in the pcre_get.c source file.
504 ph10 371
505 ph10 360 11. An option change at the start of a pattern that had top-level alternatives
506     could cause overwriting and/or a crash. This command provoked a crash in
507 ph10 371 some environments:
509     printf "/(?i)[\xc3\xa9\xc3\xbd]|[\xc3\xa9\xc3\xbdA]/8\n" | pcretest
511 ph10 360 This potential security problem was recorded as CVE-2008-2371.
512 ph10 371
513 ph10 361 12. For a pattern where the match had to start at the beginning or immediately
514     after a newline (e.g /.*anything/ without the DOTALL flag), pcre_exec() and
515 ph10 371 pcre_dfa_exec() could read past the end of the passed subject if there was
516 ph10 361 no match. To help with detecting such bugs (e.g. with valgrind), I modified
517     pcretest so that it places the subject at the end of its malloc-ed buffer.
518 ph10 371
519 ph10 364 13. The change to pcretest in 12 above threw up a couple more cases when pcre_
520 ph10 371 exec() might read past the end of the data buffer in UTF-8 mode.
522 ph10 364 14. A similar bug to 7.3/2 existed when the PCRE_FIRSTLINE option was set and
523 ph10 371 the data contained the byte 0x85 as part of a UTF-8 character within its
524     first line. This applied both to normal and DFA matching.
526 ph10 366 15. Lazy qualifiers were not working in some cases in UTF-8 mode. For example,
527 ph10 371 /^[^d]*?$/8 failed to match "abc".
528 ph10 351
529 ph10 371 16. Added a missing copyright notice to pcrecpp_internal.h.
531     17. Make it more clear in the documentation that values returned from
532 ph10 368 pcre_exec() in ovector are byte offsets, not character counts.
533 ph10 357
534 ph10 371 18. Tidied a few places to stop certain compilers from issuing warnings.
535 ph10 368
536 ph10 373 19. Updated the Virtual Pascal + BCC files to compile the latest v7.7, as
537     supplied by Stefan Weber. I made a further small update for 7.8 because
538 ph10 374 there is a change of source arrangements: the pcre_searchfuncs.c module is
539     replaced by pcre_ucd.c.
540 ph10 371
541 ph10 373
542 ph10 347 Version 7.7 07-May-08
543 ph10 321 ---------------------
545     1. Applied Craig's patch to sort out a long long problem: "If we can't convert
546 ph10 345 a string to a long long, pretend we don't even have a long long." This is
547 ph10 321 done by checking for the strtoq, strtoll, and _strtoi64 functions.
548 ph10 345
549 ph10 322 2. Applied Craig's patch to pcrecpp.cc to restore ABI compatibility with
550     pre-7.6 versions, which defined a global no_arg variable instead of putting
551 ph10 332 it in the RE class. (See also #8 below.)
552 ph10 345
553     3. Remove a line of dead code, identified by coverity and reported by Nuno
554     Lopes.
556 ph10 324 4. Fixed two related pcregrep bugs involving -r with --include or --exclude:
557 ph10 321
558 ph10 324 (1) The include/exclude patterns were being applied to the whole pathnames
559 ph10 345 of files, instead of just to the final components.
561 ph10 324 (2) If there was more than one level of directory, the subdirectories were
562     skipped unless they satisfied the include/exclude conditions. This is
563     inconsistent with GNU grep (and could even be seen as contrary to the
564     pcregrep specification - which I improved to make it absolutely clear).
565     The action now is always to scan all levels of directory, and just
566     apply the include/exclude patterns to regular files.
567 ph10 345
568 ph10 325 5. Added the --include_dir and --exclude_dir patterns to pcregrep, and used
569 ph10 345 --exclude_dir in the tests to avoid scanning .svn directories.
571 ph10 326 6. Applied Craig's patch to the QuoteMeta function so that it escapes the
572 ph10 345 NUL character as backslash + 0 rather than backslash + NUL, because PCRE
573 ph10 326 doesn't support NULs in patterns.
574 ph10 345
575     7. Added some missing "const"s to declarations of static tables in
576     pcre_compile.c and pcre_dfa_exec.c.
578 ph10 328 8. Applied Craig's patch to pcrecpp.cc to fix a problem in OS X that was
579 ph10 329 caused by fix #2 above. (Subsequently also a second patch to fix the
580 ph10 345 first patch. And a third patch - this was a messy problem.)
581 ph10 321
582 ph10 345 9. Applied Craig's patch to remove the use of push_back().
584 ph10 332 10. Applied Alan Lehotsky's patch to add REG_STARTEND support to the POSIX
585     matching function regexec().
586 ph10 345
587     11. Added support for the Oniguruma syntax \g<name>, \g<n>, \g'name', \g'n',
588     which, however, unlike Perl's \g{...}, are subroutine calls, not back
589 ph10 333 references. PCRE supports relative numbers with this syntax (I don't think
590     Oniguruma does).
591 ph10 345
592     12. Previously, a group with a zero repeat such as (...){0} was completely
593 ph10 335 omitted from the compiled regex. However, this means that if the group
594     was called as a subroutine from elsewhere in the pattern, things went wrong
595 ph10 345 (an internal error was given). Such groups are now left in the compiled
596     pattern, with a new opcode that causes them to be skipped at execution
597 ph10 335 time.
598 ph10 345
599 ph10 341 13. Added the PCRE_JAVASCRIPT_COMPAT option. This makes the following changes
600     to the way PCRE behaves:
601 ph10 345
602 ph10 336 (a) A lone ] character is dis-allowed (Perl treats it as data).
603 ph10 345
604     (b) A back reference to an unmatched subpattern matches an empty string
605 ph10 336 (Perl fails the current match path).
606 ph10 345
607 ph10 341 (c) A data ] in a character class must be notated as \] because if the
608 ph10 345 first data character in a class is ], it defines an empty class. (In
609 ph10 341 Perl it is not possible to have an empty class.) The empty class []
610     never matches; it forces failure and is equivalent to (*FAIL) or (?!).
611     The negative empty class [^] matches any one character, independently
612     of the DOTALL setting.
613 ph10 345
614     14. A pattern such as /(?2)[]a()b](abc)/ which had a forward reference to a
615 ph10 340 non-existent subpattern following a character class starting with ']' and
616     containing () gave an internal compiling error instead of "reference to
617     non-existent subpattern". Fortunately, when the pattern did exist, the
618 ph10 345 compiled code was correct. (When scanning forwards to check for the
619 ph10 340 existencd of the subpattern, it was treating the data ']' as terminating
620 ph10 345 the class, so got the count wrong. When actually compiling, the reference
621 ph10 340 was subsequently set up correctly.)
622 ph10 345
623 ph10 341 15. The "always fail" assertion (?!) is optimzed to (*FAIL) by pcre_compile;
624 ph10 345 it was being rejected as not supported by pcre_dfa_exec(), even though
625     other assertions are supported. I have made pcre_dfa_exec() support
626     (*FAIL).
628 ph10 342 16. The implementation of 13c above involved the invention of a new opcode,
629 ph10 345 OP_ALLANY, which is like OP_ANY but doesn't check the /s flag. Since /s
630     cannot be changed at match time, I realized I could make a small
631     improvement to matching performance by compiling OP_ALLANY instead of
632     OP_ANY for "." when DOTALL was set, and then removing the runtime tests
633     on the OP_ANY path.
635     17. Compiling pcretest on Windows with readline support failed without the
636 ph10 343 following two fixes: (1) Make the unistd.h include conditional on
637 ph10 345 HAVE_UNISTD_H; (2) #define isatty and fileno as _isatty and _fileno.
639 ph10 344 18. Changed CMakeLists.txt and cmake/FindReadline.cmake to arrange for the
640 ph10 345 ncurses library to be included for pcretest when ReadLine support is
641     requested, but also to allow for it to be overridden. This patch came from
642     Daniel Bergström.
643 ph10 347
644 ph10 346 19. There was a typo in the file ucpinternal.h where f0_rangeflag was defined
645 ph10 347 as 0x00f00000 instead of 0x00800000. Luckily, this would not have caused
646     any errors with the current Unicode tables. Thanks to Peter Kankowski for
647     spotting this.
648 ph10 324
649 ph10 332
650 ph10 319 Version 7.6 28-Jan-08
651 ph10 300 ---------------------
653 ph10 302 1. A character class containing a very large number of characters with
654     codepoints greater than 255 (in UTF-8 mode, of course) caused a buffer
655     overflow.
656 ph10 309
657     2. Patch to cut out the "long long" test in pcrecpp_unittest when
658     HAVE_LONG_LONG is not defined.
660 ph10 303 3. Applied Christian Ehrlicher's patch to update the CMake build files to
661 ph10 304 bring them up to date and include new features. This patch includes:
662 ph10 309
663 ph10 304 - Fixed PH's badly added libz and libbz2 support.
664     - Fixed a problem with static linking.
665 ph10 312 - Added pcredemo. [But later removed - see 7 below.]
666 ph10 304 - Fixed dftables problem and added an option.
667     - Added a number of HAVE_XXX tests, including HAVE_WINDOWS_H and
669     - Added readline support for pcretest.
670 ph10 309 - Added an listing of the option settings after cmake has run.
672 ph10 314 4. A user submitted a patch to Makefile that makes it easy to create
673     "pcre.dll" under mingw when using Configure/Make. I added stuff to
674     Makefile.am that cause it to include this special target, without
675     affecting anything else. Note that the same mingw target plus all
676     the other distribution libraries and programs are now supported
677     when configuring with CMake (see 6 below) instead of with
678     Configure/Make.
679 ph10 309
680 ph10 308 5. Applied Craig's patch that moves no_arg into the RE class in the C++ code.
681     This is an attempt to solve the reported problem "pcrecpp::no_arg is not
682 ph10 309 exported in the Windows port". It has not yet been confirmed that the patch
683     solves the problem, but it does no harm.
684 ph10 313
685 ph10 311 6. Applied Sheri's patch to CMakeLists.txt to add NON_STANDARD_LIB_PREFIX and
686 ph10 319 NON_STANDARD_LIB_SUFFIX for dll names built with mingw when configured
687 ph10 314 with CMake, and also correct the comment about stack recursion.
688 ph10 313
689 ph10 312 7. Remove the automatic building of pcredemo from the ./configure system and
690 ph10 313 from CMakeLists.txt. The whole idea of pcredemo.c is that it is an example
691     of a program that users should build themselves after PCRE is installed, so
692     building it automatically is not really right. What is more, it gave
693 ph10 312 trouble in some build environments.
694 ph10 300
695 ph10 319 8. Further tidies to CMakeLists.txt from Sheri and Christian.
696 ph10 308
697 ph10 319
698 ph10 298 Version 7.5 10-Jan-08
699 ph10 263 ---------------------
701     1. Applied a patch from Craig: "This patch makes it possible to 'ignore'
702     values in parens when parsing an RE using the C++ wrapper."
703 ph10 286
704 ph10 264 2. Negative specials like \S did not work in character classes in UTF-8 mode.
705     Characters greater than 255 were excluded from the class instead of being
706     included.
707 ph10 286
708     3. The same bug as (2) above applied to negated POSIX classes such as
709 ph10 265 [:^space:].
710 ph10 286
711 ph10 267 4. PCRECPP_STATIC was referenced in pcrecpp_internal.h, but nowhere was it
712     defined or documented. It seems to have been a typo for PCRE_STATIC, so
713 ph10 286 I have changed it.
715 ph10 268 5. The construct (?&) was not diagnosed as a syntax error (it referenced the
716 ph10 286 first named subpattern) and a construct such as (?&a) would reference the
717     first named subpattern whose name started with "a" (in other words, the
718 ph10 272 length check was missing). Both these problems are fixed. "Subpattern name
719     expected" is now given for (?&) (a zero-length name), and this patch also
720     makes it give the same error for \k'' (previously it complained that that
721     was a reference to a non-existent subpattern).
722 ph10 286
723 ph10 269 6. The erroneous patterns (?+-a) and (?-+a) give different error messages;
724 ph10 286 this is right because (?- can be followed by option settings as well as by
725 ph10 269 digits. I have, however, made the messages clearer.
726 ph10 286
727 ph10 270 7. Patterns such as (?(1)a|b) (a pattern that contains fewer subpatterns
728     than the number used in the conditional) now cause a compile-time error.
729     This is actually not compatible with Perl, which accepts such patterns, but
730     treats the conditional as always being FALSE (as PCRE used to), but it
731     seems to me that giving a diagnostic is better.
732 ph10 286
733 ph10 274 8. Change "alphameric" to the more common word "alphanumeric" in comments
734 ph10 275 and messages.
735 ph10 286
736     9. Fix two occurrences of "backslash" in comments that should have been
737     "backspace".
739     10. Remove two redundant lines of code that can never be obeyed (their function
740     was moved elsewhere).
742     11. The program that makes PCRE's Unicode character property table had a bug
743     which caused it to generate incorrect table entries for sequences of
744 ph10 277 characters that have the same character type, but are in different scripts.
745 ph10 286 It amalgamated them into a single range, with the script of the first of
746 ph10 277 them. In other words, some characters were in the wrong script. There were
747     thirteen such cases, affecting characters in the following ranges:
748 ph10 286
749 ph10 277 U+002b0 - U+002c1
750     U+0060c - U+0060d
751 ph10 286 U+0061e - U+00612
752 ph10 277 U+0064b - U+0065e
753     U+0074d - U+0076d
754     U+01800 - U+01805
755     U+01d00 - U+01d77
756     U+01d9b - U+01dbf
757     U+0200b - U+0200f
758     U+030fc - U+030fe
759     U+03260 - U+0327f
760     U+0fb46 - U+0fbb1
761     U+10450 - U+1049d
762 ph10 286
763 ph10 279 12. The -o option (show only the matching part of a line) for pcregrep was not
764     compatible with GNU grep in that, if there was more than one match in a
765     line, it showed only the first of them. It now behaves in the same way as
766     GNU grep.
767 ph10 286
768     13. If the -o and -v options were combined for pcregrep, it printed a blank
769     line for every non-matching line. GNU grep prints nothing, and pcregrep now
770     does the same. The return code can be used to tell if there were any
771     non-matching lines.
773 ph10 289 14. Added --file-offsets and --line-offsets to pcregrep.
775     15. The pattern (?=something)(?R) was not being diagnosed as a potentially
776 ph10 282 infinitely looping recursion. The bug was that positive lookaheads were not
777 ph10 286 being skipped when checking for a possible empty match (negative lookaheads
778     and both kinds of lookbehind were skipped).
780 ph10 289 16. Fixed two typos in the Windows-only code in pcregrep.c, and moved the
781 ph10 284 inclusion of <windows.h> to before rather than after the definition of
782     INVALID_FILE_ATTRIBUTES (patch from David Byron).
783 ph10 263
784 ph10 289 17. Specifying a possessive quantifier with a specific limit for a Unicode
785 ph10 286 character property caused pcre_compile() to compile bad code, which led at
786     runtime to PCRE_ERROR_INTERNAL (-14). Examples of patterns that caused this
787     are: /\p{Zl}{2,3}+/8 and /\p{Cc}{2}+/8. It was the possessive "+" that
788     caused the error; without that there was no problem.
789 ph10 263
790 ph10 289 18. Added --enable-pcregrep-libz and --enable-pcregrep-libbz2.
791 ph10 286
792 ph10 289 19. Added --enable-pcretest-libreadline.
793 ph10 286
794 ph10 289 20. In pcrecpp.cc, the variable 'count' was incremented twice in
795 ph10 288 RE::GlobalReplace(). As a result, the number of replacements returned was
796 ph10 298 double what it should be. I removed one of the increments, but Craig sent a
797     later patch that removed the other one (the right fix) and added unit tests
798     that check the return values (which was not done before).
799 ph10 292
800 ph10 291 21. Several CMake things:
801 ph10 287
802 ph10 291 (1) Arranged that, when cmake is used on Unix, the libraries end up with
803 ph10 292 the names libpcre and libpcreposix, not just pcre and pcreposix.
804 ph10 288
805 ph10 292 (2) The above change means that pcretest and pcregrep are now correctly
806     linked with the newly-built libraries, not previously installed ones.
807 ph10 291
809 ph10 298
810     22. In UTF-8 mode, with newline set to "any", a pattern such as .*a.*=.b.*
811     crashed when matching a string such as a\x{2029}b (note that \x{2029} is a
812     UTF-8 newline character). The key issue is that the pattern starts .*;
813     this means that the match must be either at the beginning, or after a
814     newline. The bug was in the code for advancing after a failed match and
815     checking that the new position followed a newline. It was not taking
816 ph10 294 account of UTF-8 characters correctly.
817 ph10 298
818     23. PCRE was behaving differently from Perl in the way it recognized POSIX
819     character classes. PCRE was not treating the sequence [:...:] as a
820     character class unless the ... were all letters. Perl, however, seems to
821     allow any characters between [: and :], though of course it rejects as
822     unknown any "names" that contain non-letters, because all the known class
823     names consist only of letters. Thus, Perl gives an error for [[:1234:]],
824     for example, whereas PCRE did not - it did not recognize a POSIX character
825     class. This seemed a bit dangerous, so the code has been changed to be
826     closer to Perl. The behaviour is not identical to Perl, because PCRE will
827     diagnose an unknown class for, for example, [[:l\ower:]] where Perl will
828     treat it as [[:lower:]]. However, PCRE does now give "unknown" errors where
829     Perl does, and where it didn't before.
831 ph10 296 24. Rewrite so as to remove the single use of %n from pcregrep because in some
832     Windows environments %n is disabled by default.
833 ph10 292
835 ph10 260 Version 7.4 21-Sep-07
836 ph10 230 ---------------------
838     1. Change 7.3/28 was implemented for classes by looking at the bitmap. This
839 ph10 231 means that a class such as [\s] counted as "explicit reference to CR or
840     LF". That isn't really right - the whole point of the change was to try to
841     help when there was an actual mention of one of the two characters. So now
842     the change happens only if \r or \n (or a literal CR or LF) character is
843 ph10 230 encountered.
844 ph10 231
845     2. The 32-bit options word was also used for 6 internal flags, but the numbers
846     of both had grown to the point where there were only 3 bits left.
847     Fortunately, there was spare space in the data structure, and so I have
848     moved the internal flags into a new 16-bit field to free up more option
849 ph10 230 bits.
850 ph10 231
851     3. The appearance of (?J) at the start of a pattern set the DUPNAMES option,
852     but did not set the internal JCHANGED flag - either of these is enough to
853     control the way the "get" function works - but the PCRE_INFO_JCHANGED
854     facility is supposed to tell if (?J) was ever used, so now (?J) at the
855 ph10 230 start sets both bits.
857 ph10 231 4. Added options (at build time, compile time, exec time) to change \R from
858     matching any Unicode line ending sequence to just matching CR, LF, or CRLF.
859 ph10 230
860 ph10 243 5. doc/pcresyntax.html was missing from the distribution.
862     6. Put back the definition of PCRE_ERROR_NULLWSLIMIT, for backward
863 ph10 233 compatibility, even though it is no longer used.
864 ph10 243
865     7. Added macro for snprintf to pcrecpp_unittest.cc and also for strtoll and
866 ph10 254 strtoull to pcrecpp.cc to select the available functions in WIN32 when the
867 ph10 259 windows.h file is present (where different names are used). [This was
868 ph10 257 reversed later after testing - see 16 below.]
869 ph10 243
870     8. Changed all #include <config.h> to #include "config.h". There were also
871 ph10 236 some further <pcre.h> cases that I changed to "pcre.h".
872 ph10 243
873 ph10 239 9. When pcregrep was used with the --colour option, it missed the line ending
874 ph10 243 sequence off the lines that it output.
876     10. It was pointed out to me that arrays of string pointers cause lots of
877     relocations when a shared library is dynamically loaded. A technique of
878     using a single long string with a table of offsets can drastically reduce
879     these. I have refactored PCRE in four places to do this. The result is
880 ph10 240 dramatic:
881 ph10 243
882 ph10 240 Originally: 290
883     After changing UCP table: 187
884 ph10 243 After changing error message table: 43
885 ph10 240 After changing table of "verbs" 36
886     After changing table of Posix names 22
887 ph10 243
888 ph10 240 Thanks to the folks working on Gregex for glib for this insight.
889 ph10 247
890 ph10 244 11. --disable-stack-for-recursion caused compiling to fail unless -enable-
891 ph10 247 unicode-properties was also set.
893 ph10 248 12. Updated the tests so that they work when \R is defaulted to ANYCRLF.
894 ph10 231
895 ph10 253 13. Added checks for ANY and ANYCRLF to pcrecpp.cc where it previously
896     checked only for CRLF.
897 ph10 233
898 ph10 259 14. Added casts to pcretest.c to avoid compiler warnings.
900 ph10 256 15. Added Craig's patch to various pcrecpp modules to avoid compiler warnings.
901 ph10 253
902 ph10 257 16. Added Craig's patch to remove the WINDOWS_H tests, that were not working,
903 ph10 259 and instead check for _strtoi64 explicitly, and avoid the use of snprintf()
904     entirely. This removes changes made in 7 above.
905 ph10 256
906 ph10 261 17. The CMake files have been updated, and there is now more information about
907     building with CMake in the NON-UNIX-USE document.
908 ph10 257
909 ph10 261
910 ph10 228 Version 7.3 28-Aug-07
911 ph10 157 ---------------------
913 ph10 189 1. In the rejigging of the build system that eventually resulted in 7.1, the
914     line "#include <pcre.h>" was included in pcre_internal.h. The use of angle
915     brackets there is not right, since it causes compilers to look for an
916     installed pcre.h, not the version that is in the source that is being
917     compiled (which of course may be different). I have changed it back to:
918 ph10 197
919 ph10 189 #include "pcre.h"
920 ph10 197
921     I have a vague recollection that the change was concerned with compiling in
922     different directories, but in the new build system, that is taken care of
923     by the VPATH setting the Makefile.
925 ph10 190 2. The pattern .*$ when run in not-DOTALL UTF-8 mode with newline=any failed
926     when the subject happened to end in the byte 0x85 (e.g. if the last
927     character was \x{1ec5}). *Character* 0x85 is one of the "any" newline
928     characters but of course it shouldn't be taken as a newline when it is part
929     of another character. The bug was that, for an unlimited repeat of . in
930     not-DOTALL UTF-8 mode, PCRE was advancing by bytes rather than by
931     characters when looking for a newline.
932 ph10 197
933     3. A small performance improvement in the DOTALL UTF-8 mode .* case.
935     4. Debugging: adjusted the names of opcodes for different kinds of parentheses
936     in debug output.
938 ph10 201 5. Arrange to use "%I64d" instead of "%lld" and "%I64u" instead of "%llu" for
939     long printing in the pcrecpp unittest when running under MinGW.
940 ph10 189
941 ph10 197 6. ESC_K was left out of the EBCDIC table.
942 ph10 189
943 ph10 197 7. Change 7.0/38 introduced a new limit on the number of nested non-capturing
944     parentheses; I made it 1000, which seemed large enough. Unfortunately, the
945     limit also applies to "virtual nesting" when a pattern is recursive, and in
946     this case 1000 isn't so big. I have been able to remove this limit at the
947     expense of backing off one optimization in certain circumstances. Normally,
948     when pcre_exec() would call its internal match() function recursively and
949     immediately return the result unconditionally, it uses a "tail recursion"
950     feature to save stack. However, when a subpattern that can match an empty
951     string has an unlimited repetition quantifier, it no longer makes this
952     optimization. That gives it a stack frame in which to save the data for
953     checking that an empty string has been matched. Previously this was taken
954     from the 1000-entry workspace that had been reserved. So now there is no
955     explicit limit, but more stack is used.
956 ph10 189
957 ph10 200 8. Applied Daniel's patches to solve problems with the import/export magic
958     syntax that is required for Windows, and which was going wrong for the
959     pcreposix and pcrecpp parts of the library. These were overlooked when this
960     problem was solved for the main library.
961 ph10 197
962 ph10 202 9. There were some crude static tests to avoid integer overflow when computing
963     the size of patterns that contain repeated groups with explicit upper
964     limits. As the maximum quantifier is 65535, the maximum group length was
965     set at 30,000 so that the product of these two numbers did not overflow a
966     32-bit integer. However, it turns out that people want to use groups that
967     are longer than 30,000 bytes (though not repeat them that many times).
968     Change 7.0/17 (the refactoring of the way the pattern size is computed) has
969     made it possible to implement the integer overflow checks in a much more
970     dynamic way, which I have now done. The artificial limitation on group
971     length has been removed - we now have only the limit on the total length of
972     the compiled pattern, which depends on the LINK_SIZE setting.
973 ph10 208
974     10. Fixed a bug in the documentation for get/copy named substring when
975     duplicate names are permitted. If none of the named substrings are set, the
976     functions return PCRE_ERROR_NOSUBSTRING (7); the doc said they returned an
977     empty string.
979     11. Because Perl interprets \Q...\E at a high level, and ignores orphan \E
980     instances, patterns such as [\Q\E] or [\E] or even [^\E] cause an error,
981     because the ] is interpreted as the first data character and the
982     terminating ] is not found. PCRE has been made compatible with Perl in this
983     regard. Previously, it interpreted [\Q\E] as an empty class, and [\E] could
984     cause memory overwriting.
986 ph10 206 10. Like Perl, PCRE automatically breaks an unlimited repeat after an empty
987     string has been matched (to stop an infinite loop). It was not recognizing
988 ph10 208 a conditional subpattern that could match an empty string if that
989 ph10 206 subpattern was within another subpattern. For example, it looped when
990 ph10 208 trying to match (((?(1)X|))*) but it was OK with ((?(1)X|)*) where the
991 ph10 206 condition was not nested. This bug has been fixed.
992 ph10 208
993 ph10 207 12. A pattern like \X?\d or \P{L}?\d in non-UTF-8 mode could cause a backtrack
994     past the start of the subject in the presence of bytes with the top bit
995     set, for example "\x8aBCD".
996 ph10 211
997 ph10 210 13. Added Perl 5.10 experimental backtracking controls (*FAIL), (*F), (*PRUNE),
998     (*SKIP), (*THEN), (*COMMIT), and (*ACCEPT).
999 ph10 200
1000 ph10 211 14. Optimized (?!) to (*FAIL).
1001 ph10 202
1002 ph10 212 15. Updated the test for a valid UTF-8 string to conform to the later RFC 3629.
1003     This restricts code points to be within the range 0 to 0x10FFFF, excluding
1004     the "low surrogate" sequence 0xD800 to 0xDFFF. Previously, PCRE allowed the
1005     full range 0 to 0x7FFFFFFF, as defined by RFC 2279. Internally, it still
1006     does: it's just the validity check that is more restrictive.
1007 ph10 220
1008     16. Inserted checks for integer overflows during escape sequence (backslash)
1009     processing, and also fixed erroneous offset values for syntax errors during
1010     backslash processing.
1012 ph10 214 17. Fixed another case of looking too far back in non-UTF-8 mode (cf 12 above)
1013 ph10 220 for patterns like [\PPP\x8a]{1,}\x80 with the subject "A\x80".
1015 ph10 215 18. An unterminated class in a pattern like (?1)\c[ with a "forward reference"
1016     caused an overrun.
1017 ph10 220
1018     19. A pattern like (?:[\PPa*]*){8,} which had an "extended class" (one with
1019     something other than just ASCII characters) inside a group that had an
1020     unlimited repeat caused a loop at compile time (while checking to see
1021     whether the group could match an empty string).
1023 ph10 217 20. Debugging a pattern containing \p or \P could cause a crash. For example,
1024     [\P{Any}] did so. (Error in the code for printing property names.)
1025 ph10 210
1026 ph10 220 21. An orphan \E inside a character class could cause a crash.
1028     22. A repeated capturing bracket such as (A)? could cause a wild memory
1029 ph10 218 reference during compilation.
1030 ph10 220
1031     23. There are several functions in pcre_compile() that scan along a compiled
1032     expression for various reasons (e.g. to see if it's fixed length for look
1033 ph10 218 behind). There were bugs in these functions when a repeated \p or \P was
1034 ph10 220 present in the pattern. These operators have additional parameters compared
1035     with \d, etc, and these were not being taken into account when moving along
1036 ph10 218 the compiled data. Specifically:
1037 ph10 220
1038     (a) A item such as \p{Yi}{3} in a lookbehind was not treated as fixed
1039     length.
1041     (b) An item such as \pL+ within a repeated group could cause crashes or
1042 ph10 218 loops.
1043 ph10 220
1044 ph10 218 (c) A pattern such as \p{Yi}+(\P{Yi}+)(?1) could give an incorrect
1045 ph10 220 "reference to non-existent subpattern" error.
1047 ph10 221 (d) A pattern like (\P{Yi}{2}\277)? could loop at compile time.
1049 ph10 219 24. A repeated \S or \W in UTF-8 mode could give wrong answers when multibyte
1050 ph10 220 characters were involved (for example /\S{2}/8g with "A\x{a3}BC").
1051 ph10 211
1052 ph10 222 25. Using pcregrep in multiline, inverted mode (-Mv) caused it to loop.
1053 ph10 218
1054 ph10 227 26. Patterns such as [\P{Yi}A] which include \p or \P and just one other
1055 ph10 223 character were causing crashes (broken optimization).
1056 ph10 227
1057 ph10 224 27. Patterns such as (\P{Yi}*\277)* (group with possible zero repeat containing
1058 ph10 227 \p or \P) caused a compile-time loop.
1060 ph10 226 28. More problems have arisen in unanchored patterns when CRLF is a valid line
1061     break. For example, the unstudied pattern [\r\n]A does not match the string
1062     "\r\nA" because change 7.0/46 below moves the current point on by two
1063     characters after failing to match at the start. However, the pattern \nA
1064 ph10 227 *does* match, because it doesn't start till \n, and if [\r\n]A is studied,
1065     the same is true. There doesn't seem any very clean way out of this, but
1066     what I have chosen to do makes the common cases work: PCRE now takes note
1067     of whether there can be an explicit match for \r or \n anywhere in the
1068     pattern, and if so, 7.0/46 no longer applies. As part of this change,
1069     there's a new PCRE_INFO_HASCRORLF option for finding out whether a compiled
1070     pattern has explicit CR or LF references.
1071 ph10 222
1072 ph10 227 29. Added (*CR) etc for changing newline setting at start of pattern.
1073 ph10 223
1074 ph10 227
1075 ph10 189 Version 7.2 19-Jun-07
1076     ---------------------
1078 ph10 157 1. If the fr_FR locale cannot be found for test 3, try the "french" locale,
1079     which is apparently normally available under Windows.
1080 ph10 159
1081     2. Re-jig the pcregrep tests with different newline settings in an attempt
1082     to make them independent of the local environment's newline setting.
1083 ph10 157
1084 ph10 160 3. Add code to configure.ac to remove -g from the CFLAGS default settings.
1085 ph10 165
1086 ph10 161 4. Some of the "internals" tests were previously cut out when the link size
1087     was not 2, because the output contained actual offsets. The recent new
1088 ph10 165 "Z" feature of pcretest means that these can be cut out, making the tests
1089     usable with all link sizes.
1091 ph10 164 5. Implemented Stan Switzer's goto replacement for longjmp() when not using
1092     stack recursion. This gives a massive performance boost under BSD, but just
1093 ph10 165 a small improvement under Linux. However, it saves one field in the frame
1094 ph10 164 in all cases.
1095 ph10 172
1096 ph10 166 6. Added more features from the forthcoming Perl 5.10:
1097 ph10 172
1098 ph10 166 (a) (?-n) (where n is a string of digits) is a relative subroutine or
1099     recursion call. It refers to the nth most recently opened parentheses.
1100 ph10 172
1101 ph10 166 (b) (?+n) is also a relative subroutine call; it refers to the nth next
1102 ph10 172 to be opened parentheses.
1104     (c) Conditions that refer to capturing parentheses can be specified
1105 ph10 167 relatively, for example, (?(-2)... or (?(+3)...
1106 ph10 172
1107 ph10 168 (d) \K resets the start of the current match so that everything before
1108 ph10 172 is not part of it.
1110 ph10 171 (e) \k{name} is synonymous with \k<name> and \k'name' (.NET compatible).
1111 ph10 172
1112 ph10 171 (f) \g{name} is another synonym - part of Perl 5.10's unification of
1113 ph10 172 reference syntax.
1114 ph10 182
1115 ph10 175 (g) (?| introduces a group in which the numbering of parentheses in each
1116 ph10 182 alternative starts with the same number.
1117 ph10 172
1118 ph10 182 (h) \h, \H, \v, and \V match horizontal and vertical whitespace.
1120 ph10 172 7. Added two new calls to pcre_fullinfo(): PCRE_INFO_OKPARTIAL and
1123     8. A pattern such as (.*(.)?)* caused pcre_exec() to fail by either not
1124     terminating or by crashing. Diagnosed by Viktor Griph; it was in the code
1125 ph10 170 for detecting groups that can match an empty string.
1126 ph10 159
1127 ph10 172 9. A pattern with a very large number of alternatives (more than several
1128     hundred) was running out of internal workspace during the pre-compile
1129     phase, where pcre_compile() figures out how much memory will be needed. A
1130     bit of new cunning has reduced the workspace needed for groups with
1131     alternatives. The 1000-alternative test pattern now uses 12 bytes of
1132     workspace instead of running out of the 4096 that are available.
1133 ph10 182
1134 ph10 176 10. Inserted some missing (unsigned int) casts to get rid of compiler warnings.
1135 ph10 172
1136 ph10 179 11. Applied patch from Google to remove an optimization that didn't quite work.
1137     The report of the bug said:
1138 ph10 182
1139 ph10 179 pcrecpp::RE("a*").FullMatch("aaa") matches, while
1140     pcrecpp::RE("a*?").FullMatch("aaa") does not, and
1141 ph10 182 pcrecpp::RE("a*?\\z").FullMatch("aaa") does again.
1142 ph10 185
1143 ph10 184 12. If \p or \P was used in non-UTF-8 mode on a character greater than 127
1144 ph10 185 it matched the wrong number of bytes.
1145 ph10 172
1146 ph10 179
1147 ph10 155 Version 7.1 24-Apr-07
1148 ph10 98 ---------------------
1150 ph10 111 1. Applied Bob Rossi and Daniel G's patches to convert the build system to one
1151 ph10 122 that is more "standard", making use of automake and other Autotools. There
1152 ph10 99 is some re-arrangement of the files and adjustment of comments consequent
1153     on this.
1154 ph10 111
1155     2. Part of the patch fixed a problem with the pcregrep tests. The test of -r
1156     for recursive directory scanning broke on some systems because the files
1157     are not scanned in any specific order and on different systems the order
1158     was different. A call to "sort" has been inserted into RunGrepTest for the
1159     approprate test as a short-term fix. In the longer term there may be an
1160 ph10 100 alternative.
1161 ph10 111
1162 ph10 100 3. I had an email from Eric Raymond about problems translating some of PCRE's
1163 ph10 111 man pages to HTML (despite the fact that I distribute HTML pages, some
1164     people do their own conversions for various reasons). The problems
1165     concerned the use of low-level troff macros .br and .in. I have therefore
1166     removed all such uses from the man pages (some were redundant, some could
1167 ph10 113 be replaced by .nf/.fi pairs). The 132html script that I use to generate
1168     HTML has been updated to handle .nf/.fi and to complain if it encounters
1169     .br or .in.
1170 ph10 111
1171 ph10 100 4. Updated comments in configure.ac that get placed in config.h.in and also
1172 ph10 123 arranged for config.h to be included in the distribution, with the name
1173 ph10 111 config.h.generic, for the benefit of those who have to compile without
1174     Autotools (compare pcre.h, which is now distributed as pcre.h.generic).
1176     5. Updated the support (such as it is) for Virtual Pascal, thanks to Stefan
1177     Weber: (1) pcre_internal.h was missing some function renames; (2) updated
1178 ph10 127 makevp.bat for the current PCRE, using the additional files
1179 ph10 135 makevp_c.txt, makevp_l.txt, and pcregexp.pas.
1180 ph10 111
1181     6. A Windows user reported a minor discrepancy with test 2, which turned out
1182     to be caused by a trailing space on an input line that had got lost in his
1183 ph10 102 copy. The trailing space was an accident, so I've just removed it.
1184 ph10 111
1185 ph10 104 7. Add -Wl,-R... flags in pcre-config.in for *BSD* systems, as I'm told
1186 ph10 111 that is needed.
1188 ph10 105 8. Mark ucp_table (in ucptable.h) and ucp_gentype (in pcre_ucp_searchfuncs.c)
1189     as "const" (a) because they are and (b) because it helps the PHP
1190     maintainers who have recently made a script to detect big data structures
1191 ph10 111 in the php code that should be moved to the .rodata section. I remembered
1192     to update Builducptable as well, so it won't revert if ucptable.h is ever
1193 ph10 105 re-created.
1194 ph10 111
1195     9. Added some extra #ifdef SUPPORT_UTF8 conditionals into pcretest.c,
1196 ph10 107 pcre_printint.src, pcre_compile.c, pcre_study.c, and pcre_tables.c, in
1197 ph10 111 order to be able to cut out the UTF-8 tables in the latter when UTF-8
1198     support is not required. This saves 1.5-2K of code, which is important in
1199 ph10 107 some applications.
1200 ph10 111
1201 ph10 110 Later: more #ifdefs are needed in pcre_ord2utf8.c and pcre_valid_utf8.c
1202 ph10 111 so as not to refer to the tables, even though these functions will never be
1203     called when UTF-8 support is disabled. Otherwise there are problems with a
1204     shared library.
1206 ph10 118 10. Fixed two bugs in the emulated memmove() function in pcre_internal.h:
1208     (a) It was defining its arguments as char * instead of void *.
1210 ph10 123 (b) It was assuming that all moves were upwards in memory; this was true
1211     a long time ago when I wrote it, but is no longer the case.
1213 ph10 118 The emulated memove() is provided for those environments that have neither
1214 ph10 123 memmove() nor bcopy(). I didn't think anyone used it these days, but that
1215 ph10 118 is clearly not the case, as these two bugs were recently reported.
1216 ph10 123
1217 ph10 111 11. The script PrepareRelease is now distributed: it calls 132html, CleanTxt,
1218 ph10 123 and Detrail to create the HTML documentation, the .txt form of the man
1219     pages, and it removes trailing spaces from listed files. It also creates
1220     pcre.h.generic and config.h.generic from pcre.h and config.h. In the latter
1221     case, it wraps all the #defines with #ifndefs. This script should be run
1222 ph10 111 before "make dist".
1223 ph10 123
1224 ph10 115 12. Fixed two fairly obscure bugs concerned with quantified caseless matching
1225     with Unicode property support.
1226 ph10 123
1227     (a) For a maximizing quantifier, if the two different cases of the
1228     character were of different lengths in their UTF-8 codings (there are
1229     some cases like this - I found 11), and the matching function had to
1230 ph10 115 back up over a mixture of the two cases, it incorrectly assumed they
1231     were both the same length.
1232 ph10 123
1233     (b) When PCRE was configured to use the heap rather than the stack for
1234     recursion during matching, it was not correctly preserving the data for
1235     the other case of a UTF-8 character when checking ahead for a match
1236     while processing a minimizing repeat. If the check also involved
1237 ph10 115 matching a wide character, but failed, corruption could cause an
1238     erroneous result when trying to check for a repeat of the original
1239     character.
1240 ph10 123
1241 ph10 116 13. Some tidying changes to the testing mechanism:
1242 ph10 98
1243 ph10 116 (a) The RunTest script now detects the internal link size and whether there
1244     is UTF-8 and UCP support by running ./pcretest -C instead of relying on
1245 ph10 123 values substituted by "configure". (The RunGrepTest script already did
1246     this for UTF-8.) The configure.ac script no longer substitutes the
1247     relevant variables.
1249 ph10 116 (b) The debugging options /B and /D in pcretest show the compiled bytecode
1250     with length and offset values. This means that the output is different
1251     for different internal link sizes. Test 2 is skipped for link sizes
1252     other than 2 because of this, bypassing the problem. Unfortunately,
1253     there was also a test in test 3 (the locale tests) that used /B and
1254 ph10 123 failed for link sizes other than 2. Rather than cut the whole test out,
1255     I have added a new /Z option to pcretest that replaces the length and
1256     offset values with spaces. This is now used to make test 3 independent
1257 ph10 122 of link size. (Test 2 will be tidied up later.)
1258 ph10 123
1259     14. If erroroffset was passed as NULL to pcre_compile, it provoked a
1260 ph10 122 segmentation fault instead of returning the appropriate error message.
1261 ph10 134
1262 ph10 131 15. In multiline mode when the newline sequence was set to "any", the pattern
1263 ph10 134 ^$ would give a match between the \r and \n of a subject such as "A\r\nB".
1264     This doesn't seem right; it now treats the CRLF combination as the line
1265     ending, and so does not match in that case. It's only a pattern such as ^$
1266     that would hit this one: something like ^ABC$ would have failed after \r
1267     and then tried again after \r\n.
1269 ph10 131 16. Changed the comparison command for RunGrepTest from "diff -u" to "diff -ub"
1270 ph10 134 in an attempt to make files that differ only in their line terminators
1271     compare equal. This works on Linux.
1272 ph10 142
1273 ph10 141 17. Under certain error circumstances pcregrep might try to free random memory
1274     as it exited. This is now fixed, thanks to valgrind.
1275 ph10 142
1276 ph10 141 19. In pcretest, if the pattern /(?m)^$/g<any> was matched against the string
1277 ph10 142 "abc\r\n\r\n", it found an unwanted second match after the second \r. This
1278     was because its rules for how to advance for /g after matching an empty
1279 ph10 143 string at the end of a line did not allow for this case. They now check for
1280     it specially.
1281 ph10 150
1282     20. pcretest is supposed to handle patterns and data of any length, by
1283     extending its buffers when necessary. It was getting this wrong when the
1284 ph10 147 buffer for a data line had to be extended.
1285 ph10 150
1286 ph10 149 21. Added PCRE_NEWLINE_ANYCRLF which is like ANY, but matches only CR, LF, or
1287 ph10 150 CRLF as a newline sequence.
1288 ph10 152
1289 ph10 151 22. Code for handling Unicode properties in pcre_dfa_exec() wasn't being cut
1290 ph10 152 out by #ifdef SUPPORT_UCP. This did no harm, as it could never be used, but
1291     I have nevertheless tidied it up.
1292 ph10 111
1293 ph10 152 23. Added some casts to kill warnings from HP-UX ia64 compiler.
1294 ph10 123
1295 ph10 153 24. Added a man page for pcre-config.
1296 ph10 152
1297 ph10 153
1298 nigel 93 Version 7.0 19-Dec-06
1299     ---------------------
1301     1. Fixed a signed/unsigned compiler warning in pcre_compile.c, shown up by
1302     moving to gcc 4.1.1.
1304     2. The -S option for pcretest uses setrlimit(); I had omitted to #include
1305     sys/time.h, which is documented as needed for this function. It doesn't
1306     seem to matter on Linux, but it showed up on some releases of OS X.
1308     3. It seems that there are systems where bytes whose values are greater than
1309     127 match isprint() in the "C" locale. The "C" locale should be the
1310     default when a C program starts up. In most systems, only ASCII printing
1311     characters match isprint(). This difference caused the output from pcretest
1312     to vary, making some of the tests fail. I have changed pcretest so that:
1314     (a) When it is outputting text in the compiled version of a pattern, bytes
1315     other than 32-126 are always shown as hex escapes.
1317     (b) When it is outputting text that is a matched part of a subject string,
1318     it does the same, unless a different locale has been set for the match
1319     (using the /L modifier). In this case, it uses isprint() to decide.
1321     4. Fixed a major bug that caused incorrect computation of the amount of memory
1322     required for a compiled pattern when options that changed within the
1323     pattern affected the logic of the preliminary scan that determines the
1324     length. The relevant options are -x, and -i in UTF-8 mode. The result was
1325     that the computed length was too small. The symptoms of this bug were
1326     either the PCRE error "internal error: code overflow" from pcre_compile(),
1327     or a glibc crash with a message such as "pcretest: free(): invalid next
1328     size (fast)". Examples of patterns that provoked this bug (shown in
1329     pcretest format) are:
1331     /(?-x: )/x
1332     /(?x)(?-x: \s*#\s*)/
1333     /((?i)[\x{c0}])/8
1334     /(?i:[\x{c0}])/8
1336     HOWEVER: Change 17 below makes this fix obsolete as the memory computation
1337     is now done differently.
1339     5. Applied patches from Google to: (a) add a QuoteMeta function to the C++
1340     wrapper classes; (b) implement a new function in the C++ scanner that is
1341     more efficient than the old way of doing things because it avoids levels of
1342     recursion in the regex matching; (c) add a paragraph to the documentation
1343     for the FullMatch() function.
1345     6. The escape sequence \n was being treated as whatever was defined as
1346     "newline". Not only was this contrary to the documentation, which states
1347     that \n is character 10 (hex 0A), but it also went horribly wrong when
1348     "newline" was defined as CRLF. This has been fixed.
1350     7. In pcre_dfa_exec.c the value of an unsigned integer (the variable called c)
1351     was being set to -1 for the "end of line" case (supposedly a value that no
1352     character can have). Though this value is never used (the check for end of
1353     line is "zero bytes in current character"), it caused compiler complaints.
1354     I've changed it to 0xffffffff.
1356     8. In pcre_version.c, the version string was being built by a sequence of
1357     C macros that, in the event of PCRE_PRERELEASE being defined as an empty
1358     string (as it is for production releases) called a macro with an empty
1359     argument. The C standard says the result of this is undefined. The gcc
1360     compiler treats it as an empty string (which was what was wanted) but it is
1361     reported that Visual C gives an error. The source has been hacked around to
1362     avoid this problem.
1364     9. On the advice of a Windows user, included <io.h> and <fcntl.h> in Windows
1365     builds of pcretest, and changed the call to _setmode() to use _O_BINARY
1366     instead of 0x8000. Made all the #ifdefs test both _WIN32 and WIN32 (not all
1367     of them did).
1369     10. Originally, pcretest opened its input and output without "b"; then I was
1370     told that "b" was needed in some environments, so it was added for release
1371     5.0 to both the input and output. (It makes no difference on Unix-like
1372     systems.) Later I was told that it is wrong for the input on Windows. I've
1373     now abstracted the modes into two macros, to make it easier to fiddle with
1374     them, and removed "b" from the input mode under Windows.
1376     11. Added pkgconfig support for the C++ wrapper library, libpcrecpp.
1378     12. Added -help and --help to pcretest as an official way of being reminded
1379     of the options.
1381     13. Removed some redundant semicolons after macro calls in pcrecpparg.h.in
1382     and pcrecpp.cc because they annoy compilers at high warning levels.
1384     14. A bit of tidying/refactoring in pcre_exec.c in the main bumpalong loop.
1386     15. Fixed an occurrence of == in configure.ac that should have been = (shell
1387     scripts are not C programs :-) and which was not noticed because it works
1388     on Linux.
1390     16. pcretest is supposed to handle any length of pattern and data line (as one
1391     line or as a continued sequence of lines) by extending its input buffer if
1392     necessary. This feature was broken for very long pattern lines, leading to
1393     a string of junk being passed to pcre_compile() if the pattern was longer
1394     than about 50K.
1396     17. I have done a major re-factoring of the way pcre_compile() computes the
1397     amount of memory needed for a compiled pattern. Previously, there was code
1398     that made a preliminary scan of the pattern in order to do this. That was
1399     OK when PCRE was new, but as the facilities have expanded, it has become
1400     harder and harder to keep it in step with the real compile phase, and there
1401     have been a number of bugs (see for example, 4 above). I have now found a
1402     cunning way of running the real compile function in a "fake" mode that
1403     enables it to compute how much memory it would need, while actually only
1404     ever using a few hundred bytes of working memory and without too many
1405     tests of the mode. This should make future maintenance and development
1406     easier. A side effect of this work is that the limit of 200 on the nesting
1407     depth of parentheses has been removed (though this was never a serious
1408     limitation, I suspect). However, there is a downside: pcre_compile() now
1409     runs more slowly than before (30% or more, depending on the pattern). I
1410     hope this isn't a big issue. There is no effect on runtime performance.
1412     18. Fixed a minor bug in pcretest: if a pattern line was not terminated by a
1413     newline (only possible for the last line of a file) and it was a
1414     pattern that set a locale (followed by /Lsomething), pcretest crashed.
1416     19. Added additional timing features to pcretest. (1) The -tm option now times
1417     matching only, not compiling. (2) Both -t and -tm can be followed, as a
1418     separate command line item, by a number that specifies the number of
1419     repeats to use when timing. The default is 50000; this gives better
1420     precision, but takes uncomfortably long for very large patterns.
1422     20. Extended pcre_study() to be more clever in cases where a branch of a
1423     subpattern has no definite first character. For example, (a*|b*)[cd] would
1424     previously give no result from pcre_study(). Now it recognizes that the
1425     first character must be a, b, c, or d.
1427     21. There was an incorrect error "recursive call could loop indefinitely" if
1428     a subpattern (or the entire pattern) that was being tested for matching an
1429     empty string contained only one non-empty item after a nested subpattern.
1430     For example, the pattern (?>\x{100}*)\d(?R) provoked this error
1431     incorrectly, because the \d was being skipped in the check.
1433     22. The pcretest program now has a new pattern option /B and a command line
1434     option -b, which is equivalent to adding /B to every pattern. This causes
1435     it to show the compiled bytecode, without the additional information that
1436     -d shows. The effect of -d is now the same as -b with -i (and similarly, /D
1437     is the same as /B/I).
1439     23. A new optimization is now able automatically to treat some sequences such
1440     as a*b as a*+b. More specifically, if something simple (such as a character
1441     or a simple class like \d) has an unlimited quantifier, and is followed by
1442     something that cannot possibly match the quantified thing, the quantifier
1443     is automatically "possessified".
1445     24. A recursive reference to a subpattern whose number was greater than 39
1446     went wrong under certain circumstances in UTF-8 mode. This bug could also
1447     have affected the operation of pcre_study().
1449     25. Realized that a little bit of performance could be had by replacing
1450     (c & 0xc0) == 0xc0 with c >= 0xc0 when processing UTF-8 characters.
1452     26. Timing data from pcretest is now shown to 4 decimal places instead of 3.
1454     27. Possessive quantifiers such as a++ were previously implemented by turning
1455     them into atomic groups such as ($>a+). Now they have their own opcodes,
1456     which improves performance. This includes the automatically created ones
1457     from 23 above.
1459     28. A pattern such as (?=(\w+))\1: which simulates an atomic group using a
1460     lookahead was broken if it was not anchored. PCRE was mistakenly expecting
1461     the first matched character to be a colon. This applied both to named and
1462     numbered groups.
1464     29. The ucpinternal.h header file was missing its idempotency #ifdef.
1466     30. I was sent a "project" file called libpcre.a.dev which I understand makes
1467     building PCRE on Windows easier, so I have included it in the distribution.
1469     31. There is now a check in pcretest against a ridiculously large number being
1470     returned by pcre_exec() or pcre_dfa_exec(). If this happens in a /g or /G
1471     loop, the loop is abandoned.
1473     32. Forward references to subpatterns in conditions such as (?(2)...) where
1474     subpattern 2 is defined later cause pcre_compile() to search forwards in
1475     the pattern for the relevant set of parentheses. This search went wrong
1476     when there were unescaped parentheses in a character class, parentheses
1477     escaped with \Q...\E, or parentheses in a #-comment in /x mode.
1479     33. "Subroutine" calls and backreferences were previously restricted to
1480     referencing subpatterns earlier in the regex. This restriction has now
1481     been removed.
1483     34. Added a number of extra features that are going to be in Perl 5.10. On the
1484     whole, these are just syntactic alternatives for features that PCRE had
1485     previously implemented using the Python syntax or my own invention. The
1486     other formats are all retained for compatibility.
1488     (a) Named groups can now be defined as (?<name>...) or (?'name'...) as well
1489     as (?P<name>...). The new forms, as well as being in Perl 5.10, are
1490     also .NET compatible.
1492     (b) A recursion or subroutine call to a named group can now be defined as
1493     (?&name) as well as (?P>name).
1495     (c) A backreference to a named group can now be defined as \k<name> or
1496     \k'name' as well as (?P=name). The new forms, as well as being in Perl
1497     5.10, are also .NET compatible.
1499     (d) A conditional reference to a named group can now use the syntax
1500     (?(<name>) or (?('name') as well as (?(name).
1502     (e) A "conditional group" of the form (?(DEFINE)...) can be used to define
1503     groups (named and numbered) that are never evaluated inline, but can be
1504     called as "subroutines" from elsewhere. In effect, the DEFINE condition
1505     is always false. There may be only one alternative in such a group.
1507     (f) A test for recursion can be given as (?(R1).. or (?(R&name)... as well
1508     as the simple (?(R). The condition is true only if the most recent
1509     recursion is that of the given number or name. It does not search out
1510     through the entire recursion stack.
1512     (g) The escape \gN or \g{N} has been added, where N is a positive or
1513     negative number, specifying an absolute or relative reference.
1515     35. Tidied to get rid of some further signed/unsigned compiler warnings and
1516     some "unreachable code" warnings.
1518     36. Updated the Unicode property tables to Unicode version 5.0.0. Amongst other
1519     things, this adds five new scripts.
1521     37. Perl ignores orphaned \E escapes completely. PCRE now does the same.
1522     There were also incompatibilities regarding the handling of \Q..\E inside
1523     character classes, for example with patterns like [\Qa\E-\Qz\E] where the
1524     hyphen was adjacent to \Q or \E. I hope I've cleared all this up now.
1526     38. Like Perl, PCRE detects when an indefinitely repeated parenthesized group
1527     matches an empty string, and forcibly breaks the loop. There were bugs in
1528     this code in non-simple cases. For a pattern such as ^(a()*)* matched
1529     against aaaa the result was just "a" rather than "aaaa", for example. Two
1530     separate and independent bugs (that affected different cases) have been
1531     fixed.
1533     39. Refactored the code to abolish the use of different opcodes for small
1534     capturing bracket numbers. This is a tidy that I avoided doing when I
1535     removed the limit on the number of capturing brackets for 3.5 back in 2001.
1536     The new approach is not only tidier, it makes it possible to reduce the
1537     memory needed to fix the previous bug (38).
1539     40. Implemented PCRE_NEWLINE_ANY to recognize any of the Unicode newline
1540     sequences (http://unicode.org/unicode/reports/tr18/) as "newline" when
1541     processing dot, circumflex, or dollar metacharacters, or #-comments in /x
1542     mode.
1544     41. Add \R to match any Unicode newline sequence, as suggested in the Unicode
1545     report.
1547     42. Applied patch, originally from Ari Pollak, modified by Google, to allow
1548     copy construction and assignment in the C++ wrapper.
1550     43. Updated pcregrep to support "--newline=any". In the process, I fixed a
1551     couple of bugs that could have given wrong results in the "--newline=crlf"
1552     case.
1554     44. Added a number of casts and did some reorganization of signed/unsigned int
1555     variables following suggestions from Dair Grant. Also renamed the variable
1556     "this" as "item" because it is a C++ keyword.
1558     45. Arranged for dftables to add
1560     #include "pcre_internal.h"
1562     to pcre_chartables.c because without it, gcc 4.x may remove the array
1563     definition from the final binary if PCRE is built into a static library and
1564     dead code stripping is activated.
1566     46. For an unanchored pattern, if a match attempt fails at the start of a
1567     newline sequence, and the newline setting is CRLF or ANY, and the next two
1568     characters are CRLF, advance by two characters instead of one.
1571 nigel 91 Version 6.7 04-Jul-06
1572     ---------------------
1574     1. In order to handle tests when input lines are enormously long, pcretest has
1575     been re-factored so that it automatically extends its buffers when
1576     necessary. The code is crude, but this _is_ just a test program. The
1577     default size has been increased from 32K to 50K.
1579     2. The code in pcre_study() was using the value of the re argument before
1580     testing it for NULL. (Of course, in any sensible call of the function, it
1581     won't be NULL.)
1583     3. The memmove() emulation function in pcre_internal.h, which is used on
1584     systems that lack both memmove() and bcopy() - that is, hardly ever -
1585     was missing a "static" storage class specifier.
1587     4. When UTF-8 mode was not set, PCRE looped when compiling certain patterns
1588     containing an extended class (one that cannot be represented by a bitmap
1589     because it contains high-valued characters or Unicode property items, e.g.
1590     [\pZ]). Almost always one would set UTF-8 mode when processing such a
1591     pattern, but PCRE should not loop if you do not (it no longer does).
1592     [Detail: two cases were found: (a) a repeated subpattern containing an
1593     extended class; (b) a recursive reference to a subpattern that followed a
1594     previous extended class. It wasn't skipping over the extended class
1595     correctly when UTF-8 mode was not set.]
1597     5. A negated single-character class was not being recognized as fixed-length
1598     in lookbehind assertions such as (?<=[^f]), leading to an incorrect
1599     compile error "lookbehind assertion is not fixed length".
1601     6. The RunPerlTest auxiliary script was showing an unexpected difference
1602     between PCRE and Perl for UTF-8 tests. It turns out that it is hard to
1603     write a Perl script that can interpret lines of an input file either as
1604     byte characters or as UTF-8, which is what "perltest" was being required to
1605     do for the non-UTF-8 and UTF-8 tests, respectively. Essentially what you
1606     can't do is switch easily at run time between having the "use utf8;" pragma
1607     or not. In the end, I fudged it by using the RunPerlTest script to insert
1608     "use utf8;" explicitly for the UTF-8 tests.
1610     7. In multiline (/m) mode, PCRE was matching ^ after a terminating newline at
1611     the end of the subject string, contrary to the documentation and to what
1612     Perl does. This was true of both matching functions. Now it matches only at
1613     the start of the subject and immediately after *internal* newlines.
1615     8. A call of pcre_fullinfo() from pcretest to get the option bits was passing
1616     a pointer to an int instead of a pointer to an unsigned long int. This
1617     caused problems on 64-bit systems.
1619     9. Applied a patch from the folks at Google to pcrecpp.cc, to fix "another
1620     instance of the 'standard' template library not being so standard".
1622     10. There was no check on the number of named subpatterns nor the maximum
1623     length of a subpattern name. The product of these values is used to compute
1624     the size of the memory block for a compiled pattern. By supplying a very
1625     long subpattern name and a large number of named subpatterns, the size
1626     computation could be caused to overflow. This is now prevented by limiting
1627     the length of names to 32 characters, and the number of named subpatterns
1628     to 10,000.
1630     11. Subpatterns that are repeated with specific counts have to be replicated in
1631     the compiled pattern. The size of memory for this was computed from the
1632     length of the subpattern and the repeat count. The latter is limited to
1633     65535, but there was no limit on the former, meaning that integer overflow
1634     could in principle occur. The compiled length of a repeated subpattern is
1635     now limited to 30,000 bytes in order to prevent this.
1637     12. Added the optional facility to have named substrings with the same name.
1639     13. Added the ability to use a named substring as a condition, using the
1640     Python syntax: (?(name)yes|no). This overloads (?(R)... and names that
1641     are numbers (not recommended). Forward references are permitted.
1643     14. Added forward references in named backreferences (if you see what I mean).
1645     15. In UTF-8 mode, with the PCRE_DOTALL option set, a quantified dot in the
1646     pattern could run off the end of the subject. For example, the pattern
1647     "(?s)(.{1,5})"8 did this with the subject "ab".
1649     16. If PCRE_DOTALL or PCRE_MULTILINE were set, pcre_dfa_exec() behaved as if
1650     PCRE_CASELESS was set when matching characters that were quantified with ?
1651     or *.
1653     17. A character class other than a single negated character that had a minimum
1654     but no maximum quantifier - for example [ab]{6,} - was not handled
1655     correctly by pce_dfa_exec(). It would match only one character.
1657     18. A valid (though odd) pattern that looked like a POSIX character
1658     class but used an invalid character after [ (for example [[,abc,]]) caused
1659     pcre_compile() to give the error "Failed: internal error: code overflow" or
1660     in some cases to crash with a glibc free() error. This could even happen if
1661     the pattern terminated after [[ but there just happened to be a sequence of
1662     letters, a binary zero, and a closing ] in the memory that followed.
1664     19. Perl's treatment of octal escapes in the range \400 to \777 has changed
1665     over the years. Originally (before any Unicode support), just the bottom 8
1666     bits were taken. Thus, for example, \500 really meant \100. Nowadays the
1667     output from "man perlunicode" includes this:
1669     The regular expression compiler produces polymorphic opcodes. That
1670     is, the pattern adapts to the data and automatically switches to
1671     the Unicode character scheme when presented with Unicode data--or
1672     instead uses a traditional byte scheme when presented with byte
1673     data.
1675     Sadly, a wide octal escape does not cause a switch, and in a string with
1676     no other multibyte characters, these octal escapes are treated as before.
1677     Thus, in Perl, the pattern /\500/ actually matches \100 but the pattern
1678     /\500|\x{1ff}/ matches \500 or \777 because the whole thing is treated as a
1679     Unicode string.
1681     I have not perpetrated such confusion in PCRE. Up till now, it took just
1682     the bottom 8 bits, as in old Perl. I have now made octal escapes with
1683     values greater than \377 illegal in non-UTF-8 mode. In UTF-8 mode they
1684     translate to the appropriate multibyte character.
1686     29. Applied some refactoring to reduce the number of warnings from Microsoft
1687     and Borland compilers. This has included removing the fudge introduced
1688     seven years ago for the OS/2 compiler (see 2.02/2 below) because it caused
1689     a warning about an unused variable.
1691     21. PCRE has not included VT (character 0x0b) in the set of whitespace
1692     characters since release 4.0, because Perl (from release 5.004) does not.
1693     [Or at least, is documented not to: some releases seem to be in conflict
1694     with the documentation.] However, when a pattern was studied with
1695     pcre_study() and all its branches started with \s, PCRE still included VT
1696     as a possible starting character. Of course, this did no harm; it just
1697     caused an unnecessary match attempt.
1699     22. Removed a now-redundant internal flag bit that recorded the fact that case
1700     dependency changed within the pattern. This was once needed for "required
1701     byte" processing, but is no longer used. This recovers a now-scarce options
1702     bit. Also moved the least significant internal flag bit to the most-
1703     significant bit of the word, which was not previously used (hangover from
1704     the days when it was an int rather than a uint) to free up another bit for
1705     the future.
1707     23. Added support for CRLF line endings as well as CR and LF. As well as the
1708     default being selectable at build time, it can now be changed at runtime
1709     via the PCRE_NEWLINE_xxx flags. There are now options for pcregrep to
1710     specify that it is scanning data with non-default line endings.
1712     24. Changed the definition of CXXLINK to make it agree with the definition of
1713     LINK in the Makefile, by replacing LDFLAGS to CXXFLAGS.
1715     25. Applied Ian Taylor's patches to avoid using another stack frame for tail
1716     recursions. This makes a big different to stack usage for some patterns.
1718     26. If a subpattern containing a named recursion or subroutine reference such
1719     as (?P>B) was quantified, for example (xxx(?P>B)){3}, the calculation of
1720     the space required for the compiled pattern went wrong and gave too small a
1721     value. Depending on the environment, this could lead to "Failed: internal
1722     error: code overflow at offset 49" or "glibc detected double free or
1723     corruption" errors.
1725     27. Applied patches from Google (a) to support the new newline modes and (b) to
1726     advance over multibyte UTF-8 characters in GlobalReplace.
1728     28. Change free() to pcre_free() in pcredemo.c. Apparently this makes a
1729     difference for some implementation of PCRE in some Windows version.
1731     29. Added some extra testing facilities to pcretest:
1733     \q<number> in a data line sets the "match limit" value
1734     \Q<number> in a data line sets the "match recursion limt" value
1735     -S <number> sets the stack size, where <number> is in megabytes
1737     The -S option isn't available for Windows.
1740 nigel 89 Version 6.6 06-Feb-06
1741     ---------------------
1743     1. Change 16(a) for 6.5 broke things, because PCRE_DATA_SCOPE was not defined
1744     in pcreposix.h. I have copied the definition from pcre.h.
1746     2. Change 25 for 6.5 broke compilation in a build directory out-of-tree
1747     because pcre.h is no longer a built file.
1749     3. Added Jeff Friedl's additional debugging patches to pcregrep. These are
1750     not normally included in the compiled code.
1753 nigel 87 Version 6.5 01-Feb-06
1754     ---------------------
1756     1. When using the partial match feature with pcre_dfa_exec(), it was not
1757     anchoring the second and subsequent partial matches at the new starting
1758     point. This could lead to incorrect results. For example, with the pattern
1759     /1234/, partially matching against "123" and then "a4" gave a match.
1761     2. Changes to pcregrep:
1763     (a) All non-match returns from pcre_exec() were being treated as failures
1764     to match the line. Now, unless the error is PCRE_ERROR_NOMATCH, an
1765     error message is output. Some extra information is given for the
1767     probably the only errors that are likely to be caused by users (by
1768     specifying a regex that has nested indefinite repeats, for instance).
1769     If there are more than 20 of these errors, pcregrep is abandoned.
1771     (b) A binary zero was treated as data while matching, but terminated the
1772     output line if it was written out. This has been fixed: binary zeroes
1773     are now no different to any other data bytes.
1775     (c) Whichever of the LC_ALL or LC_CTYPE environment variables is set is
1776     used to set a locale for matching. The --locale=xxxx long option has
1777     been added (no short equivalent) to specify a locale explicitly on the
1778     pcregrep command, overriding the environment variables.
1780     (d) When -B was used with -n, some line numbers in the output were one less
1781     than they should have been.
1783     (e) Added the -o (--only-matching) option.
1785     (f) If -A or -C was used with -c (count only), some lines of context were
1786     accidentally printed for the final match.
1788     (g) Added the -H (--with-filename) option.
1790     (h) The combination of options -rh failed to suppress file names for files
1791     that were found from directory arguments.
1793     (i) Added the -D (--devices) and -d (--directories) options.
1795     (j) Added the -F (--fixed-strings) option.
1797     (k) Allow "-" to be used as a file name for -f as well as for a data file.
1799     (l) Added the --colo(u)r option.
1801     (m) Added Jeffrey Friedl's -S testing option, but within #ifdefs so that it
1802     is not present by default.
1804     3. A nasty bug was discovered in the handling of recursive patterns, that is,
1805     items such as (?R) or (?1), when the recursion could match a number of
1806     alternatives. If it matched one of the alternatives, but subsequently,
1807     outside the recursion, there was a failure, the code tried to back up into
1808     the recursion. However, because of the way PCRE is implemented, this is not
1809     possible, and the result was an incorrect result from the match.
1811     In order to prevent this happening, the specification of recursion has
1812     been changed so that all such subpatterns are automatically treated as
1813     atomic groups. Thus, for example, (?R) is treated as if it were (?>(?R)).
1815     4. I had overlooked the fact that, in some locales, there are characters for
1816     which isalpha() is true but neither isupper() nor islower() are true. In
1817     the fr_FR locale, for instance, the \xAA and \xBA characters (ordmasculine
1818     and ordfeminine) are like this. This affected the treatment of \w and \W
1819     when they appeared in character classes, but not when they appeared outside
1820     a character class. The bit map for "word" characters is now created
1821     separately from the results of isalnum() instead of just taking it from the
1822     upper, lower, and digit maps. (Plus the underscore character, of course.)
1824     5. The above bug also affected the handling of POSIX character classes such as
1825     [[:alpha:]] and [[:alnum:]]. These do not have their own bit maps in PCRE's
1826     permanent tables. Instead, the bit maps for such a class were previously
1827     created as the appropriate unions of the upper, lower, and digit bitmaps.
1828     Now they are created by subtraction from the [[:word:]] class, which has
1829     its own bitmap.
1831     6. The [[:blank:]] character class matches horizontal, but not vertical space.
1832     It is created by subtracting the vertical space characters (\x09, \x0a,
1833     \x0b, \x0c) from the [[:space:]] bitmap. Previously, however, the
1834     subtraction was done in the overall bitmap for a character class, meaning
1835     that a class such as [\x0c[:blank:]] was incorrect because \x0c would not
1836     be recognized. This bug has been fixed.
1838     7. Patches from the folks at Google:
1840     (a) pcrecpp.cc: "to handle a corner case that may or may not happen in
1841     real life, but is still worth protecting against".
1843     (b) pcrecpp.cc: "corrects a bug when negative radixes are used with
1844     regular expressions".
1846     (c) pcre_scanner.cc: avoid use of std::count() because not all systems
1847     have it.
1849     (d) Split off pcrecpparg.h from pcrecpp.h and had the former built by
1850     "configure" and the latter not, in order to fix a problem somebody had
1851     with compiling the Arg class on HP-UX.
1853     (e) Improve the error-handling of the C++ wrapper a little bit.
1855     (f) New tests for checking recursion limiting.
1857     8. The pcre_memmove() function, which is used only if the environment does not
1858     have a standard memmove() function (and is therefore rarely compiled),
1859     contained two bugs: (a) use of int instead of size_t, and (b) it was not
1860     returning a result (though PCRE never actually uses the result).
1862     9. In the POSIX regexec() interface, if nmatch is specified as a ridiculously
1863     large number - greater than INT_MAX/(3*sizeof(int)) - REG_ESPACE is
1864     returned instead of calling malloc() with an overflowing number that would
1865     most likely cause subsequent chaos.
1867     10. The debugging option of pcretest was not showing the NO_AUTO_CAPTURE flag.
1869     11. The POSIX flag REG_NOSUB is now supported. When a pattern that was compiled
1870     with this option is matched, the nmatch and pmatch options of regexec() are
1871     ignored.
1873     12. Added REG_UTF8 to the POSIX interface. This is not defined by POSIX, but is
1874     provided in case anyone wants to the the POSIX interface with UTF-8
1875     strings.
1877     13. Added CXXLDFLAGS to the Makefile parameters to provide settings only on the
1878     C++ linking (needed for some HP-UX environments).
1880     14. Avoid compiler warnings in get_ucpname() when compiled without UCP support
1881     (unused parameter) and in the pcre_printint() function (omitted "default"
1882     switch label when the default is to do nothing).
1884     15. Added some code to make it possible, when PCRE is compiled as a C++
1885     library, to replace subject pointers for pcre_exec() with a smart pointer
1886     class, thus making it possible to process discontinuous strings.
1888     16. The two macros PCRE_EXPORT and PCRE_DATA_SCOPE are confusing, and perform
1889     much the same function. They were added by different people who were trying
1890     to make PCRE easy to compile on non-Unix systems. It has been suggested
1891     that PCRE_EXPORT be abolished now that there is more automatic apparatus
1892     for compiling on Windows systems. I have therefore replaced it with
1893     PCRE_DATA_SCOPE. This is set automatically for Windows; if not set it
1894     defaults to "extern" for C or "extern C" for C++, which works fine on
1895     Unix-like systems. It is now possible to override the value of PCRE_DATA_
1896     SCOPE with something explicit in config.h. In addition:
1898     (a) pcreposix.h still had just "extern" instead of either of these macros;
1899     I have replaced it with PCRE_DATA_SCOPE.
1901     (b) Functions such as _pcre_xclass(), which are internal to the library,
1902     but external in the C sense, all had PCRE_EXPORT in their definitions.
1903     This is apparently wrong for the Windows case, so I have removed it.
1904     (It makes no difference on Unix-like systems.)
1906     17. Added a new limit, MATCH_LIMIT_RECURSION, which limits the depth of nesting
1907     of recursive calls to match(). This is different to MATCH_LIMIT because
1908     that limits the total number of calls to match(), not all of which increase
1909     the depth of recursion. Limiting the recursion depth limits the amount of
1910     stack (or heap if NO_RECURSE is set) that is used. The default can be set
1911     when PCRE is compiled, and changed at run time. A patch from Google adds
1912     this functionality to the C++ interface.
1914     18. Changes to the handling of Unicode character properties:
1916     (a) Updated the table to Unicode 4.1.0.
1918     (b) Recognize characters that are not in the table as "Cn" (undefined).
1920     (c) I revised the way the table is implemented to a much improved format
1921     which includes recognition of ranges. It now supports the ranges that
1922     are defined in UnicodeData.txt, and it also amalgamates other
1923     characters into ranges. This has reduced the number of entries in the
1924     table from around 16,000 to around 3,000, thus reducing its size
1925     considerably. I realized I did not need to use a tree structure after
1926     all - a binary chop search is just as efficient. Having reduced the
1927     number of entries, I extended their size from 6 bytes to 8 bytes to
1928     allow for more data.
1930     (d) Added support for Unicode script names via properties such as \p{Han}.
1932     19. In UTF-8 mode, a backslash followed by a non-Ascii character was not
1933     matching that character.
1935     20. When matching a repeated Unicode property with a minimum greater than zero,
1936     (for example \pL{2,}), PCRE could look past the end of the subject if it
1937     reached it while seeking the minimum number of characters. This could
1938     happen only if some of the characters were more than one byte long, because
1939     there is a check for at least the minimum number of bytes.
1941     21. Refactored the implementation of \p and \P so as to be more general, to
1942     allow for more different types of property in future. This has changed the
1943     compiled form incompatibly. Anybody with saved compiled patterns that use
1944     \p or \P will have to recompile them.
1946     22. Added "Any" and "L&" to the supported property types.
1948     23. Recognize \x{...} as a code point specifier, even when not in UTF-8 mode,
1949     but give a compile time error if the value is greater than 0xff.
1951     24. The man pages for pcrepartial, pcreprecompile, and pcre_compile2 were
1952     accidentally not being installed or uninstalled.
1954     25. The pcre.h file was built from pcre.h.in, but the only changes that were
1955     made were to insert the current release number. This seemed silly, because
1956     it made things harder for people building PCRE on systems that don't run
1957     "configure". I have turned pcre.h into a distributed file, no longer built
1958     by "configure", with the version identification directly included. There is
1959     no longer a pcre.h.in file.
1961     However, this change necessitated a change to the pcre-config script as
1962     well. It is built from pcre-config.in, and one of the substitutions was the
1963     release number. I have updated configure.ac so that ./configure now finds
1964     the release number by grepping pcre.h.
1966     26. Added the ability to run the tests under valgrind.
1969 nigel 85 Version 6.4 05-Sep-05
1970     ---------------------
1972     1. Change 6.0/10/(l) to pcregrep introduced a bug that caused separator lines
1973     "--" to be printed when multiple files were scanned, even when none of the
1974     -A, -B, or -C options were used. This is not compatible with Gnu grep, so I
1975     consider it to be a bug, and have restored the previous behaviour.
1977     2. A couple of code tidies to get rid of compiler warnings.
1979     3. The pcretest program used to cheat by referring to symbols in the library
1980     whose names begin with _pcre_. These are internal symbols that are not
1981     really supposed to be visible externally, and in some environments it is
1982     possible to suppress them. The cheating is now confined to including
1983     certain files from the library's source, which is a bit cleaner.
1985     4. Renamed pcre.in as pcre.h.in to go with pcrecpp.h.in; it also makes the
1986     file's purpose clearer.
1988     5. Reorganized pcre_ucp_findchar().
1991 nigel 83 Version 6.3 15-Aug-05
1992     ---------------------
1994     1. The file libpcre.pc.in did not have general read permission in the tarball.
1996     2. There were some problems when building without C++ support:
1998     (a) If C++ support was not built, "make install" and "make test" still
1999     tried to test it.
2001     (b) There were problems when the value of CXX was explicitly set. Some
2002     changes have been made to try to fix these, and ...
2004     (c) --disable-cpp can now be used to explicitly disable C++ support.
2006     (d) The use of @CPP_OBJ@ directly caused a blank line preceded by a
2007     backslash in a target when C++ was disabled. This confuses some
2008     versions of "make", apparently. Using an intermediate variable solves
2009     this. (Same for CPP_LOBJ.)
2011     3. $(LINK_FOR_BUILD) now includes $(CFLAGS_FOR_BUILD) and $(LINK)
2012     (non-Windows) now includes $(CFLAGS) because these flags are sometimes
2013     necessary on certain architectures.
2015     4. Added a setting of -export-symbols-regex to the link command to remove
2016     those symbols that are exported in the C sense, but actually are local
2017     within the library, and not documented. Their names all begin with
2018     "_pcre_". This is not a perfect job, because (a) we have to except some
2019     symbols that pcretest ("illegally") uses, and (b) the facility isn't always
2020     available (and never for static libraries). I have made a note to try to
2021     find a way round (a) in the future.
2024 nigel 81 Version 6.2 01-Aug-05
2025     ---------------------
2027     1. There was no test for integer overflow of quantifier values. A construction
2028     such as {1111111111111111} would give undefined results. What is worse, if
2029     a minimum quantifier for a parenthesized subpattern overflowed and became
2030     negative, the calculation of the memory size went wrong. This could have
2031     led to memory overwriting.
2033     2. Building PCRE using VPATH was broken. Hopefully it is now fixed.
2035     3. Added "b" to the 2nd argument of fopen() in dftables.c, for non-Unix-like
2036     operating environments where this matters.
2038     4. Applied Giuseppe Maxia's patch to add additional features for controlling
2039     PCRE options from within the C++ wrapper.
2041     5. Named capturing subpatterns were not being correctly counted when a pattern
2042     was compiled. This caused two problems: (a) If there were more than 100
2043     such subpatterns, the calculation of the memory needed for the whole
2044     compiled pattern went wrong, leading to an overflow error. (b) Numerical
2045     back references of the form \12, where the number was greater than 9, were
2046     not recognized as back references, even though there were sufficient
2047     previous subpatterns.
2049     6. Two minor patches to pcrecpp.cc in order to allow it to compile on older
2050     versions of gcc, e.g. 2.95.4.
2053 nigel 79 Version 6.1 21-Jun-05
2054     ---------------------
2056     1. There was one reference to the variable "posix" in pcretest.c that was not
2057     surrounded by "#if !defined NOPOSIX".
2059     2. Make it possible to compile pcretest without DFA support, UTF8 support, or
2060     the cross-check on the old pcre_info() function, for the benefit of the
2061     cut-down version of PCRE that is currently imported into Exim.
2063     3. A (silly) pattern starting with (?i)(?-i) caused an internal space
2064     allocation error. I've done the easy fix, which wastes 2 bytes for sensible
2065     patterns that start (?i) but I don't think that matters. The use of (?i) is
2066     just an example; this all applies to the other options as well.
2068     4. Since libtool seems to echo the compile commands it is issuing, the output
2069     from "make" can be reduced a bit by putting "@" in front of each libtool
2070     compile command.
2072     5. Patch from the folks at Google for configure.in to be a bit more thorough
2073     in checking for a suitable C++ installation before trying to compile the
2074     C++ stuff. This should fix a reported problem when a compiler was present,
2075     but no suitable headers.
2077     6. The man pages all had just "PCRE" as their title. I have changed them to
2078     be the relevant file name. I have also arranged that these names are
2079     retained in the file doc/pcre.txt, which is a concatenation in text format
2080     of all the man pages except the little individual ones for each function.
2082     7. The NON-UNIX-USE file had not been updated for the different set of source
2083     files that come with release 6. I also added a few comments about the C++
2084     wrapper.
2087 nigel 77 Version 6.0 07-Jun-05
2088     ---------------------
2090     1. Some minor internal re-organization to help with my DFA experiments.
2092     2. Some missing #ifdef SUPPORT_UCP conditionals in pcretest and printint that
2093     didn't matter for the library itself when fully configured, but did matter
2094     when compiling without UCP support, or within Exim, where the ucp files are
2095     not imported.
2097     3. Refactoring of the library code to split up the various functions into
2098     different source modules. The addition of the new DFA matching code (see
2099     below) to a single monolithic source would have made it really too
2100     unwieldy, quite apart from causing all the code to be include in a
2101     statically linked application, when only some functions are used. This is
2102     relevant even without the DFA addition now that patterns can be compiled in
2103     one application and matched in another.
2105     The downside of splitting up is that there have to be some external
2106     functions and data tables that are used internally in different modules of
2107     the library but which are not part of the API. These have all had their
2108     names changed to start with "_pcre_" so that they are unlikely to clash
2109     with other external names.
2111     4. Added an alternate matching function, pcre_dfa_exec(), which matches using
2112     a different (DFA) algorithm. Although it is slower than the original
2113     function, it does have some advantages for certain types of matching
2114     problem.
2116     5. Upgrades to pcretest in order to test the features of pcre_dfa_exec(),
2117     including restarting after a partial match.
2119     6. A patch for pcregrep that defines INVALID_FILE_ATTRIBUTES if it is not
2120     defined when compiling for Windows was sent to me. I have put it into the
2121     code, though I have no means of testing or verifying it.
2123     7. Added the pcre_refcount() auxiliary function.
2125     8. Added the PCRE_FIRSTLINE option. This constrains an unanchored pattern to
2126     match before or at the first newline in the subject string. In pcretest,
2127     the /f option on a pattern can be used to set this.
2129     9. A repeated \w when used in UTF-8 mode with characters greater than 256
2130     would behave wrongly. This has been present in PCRE since release 4.0.
2132     10. A number of changes to the pcregrep command:
2134     (a) Refactored how -x works; insert ^(...)$ instead of setting
2135     PCRE_ANCHORED and checking the length, in preparation for adding
2136     something similar for -w.
2138     (b) Added the -w (match as a word) option.
2140     (c) Refactored the way lines are read and buffered so as to have more
2141     than one at a time available.
2143     (d) Implemented a pcregrep test script.
2145     (e) Added the -M (multiline match) option. This allows patterns to match
2146     over several lines of the subject. The buffering ensures that at least
2147     8K, or the rest of the document (whichever is the shorter) is available
2148     for matching (and similarly the previous 8K for lookbehind assertions).
2150     (f) Changed the --help output so that it now says
2152     -w, --word-regex(p)
2154     instead of two lines, one with "regex" and the other with "regexp"
2155     because that confused at least one person since the short forms are the
2156     same. (This required a bit of code, as the output is generated
2157     automatically from a table. It wasn't just a text change.)
2159     (g) -- can be used to terminate pcregrep options if the next thing isn't an
2160     option but starts with a hyphen. Could be a pattern or a path name
2161     starting with a hyphen, for instance.
2163     (h) "-" can be given as a file name to represent stdin.
2165     (i) When file names are being printed, "(standard input)" is used for
2166     the standard input, for compatibility with GNU grep. Previously
2167     "<stdin>" was used.
2169     (j) The option --label=xxx can be used to supply a name to be used for
2170     stdin when file names are being printed. There is no short form.
2172     (k) Re-factored the options decoding logic because we are going to add
2173     two more options that take data. Such options can now be given in four
2174     different ways, e.g. "-fname", "-f name", "--file=name", "--file name".
2176     (l) Added the -A, -B, and -C options for requesting that lines of context
2177     around matches be printed.
2179     (m) Added the -L option to print the names of files that do not contain
2180     any matching lines, that is, the complement of -l.
2182     (n) The return code is 2 if any file cannot be opened, but pcregrep does
2183     continue to scan other files.
2185     (o) The -s option was incorrectly implemented. For compatibility with other
2186     greps, it now suppresses the error message for a non-existent or non-
2187     accessible file (but not the return code). There is a new option called
2188     -q that suppresses the output of matching lines, which was what -s was
2189     previously doing.
2191     (p) Added --include and --exclude options to specify files for inclusion
2192     and exclusion when recursing.
2194     11. The Makefile was not using the Autoconf-supported LDFLAGS macro properly.
2195     Hopefully, it now does.
2197     12. Missing cast in pcre_study().
2199     13. Added an "uninstall" target to the makefile.
2201     14. Replaced "extern" in the function prototypes in Makefile.in with
2202     "PCRE_DATA_SCOPE", which defaults to 'extern' or 'extern "C"' in the Unix
2203     world, but is set differently for Windows.
2205     15. Added a second compiling function called pcre_compile2(). The only
2206     difference is that it has an extra argument, which is a pointer to an
2207     integer error code. When there is a compile-time failure, this is set
2208     non-zero, in addition to the error test pointer being set to point to an
2209     error message. The new argument may be NULL if no error number is required
2210     (but then you may as well call pcre_compile(), which is now just a
2211     wrapper). This facility is provided because some applications need a
2212     numeric error indication, but it has also enabled me to tidy up the way
2213     compile-time errors are handled in the POSIX wrapper.
2215     16. Added VPATH=.libs to the makefile; this should help when building with one
2216     prefix path and installing with another. (Or so I'm told by someone who
2217     knows more about this stuff than I do.)
2219     17. Added a new option, REG_DOTALL, to the POSIX function regcomp(). This
2220     passes PCRE_DOTALL to the pcre_compile() function, making the "." character
2221     match everything, including newlines. This is not POSIX-compatible, but
2222     somebody wanted the feature. From pcretest it can be activated by using
2223     both the P and the s flags.
2225     18. AC_PROG_LIBTOOL appeared twice in Makefile.in. Removed one.
2227     19. libpcre.pc was being incorrectly installed as executable.
2229     20. A couple of places in pcretest check for end-of-line by looking for '\n';
2230     it now also looks for '\r' so that it will work unmodified on Windows.
2232     21. Added Google's contributed C++ wrapper to the distribution.
2234     22. Added some untidy missing memory free() calls in pcretest, to keep
2235     Electric Fence happy when testing.
2239 nigel 75 Version 5.0 13-Sep-04
2240     ---------------------
2242     1. Internal change: literal characters are no longer packed up into items
2243     containing multiple characters in a single byte-string. Each character
2244     is now matched using a separate opcode. However, there may be more than one
2245     byte in the character in UTF-8 mode.
2247     2. The pcre_callout_block structure has two new fields: pattern_position and
2248     next_item_length. These contain the offset in the pattern to the next match
2249     item, and its length, respectively.
2251     3. The PCRE_AUTO_CALLOUT option for pcre_compile() requests the automatic
2252     insertion of callouts before each pattern item. Added the /C option to
2253     pcretest to make use of this.
2255     4. On the advice of a Windows user, the lines
2257     #if defined(_WIN32) || defined(WIN32)
2258     _setmode( _fileno( stdout ), 0x8000 );
2259     #endif /* defined(_WIN32) || defined(WIN32) */
2261     have been added to the source of pcretest. This apparently does useful
2262     magic in relation to line terminators.
2264     5. Changed "r" and "w" in the calls to fopen() in pcretest to "rb" and "wb"
2265     for the benefit of those environments where the "b" makes a difference.
2267     6. The icc compiler has the same options as gcc, but "configure" doesn't seem
2268     to know about it. I have put a hack into configure.in that adds in code
2269     to set GCC=yes if CC=icc. This seems to end up at a point in the
2270     generated configure script that is early enough to affect the setting of
2271     compiler options, which is what is needed, but I have no means of testing
2272     whether it really works. (The user who reported this had patched the
2273     generated configure script, which of course I cannot do.)
2275     LATER: After change 22 below (new libtool files), the configure script
2276     seems to know about icc (and also ecc). Therefore, I have commented out
2277     this hack in configure.in.
2279     7. Added support for pkg-config (2 patches were sent in).
2281     8. Negated POSIX character classes that used a combination of internal tables
2282     were completely broken. These were [[:^alpha:]], [[:^alnum:]], and
2283     [[:^ascii]]. Typically, they would match almost any characters. The other
2284     POSIX classes were not broken in this way.
2286     9. Matching the pattern "\b.*?" against "ab cd", starting at offset 1, failed
2287     to find the match, as PCRE was deluded into thinking that the match had to
2288     start at the start point or following a newline. The same bug applied to
2289     patterns with negative forward assertions or any backward assertions
2290     preceding ".*" at the start, unless the pattern required a fixed first
2291     character. This was a failing pattern: "(?!.bcd).*". The bug is now fixed.
2293     10. In UTF-8 mode, when moving forwards in the subject after a failed match
2294     starting at the last subject character, bytes beyond the end of the subject
2295     string were read.
2297     11. Renamed the variable "class" as "classbits" to make life easier for C++
2298     users. (Previously there was a macro definition, but it apparently wasn't
2299     enough.)
2301     12. Added the new field "tables" to the extra data so that tables can be passed
2302     in at exec time, or the internal tables can be re-selected. This allows
2303     a compiled regex to be saved and re-used at a later time by a different
2304     program that might have everything at different addresses.
2306     13. Modified the pcre-config script so that, when run on Solaris, it shows a
2307     -R library as well as a -L library.
2309     14. The debugging options of pcretest (-d on the command line or D on a
2310     pattern) showed incorrect output for anything following an extended class
2311     that contained multibyte characters and which was followed by a quantifier.
2313     15. Added optional support for general category Unicode character properties
2314     via the \p, \P, and \X escapes. Unicode property support implies UTF-8
2315     support. It adds about 90K to the size of the library. The meanings of the
2316     inbuilt class escapes such as \d and \s have NOT been changed.
2318     16. Updated pcredemo.c to include calls to free() to release the memory for the
2319     compiled pattern.
2321     17. The generated file chartables.c was being created in the source directory
2322     instead of in the building directory. This caused the build to fail if the
2323     source directory was different from the building directory, and was
2324     read-only.
2326     18. Added some sample Win commands from Mark Tetrode into the NON-UNIX-USE
2327     file. No doubt somebody will tell me if they don't make sense... Also added
2328     Dan Mooney's comments about building on OpenVMS.
2330     19. Added support for partial matching via the PCRE_PARTIAL option for
2331     pcre_exec() and the \P data escape in pcretest.
2333     20. Extended pcretest with 3 new pattern features:
2335     (i) A pattern option of the form ">rest-of-line" causes pcretest to
2336     write the compiled pattern to the file whose name is "rest-of-line".
2337     This is a straight binary dump of the data, with the saved pointer to
2338     the character tables forced to be NULL. The study data, if any, is
2339     written too. After writing, pcretest reads a new pattern.
2341     (ii) If, instead of a pattern, "<rest-of-line" is given, pcretest reads a
2342     compiled pattern from the given file. There must not be any
2343     occurrences of "<" in the file name (pretty unlikely); if there are,
2344     pcretest will instead treat the initial "<" as a pattern delimiter.
2345     After reading in the pattern, pcretest goes on to read data lines as
2346     usual.
2348     (iii) The F pattern option causes pcretest to flip the bytes in the 32-bit
2349     and 16-bit fields in a compiled pattern, to simulate a pattern that
2350     was compiled on a host of opposite endianness.
2352     21. The pcre-exec() function can now cope with patterns that were compiled on
2353     hosts of opposite endianness, with this restriction:
2355     As for any compiled expression that is saved and used later, the tables
2356     pointer field cannot be preserved; the extra_data field in the arguments
2357     to pcre_exec() should be used to pass in a tables address if a value
2358     other than the default internal tables were used at compile time.
2360     22. Calling pcre_exec() with a negative value of the "ovecsize" parameter is
2361     now diagnosed as an error. Previously, most of the time, a negative number
2362     would have been treated as zero, but if in addition "ovector" was passed as
2363     NULL, a crash could occur.
2365     23. Updated the files ltmain.sh, config.sub, config.guess, and aclocal.m4 with
2366     new versions from the libtool 1.5 distribution (the last one is a copy of
2367     a file called libtool.m4). This seems to have fixed the need to patch
2368     "configure" to support Darwin 1.3 (which I used to do). However, I still
2369     had to patch ltmain.sh to ensure that ${SED} is set (it isn't on my
2370     workstation).
2372     24. Changed the PCRE licence to be the more standard "BSD" licence.
2375 nigel 73 Version 4.5 01-Dec-03
2376     ---------------------
2378     1. There has been some re-arrangement of the code for the match() function so
2379     that it can be compiled in a version that does not call itself recursively.
2380     Instead, it keeps those local variables that need separate instances for
2381     each "recursion" in a frame on the heap, and gets/frees frames whenever it
2382     needs to "recurse". Keeping track of where control must go is done by means
2383     of setjmp/longjmp. The whole thing is implemented by a set of macros that
2384     hide most of the details from the main code, and operates only if
2385     NO_RECURSE is defined while compiling pcre.c. If PCRE is built using the
2386     "configure" mechanism, "--disable-stack-for-recursion" turns on this way of
2387     operating.
2389     To make it easier for callers to provide specially tailored get/free
2390     functions for this usage, two new functions, pcre_stack_malloc, and
2391     pcre_stack_free, are used. They are always called in strict stacking order,
2392     and the size of block requested is always the same.
2394     The PCRE_CONFIG_STACKRECURSE info parameter can be used to find out whether
2395     PCRE has been compiled to use the stack or the heap for recursion. The
2396     -C option of pcretest uses this to show which version is compiled.
2398     A new data escape \S, is added to pcretest; it causes the amounts of store
2399     obtained and freed by both kinds of malloc/free at match time to be added
2400     to the output.
2402     2. Changed the locale test to use "fr_FR" instead of "fr" because that's
2403     what's available on my current Linux desktop machine.
2405     3. When matching a UTF-8 string, the test for a valid string at the start has
2406     been extended. If start_offset is not zero, PCRE now checks that it points
2407     to a byte that is the start of a UTF-8 character. If not, it returns
2408     PCRE_ERROR_BADUTF8_OFFSET (-11). Note: the whole string is still checked;
2409     this is necessary because there may be backward assertions in the pattern.
2410     When matching the same subject several times, it may save resources to use
2411     PCRE_NO_UTF8_CHECK on all but the first call if the string is long.
2413     4. The code for checking the validity of UTF-8 strings has been tightened so
2414     that it rejects (a) strings containing 0xfe or 0xff bytes and (b) strings
2415     containing "overlong sequences".
2417     5. Fixed a bug (appearing twice) that I could not find any way of exploiting!
2418     I had written "if ((digitab[*p++] && chtab_digit) == 0)" where the "&&"
2419     should have been "&", but it just so happened that all the cases this let
2420     through by mistake were picked up later in the function.
2422     6. I had used a variable called "isblank" - this is a C99 function, causing
2423     some compilers to warn. To avoid this, I renamed it (as "blankclass").
2425     7. Cosmetic: (a) only output another newline at the end of pcretest if it is
2426     prompting; (b) run "./pcretest /dev/null" at the start of the test script
2427     so the version is shown; (c) stop "make test" echoing "./RunTest".
2429     8. Added patches from David Burgess to enable PCRE to run on EBCDIC systems.
2431     9. The prototype for memmove() for systems that don't have it was using
2432     size_t, but the inclusion of the header that defines size_t was later. I've
2433     moved the #includes for the C headers earlier to avoid this.
2435     10. Added some adjustments to the code to make it easier to compiler on certain
2436     special systems:
2438     (a) Some "const" qualifiers were missing.
2439     (b) Added the macro EXPORT before all exported functions; by default this
2440     is defined to be empty.
2441     (c) Changed the dftables auxiliary program (that builds chartables.c) so
2442     that it reads its output file name as an argument instead of writing
2443     to the standard output and assuming this can be redirected.
2445     11. In UTF-8 mode, if a recursive reference (e.g. (?1)) followed a character
2446     class containing characters with values greater than 255, PCRE compilation
2447     went into a loop.
2449     12. A recursive reference to a subpattern that was within another subpattern
2450     that had a minimum quantifier of zero caused PCRE to crash. For example,
2451     (x(y(?2))z)? provoked this bug with a subject that got as far as the
2452     recursion. If the recursively-called subpattern itself had a zero repeat,
2453     that was OK.
2455     13. In pcretest, the buffer for reading a data line was set at 30K, but the
2456     buffer into which it was copied (for escape processing) was still set at
2457     1024, so long lines caused crashes.
2459     14. A pattern such as /[ab]{1,3}+/ failed to compile, giving the error
2460     "internal error: code overflow...". This applied to any character class
2461     that was followed by a possessive quantifier.
2463     15. Modified the Makefile to add libpcre.la as a prerequisite for
2464     libpcreposix.la because I was told this is needed for a parallel build to
2465     work.
2467     16. If a pattern that contained .* following optional items at the start was
2468     studied, the wrong optimizing data was generated, leading to matching
2469     errors. For example, studying /[ab]*.*c/ concluded, erroneously, that any
2470     matching string must start with a or b or c. The correct conclusion for
2471     this pattern is that a match can start with any character.
2474 nigel 71 Version 4.4 13-Aug-03
2475     ---------------------
2477     1. In UTF-8 mode, a character class containing characters with values between
2478     127 and 255 was not handled correctly if the compiled pattern was studied.
2479     In fixing this, I have also improved the studying algorithm for such
2480     classes (slightly).
2482     2. Three internal functions had redundant arguments passed to them. Removal
2483     might give a very teeny performance improvement.
2485     3. Documentation bug: the value of the capture_top field in a callout is *one
2486     more than* the number of the hightest numbered captured substring.
2488     4. The Makefile linked pcretest and pcregrep with -lpcre, which could result
2489     in incorrectly linking with a previously installed version. They now link
2490     explicitly with libpcre.la.
2492     5. configure.in no longer needs to recognize Cygwin specially.
2494     6. A problem in pcre.in for Windows platforms is fixed.
2496     7. If a pattern was successfully studied, and the -d (or /D) flag was given to
2497     pcretest, it used to include the size of the study block as part of its
2498     output. Unfortunately, the structure contains a field that has a different
2499     size on different hardware architectures. This meant that the tests that
2500     showed this size failed. As the block is currently always of a fixed size,
2501     this information isn't actually particularly useful in pcretest output, so
2502     I have just removed it.
2504     8. Three pre-processor statements accidentally did not start in column 1.
2505     Sadly, there are *still* compilers around that complain, even though
2506     standard C has not required this for well over a decade. Sigh.
2508     9. In pcretest, the code for checking callouts passed small integers in the
2509     callout_data field, which is a void * field. However, some picky compilers
2510     complained about the casts involved for this on 64-bit systems. Now
2511     pcretest passes the address of the small integer instead, which should get
2512     rid of the warnings.
2514     10. By default, when in UTF-8 mode, PCRE now checks for valid UTF-8 strings at
2515     both compile and run time, and gives an error if an invalid UTF-8 sequence
2516     is found. There is a option for disabling this check in cases where the
2517     string is known to be correct and/or the maximum performance is wanted.
2519     11. In response to a bug report, I changed one line in Makefile.in from
2521     -Wl,--out-implib,.libs/lib@WIN_PREFIX@pcreposix.dll.a \
2522     to
2523     -Wl,--out-implib,.libs/@WIN_PREFIX@libpcreposix.dll.a \
2525     to look similar to other lines, but I have no way of telling whether this
2526     is the right thing to do, as I do not use Windows. No doubt I'll get told
2527     if it's wrong...
2530 nigel 69 Version 4.3 21-May-03
2531     ---------------------
2533     1. Two instances of @WIN_PREFIX@ omitted from the Windows targets in the
2534     Makefile.
2536     2. Some refactoring to improve the quality of the code:
2538     (i) The utf8_table... variables are now declared "const".
2540     (ii) The code for \cx, which used the "case flipping" table to upper case
2541     lower case letters, now just substracts 32. This is ASCII-specific,
2542     but the whole concept of \cx is ASCII-specific, so it seems
2543     reasonable.
2545     (iii) PCRE was using its character types table to recognize decimal and
2546     hexadecimal digits in the pattern. This is silly, because it handles
2547     only 0-9, a-f, and A-F, but the character types table is locale-
2548     specific, which means strange things might happen. A private
2549     table is now used for this - though it costs 256 bytes, a table is
2550     much faster than multiple explicit tests. Of course, the standard
2551     character types table is still used for matching digits in subject
2552     strings against \d.
2554     (iv) Strictly, the identifier ESC_t is reserved by POSIX (all identifiers
2555     ending in _t are). So I've renamed it as ESC_tee.
2557     3. The first argument for regexec() in the POSIX wrapper should have been
2558     defined as "const".
2560     4. Changed pcretest to use malloc() for its buffers so that they can be
2561     Electric Fenced for debugging.
2563     5. There were several places in the code where, in UTF-8 mode, PCRE would try
2564     to read one or more bytes before the start of the subject string. Often this
2565     had no effect on PCRE's behaviour, but in some circumstances it could
2566     provoke a segmentation fault.
2568     6. A lookbehind at the start of a pattern in UTF-8 mode could also cause PCRE
2569     to try to read one or more bytes before the start of the subject string.
2571     7. A lookbehind in a pattern matched in non-UTF-8 mode on a PCRE compiled with
2572     UTF-8 support could misbehave in various ways if the subject string
2573     contained bytes with the 0x80 bit set and the 0x40 bit unset in a lookbehind
2574     area. (PCRE was not checking for the UTF-8 mode flag, and trying to move
2575     back over UTF-8 characters.)
2578 nigel 67 Version 4.2 14-Apr-03
2579     ---------------------
2581     1. Typo "#if SUPPORT_UTF8" instead of "#ifdef SUPPORT_UTF8" fixed.
2583     2. Changes to the building process, supplied by Ronald Landheer-Cieslak
2584     [ON_WINDOWS]: new variable, "#" on non-Windows platforms
2585     [NOT_ON_WINDOWS]: new variable, "#" on Windows platforms
2586     [WIN_PREFIX]: new variable, "cyg" for Cygwin
2587     * Makefile.in: use autoconf substitution for OBJEXT, EXEEXT, BUILD_OBJEXT
2588     and BUILD_EXEEXT
2589     Note: automatic setting of the BUILD variables is not yet working
2590     set CPPFLAGS and BUILD_CPPFLAGS (but don't use yet) - should be used at
2591     compile-time but not at link-time
2592     [LINK]: use for linking executables only
2593     make different versions for Windows and non-Windows
2594     [LINKLIB]: new variable, copy of UNIX-style LINK, used for linking
2595     libraries
2596     [LINK_FOR_BUILD]: new variable
2597     [OBJEXT]: use throughout
2598     [EXEEXT]: use throughout
2599     <winshared>: new target
2600     <wininstall>: new target
2601     <dftables.o>: use native compiler
2602     <dftables>: use native linker
2603     <install>: handle Windows platform correctly
2604     <clean>: ditto
2605     <check>: ditto
2606     copy DLL to top builddir before testing
2608     As part of these changes, -no-undefined was removed again. This was reported
2609     to give trouble on HP-UX 11.0, so getting rid of it seems like a good idea
2610     in any case.
2612     3. Some tidies to get rid of compiler warnings:
2614     . In the match_data structure, match_limit was an unsigned long int, whereas
2615     match_call_count was an int. I've made them both unsigned long ints.
2617     . In pcretest the fact that a const uschar * doesn't automatically cast to
2618     a void * provoked a warning.
2620     . Turning on some more compiler warnings threw up some "shadow" variables
2621     and a few more missing casts.
2623     4. If PCRE was complied with UTF-8 support, but called without the PCRE_UTF8
2624     option, a class that contained a single character with a value between 128
2625     and 255 (e.g. /[\xFF]/) caused PCRE to crash.
2627     5. If PCRE was compiled with UTF-8 support, but called without the PCRE_UTF8
2628     option, a class that contained several characters, but with at least one
2629     whose value was between 128 and 255 caused PCRE to crash.
2632 nigel 65 Version 4.1 12-Mar-03
2633     ---------------------
2634 nigel 63
2635 nigel 65 1. Compiling with gcc -pedantic found a couple of places where casts were
2636     needed, and a string in dftables.c that was longer than standard compilers are
2637     required to support.
2639     2. Compiling with Sun's compiler found a few more places where the code could
2640     be tidied up in order to avoid warnings.
2642     3. The variables for cross-compiling were called HOST_CC and HOST_CFLAGS; the
2643     first of these names is deprecated in the latest Autoconf in favour of the name
2644     CC_FOR_BUILD, because "host" is typically used to mean the system on which the
2645     compiled code will be run. I can't find a reference for HOST_CFLAGS, but by
2646     analogy I have changed it to CFLAGS_FOR_BUILD.
2648     4. Added -no-undefined to the linking command in the Makefile, because this is
2649     apparently helpful for Windows. To make it work, also added "-L. -lpcre" to the
2650     linking step for the pcreposix library.
2652     5. PCRE was failing to diagnose the case of two named groups with the same
2653     name.
2655     6. A problem with one of PCRE's optimizations was discovered. PCRE remembers a
2656     literal character that is needed in the subject for a match, and scans along to
2657     ensure that it is present before embarking on the full matching process. This
2658     saves time in cases of nested unlimited repeats that are never going to match.
2659     Problem: the scan can take a lot of time if the subject is very long (e.g.
2660     megabytes), thus penalizing straightforward matches. It is now done only if the
2661     amount of subject to be scanned is less than 1000 bytes.
2663     7. A lesser problem with the same optimization is that it was recording the
2664     first character of an anchored pattern as "needed", thus provoking a search
2665     right along the subject, even when the first match of the pattern was going to
2666     fail. The "needed" character is now not set for anchored patterns, unless it
2667     follows something in the pattern that is of non-fixed length. Thus, it still
2668     fulfils its original purpose of finding quick non-matches in cases of nested
2669     unlimited repeats, but isn't used for simple anchored patterns such as /^abc/.
2672     Version 4.0 17-Feb-03
2673     ---------------------
2675 nigel 63 1. If a comment in an extended regex that started immediately after a meta-item
2676     extended to the end of string, PCRE compiled incorrect data. This could lead to
2677     all kinds of weird effects. Example: /#/ was bad; /()#/ was bad; /a#/ was not.
2679     2. Moved to autoconf 2.53 and libtool 1.4.2.
2681     3. Perl 5.8 no longer needs "use utf8" for doing UTF-8 things. Consequently,
2682     the special perltest8 script is no longer needed - all the tests can be run
2683     from a single perltest script.
2685     4. From 5.004, Perl has not included the VT character (0x0b) in the set defined
2686     by \s. It has now been removed in PCRE. This means it isn't recognized as
2687     whitespace in /x regexes too, which is the same as Perl. Note that the POSIX
2688     class [:space:] *does* include VT, thereby creating a mess.
2690     5. Added the class [:blank:] (a GNU extension from Perl 5.8) to match only
2691     space and tab.
2693     6. Perl 5.005 was a long time ago. It's time to amalgamate the tests that use
2694     its new features into the main test script, reducing the number of scripts.
2696     7. Perl 5.8 has changed the meaning of patterns like /a(?i)b/. Earlier versions
2697     were backward compatible, and made the (?i) apply to the whole pattern, as if
2698     /i were given. Now it behaves more logically, and applies the option setting
2699     only to what follows. PCRE has been changed to follow suit. However, if it
2700     finds options settings right at the start of the pattern, it extracts them into
2701     the global options, as before. Thus, they show up in the info data.
2703     8. Added support for the \Q...\E escape sequence. Characters in between are
2704     treated as literals. This is slightly different from Perl in that $ and @ are
2705     also handled as literals inside the quotes. In Perl, they will cause variable
2706     interpolation. Note the following examples:
2708     Pattern PCRE matches Perl matches
2710     \Qabc$xyz\E abc$xyz abc followed by the contents of $xyz
2711     \Qabc\$xyz\E abc\$xyz abc\$xyz
2712     \Qabc\E\$\Qxyz\E abc$xyz abc$xyz
2714     For compatibility with Perl, \Q...\E sequences are recognized inside character
2715     classes as well as outside them.
2717     9. Re-organized 3 code statements in pcretest to avoid "overflow in
2718     floating-point constant arithmetic" warnings from a Microsoft compiler. Added a
2719     (size_t) cast to one statement in pcretest and one in pcreposix to avoid
2720     signed/unsigned warnings.
2722     10. SunOS4 doesn't have strtoul(). This was used only for unpicking the -o
2723     option for pcretest, so I've replaced it by a simple function that does just
2724     that job.
2726     11. pcregrep was ending with code 0 instead of 2 for the commands "pcregrep" or
2727     "pcregrep -".
2729     12. Added "possessive quantifiers" ?+, *+, ++, and {,}+ which come from Sun's
2730     Java package. This provides some syntactic sugar for simple cases of what my
2731     documentation calls "once-only subpatterns". A pattern such as x*+ is the same
2732     as (?>x*). In other words, if what is inside (?>...) is just a single repeated
2733     item, you can use this simplified notation. Note that only makes sense with
2734     greedy quantifiers. Consequently, the use of the possessive quantifier forces
2735     greediness, whatever the setting of the PCRE_UNGREEDY option.
2737     13. A change of greediness default within a pattern was not taking effect at
2738     the current level for patterns like /(b+(?U)a+)/. It did apply to parenthesized
2739     subpatterns that followed. Patterns like /b+(?U)a+/ worked because the option
2740     was abstracted outside.
2742     14. PCRE now supports the \G assertion. It is true when the current matching
2743     position is at the start point of the match. This differs from \A when the
2744     starting offset is non-zero. Used with the /g option of pcretest (or similar
2745     code), it works in the same way as it does for Perl's /g option. If all
2746     alternatives of a regex begin with \G, the expression is anchored to the start
2747     match position, and the "anchored" flag is set in the compiled expression.
2749     15. Some bugs concerning the handling of certain option changes within patterns
2750     have been fixed. These applied to options other than (?ims). For example,
2751     "a(?x: b c )d" did not match "XabcdY" but did match "Xa b c dY". It should have
2752     been the other way round. Some of this was related to change 7 above.
2754     16. PCRE now gives errors for /[.x.]/ and /[=x=]/ as unsupported POSIX
2755     features, as Perl does. Previously, PCRE gave the warnings only for /[[.x.]]/
2756     and /[[=x=]]/. PCRE now also gives an error for /[:name:]/ because it supports
2757     POSIX classes only within a class (e.g. /[[:alpha:]]/).
2759     17. Added support for Perl's \C escape. This matches one byte, even in UTF8
2760     mode. Unlike ".", it always matches newline, whatever the setting of
2761     PCRE_DOTALL. However, PCRE does not permit \C to appear in lookbehind
2762     assertions. Perl allows it, but it doesn't (in general) work because it can't
2763     calculate the length of the lookbehind. At least, that's the case for Perl
2764     5.8.0 - I've been told they are going to document that it doesn't work in
2765     future.
2767     18. Added an error diagnosis for escapes that PCRE does not support: these are
2768     \L, \l, \N, \P, \p, \U, \u, and \X.
2770     19. Although correctly diagnosing a missing ']' in a character class, PCRE was
2771     reading past the end of the pattern in cases such as /[abcd/.
2773     20. PCRE was getting more memory than necessary for patterns with classes that
2774     contained both POSIX named classes and other characters, e.g. /[[:space:]abc/.
2776     21. Added some code, conditional on #ifdef VPCOMPAT, to make life easier for
2777     compiling PCRE for use with Virtual Pascal.
2779     22. Small fix to the Makefile to make it work properly if the build is done
2780     outside the source tree.
2782     23. Added a new extension: a condition to go with recursion. If a conditional
2783     subpattern starts with (?(R) the "true" branch is used if recursion has
2784     happened, whereas the "false" branch is used only at the top level.
2786     24. When there was a very long string of literal characters (over 255 bytes
2787     without UTF support, over 250 bytes with UTF support), the computation of how
2788     much memory was required could be incorrect, leading to segfaults or other
2789     strange effects.
2791     25. PCRE was incorrectly assuming anchoring (either to start of subject or to
2792     start of line for a non-DOTALL pattern) when a pattern started with (.*) and
2793     there was a subsequent back reference to those brackets. This meant that, for
2794     example, /(.*)\d+\1/ failed to match "abc123bc". Unfortunately, it isn't
2795     possible to check for precisely this case. All we can do is abandon the
2796     optimization if .* occurs inside capturing brackets when there are any back
2797     references whatsoever. (See below for a better fix that came later.)
2799     26. The handling of the optimization for finding the first character of a
2800     non-anchored pattern, and for finding a character that is required later in the
2801     match were failing in some cases. This didn't break the matching; it just
2802     failed to optimize when it could. The way this is done has been re-implemented.
2804     27. Fixed typo in error message for invalid (?R item (it said "(?p").
2806     28. Added a new feature that provides some of the functionality that Perl
2807     provides with (?{...}). The facility is termed a "callout". The way it is done
2808     in PCRE is for the caller to provide an optional function, by setting
2809     pcre_callout to its entry point. Like pcre_malloc and pcre_free, this is a
2810     global variable. By default it is unset, which disables all calling out. To get
2811     the function called, the regex must include (?C) at appropriate points. This
2812     is, in fact, equivalent to (?C0), and any number <= 255 may be given with (?C).
2813     This provides a means of identifying different callout points. When PCRE
2814     reaches such a point in the regex, if pcre_callout has been set, the external
2815     function is called. It is provided with data in a structure called
2816     pcre_callout_block, which is defined in pcre.h. If the function returns 0,
2817     matching continues; if it returns a non-zero value, the match at the current
2818     point fails. However, backtracking will occur if possible. [This was changed
2819     later and other features added - see item 49 below.]
2821     29. pcretest is upgraded to test the callout functionality. It provides a
2822     callout function that displays information. By default, it shows the start of
2823     the match and the current position in the text. There are some new data escapes
2824     to vary what happens:
2826     \C+ in addition, show current contents of captured substrings
2827     \C- do not supply a callout function
2828     \C!n return 1 when callout number n is reached
2829     \C!n!m return 1 when callout number n is reached for the mth time
2831     30. If pcregrep was called with the -l option and just a single file name, it
2832     output "<stdin>" if a match was found, instead of the file name.
2834     31. Improve the efficiency of the POSIX API to PCRE. If the number of capturing
2835     slots is less than POSIX_MALLOC_THRESHOLD, use a block on the stack to pass to
2836     pcre_exec(). This saves a malloc/free per call. The default value of
2837     POSIX_MALLOC_THRESHOLD is 10; it can be changed by --with-posix-malloc-threshold
2838     when configuring.
2840     32. The default maximum size of a compiled pattern is 64K. There have been a
2841     few cases of people hitting this limit. The code now uses macros to handle the
2842     storing of links as offsets within the compiled pattern. It defaults to 2-byte
2843     links, but this can be changed to 3 or 4 bytes by --with-link-size when
2844     configuring. Tests 2 and 5 work only with 2-byte links because they output
2845     debugging information about compiled patterns.
2847     33. Internal code re-arrangements:
2849     (a) Moved the debugging function for printing out a compiled regex into
2850     its own source file (printint.c) and used #include to pull it into
2851     pcretest.c and, when DEBUG is defined, into pcre.c, instead of having two
2852     separate copies.
2854     (b) Defined the list of op-code names for debugging as a macro in
2855     internal.h so that it is next to the definition of the opcodes.
2857     (c) Defined a table of op-code lengths for simpler skipping along compiled
2858     code. This is again a macro in internal.h so that it is next to the
2859     definition of the opcodes.
2861     34. Added support for recursive calls to individual subpatterns, along the
2862     lines of Robin Houston's patch (but implemented somewhat differently).
2864     35. Further mods to the Makefile to help Win32. Also, added code to pcregrep to
2865     allow it to read and process whole directories in Win32. This code was
2866     contributed by Lionel Fourquaux; it has not been tested by me.
2868     36. Added support for named subpatterns. The Python syntax (?P<name>...) is
2869     used to name a group. Names consist of alphanumerics and underscores, and must
2870     be unique. Back references use the syntax (?P=name) and recursive calls use
2871     (?P>name) which is a PCRE extension to the Python extension. Groups still have
2872     numbers. The function pcre_fullinfo() can be used after compilation to extract
2873     a name/number map. There are three relevant calls:
2875     PCRE_INFO_NAMEENTRYSIZE yields the size of each entry in the map
2876     PCRE_INFO_NAMECOUNT yields the number of entries
2877     PCRE_INFO_NAMETABLE yields a pointer to the map.
2879     The map is a vector of fixed-size entries. The size of each entry depends on
2880     the length of the longest name used. The first two bytes of each entry are the
2881     group number, most significant byte first. There follows the corresponding
2882     name, zero terminated. The names are in alphabetical order.
2884     37. Make the maximum literal string in the compiled code 250 for the non-UTF-8
2885     case instead of 255. Making it the same both with and without UTF-8 support
2886     means that the same test output works with both.
2888     38. There was a case of malloc(0) in the POSIX testing code in pcretest. Avoid
2889     calling malloc() with a zero argument.
2891     39. Change 25 above had to resort to a heavy-handed test for the .* anchoring
2892     optimization. I've improved things by keeping a bitmap of backreferences with
2893     numbers 1-31 so that if .* occurs inside capturing brackets that are not in
2894     fact referenced, the optimization can be applied. It is unlikely that a
2895     relevant occurrence of .* (i.e. one which might indicate anchoring or forcing
2896     the match to follow \n) will appear inside brackets with a number greater than
2897     31, but if it does, any back reference > 31 suppresses the optimization.
2899     40. Added a new compile-time option PCRE_NO_AUTO_CAPTURE. This has the effect
2900     of disabling numbered capturing parentheses. Any opening parenthesis that is
2901     not followed by ? behaves as if it were followed by ?: but named parentheses
2902     can still be used for capturing (and they will acquire numbers in the usual
2903     way).
2905     41. Redesigned the return codes from the match() function into yes/no/error so
2906     that errors can be passed back from deep inside the nested calls. A malloc
2907     failure while inside a recursive subpattern call now causes the
2908     PCRE_ERROR_NOMEMORY return instead of quietly going wrong.
2910     42. It is now possible to set a limit on the number of times the match()
2911     function is called in a call to pcre_exec(). This facility makes it possible to
2912     limit the amount of recursion and backtracking, though not in a directly
2913     obvious way, because the match() function is used in a number of different
2914     circumstances. The count starts from zero for each position in the subject
2915     string (for non-anchored patterns). The default limit is, for compatibility, a
2916     large number, namely 10 000 000. You can change this in two ways:
2918     (a) When configuring PCRE before making, you can use --with-match-limit=n
2919     to set a default value for the compiled library.
2921     (b) For each call to pcre_exec(), you can pass a pcre_extra block in which
2922     a different value is set. See 45 below.
2924     If the limit is exceeded, pcre_exec() returns PCRE_ERROR_MATCHLIMIT.
2926     43. Added a new function pcre_config(int, void *) to enable run-time extraction
2927     of things that can be changed at compile time. The first argument specifies
2928     what is wanted and the second points to where the information is to be placed.
2929     The current list of available information is:
2933     The output is an integer that is set to one if UTF-8 support is available;
2934     otherwise it is set to zero.
2938     The output is an integer that it set to the value of the code that is used for
2939     newline. It is either LF (10) or CR (13).
2943     The output is an integer that contains the number of bytes used for internal
2944     linkage in compiled expressions. The value is 2, 3, or 4. See item 32 above.
2948     The output is an integer that contains the threshold above which the POSIX
2949     interface uses malloc() for output vectors. See item 31 above.
2953     The output is an unsigned integer that contains the default limit of the number
2954     of match() calls in a pcre_exec() execution. See 42 above.
2956     44. pcretest has been upgraded by the addition of the -C option. This causes it
2957     to extract all the available output from the new pcre_config() function, and to
2958     output it. The program then exits immediately.
2960     45. A need has arisen to pass over additional data with calls to pcre_exec() in
2961     order to support additional features. One way would have been to define
2962     pcre_exec2() (for example) with extra arguments, but this would not have been
2963     extensible, and would also have required all calls to the original function to
2964     be mapped to the new one. Instead, I have chosen to extend the mechanism that
2965     is used for passing in "extra" data from pcre_study().
2967     The pcre_extra structure is now exposed and defined in pcre.h. It currently
2968     contains the following fields:
2970     flags a bitmap indicating which of the following fields are set
2971     study_data opaque data from pcre_study()
2972     match_limit a way of specifying a limit on match() calls for a specific
2973     call to pcre_exec()
2974     callout_data data for callouts (see 49 below)
2976     The flag bits are also defined in pcre.h, and are
2982     The pcre_study() function now returns one of these new pcre_extra blocks, with
2983     the actual study data pointed to by the study_data field, and the
2984     PCRE_EXTRA_STUDY_DATA flag set. This can be passed directly to pcre_exec() as
2985     before. That is, this change is entirely upwards-compatible and requires no
2986     change to existing code.
2988     If you want to pass in additional data to pcre_exec(), you can either place it
2989     in a pcre_extra block provided by pcre_study(), or create your own pcre_extra
2990     block.
2992     46. pcretest has been extended to test the PCRE_EXTRA_MATCH_LIMIT feature. If a
2993     data string contains the escape sequence \M, pcretest calls pcre_exec() several
2994     times with different match limits, until it finds the minimum value needed for
2995     pcre_exec() to complete. The value is then output. This can be instructive; for
2996     most simple matches the number is quite small, but for pathological cases it
2997     gets very large very quickly.
2999     47. There's a new option for pcre_fullinfo() called PCRE_INFO_STUDYSIZE. It
3000     returns the size of the data block pointed to by the study_data field in a
3001     pcre_extra block, that is, the value that was passed as the argument to
3002     pcre_malloc() when PCRE was getting memory in which to place the information
3003     created by pcre_study(). The fourth argument should point to a size_t variable.
3004     pcretest has been extended so that this information is shown after a successful
3005     pcre_study() call when information about the compiled regex is being displayed.
3007     48. Cosmetic change to Makefile: there's no need to have / after $(DESTDIR)
3008     because what follows is always an absolute path. (Later: it turns out that this
3009     is more than cosmetic for MinGW, because it doesn't like empty path
3010     components.)
3012     49. Some changes have been made to the callout feature (see 28 above):
3014     (i) A callout function now has three choices for what it returns:
3016     0 => success, carry on matching
3017     > 0 => failure at this point, but backtrack if possible
3018     < 0 => serious error, return this value from pcre_exec()
3020     Negative values should normally be chosen from the set of PCRE_ERROR_xxx
3021     values. In particular, returning PCRE_ERROR_NOMATCH forces a standard
3022     "match failed" error. The error number PCRE_ERROR_CALLOUT is reserved for
3023     use by callout functions. It will never be used by PCRE itself.
3025     (ii) The pcre_extra structure (see 45 above) has a void * field called
3026     callout_data, with corresponding flag bit PCRE_EXTRA_CALLOUT_DATA. The
3027     pcre_callout_block structure has a field of the same name. The contents of
3028     the field passed in the pcre_extra structure are passed to the callout
3029     function in the corresponding field in the callout block. This makes it
3030     easier to use the same callout-containing regex from multiple threads. For
3031     testing, the pcretest program has a new data escape
3033     \C*n pass the number n (may be negative) as callout_data
3035     If the callout function in pcretest receives a non-zero value as
3036     callout_data, it returns that value.
3038     50. Makefile wasn't handling CFLAGS properly when compiling dftables. Also,
3039     there were some redundant $(CFLAGS) in commands that are now specified as
3040     $(LINK), which already includes $(CFLAGS).
3042     51. Extensions to UTF-8 support are listed below. These all apply when (a) PCRE
3043     has been compiled with UTF-8 support *and* pcre_compile() has been compiled
3044     with the PCRE_UTF8 flag. Patterns that are compiled without that flag assume
3045     one-byte characters throughout. Note that case-insensitive matching applies
3046     only to characters whose values are less than 256. PCRE doesn't support the
3047     notion of cases for higher-valued characters.
3049     (i) A character class whose characters are all within 0-255 is handled as
3050     a bit map, and the map is inverted for negative classes. Previously, a
3051     character > 255 always failed to match such a class; however it should
3052     match if the class was a negative one (e.g. [^ab]). This has been fixed.
3054     (ii) A negated character class with a single character < 255 is coded as
3055     "not this character" (OP_NOT). This wasn't working properly when the test
3056     character was multibyte, either singly or repeated.
3058     (iii) Repeats of multibyte characters are now handled correctly in UTF-8
3059     mode, for example: \x{100}{2,3}.
3061     (iv) The character escapes \b, \B, \d, \D, \s, \S, \w, and \W (either
3062     singly or repeated) now correctly test multibyte characters. However,
3063     PCRE doesn't recognize any characters with values greater than 255 as
3064     digits, spaces, or word characters. Such characters always match \D, \S,
3065     and \W, and never match \d, \s, or \w.
3067     (v) Classes may now contain characters and character ranges with values
3068     greater than 255. For example: [ab\x{100}-\x{400}].
3070     (vi) pcregrep now has a --utf-8 option (synonym -u) which makes it call
3071     PCRE in UTF-8 mode.
3073     52. The info request value PCRE_INFO_FIRSTCHAR has been renamed
3074     PCRE_INFO_FIRSTBYTE because it is a byte value. However, the old name is
3075     retained for backwards compatibility. (Note that LASTLITERAL is also a byte
3076     value.)
3078     53. The single man page has become too large. I have therefore split it up into
3079     a number of separate man pages. These also give rise to individual HTML pages;
3080     these are now put in a separate directory, and there is an index.html page that
3081     lists them all. Some hyperlinking between the pages has been installed.
3083     54. Added convenience functions for handling named capturing parentheses.
3085     55. Unknown escapes inside character classes (e.g. [\M]) and escapes that
3086     aren't interpreted therein (e.g. [\C]) are literals in Perl. This is now also
3087     true in PCRE, except when the PCRE_EXTENDED option is set, in which case they
3088     are faulted.
3090     56. Introduced HOST_CC and HOST_CFLAGS which can be set in the environment when
3091     calling configure. These values are used when compiling the dftables.c program
3092     which is run to generate the source of the default character tables. They
3093     default to the values of CC and CFLAGS. If you are cross-compiling PCRE,
3094     you will need to set these values.
3096     57. Updated the building process for Windows DLL, as provided by Fred Cox.
3099     Version 3.9 02-Jan-02
3100 nigel 61 ---------------------
3102     1. A bit of extraneous text had somehow crept into the pcregrep documentation.
3104     2. If --disable-static was given, the building process failed when trying to
3105     build pcretest and pcregrep. (For some reason it was using libtool to compile
3106     them, which is not right, as they aren't part of the library.)
3109 nigel 59 Version 3.8 18-Dec-01
3110     ---------------------
3112     1. The experimental UTF-8 code was completely screwed up. It was packing the
3113     bytes in the wrong order. How dumb can you get?
3116 nigel 57 Version 3.7 29-Oct-01
3117     ---------------------
3119     1. In updating pcretest to check change 1 of version 3.6, I screwed up.
3120     This caused pcretest, when used on the test data, to segfault. Unfortunately,
3121     this didn't happen under Solaris 8, where I normally test things.
3123 nigel 59 2. The Makefile had to be changed to make it work on BSD systems, where 'make'
3124     doesn't seem to recognize that ./xxx and xxx are the same file. (This entry
3125     isn't in ChangeLog distributed with 3.7 because I forgot when I hastily made
3126     this fix an hour or so after the initial 3.7 release.)
3127 nigel 57
3128 nigel 59
3129 nigel 55 Version 3.6 23-Oct-01
3130     ---------------------
3132     1. Crashed with /(sens|respons)e and \1ibility/ and "sense and sensibility" if
3133     offsets passed as NULL with zero offset count.
3135     2. The config.guess and config.sub files had not been updated when I moved to
3136     the latest autoconf.
3139 nigel 53 Version 3.5 15-Aug-01
3140     ---------------------
3141 nigel 5
3142 nigel 53 1. Added some missing #if !defined NOPOSIX conditionals in pcretest.c that
3143     had been forgotten.
3145     2. By using declared but undefined structures, we can avoid using "void"
3146     definitions in pcre.h while keeping the internal definitions of the structures
3147     private.
3149     3. The distribution is now built using autoconf 2.50 and libtool 1.4. From a
3150     user point of view, this means that both static and shared libraries are built
3151     by default, but this can be individually controlled. More of the work of
3152     handling this static/shared cases is now inside libtool instead of PCRE's make
3153     file.
3155     4. The pcretest utility is now installed along with pcregrep because it is
3156     useful for users (to test regexs) and by doing this, it automatically gets
3157     relinked by libtool. The documentation has been turned into a man page, so
3158     there are now .1, .txt, and .html versions in /doc.
3160     5. Upgrades to pcregrep:
3161     (i) Added long-form option names like gnu grep.
3162     (ii) Added --help to list all options with an explanatory phrase.
3163     (iii) Added -r, --recursive to recurse into sub-directories.
3164     (iv) Added -f, --file to read patterns from a file.
3166     6. pcre_exec() was referring to its "code" argument before testing that
3167     argument for NULL (and giving an error if it was NULL).
3169     7. Upgraded Makefile.in to allow for compiling in a different directory from
3170     the source directory.
3172     8. Tiny buglet in pcretest: when pcre_fullinfo() was called to retrieve the
3173     options bits, the pointer it was passed was to an int instead of to an unsigned
3174     long int. This mattered only on 64-bit systems.
3176     9. Fixed typo (3.4/1) in pcre.h again. Sigh. I had changed pcre.h (which is
3177     generated) instead of pcre.in, which it its source. Also made the same change
3178     in several of the .c files.
3180     10. A new release of gcc defines printf() as a macro, which broke pcretest
3181     because it had an ifdef in the middle of a string argument for printf(). Fixed
3182     by using separate calls to printf().
3184     11. Added --enable-newline-is-cr and --enable-newline-is-lf to the configure
3185     script, to force use of CR or LF instead of \n in the source. On non-Unix
3186     systems, the value can be set in config.h.
3188     12. The limit of 200 on non-capturing parentheses is a _nesting_ limit, not an
3189     absolute limit. Changed the text of the error message to make this clear, and
3190     likewise updated the man page.
3192     13. The limit of 99 on the number of capturing subpatterns has been removed.
3193     The new limit is 65535, which I hope will not be a "real" limit.
3196 nigel 51 Version 3.4 22-Aug-00
3197     ---------------------
3199     1. Fixed typo in pcre.h: unsigned const char * changed to const unsigned char *.
3201     2. Diagnose condition (?(0) as an error instead of crashing on matching.
3204 nigel 49 Version 3.3 01-Aug-00
3205     ---------------------
3207     1. If an octal character was given, but the value was greater than \377, it
3208     was not getting masked to the least significant bits, as documented. This could
3209     lead to crashes in some systems.
3211     2. Perl 5.6 (if not earlier versions) accepts classes like [a-\d] and treats
3212     the hyphen as a literal. PCRE used to give an error; it now behaves like Perl.
3214     3. Added the functions pcre_free_substring() and pcre_free_substring_list().
3215     These just pass their arguments on to (pcre_free)(), but they are provided
3216     because some uses of PCRE bind it to non-C systems that can call its functions,
3217     but cannot call free() or pcre_free() directly.
3219     4. Add "make test" as a synonym for "make check". Corrected some comments in
3220     the Makefile.
3222     5. Add $(DESTDIR)/ in front of all the paths in the "install" target in the
3223     Makefile.
3225     6. Changed the name of pgrep to pcregrep, because Solaris has introduced a
3226     command called pgrep for grepping around the active processes.
3228     7. Added the beginnings of support for UTF-8 character strings.
3230     8. Arranged for the Makefile to pass over the settings of CC, CFLAGS, and
3231     RANLIB to ./ltconfig so that they are used by libtool. I think these are all
3232     the relevant ones. (AR is not passed because ./ltconfig does its own figuring
3233     out for the ar command.)
3236 nigel 47 Version 3.2 12-May-00
3237     ---------------------
3239     This is purely a bug fixing release.
3241     1. If the pattern /((Z)+|A)*/ was matched agained ZABCDEFG it matched Z instead
3242     of ZA. This was just one example of several cases that could provoke this bug,
3243     which was introduced by change 9 of version 2.00. The code for breaking
3244     infinite loops after an iteration that matches an empty string was't working
3245     correctly.
3247     2. The pcretest program was not imitating Perl correctly for the pattern /a*/g
3248     when matched against abbab (for example). After matching an empty string, it
3249     wasn't forcing anchoring when setting PCRE_NOTEMPTY for the next attempt; this
3250     caused it to match further down the string than it should.
3252     3. The code contained an inclusion of sys/types.h. It isn't clear why this
3253     was there because it doesn't seem to be needed, and it causes trouble on some
3254     systems, as it is not a Standard C header. It has been removed.
3256     4. Made 4 silly changes to the source to avoid stupid compiler warnings that
3257     were reported on the Macintosh. The changes were from
3259     while ((c = *(++ptr)) != 0 && c != '\n');
3260     to
3261     while ((c = *(++ptr)) != 0 && c != '\n') ;
3263     Totally extraordinary, but if that's what it takes...
3265     5. PCRE is being used in one environment where neither memmove() nor bcopy() is
3266     available. Added HAVE_BCOPY and an autoconf test for it; if neither
3267     HAVE_MEMMOVE nor HAVE_BCOPY is set, use a built-in emulation function which
3268     assumes the way PCRE uses memmove() (always moving upwards).
3270     6. PCRE is being used in one environment where strchr() is not available. There
3271     was only one use in pcre.c, and writing it out to avoid strchr() probably gives
3272     faster code anyway.
3275 nigel 45 Version 3.1 09-Feb-00
3276     ---------------------
3278     The only change in this release is the fixing of some bugs in Makefile.in for
3279     the "install" target:
3281     (1) It was failing to install pcreposix.h.
3283     (2) It was overwriting the pcre.3 man page with the pcreposix.3 man page.
3286 nigel 43 Version 3.0 01-Feb-00
3287     ---------------------
3288 nigel 41
3289     1. Add support for the /+ modifier to perltest (to output $` like it does in
3290     pcretest).
3292     2. Add support for the /g modifier to perltest.
3294     3. Fix pcretest so that it behaves even more like Perl for /g when the pattern
3295     matches null strings.
3297     4. Fix perltest so that it doesn't do unwanted things when fed an empty
3298     pattern. Perl treats empty patterns specially - it reuses the most recent
3299     pattern, which is not what we want. Replace // by /(?#)/ in order to avoid this
3300     effect.
3302     5. The POSIX interface was broken in that it was just handing over the POSIX
3303     captured string vector to pcre_exec(), but (since release 2.00) PCRE has
3304     required a bigger vector, with some working space on the end. This means that
3305     the POSIX wrapper now has to get and free some memory, and copy the results.
3307 nigel 43 6. Added some simple autoconf support, placing the test data and the
3308     documentation in separate directories, re-organizing some of the
3309     information files, and making it build pcre-config (a GNU standard). Also added
3310     libtool support for building PCRE as a shared library, which is now the
3311     default.
3312 nigel 41
3313 nigel 43 7. Got rid of the leading zero in the definition of PCRE_MINOR because 08 and
3314     09 are not valid octal constants. Single digits will be used for minor values
3315     less than 10.
3317     8. Defined REG_EXTENDED and REG_NOSUB as zero in the POSIX header, so that
3318     existing programs that set these in the POSIX interface can use PCRE without
3319     modification.
3321     9. Added a new function, pcre_fullinfo() with an extensible interface. It can
3322     return all that pcre_info() returns, plus additional data. The pcre_info()
3323     function is retained for compatibility, but is considered to be obsolete.
3325     10. Added experimental recursion feature (?R) to handle one common case that
3326     Perl 5.6 will be able to do with (?p{...}).
3328     11. Added support for POSIX character classes like [:alpha:], which Perl is
3329     adopting.
3332 nigel 39 Version 2.08 31-Aug-99
3333     ----------------------
3335     1. When startoffset was not zero and the pattern began with ".*", PCRE was not
3336     trying to match at the startoffset position, but instead was moving forward to
3337     the next newline as if a previous match had failed.
3339     2. pcretest was not making use of PCRE_NOTEMPTY when repeating for /g and /G,
3340     and could get into a loop if a null string was matched other than at the start
3341     of the subject.
3343     3. Added definitions of PCRE_MAJOR and PCRE_MINOR to pcre.h so the version can
3344     be distinguished at compile time, and for completeness also added PCRE_DATE.
3346     5. Added Paul Sokolovsky's minor changes to make it easy to compile a Win32 DLL
3347     in GnuWin32 environments.
3350 nigel 37 Version 2.07 29-Jul-99
3351     ----------------------
3353     1. The documentation is now supplied in plain text form and HTML as well as in
3354     the form of man page sources.
3356     2. C++ compilers don't like assigning (void *) values to other pointer types.
3357     In particular this affects malloc(). Although there is no problem in Standard
3358     C, I've put in casts to keep C++ compilers happy.
3360     3. Typo on pcretest.c; a cast of (unsigned char *) in the POSIX regexec() call
3361     should be (const char *).
3363     4. If NOPOSIX is defined, pcretest.c compiles without POSIX support. This may
3364     be useful for non-Unix systems who don't want to bother with the POSIX stuff.
3365     However, I haven't made this a standard facility. The documentation doesn't
3366     mention it, and the Makefile doesn't support it.
3368     5. The Makefile now contains an "install" target, with editable destinations at
3369     the top of the file. The pcretest program is not installed.
3371     6. pgrep -V now gives the PCRE version number and date.
3373     7. Fixed bug: a zero repetition after a literal string (e.g. /abcde{0}/) was
3374     causing the entire string to be ignored, instead of just the last character.
3376     8. If a pattern like /"([^\\"]+|\\.)*"/ is applied in the normal way to a
3377     non-matching string, it can take a very, very long time, even for strings of
3378     quite modest length, because of the nested recursion. PCRE now does better in
3379     some of these cases. It does this by remembering the last required literal
3380     character in the pattern, and pre-searching the subject to ensure it is present
3381     before running the real match. In other words, it applies a heuristic to detect
3382     some types of certain failure quickly, and in the above example, if presented
3383     with a string that has no trailing " it gives "no match" very quickly.
3385     9. A new runtime option PCRE_NOTEMPTY causes null string matches to be ignored;
3386     other alternatives are tried instead.
3389 nigel 35 Version 2.06 09-Jun-99
3390     ----------------------
3392     1. Change pcretest's output for amount of store used to show just the code
3393     space, because the remainder (the data block) varies in size between 32-bit and
3394     64-bit systems.
3396     2. Added an extra argument to pcre_exec() to supply an offset in the subject to
3397     start matching at. This allows lookbehinds to work when searching for multiple
3398     occurrences in a string.
3400     3. Added additional options to pcretest for testing multiple occurrences:
3402     /+ outputs the rest of the string that follows a match
3403     /g loops for multiple occurrences, using the new startoffset argument
3404     /G loops for multiple occurrences by passing an incremented pointer
3406     4. PCRE wasn't doing the "first character" optimization for patterns starting
3407     with \b or \B, though it was doing it for other lookbehind assertions. That is,
3408     it wasn't noticing that a match for a pattern such as /\bxyz/ has to start with
3409     the letter 'x'. On long subject strings, this gives a significant speed-up.
3412 nigel 33 Version 2.05 21-Apr-99
3413     ----------------------
3415     1. Changed the type of magic_number from int to long int so that it works
3416     properly on 16-bit systems.
3418     2. Fixed a bug which caused patterns starting with .* not to work correctly
3419     when the subject string contained newline characters. PCRE was assuming
3420     anchoring for such patterns in all cases, which is not correct because .* will
3421     not pass a newline unless PCRE_DOTALL is set. It now assumes anchoring only if
3422     DOTALL is set at top level; otherwise it knows that patterns starting with .*
3423     must be retried after every newline in the subject.
3426 nigel 31 Version 2.04 18-Feb-99
3427     ----------------------
3429     1. For parenthesized subpatterns with repeats whose minimum was zero, the
3430     computation of the store needed to hold the pattern was incorrect (too large).
3431     If such patterns were nested a few deep, this could multiply and become a real
3432     problem.
3434     2. Added /M option to pcretest to show the memory requirement of a specific
3435     pattern. Made -m a synonym of -s (which does this globally) for compatibility.
3437     3. Subpatterns of the form (regex){n,m} (i.e. limited maximum) were being
3438     compiled in such a way that the backtracking after subsequent failure was
3439     pessimal. Something like (a){0,3} was compiled as (a)?(a)?(a)? instead of
3440     ((a)((a)(a)?)?)? with disastrous performance if the maximum was of any size.
3443 nigel 29 Version 2.03 02-Feb-99
3444     ----------------------
3446     1. Fixed typo and small mistake in man page.
3448 nigel 31 2. Added 4th condition (GPL supersedes if conflict) and created separate
3449     LICENCE file containing the conditions.
3450 nigel 29
3451     3. Updated pcretest so that patterns such as /abc\/def/ work like they do in
3452     Perl, that is the internal \ allows the delimiter to be included in the
3453     pattern. Locked out the use of \ as a delimiter. If \ immediately follows
3454     the final delimiter, add \ to the end of the pattern (to test the error).
3456     4. Added the convenience functions for extracting substrings after a successful
3457     match. Updated pcretest to make it able to test these functions.
3460 nigel 27 Version 2.02 14-Jan-99
3461     ----------------------
3463     1. Initialized the working variables associated with each extraction so that
3464     their saving and restoring doesn't refer to uninitialized store.
3466     2. Put dummy code into study.c in order to trick the optimizer of the IBM C
3467     compiler for OS/2 into generating correct code. Apparently IBM isn't going to
3468     fix the problem.
3470     3. Pcretest: the timing code wasn't using LOOPREPEAT for timing execution
3471     calls, and wasn't printing the correct value for compiling calls. Increased the
3472     default value of LOOPREPEAT, and the number of significant figures in the
3473     times.
3475     4. Changed "/bin/rm" in the Makefile to "-rm" so it works on Windows NT.
3477     5. Renamed "deftables" as "dftables" to get it down to 8 characters, to avoid
3478     a building problem on Windows NT with a FAT file system.
3481 nigel 25 Version 2.01 21-Oct-98
3482     ----------------------
3484     1. Changed the API for pcre_compile() to allow for the provision of a pointer
3485     to character tables built by pcre_maketables() in the current locale. If NULL
3486     is passed, the default tables are used.
3489 nigel 23 Version 2.00 24-Sep-98
3490 nigel 21 ----------------------
3492 nigel 23 1. Since the (>?) facility is in Perl 5.005, don't require PCRE_EXTRA to enable
3493     it any more.
3495     2. Allow quantification of (?>) groups, and make it work correctly.
3497     3. The first character computation wasn't working for (?>) groups.
3499     4. Correct the implementation of \Z (it is permitted to match on the \n at the
3500     end of the subject) and add 5.005's \z, which really does match only at the
3501     very end of the subject.
3503     5. Remove the \X "cut" facility; Perl doesn't have it, and (?> is neater.
3505     6. Remove the ability to specify CASELESS, MULTILINE, DOTALL, and
3506     DOLLAR_END_ONLY at runtime, to make it possible to implement the Perl 5.005
3507     localized options. All options to pcre_study() were also removed.
3509     7. Add other new features from 5.005:
3511     $(?<= positive lookbehind
3512     $(?<! negative lookbehind
3513     (?imsx-imsx) added the unsetting capability
3514     such a setting is global if at outer level; local otherwise
3515     (?imsx-imsx:) non-capturing groups with option setting
3516     (?(cond)re|re) conditional pattern matching
3518     A backreference to itself in a repeated group matches the previous
3519     captured string.
3521     8. General tidying up of studying (both automatic and via "study")
3522     consequential on the addition of new assertions.
3524     9. As in 5.005, unlimited repeated groups that could match an empty substring
3525     are no longer faulted at compile time. Instead, the loop is forcibly broken at
3526     runtime if any iteration does actually match an empty substring.
3528     10. Include the RunTest script in the distribution.
3530     11. Added tests from the Perl 5.005_02 distribution. This showed up a few
3531     discrepancies, some of which were old and were also with respect to 5.004. They
3532     have now been fixed.
3535     Version 1.09 28-Apr-98
3536     ----------------------
3538 nigel 21 1. A negated single character class followed by a quantifier with a minimum
3539     value of one (e.g. [^x]{1,6} ) was not compiled correctly. This could lead to
3540     program crashes, or just wrong answers. This did not apply to negated classes
3541     containing more than one character, or to minima other than one.
3544 nigel 19 Version 1.08 27-Mar-98
3545     ----------------------
3547     1. Add PCRE_UNGREEDY to invert the greediness of quantifiers.
3549     2. Add (?U) and (?X) to set PCRE_UNGREEDY and PCRE_EXTRA respectively. The
3550     latter must appear before anything that relies on it in the pattern.
3553 nigel 17 Version 1.07 16-Feb-98
3554     ----------------------
3556     1. A pattern such as /((a)*)*/ was not being diagnosed as in error (unlimited
3557     repeat of a potentially empty string).
3560 nigel 15 Version 1.06 23-Jan-98
3561     ----------------------
3563     1. Added Markus Oberhumer's little patches for C++.
3565     2. Literal strings longer than 255 characters were broken.
3568 nigel 13 Version 1.05 23-Dec-97
3569     ----------------------
3571     1. Negated character classes containing more than one character were failing if
3572     PCRE_CASELESS was set at run time.
3575 nigel 11 Version 1.04 19-Dec-97
3576     ----------------------
3578     1. Corrected the man page, where some "const" qualifiers had been omitted.
3580     2. Made debugging output print "{0,xxx}" instead of just "{,xxx}" to agree with
3581     input syntax.
3583     3. Fixed memory leak which occurred when a regex with back references was
3584     matched with an offsets vector that wasn't big enough. The temporary memory
3585     that is used in this case wasn't being freed if the match failed.
3587     4. Tidied pcretest to ensure it frees memory that it gets.
3589     5. Temporary memory was being obtained in the case where the passed offsets
3590     vector was exactly big enough.
3592     6. Corrected definition of offsetof() from change 5 below.
3594     7. I had screwed up change 6 below and broken the rules for the use of
3595     setjmp(). Now fixed.
3598 nigel 9 Version 1.03 18-Dec-97
3599     ----------------------
3601     1. A erroneous regex with a missing opening parenthesis was correctly
3602     diagnosed, but PCRE attempted to access brastack[-1], which could cause crashes
3603     on some systems.
3605     2. Replaced offsetof(real_pcre, code) by offsetof(real_pcre, code[0]) because
3606     it was reported that one broken compiler failed on the former because "code" is
3607     also an independent variable.
3609     3. The erroneous regex a[]b caused an array overrun reference.
3611     4. A regex ending with a one-character negative class (e.g. /[^k]$/) did not
3612     fail on data ending with that character. (It was going on too far, and checking
3613     the next character, typically a binary zero.) This was specific to the
3614     optimized code for single-character negative classes.
3616     5. Added a contributed patch from the TIN world which does the following:
3618     + Add an undef for memmove, in case the the system defines a macro for it.
3620     + Add a definition of offsetof(), in case there isn't one. (I don't know
3621     the reason behind this - offsetof() is part of the ANSI standard - but
3622     it does no harm).
3624     + Reduce the ifdef's in pcre.c using macro DPRINTF, thereby eliminating
3625     most of the places where whitespace preceded '#'. I have given up and
3626     allowed the remaining 2 cases to be at the margin.
3628     + Rename some variables in pcre to eliminate shadowing. This seems very
3629     pedantic, but does no harm, of course.
3631     6. Moved the call to setjmp() into its own function, to get rid of warnings
3632     from gcc -Wall, and avoided calling it at all unless PCRE_EXTRA is used.
3634     7. Constructs such as \d{8,} were compiling into the equivalent of
3635 nigel 11 \d{8}\d{0,65527} instead of \d{8}\d* which didn't make much difference to the
3636 nigel 9 outcome, but in this particular case used more store than had been allocated,
3637     which caused the bug to be discovered because it threw up an internal error.
3639     8. The debugging code in both pcre and pcretest for outputting the compiled
3640     form of a regex was going wrong in the case of back references followed by
3641     curly-bracketed repeats.
3644 nigel 7 Version 1.02 12-Dec-97
3645     ----------------------
3647     1. Typos in pcre.3 and comments in the source fixed.
3649     2. Applied a contributed patch to get rid of places where it used to remove
3650     'const' from variables, and fixed some signed/unsigned and uninitialized
3651     variable warnings.
3653     3. Added the "runtest" target to Makefile.
3655     4. Set default compiler flag to -O2 rather than just -O.
3658 nigel 5 Version 1.01 19-Nov-97
3659     ----------------------
3661     1. PCRE was failing to diagnose unlimited repeat of empty string for patterns
3662     like /([ab]*)*/, that is, for classes with more than one character in them.
3664     2. Likewise, it wasn't diagnosing patterns with "once-only" subpatterns, such
3665     as /((?>a*))*/ (a PCRE_EXTRA facility).
3668     Version 1.00 18-Nov-97
3669     ----------------------
3671     1. Added compile-time macros to support systems such as SunOS4 which don't have
3672     memmove() or strerror() but have other things that can be used instead.
3674     2. Arranged that "make clean" removes the executables.
3677 nigel 3 Version 0.99 27-Oct-97
3678     ----------------------
3680     1. Fixed bug in code for optimizing classes with only one character. It was
3681     initializing a 32-byte map regardless, which could cause it to run off the end
3682     of the memory it had got.
3684     2. Added, conditional on PCRE_EXTRA, the proposed (?>REGEX) construction.
3687     Version 0.98 22-Oct-97
3688     ----------------------
3690     1. Fixed bug in code for handling temporary memory usage when there are more
3691     back references than supplied space in the ovector. This could cause segfaults.
3694     Version 0.97 21-Oct-97
3695     ----------------------
3697     1. Added the \X "cut" facility, conditional on PCRE_EXTRA.
3699     2. Optimized negated single characters not to use a bit map.
3701     3. Brought error texts together as macro definitions; clarified some of them;
3702     fixed one that was wrong - it said "range out of order" when it meant "invalid
3703     escape sequence".
3705     4. Changed some char * arguments to const char *.
3707     5. Added PCRE_NOTBOL and PCRE_NOTEOL (from POSIX).
3709     6. Added the POSIX-style API wrapper in pcreposix.a and testing facilities in
3710     pcretest.
3713     Version 0.96 16-Oct-97
3714     ----------------------
3716     1. Added a simple "pgrep" utility to the distribution.
3718     2. Fixed an incompatibility with Perl: "{" is now treated as a normal character
3719     unless it appears in one of the precise forms "{ddd}", "{ddd,}", or "{ddd,ddd}"
3720     where "ddd" means "one or more decimal digits".
3722     3. Fixed serious bug. If a pattern had a back reference, but the call to
3723     pcre_exec() didn't supply a large enough ovector to record the related
3724     identifying subpattern, the match always failed. PCRE now remembers the number
3725     of the largest back reference, and gets some temporary memory in which to save
3726     the offsets during matching if necessary, in order to ensure that
3727     backreferences always work.
3729     4. Increased the compatibility with Perl in a number of ways:
3731     (a) . no longer matches \n by default; an option PCRE_DOTALL is provided
3732     to request this handling. The option can be set at compile or exec time.
3734     (b) $ matches before a terminating newline by default; an option
3735     PCRE_DOLLAR_ENDONLY is provided to override this (but not in multiline
3736     mode). The option can be set at compile or exec time.
3738     (c) The handling of \ followed by a digit other than 0 is now supposed to be
3739     the same as Perl's. If the decimal number it represents is less than 10
3740     or there aren't that many previous left capturing parentheses, an octal
3741     escape is read. Inside a character class, it's always an octal escape,
3742     even if it is a single digit.
3744     (d) An escaped but undefined alphabetic character is taken as a literal,
3745     unless PCRE_EXTRA is set. Currently this just reserves the remaining
3746     escapes.
3748     (e) {0} is now permitted. (The previous item is removed from the compiled
3749     pattern).
3751     5. Changed all the names of code files so that the basic parts are no longer
3752     than 10 characters, and abolished the teeny "globals.c" file.
3754     6. Changed the handling of character classes; they are now done with a 32-byte
3755     bit map always.
3757     7. Added the -d and /D options to pcretest to make it possible to look at the
3758     internals of compilation without having to recompile pcre.
3761     Version 0.95 23-Sep-97
3762     ----------------------
3764     1. Fixed bug in pre-pass concerning escaped "normal" characters such as \x5c or
3765     \x20 at the start of a run of normal characters. These were being treated as
3766     real characters, instead of the source characters being re-checked.
3769     Version 0.94 18-Sep-97
3770     ----------------------
3772     1. The functions are now thread-safe, with the caveat that the global variables
3773     containing pointers to malloc() and free() or alternative functions are the
3774     same for all threads.