ViewVC logotype

Contents of /code/trunk/ChangeLog

Parent Directory Parent Directory | Revision Log Revision Log

Revision 536 - (show annotations)
Mon Jun 7 16:39:32 2010 UTC (5 years, 3 months ago) by ph10
File size: 187985 byte(s)
Error occurred while calculating annotation data.
Reverse pcreposix.c #include patch; doesn't work on Windows.
1 ChangeLog for PCRE
2 ------------------
4 Version 8.10 03-Jun-2010
5 ------------------------
7 1. Added support for (*MARK:ARG) and for ARG additions to PRUNE, SKIP, and
10 2. (*ACCEPT) was not working when inside an atomic group.
12 3. Inside a character class, \B is treated as a literal by default, but
13 faulted if PCRE_EXTRA is set. This mimics Perl's behaviour (the -w option
14 causes the error). The code is unchanged, but I tidied the documentation.
16 4. Inside a character class, PCRE always treated \R and \X as literals,
17 whereas Perl faults them if its -w option is set. I have changed PCRE so
18 that it faults them when PCRE_EXTRA is set.
20 5. Added support for \N, which always matches any character other than
21 newline. (It is the same as "." when PCRE_DOTALL is not set.)
23 6. When compiling pcregrep with newer versions of gcc which may have
24 FORTIFY_SOURCE set, several warnings "ignoring return value of 'fwrite',
25 declared with attribute warn_unused_result" were given. Just casting the
26 result to (void) does not stop the warnings; a more elaborate fudge is
27 needed. I've used a macro to implement this.
29 7. Minor change to pcretest.c to avoid a compiler warning.
31 8. Added four artifical Unicode properties to help with an option to make
32 \s etc use properties (see next item). The new properties are: Xan
33 (alphanumeric), Xsp (Perl space), Xps (POSIX space), and Xwd (word).
35 9. Added PCRE_UCP to make \b, \d, \s, \w, and certain POSIX character classes
36 use Unicode properties. (*UCP) at the start of a pattern can be used to set
37 this option. Modified pcretest to add /W to test this facility. Added
38 REG_UCP to make it available via the POSIX interface.
40 10. Added --line-buffered to pcregrep.
42 11. In UTF-8 mode, if a pattern that was compiled with PCRE_CASELESS was
43 studied, and the match started with a letter with a code point greater than
44 127 whose first byte was different to the first byte of the other case of
45 the letter, the other case of this starting letter was not recognized
46 (#976).
48 12. If a pattern that was studied started with a repeated Unicode property
49 test, for example, \p{Nd}+, there was the theoretical possibility of
50 setting up an incorrect bitmap of starting bytes, but fortunately it could
51 not have actually happened in practice until change 8 above was made (it
52 added property types that matched character-matching opcodes).
54 13. pcre_study() now recognizes \h, \v, and \R when constructing a bit map of
55 possible starting bytes for non-anchored patterns.
57 14. Extended the "auto-possessify" feature of pcre_compile(). It now recognizes
58 \R, and also a number of cases that involve Unicode properties, both
59 explicit and implicit when PCRE_UCP is set.
61 15. If a repeated Unicode property match (e.g. \p{Lu}*) was used with non-UTF-8
62 input, it could crash or give wrong results if characters with values
63 greater than 0xc0 were present in the subject string. (Detail: it assumed
64 UTF-8 input when processing these items.)
66 16. Added a lot of (int) casts to avoid compiler warnings in systems where
67 size_t is 64-bit (#991).
69 17. Added a check for running out of memory when PCRE is compiled with
70 --disable-stack-for-recursion (#990).
74 Version 8.02 19-Mar-2010
75 ------------------------
77 1. The Unicode data tables have been updated to Unicode 5.2.0.
79 2. Added the option --libs-cpp to pcre-config, but only when C++ support is
80 configured.
82 3. Updated the licensing terms in the pcregexp.pas file, as agreed with the
83 original author of that file, following a query about its status.
85 4. On systems that do not have stdint.h (e.g. Solaris), check for and include
86 inttypes.h instead. This fixes a bug that was introduced by change 8.01/8.
88 5. A pattern such as (?&t)*+(?(DEFINE)(?<t>.)) which has a possessive
89 quantifier applied to a forward-referencing subroutine call, could compile
90 incorrect code or give the error "internal error: previously-checked
91 referenced subpattern not found".
93 6. Both MS Visual Studio and Symbian OS have problems with initializing
94 variables to point to external functions. For these systems, therefore,
95 pcre_malloc etc. are now initialized to local functions that call the
96 relevant global functions.
98 7. There were two entries missing in the vectors called coptable and poptable
99 in pcre_dfa_exec.c. This could lead to memory accesses outsize the vectors.
100 I've fixed the data, and added a kludgy way of testing at compile time that
101 the lengths are correct (equal to the number of opcodes).
103 8. Following on from 7, I added a similar kludge to check the length of the
104 eint vector in pcreposix.c.
106 9. Error texts for pcre_compile() are held as one long string to avoid too
107 much relocation at load time. To find a text, the string is searched,
108 counting zeros. There was no check for running off the end of the string,
109 which could happen if a new error number was added without updating the
110 string.
112 10. \K gave a compile-time error if it appeared in a lookbehind assersion.
114 11. \K was not working if it appeared in an atomic group or in a group that
115 was called as a "subroutine", or in an assertion. Perl 5.11 documents that
116 \K is "not well defined" if used in an assertion. PCRE now accepts it if
117 the assertion is positive, but not if it is negative.
119 12. Change 11 fortuitously reduced the size of the stack frame used in the
120 "match()" function of pcre_exec.c by one pointer. Forthcoming
121 implementation of support for (*MARK) will need an extra pointer on the
122 stack; I have reserved it now, so that the stack frame size does not
123 decrease.
125 13. A pattern such as (?P<L1>(?P<L2>0)|(?P>L2)(?P>L1)) in which the only other
126 item in branch that calls a recursion is a subroutine call - as in the
127 second branch in the above example - was incorrectly given the compile-
128 time error "recursive call could loop indefinitely" because pcre_compile()
129 was not correctly checking the subroutine for matching a non-empty string.
131 14. The checks for overrunning compiling workspace could trigger after an
132 overrun had occurred. This is a "should never occur" error, but it can be
133 triggered by pathological patterns such as hundreds of nested parentheses.
134 The checks now trigger 100 bytes before the end of the workspace.
136 15. Fix typo in configure.ac: "srtoq" should be "strtoq".
139 Version 8.01 19-Jan-2010
140 ------------------------
142 1. If a pattern contained a conditional subpattern with only one branch (in
143 particular, this includes all (*DEFINE) patterns), a call to pcre_study()
144 computed the wrong minimum data length (which is of course zero for such
145 subpatterns). This could cause incorrect "no match" results.
147 2. For patterns such as (?i)a(?-i)b|c where an option setting at the start of
148 the pattern is reset in the first branch, pcre_compile() failed with
149 "internal error: code overflow at offset...". This happened only when
150 the reset was to the original external option setting. (An optimization
151 abstracts leading options settings into an external setting, which was the
152 cause of this.)
154 3. A pattern such as ^(?!a(*SKIP)b) where a negative assertion contained one
155 of the verbs SKIP, PRUNE, or COMMIT, did not work correctly. When the
156 assertion pattern did not match (meaning that the assertion was true), it
157 was incorrectly treated as false if the SKIP had been reached during the
158 matching. This also applied to assertions used as conditions.
160 4. If an item that is not supported by pcre_dfa_exec() was encountered in an
161 assertion subpattern, including such a pattern used as a condition,
162 unpredictable results occurred, instead of the error return
165 5. The C++ GlobalReplace function was not working like Perl for the special
166 situation when an empty string is matched. It now does the fancy magic
167 stuff that is necessary.
169 6. In pcre_internal.h, obsolete includes to setjmp.h and stdarg.h have been
170 removed. (These were left over from very, very early versions of PCRE.)
172 7. Some cosmetic changes to the code to make life easier when compiling it
173 as part of something else:
175 (a) Change DEBUG to PCRE_DEBUG.
177 (b) In pcre_compile(), rename the member of the "branch_chain" structure
178 called "current" as "current_branch", to prevent a collision with the
179 Linux macro when compiled as a kernel module.
181 (c) In pcre_study(), rename the function set_bit() as set_table_bit(), to
182 prevent a collision with the Linux macro when compiled as a kernel
183 module.
185 8. In pcre_compile() there are some checks for integer overflows that used to
186 cast potentially large values to (double). This has been changed to that
187 when building, a check for int64_t is made, and if it is found, it is used
188 instead, thus avoiding the use of floating point arithmetic. (There is no
189 other use of FP in PCRE.) If int64_t is not found, the fallback is to
190 double.
192 9. Added two casts to avoid signed/unsigned warnings from VS Studio Express
193 2005 (difference between two addresses compared to an unsigned value).
195 10. Change the standard AC_CHECK_LIB test for libbz2 in configure.ac to a
196 custom one, because of the following reported problem in Windows:
198 - libbz2 uses the Pascal calling convention (WINAPI) for the functions
199 under Win32.
200 - The standard autoconf AC_CHECK_LIB fails to include "bzlib.h",
201 therefore missing the function definition.
202 - The compiler thus generates a "C" signature for the test function.
203 - The linker fails to find the "C" function.
204 - PCRE fails to configure if asked to do so against libbz2.
206 11. When running libtoolize from libtool-2.2.6b as part of autogen.sh, these
207 messages were output:
209 Consider adding `AC_CONFIG_MACRO_DIR([m4])' to configure.ac and
210 rerunning libtoolize, to keep the correct libtool macros in-tree.
211 Consider adding `-I m4' to ACLOCAL_AMFLAGS in Makefile.am.
213 I have done both of these things.
215 12. Although pcre_dfa_exec() does not use nearly as much stack as pcre_exec()
216 most of the time, it *can* run out if it is given a pattern that contains a
217 runaway infinite recursion. I updated the discussion in the pcrestack man
218 page.
220 13. Now that we have gone to the x.xx style of version numbers, the minor
221 version may start with zero. Using 08 or 09 is a bad idea because users
222 might check the value of PCRE_MINOR in their code, and 08 or 09 may be
223 interpreted as invalid octal numbers. I've updated the previous comment in
224 configure.ac, and also added a check that gives an error if 08 or 09 are
225 used.
227 14. Change 8.00/11 was not quite complete: code had been accidentally omitted,
228 causing partial matching to fail when the end of the subject matched \W
229 in a UTF-8 pattern where \W was quantified with a minimum of 3.
231 15. There were some discrepancies between the declarations in pcre_internal.h
232 of _pcre_is_newline(), _pcre_was_newline(), and _pcre_valid_utf8() and
233 their definitions. The declarations used "const uschar *" and the
234 definitions used USPTR. Even though USPTR is normally defined as "const
235 unsigned char *" (and uschar is typedeffed as "unsigned char"), it was
236 reported that: "This difference in casting confuses some C++ compilers, for
237 example, SunCC recognizes above declarations as different functions and
238 generates broken code for hbpcre." I have changed the declarations to use
239 USPTR.
241 16. GNU libtool is named differently on some systems. The autogen.sh script now
242 tries several variants such as glibtoolize (MacOSX) and libtoolize1x
243 (FreeBSD).
245 17. Applied Craig's patch that fixes an HP aCC compile error in pcre 8.00
246 (strtoXX undefined when compiling pcrecpp.cc). The patch contains this
247 comment: "Figure out how to create a longlong from a string: strtoll and
248 equivalent. It's not enough to call AC_CHECK_FUNCS: hpux has a strtoll, for
249 instance, but it only takes 2 args instead of 3!"
251 18. A subtle bug concerned with back references has been fixed by a change of
252 specification, with a corresponding code fix. A pattern such as
253 ^(xa|=?\1a)+$ which contains a back reference inside the group to which it
254 refers, was giving matches when it shouldn't. For example, xa=xaaa would
255 match that pattern. Interestingly, Perl (at least up to 5.11.3) has the
256 same bug. Such groups have to be quantified to be useful, or contained
257 inside another quantified group. (If there's no repetition, the reference
258 can never match.) The problem arises because, having left the group and
259 moved on to the rest of the pattern, a later failure that backtracks into
260 the group uses the captured value from the final iteration of the group
261 rather than the correct earlier one. I have fixed this in PCRE by forcing
262 any group that contains a reference to itself to be an atomic group; that
263 is, there cannot be any backtracking into it once it has completed. This is
264 similar to recursive and subroutine calls.
267 Version 8.00 19-Oct-09
268 ----------------------
270 1. The table for translating pcre_compile() error codes into POSIX error codes
271 was out-of-date, and there was no check on the pcre_compile() error code
272 being within the table. This could lead to an OK return being given in
273 error.
275 2. Changed the call to open a subject file in pcregrep from fopen(pathname,
276 "r") to fopen(pathname, "rb"), which fixed a problem with some of the tests
277 in a Windows environment.
279 3. The pcregrep --count option prints the count for each file even when it is
280 zero, as does GNU grep. However, pcregrep was also printing all files when
281 --files-with-matches was added. Now, when both options are given, it prints
282 counts only for those files that have at least one match. (GNU grep just
283 prints the file name in this circumstance, but including the count seems
284 more useful - otherwise, why use --count?) Also ensured that the
285 combination -clh just lists non-zero counts, with no names.
287 4. The long form of the pcregrep -F option was incorrectly implemented as
288 --fixed_strings instead of --fixed-strings. This is an incompatible change,
289 but it seems right to fix it, and I didn't think it was worth preserving
290 the old behaviour.
292 5. The command line items --regex=pattern and --regexp=pattern were not
293 recognized by pcregrep, which required --regex pattern or --regexp pattern
294 (with a space rather than an '='). The man page documented the '=' forms,
295 which are compatible with GNU grep; these now work.
297 6. No libpcreposix.pc file was created for pkg-config; there was just
298 libpcre.pc and libpcrecpp.pc. The omission has been rectified.
300 7. Added #ifndef SUPPORT_UCP into the pcre_ucd.c module, to reduce its size
301 when UCP support is not needed, by modifying the Python script that
302 generates it from Unicode data files. This should not matter if the module
303 is correctly used as a library, but I received one complaint about 50K of
304 unwanted data. My guess is that the person linked everything into his
305 program rather than using a library. Anyway, it does no harm.
307 8. A pattern such as /\x{123}{2,2}+/8 was incorrectly compiled; the trigger
308 was a minimum greater than 1 for a wide character in a possessive
309 repetition. The same bug could also affect patterns like /(\x{ff}{0,2})*/8
310 which had an unlimited repeat of a nested, fixed maximum repeat of a wide
311 character. Chaos in the form of incorrect output or a compiling loop could
312 result.
314 9. The restrictions on what a pattern can contain when partial matching is
315 requested for pcre_exec() have been removed. All patterns can now be
316 partially matched by this function. In addition, if there are at least two
317 slots in the offset vector, the offset of the earliest inspected character
318 for the match and the offset of the end of the subject are set in them when
319 PCRE_ERROR_PARTIAL is returned.
321 10. Partial matching has been split into two forms: PCRE_PARTIAL_SOFT, which is
322 synonymous with PCRE_PARTIAL, for backwards compatibility, and
323 PCRE_PARTIAL_HARD, which causes a partial match to supersede a full match,
324 and may be more useful for multi-segment matching.
326 11. Partial matching with pcre_exec() is now more intuitive. A partial match
327 used to be given if ever the end of the subject was reached; now it is
328 given only if matching could not proceed because another character was
329 needed. This makes a difference in some odd cases such as Z(*FAIL) with the
330 string "Z", which now yields "no match" instead of "partial match". In the
331 case of pcre_dfa_exec(), "no match" is given if every matching path for the
332 final character ended with (*FAIL).
334 12. Restarting a match using pcre_dfa_exec() after a partial match did not work
335 if the pattern had a "must contain" character that was already found in the
336 earlier partial match, unless partial matching was again requested. For
337 example, with the pattern /dog.(body)?/, the "must contain" character is
338 "g". If the first part-match was for the string "dog", restarting with
339 "sbody" failed. This bug has been fixed.
341 13. The string returned by pcre_dfa_exec() after a partial match has been
342 changed so that it starts at the first inspected character rather than the
343 first character of the match. This makes a difference only if the pattern
344 starts with a lookbehind assertion or \b or \B (\K is not supported by
345 pcre_dfa_exec()). It's an incompatible change, but it makes the two
346 matching functions compatible, and I think it's the right thing to do.
348 14. Added a pcredemo man page, created automatically from the pcredemo.c file,
349 so that the demonstration program is easily available in environments where
350 PCRE has not been installed from source.
352 15. Arranged to add -DPCRE_STATIC to cflags in libpcre.pc, libpcreposix.cp,
353 libpcrecpp.pc and pcre-config when PCRE is not compiled as a shared
354 library.
356 16. Added REG_UNGREEDY to the pcreposix interface, at the request of a user.
357 It maps to PCRE_UNGREEDY. It is not, of course, POSIX-compatible, but it
358 is not the first non-POSIX option to be added. Clearly some people find
359 these options useful.
361 17. If a caller to the POSIX matching function regexec() passes a non-zero
362 value for nmatch with a NULL value for pmatch, the value of
363 nmatch is forced to zero.
365 18. RunGrepTest did not have a test for the availability of the -u option of
366 the diff command, as RunTest does. It now checks in the same way as
367 RunTest, and also checks for the -b option.
369 19. If an odd number of negated classes containing just a single character
370 interposed, within parentheses, between a forward reference to a named
371 subpattern and the definition of the subpattern, compilation crashed with
372 an internal error, complaining that it could not find the referenced
373 subpattern. An example of a crashing pattern is /(?&A)(([^m])(?<A>))/.
374 [The bug was that it was starting one character too far in when skipping
375 over the character class, thus treating the ] as data rather than
376 terminating the class. This meant it could skip too much.]
378 20. Added PCRE_NOTEMPTY_ATSTART in order to be able to correctly implement the
379 /g option in pcretest when the pattern contains \K, which makes it possible
380 to have an empty string match not at the start, even when the pattern is
381 anchored. Updated pcretest and pcredemo to use this option.
383 21. If the maximum number of capturing subpatterns in a recursion was greater
384 than the maximum at the outer level, the higher number was returned, but
385 with unset values at the outer level. The correct (outer level) value is
386 now given.
388 22. If (*ACCEPT) appeared inside capturing parentheses, previous releases of
389 PCRE did not set those parentheses (unlike Perl). I have now found a way to
390 make it do so. The string so far is captured, making this feature
391 compatible with Perl.
393 23. The tests have been re-organized, adding tests 11 and 12, to make it
394 possible to check the Perl 5.10 features against Perl 5.10.
396 24. Perl 5.10 allows subroutine calls in lookbehinds, as long as the subroutine
397 pattern matches a fixed length string. PCRE did not allow this; now it
398 does. Neither allows recursion.
400 25. I finally figured out how to implement a request to provide the minimum
401 length of subject string that was needed in order to match a given pattern.
402 (It was back references and recursion that I had previously got hung up
403 on.) This code has now been added to pcre_study(); it finds a lower bound
404 to the length of subject needed. It is not necessarily the greatest lower
405 bound, but using it to avoid searching strings that are too short does give
406 some useful speed-ups. The value is available to calling programs via
407 pcre_fullinfo().
409 26. While implementing 25, I discovered to my embarrassment that pcretest had
410 not been passing the result of pcre_study() to pcre_dfa_exec(), so the
411 study optimizations had never been tested with that matching function.
412 Oops. What is worse, even when it was passed study data, there was a bug in
413 pcre_dfa_exec() that meant it never actually used it. Double oops. There
414 were also very few tests of studied patterns with pcre_dfa_exec().
416 27. If (?| is used to create subpatterns with duplicate numbers, they are now
417 allowed to have the same name, even if PCRE_DUPNAMES is not set. However,
418 on the other side of the coin, they are no longer allowed to have different
419 names, because these cannot be distinguished in PCRE, and this has caused
420 confusion. (This is a difference from Perl.)
422 28. When duplicate subpattern names are present (necessarily with different
423 numbers, as required by 27 above), and a test is made by name in a
424 conditional pattern, either for a subpattern having been matched, or for
425 recursion in such a pattern, all the associated numbered subpatterns are
426 tested, and the overall condition is true if the condition is true for any
427 one of them. This is the way Perl works, and is also more like the way
428 testing by number works.
431 Version 7.9 11-Apr-09
432 ---------------------
434 1. When building with support for bzlib/zlib (pcregrep) and/or readline
435 (pcretest), all targets were linked against these libraries. This included
436 libpcre, libpcreposix, and libpcrecpp, even though they do not use these
437 libraries. This caused unwanted dependencies to be created. This problem
438 has been fixed, and now only pcregrep is linked with bzlib/zlib and only
439 pcretest is linked with readline.
441 2. The "typedef int BOOL" in pcre_internal.h that was included inside the
442 "#ifndef FALSE" condition by an earlier change (probably 7.8/18) has been
443 moved outside it again, because FALSE and TRUE are already defined in AIX,
444 but BOOL is not.
446 3. The pcre_config() function was treating the PCRE_MATCH_LIMIT and
447 PCRE_MATCH_LIMIT_RECURSION values as ints, when they should be long ints.
449 4. The pcregrep documentation said spaces were inserted as well as colons (or
450 hyphens) following file names and line numbers when outputting matching
451 lines. This is not true; no spaces are inserted. I have also clarified the
452 wording for the --colour (or --color) option.
454 5. In pcregrep, when --colour was used with -o, the list of matching strings
455 was not coloured; this is different to GNU grep, so I have changed it to be
456 the same.
458 6. When --colo(u)r was used in pcregrep, only the first matching substring in
459 each matching line was coloured. Now it goes on to look for further matches
460 of any of the test patterns, which is the same behaviour as GNU grep.
462 7. A pattern that could match an empty string could cause pcregrep to loop; it
463 doesn't make sense to accept an empty string match in pcregrep, so I have
464 locked it out (using PCRE's PCRE_NOTEMPTY option). By experiment, this
465 seems to be how GNU grep behaves.
467 8. The pattern (?(?=.*b)b|^) was incorrectly compiled as "match must be at
468 start or after a newline", because the conditional assertion was not being
469 correctly handled. The rule now is that both the assertion and what follows
470 in the first alternative must satisfy the test.
472 9. If auto-callout was enabled in a pattern with a conditional group whose
473 condition was an assertion, PCRE could crash during matching, both with
474 pcre_exec() and pcre_dfa_exec().
476 10. The PCRE_DOLLAR_ENDONLY option was not working when pcre_dfa_exec() was
477 used for matching.
479 11. Unicode property support in character classes was not working for
480 characters (bytes) greater than 127 when not in UTF-8 mode.
482 12. Added the -M command line option to pcretest.
484 14. Added the non-standard REG_NOTEMPTY option to the POSIX interface.
486 15. Added the PCRE_NO_START_OPTIMIZE match-time option.
488 16. Added comments and documentation about mis-use of no_arg in the C++
489 wrapper.
491 17. Implemented support for UTF-8 encoding in EBCDIC environments, a patch
492 from Martin Jerabek that uses macro names for all relevant character and
493 string constants.
495 18. Added to pcre_internal.h two configuration checks: (a) If both EBCDIC and
496 SUPPORT_UTF8 are set, give an error; (b) If SUPPORT_UCP is set without
497 SUPPORT_UTF8, define SUPPORT_UTF8. The "configure" script handles both of
498 these, but not everybody uses configure.
500 19. A conditional group that had only one branch was not being correctly
501 recognized as an item that could match an empty string. This meant that an
502 enclosing group might also not be so recognized, causing infinite looping
503 (and probably a segfault) for patterns such as ^"((?(?=[a])[^"])|b)*"$
504 with the subject "ab", where knowledge that the repeated group can match
505 nothing is needed in order to break the loop.
507 20. If a pattern that was compiled with callouts was matched using pcre_dfa_
508 exec(), but without supplying a callout function, matching went wrong.
510 21. If PCRE_ERROR_MATCHLIMIT occurred during a recursion, there was a memory
511 leak if the size of the offset vector was greater than 30. When the vector
512 is smaller, the saved offsets during recursion go onto a local stack
513 vector, but for larger vectors malloc() is used. It was failing to free
514 when the recursion yielded PCRE_ERROR_MATCH_LIMIT (or any other "abnormal"
515 error, in fact).
517 22. There was a missing #ifdef SUPPORT_UTF8 round one of the variables in the
518 heapframe that is used only when UTF-8 support is enabled. This caused no
519 problem, but was untidy.
521 23. Steven Van Ingelgem's patch to CMakeLists.txt to change the name
522 CMAKE_BINARY_DIR to PROJECT_BINARY_DIR so that it works when PCRE is
523 included within another project.
525 24. Steven Van Ingelgem's patches to add more options to the CMake support,
526 slightly modified by me:
528 (a) PCRE_BUILD_TESTS can be set OFF not to build the tests, including
529 not building pcregrep.
531 (b) PCRE_BUILD_PCREGREP can be see OFF not to build pcregrep, but only
532 if PCRE_BUILD_TESTS is also set OFF, because the tests use pcregrep.
534 25. Forward references, both numeric and by name, in patterns that made use of
535 duplicate group numbers, could behave incorrectly or give incorrect errors,
536 because when scanning forward to find the reference group, PCRE was not
537 taking into account the duplicate group numbers. A pattern such as
538 ^X(?3)(a)(?|(b)|(q))(Y) is an example.
540 26. Changed a few more instances of "const unsigned char *" to USPTR, making
541 the feature of a custom pointer more persuasive (as requested by a user).
543 27. Wrapped the definitions of fileno and isatty for Windows, which appear in
544 pcretest.c, inside #ifndefs, because it seems they are sometimes already
545 pre-defined.
547 28. Added support for (*UTF8) at the start of a pattern.
549 29. Arrange for flags added by the "release type" setting in CMake to be shown
550 in the configuration summary.
553 Version 7.8 05-Sep-08
554 ---------------------
556 1. Replaced UCP searching code with optimized version as implemented for Ad
557 Muncher (http://www.admuncher.com/) by Peter Kankowski. This uses a two-
558 stage table and inline lookup instead of a function, giving speed ups of 2
559 to 5 times on some simple patterns that I tested. Permission was given to
560 distribute the MultiStage2.py script that generates the tables (it's not in
561 the tarball, but is in the Subversion repository).
563 2. Updated the Unicode datatables to Unicode 5.1.0. This adds yet more
564 scripts.
566 3. Change 12 for 7.7 introduced a bug in pcre_study() when a pattern contained
567 a group with a zero qualifier. The result of the study could be incorrect,
568 or the function might crash, depending on the pattern.
570 4. Caseless matching was not working for non-ASCII characters in back
571 references. For example, /(\x{de})\1/8i was not matching \x{de}\x{fe}.
572 It now works when Unicode Property Support is available.
574 5. In pcretest, an escape such as \x{de} in the data was always generating
575 a UTF-8 string, even in non-UTF-8 mode. Now it generates a single byte in
576 non-UTF-8 mode. If the value is greater than 255, it gives a warning about
577 truncation.
579 6. Minor bugfix in pcrecpp.cc (change "" == ... to NULL == ...).
581 7. Added two (int) casts to pcregrep when printing the difference of two
582 pointers, in case they are 64-bit values.
584 8. Added comments about Mac OS X stack usage to the pcrestack man page and to
585 test 2 if it fails.
587 9. Added PCRE_CALL_CONVENTION just before the names of all exported functions,
588 and a #define of that name to empty if it is not externally set. This is to
589 allow users of MSVC to set it if necessary.
591 10. The PCRE_EXP_DEFN macro which precedes exported functions was missing from
592 the convenience functions in the pcre_get.c source file.
594 11. An option change at the start of a pattern that had top-level alternatives
595 could cause overwriting and/or a crash. This command provoked a crash in
596 some environments:
598 printf "/(?i)[\xc3\xa9\xc3\xbd]|[\xc3\xa9\xc3\xbdA]/8\n" | pcretest
600 This potential security problem was recorded as CVE-2008-2371.
602 12. For a pattern where the match had to start at the beginning or immediately
603 after a newline (e.g /.*anything/ without the DOTALL flag), pcre_exec() and
604 pcre_dfa_exec() could read past the end of the passed subject if there was
605 no match. To help with detecting such bugs (e.g. with valgrind), I modified
606 pcretest so that it places the subject at the end of its malloc-ed buffer.
608 13. The change to pcretest in 12 above threw up a couple more cases when pcre_
609 exec() might read past the end of the data buffer in UTF-8 mode.
611 14. A similar bug to 7.3/2 existed when the PCRE_FIRSTLINE option was set and
612 the data contained the byte 0x85 as part of a UTF-8 character within its
613 first line. This applied both to normal and DFA matching.
615 15. Lazy qualifiers were not working in some cases in UTF-8 mode. For example,
616 /^[^d]*?$/8 failed to match "abc".
618 16. Added a missing copyright notice to pcrecpp_internal.h.
620 17. Make it more clear in the documentation that values returned from
621 pcre_exec() in ovector are byte offsets, not character counts.
623 18. Tidied a few places to stop certain compilers from issuing warnings.
625 19. Updated the Virtual Pascal + BCC files to compile the latest v7.7, as
626 supplied by Stefan Weber. I made a further small update for 7.8 because
627 there is a change of source arrangements: the pcre_searchfuncs.c module is
628 replaced by pcre_ucd.c.
631 Version 7.7 07-May-08
632 ---------------------
634 1. Applied Craig's patch to sort out a long long problem: "If we can't convert
635 a string to a long long, pretend we don't even have a long long." This is
636 done by checking for the strtoq, strtoll, and _strtoi64 functions.
638 2. Applied Craig's patch to pcrecpp.cc to restore ABI compatibility with
639 pre-7.6 versions, which defined a global no_arg variable instead of putting
640 it in the RE class. (See also #8 below.)
642 3. Remove a line of dead code, identified by coverity and reported by Nuno
643 Lopes.
645 4. Fixed two related pcregrep bugs involving -r with --include or --exclude:
647 (1) The include/exclude patterns were being applied to the whole pathnames
648 of files, instead of just to the final components.
650 (2) If there was more than one level of directory, the subdirectories were
651 skipped unless they satisfied the include/exclude conditions. This is
652 inconsistent with GNU grep (and could even be seen as contrary to the
653 pcregrep specification - which I improved to make it absolutely clear).
654 The action now is always to scan all levels of directory, and just
655 apply the include/exclude patterns to regular files.
657 5. Added the --include_dir and --exclude_dir patterns to pcregrep, and used
658 --exclude_dir in the tests to avoid scanning .svn directories.
660 6. Applied Craig's patch to the QuoteMeta function so that it escapes the
661 NUL character as backslash + 0 rather than backslash + NUL, because PCRE
662 doesn't support NULs in patterns.
664 7. Added some missing "const"s to declarations of static tables in
665 pcre_compile.c and pcre_dfa_exec.c.
667 8. Applied Craig's patch to pcrecpp.cc to fix a problem in OS X that was
668 caused by fix #2 above. (Subsequently also a second patch to fix the
669 first patch. And a third patch - this was a messy problem.)
671 9. Applied Craig's patch to remove the use of push_back().
673 10. Applied Alan Lehotsky's patch to add REG_STARTEND support to the POSIX
674 matching function regexec().
676 11. Added support for the Oniguruma syntax \g<name>, \g<n>, \g'name', \g'n',
677 which, however, unlike Perl's \g{...}, are subroutine calls, not back
678 references. PCRE supports relative numbers with this syntax (I don't think
679 Oniguruma does).
681 12. Previously, a group with a zero repeat such as (...){0} was completely
682 omitted from the compiled regex. However, this means that if the group
683 was called as a subroutine from elsewhere in the pattern, things went wrong
684 (an internal error was given). Such groups are now left in the compiled
685 pattern, with a new opcode that causes them to be skipped at execution
686 time.
688 13. Added the PCRE_JAVASCRIPT_COMPAT option. This makes the following changes
689 to the way PCRE behaves:
691 (a) A lone ] character is dis-allowed (Perl treats it as data).
693 (b) A back reference to an unmatched subpattern matches an empty string
694 (Perl fails the current match path).
696 (c) A data ] in a character class must be notated as \] because if the
697 first data character in a class is ], it defines an empty class. (In
698 Perl it is not possible to have an empty class.) The empty class []
699 never matches; it forces failure and is equivalent to (*FAIL) or (?!).
700 The negative empty class [^] matches any one character, independently
701 of the DOTALL setting.
703 14. A pattern such as /(?2)[]a()b](abc)/ which had a forward reference to a
704 non-existent subpattern following a character class starting with ']' and
705 containing () gave an internal compiling error instead of "reference to
706 non-existent subpattern". Fortunately, when the pattern did exist, the
707 compiled code was correct. (When scanning forwards to check for the
708 existencd of the subpattern, it was treating the data ']' as terminating
709 the class, so got the count wrong. When actually compiling, the reference
710 was subsequently set up correctly.)
712 15. The "always fail" assertion (?!) is optimzed to (*FAIL) by pcre_compile;
713 it was being rejected as not supported by pcre_dfa_exec(), even though
714 other assertions are supported. I have made pcre_dfa_exec() support
715 (*FAIL).
717 16. The implementation of 13c above involved the invention of a new opcode,
718 OP_ALLANY, which is like OP_ANY but doesn't check the /s flag. Since /s
719 cannot be changed at match time, I realized I could make a small
720 improvement to matching performance by compiling OP_ALLANY instead of
721 OP_ANY for "." when DOTALL was set, and then removing the runtime tests
722 on the OP_ANY path.
724 17. Compiling pcretest on Windows with readline support failed without the
725 following two fixes: (1) Make the unistd.h include conditional on
726 HAVE_UNISTD_H; (2) #define isatty and fileno as _isatty and _fileno.
728 18. Changed CMakeLists.txt and cmake/FindReadline.cmake to arrange for the
729 ncurses library to be included for pcretest when ReadLine support is
730 requested, but also to allow for it to be overridden. This patch came from
731 Daniel Bergström.
733 19. There was a typo in the file ucpinternal.h where f0_rangeflag was defined
734 as 0x00f00000 instead of 0x00800000. Luckily, this would not have caused
735 any errors with the current Unicode tables. Thanks to Peter Kankowski for
736 spotting this.
739 Version 7.6 28-Jan-08
740 ---------------------
742 1. A character class containing a very large number of characters with
743 codepoints greater than 255 (in UTF-8 mode, of course) caused a buffer
744 overflow.
746 2. Patch to cut out the "long long" test in pcrecpp_unittest when
747 HAVE_LONG_LONG is not defined.
749 3. Applied Christian Ehrlicher's patch to update the CMake build files to
750 bring them up to date and include new features. This patch includes:
752 - Fixed PH's badly added libz and libbz2 support.
753 - Fixed a problem with static linking.
754 - Added pcredemo. [But later removed - see 7 below.]
755 - Fixed dftables problem and added an option.
756 - Added a number of HAVE_XXX tests, including HAVE_WINDOWS_H and
758 - Added readline support for pcretest.
759 - Added an listing of the option settings after cmake has run.
761 4. A user submitted a patch to Makefile that makes it easy to create
762 "pcre.dll" under mingw when using Configure/Make. I added stuff to
763 Makefile.am that cause it to include this special target, without
764 affecting anything else. Note that the same mingw target plus all
765 the other distribution libraries and programs are now supported
766 when configuring with CMake (see 6 below) instead of with
767 Configure/Make.
769 5. Applied Craig's patch that moves no_arg into the RE class in the C++ code.
770 This is an attempt to solve the reported problem "pcrecpp::no_arg is not
771 exported in the Windows port". It has not yet been confirmed that the patch
772 solves the problem, but it does no harm.
774 6. Applied Sheri's patch to CMakeLists.txt to add NON_STANDARD_LIB_PREFIX and
775 NON_STANDARD_LIB_SUFFIX for dll names built with mingw when configured
776 with CMake, and also correct the comment about stack recursion.
778 7. Remove the automatic building of pcredemo from the ./configure system and
779 from CMakeLists.txt. The whole idea of pcredemo.c is that it is an example
780 of a program that users should build themselves after PCRE is installed, so
781 building it automatically is not really right. What is more, it gave
782 trouble in some build environments.
784 8. Further tidies to CMakeLists.txt from Sheri and Christian.
787 Version 7.5 10-Jan-08
788 ---------------------
790 1. Applied a patch from Craig: "This patch makes it possible to 'ignore'
791 values in parens when parsing an RE using the C++ wrapper."
793 2. Negative specials like \S did not work in character classes in UTF-8 mode.
794 Characters greater than 255 were excluded from the class instead of being
795 included.
797 3. The same bug as (2) above applied to negated POSIX classes such as
798 [:^space:].
800 4. PCRECPP_STATIC was referenced in pcrecpp_internal.h, but nowhere was it
801 defined or documented. It seems to have been a typo for PCRE_STATIC, so
802 I have changed it.
804 5. The construct (?&) was not diagnosed as a syntax error (it referenced the
805 first named subpattern) and a construct such as (?&a) would reference the
806 first named subpattern whose name started with "a" (in other words, the
807 length check was missing). Both these problems are fixed. "Subpattern name
808 expected" is now given for (?&) (a zero-length name), and this patch also
809 makes it give the same error for \k'' (previously it complained that that
810 was a reference to a non-existent subpattern).
812 6. The erroneous patterns (?+-a) and (?-+a) give different error messages;
813 this is right because (?- can be followed by option settings as well as by
814 digits. I have, however, made the messages clearer.
816 7. Patterns such as (?(1)a|b) (a pattern that contains fewer subpatterns
817 than the number used in the conditional) now cause a compile-time error.
818 This is actually not compatible with Perl, which accepts such patterns, but
819 treats the conditional as always being FALSE (as PCRE used to), but it
820 seems to me that giving a diagnostic is better.
822 8. Change "alphameric" to the more common word "alphanumeric" in comments
823 and messages.
825 9. Fix two occurrences of "backslash" in comments that should have been
826 "backspace".
828 10. Remove two redundant lines of code that can never be obeyed (their function
829 was moved elsewhere).
831 11. The program that makes PCRE's Unicode character property table had a bug
832 which caused it to generate incorrect table entries for sequences of
833 characters that have the same character type, but are in different scripts.
834 It amalgamated them into a single range, with the script of the first of
835 them. In other words, some characters were in the wrong script. There were
836 thirteen such cases, affecting characters in the following ranges:
838 U+002b0 - U+002c1
839 U+0060c - U+0060d
840 U+0061e - U+00612
841 U+0064b - U+0065e
842 U+0074d - U+0076d
843 U+01800 - U+01805
844 U+01d00 - U+01d77
845 U+01d9b - U+01dbf
846 U+0200b - U+0200f
847 U+030fc - U+030fe
848 U+03260 - U+0327f
849 U+0fb46 - U+0fbb1
850 U+10450 - U+1049d
852 12. The -o option (show only the matching part of a line) for pcregrep was not
853 compatible with GNU grep in that, if there was more than one match in a
854 line, it showed only the first of them. It now behaves in the same way as
855 GNU grep.
857 13. If the -o and -v options were combined for pcregrep, it printed a blank
858 line for every non-matching line. GNU grep prints nothing, and pcregrep now
859 does the same. The return code can be used to tell if there were any
860 non-matching lines.
862 14. Added --file-offsets and --line-offsets to pcregrep.
864 15. The pattern (?=something)(?R) was not being diagnosed as a potentially
865 infinitely looping recursion. The bug was that positive lookaheads were not
866 being skipped when checking for a possible empty match (negative lookaheads
867 and both kinds of lookbehind were skipped).
869 16. Fixed two typos in the Windows-only code in pcregrep.c, and moved the
870 inclusion of <windows.h> to before rather than after the definition of
871 INVALID_FILE_ATTRIBUTES (patch from David Byron).
873 17. Specifying a possessive quantifier with a specific limit for a Unicode
874 character property caused pcre_compile() to compile bad code, which led at
875 runtime to PCRE_ERROR_INTERNAL (-14). Examples of patterns that caused this
876 are: /\p{Zl}{2,3}+/8 and /\p{Cc}{2}+/8. It was the possessive "+" that
877 caused the error; without that there was no problem.
879 18. Added --enable-pcregrep-libz and --enable-pcregrep-libbz2.
881 19. Added --enable-pcretest-libreadline.
883 20. In pcrecpp.cc, the variable 'count' was incremented twice in
884 RE::GlobalReplace(). As a result, the number of replacements returned was
885 double what it should be. I removed one of the increments, but Craig sent a
886 later patch that removed the other one (the right fix) and added unit tests
887 that check the return values (which was not done before).
889 21. Several CMake things:
891 (1) Arranged that, when cmake is used on Unix, the libraries end up with
892 the names libpcre and libpcreposix, not just pcre and pcreposix.
894 (2) The above change means that pcretest and pcregrep are now correctly
895 linked with the newly-built libraries, not previously installed ones.
899 22. In UTF-8 mode, with newline set to "any", a pattern such as .*a.*=.b.*
900 crashed when matching a string such as a\x{2029}b (note that \x{2029} is a
901 UTF-8 newline character). The key issue is that the pattern starts .*;
902 this means that the match must be either at the beginning, or after a
903 newline. The bug was in the code for advancing after a failed match and
904 checking that the new position followed a newline. It was not taking
905 account of UTF-8 characters correctly.
907 23. PCRE was behaving differently from Perl in the way it recognized POSIX
908 character classes. PCRE was not treating the sequence [:...:] as a
909 character class unless the ... were all letters. Perl, however, seems to
910 allow any characters between [: and :], though of course it rejects as
911 unknown any "names" that contain non-letters, because all the known class
912 names consist only of letters. Thus, Perl gives an error for [[:1234:]],
913 for example, whereas PCRE did not - it did not recognize a POSIX character
914 class. This seemed a bit dangerous, so the code has been changed to be
915 closer to Perl. The behaviour is not identical to Perl, because PCRE will
916 diagnose an unknown class for, for example, [[:l\ower:]] where Perl will
917 treat it as [[:lower:]]. However, PCRE does now give "unknown" errors where
918 Perl does, and where it didn't before.
920 24. Rewrite so as to remove the single use of %n from pcregrep because in some
921 Windows environments %n is disabled by default.
924 Version 7.4 21-Sep-07
925 ---------------------
927 1. Change 7.3/28 was implemented for classes by looking at the bitmap. This
928 means that a class such as [\s] counted as "explicit reference to CR or
929 LF". That isn't really right - the whole point of the change was to try to
930 help when there was an actual mention of one of the two characters. So now
931 the change happens only if \r or \n (or a literal CR or LF) character is
932 encountered.
934 2. The 32-bit options word was also used for 6 internal flags, but the numbers
935 of both had grown to the point where there were only 3 bits left.
936 Fortunately, there was spare space in the data structure, and so I have
937 moved the internal flags into a new 16-bit field to free up more option
938 bits.
940 3. The appearance of (?J) at the start of a pattern set the DUPNAMES option,
941 but did not set the internal JCHANGED flag - either of these is enough to
942 control the way the "get" function works - but the PCRE_INFO_JCHANGED
943 facility is supposed to tell if (?J) was ever used, so now (?J) at the
944 start sets both bits.
946 4. Added options (at build time, compile time, exec time) to change \R from
947 matching any Unicode line ending sequence to just matching CR, LF, or CRLF.
949 5. doc/pcresyntax.html was missing from the distribution.
951 6. Put back the definition of PCRE_ERROR_NULLWSLIMIT, for backward
952 compatibility, even though it is no longer used.
954 7. Added macro for snprintf to pcrecpp_unittest.cc and also for strtoll and
955 strtoull to pcrecpp.cc to select the available functions in WIN32 when the
956 windows.h file is present (where different names are used). [This was
957 reversed later after testing - see 16 below.]
959 8. Changed all #include <config.h> to #include "config.h". There were also
960 some further <pcre.h> cases that I changed to "pcre.h".
962 9. When pcregrep was used with the --colour option, it missed the line ending
963 sequence off the lines that it output.
965 10. It was pointed out to me that arrays of string pointers cause lots of
966 relocations when a shared library is dynamically loaded. A technique of
967 using a single long string with a table of offsets can drastically reduce
968 these. I have refactored PCRE in four places to do this. The result is
969 dramatic:
971 Originally: 290
972 After changing UCP table: 187
973 After changing error message table: 43
974 After changing table of "verbs" 36
975 After changing table of Posix names 22
977 Thanks to the folks working on Gregex for glib for this insight.
979 11. --disable-stack-for-recursion caused compiling to fail unless -enable-
980 unicode-properties was also set.
982 12. Updated the tests so that they work when \R is defaulted to ANYCRLF.
984 13. Added checks for ANY and ANYCRLF to pcrecpp.cc where it previously
985 checked only for CRLF.
987 14. Added casts to pcretest.c to avoid compiler warnings.
989 15. Added Craig's patch to various pcrecpp modules to avoid compiler warnings.
991 16. Added Craig's patch to remove the WINDOWS_H tests, that were not working,
992 and instead check for _strtoi64 explicitly, and avoid the use of snprintf()
993 entirely. This removes changes made in 7 above.
995 17. The CMake files have been updated, and there is now more information about
996 building with CMake in the NON-UNIX-USE document.
999 Version 7.3 28-Aug-07
1000 ---------------------
1002 1. In the rejigging of the build system that eventually resulted in 7.1, the
1003 line "#include <pcre.h>" was included in pcre_internal.h. The use of angle
1004 brackets there is not right, since it causes compilers to look for an
1005 installed pcre.h, not the version that is in the source that is being
1006 compiled (which of course may be different). I have changed it back to:
1008 #include "pcre.h"
1010 I have a vague recollection that the change was concerned with compiling in
1011 different directories, but in the new build system, that is taken care of
1012 by the VPATH setting the Makefile.
1014 2. The pattern .*$ when run in not-DOTALL UTF-8 mode with newline=any failed
1015 when the subject happened to end in the byte 0x85 (e.g. if the last
1016 character was \x{1ec5}). *Character* 0x85 is one of the "any" newline
1017 characters but of course it shouldn't be taken as a newline when it is part
1018 of another character. The bug was that, for an unlimited repeat of . in
1019 not-DOTALL UTF-8 mode, PCRE was advancing by bytes rather than by
1020 characters when looking for a newline.
1022 3. A small performance improvement in the DOTALL UTF-8 mode .* case.
1024 4. Debugging: adjusted the names of opcodes for different kinds of parentheses
1025 in debug output.
1027 5. Arrange to use "%I64d" instead of "%lld" and "%I64u" instead of "%llu" for
1028 long printing in the pcrecpp unittest when running under MinGW.
1030 6. ESC_K was left out of the EBCDIC table.
1032 7. Change 7.0/38 introduced a new limit on the number of nested non-capturing
1033 parentheses; I made it 1000, which seemed large enough. Unfortunately, the
1034 limit also applies to "virtual nesting" when a pattern is recursive, and in
1035 this case 1000 isn't so big. I have been able to remove this limit at the
1036 expense of backing off one optimization in certain circumstances. Normally,
1037 when pcre_exec() would call its internal match() function recursively and
1038 immediately return the result unconditionally, it uses a "tail recursion"
1039 feature to save stack. However, when a subpattern that can match an empty
1040 string has an unlimited repetition quantifier, it no longer makes this
1041 optimization. That gives it a stack frame in which to save the data for
1042 checking that an empty string has been matched. Previously this was taken
1043 from the 1000-entry workspace that had been reserved. So now there is no
1044 explicit limit, but more stack is used.
1046 8. Applied Daniel's patches to solve problems with the import/export magic
1047 syntax that is required for Windows, and which was going wrong for the
1048 pcreposix and pcrecpp parts of the library. These were overlooked when this
1049 problem was solved for the main library.
1051 9. There were some crude static tests to avoid integer overflow when computing
1052 the size of patterns that contain repeated groups with explicit upper
1053 limits. As the maximum quantifier is 65535, the maximum group length was
1054 set at 30,000 so that the product of these two numbers did not overflow a
1055 32-bit integer. However, it turns out that people want to use groups that
1056 are longer than 30,000 bytes (though not repeat them that many times).
1057 Change 7.0/17 (the refactoring of the way the pattern size is computed) has
1058 made it possible to implement the integer overflow checks in a much more
1059 dynamic way, which I have now done. The artificial limitation on group
1060 length has been removed - we now have only the limit on the total length of
1061 the compiled pattern, which depends on the LINK_SIZE setting.
1063 10. Fixed a bug in the documentation for get/copy named substring when
1064 duplicate names are permitted. If none of the named substrings are set, the
1065 functions return PCRE_ERROR_NOSUBSTRING (7); the doc said they returned an
1066 empty string.
1068 11. Because Perl interprets \Q...\E at a high level, and ignores orphan \E
1069 instances, patterns such as [\Q\E] or [\E] or even [^\E] cause an error,
1070 because the ] is interpreted as the first data character and the
1071 terminating ] is not found. PCRE has been made compatible with Perl in this
1072 regard. Previously, it interpreted [\Q\E] as an empty class, and [\E] could
1073 cause memory overwriting.
1075 10. Like Perl, PCRE automatically breaks an unlimited repeat after an empty
1076 string has been matched (to stop an infinite loop). It was not recognizing
1077 a conditional subpattern that could match an empty string if that
1078 subpattern was within another subpattern. For example, it looped when
1079 trying to match (((?(1)X|))*) but it was OK with ((?(1)X|)*) where the
1080 condition was not nested. This bug has been fixed.
1082 12. A pattern like \X?\d or \P{L}?\d in non-UTF-8 mode could cause a backtrack
1083 past the start of the subject in the presence of bytes with the top bit
1084 set, for example "\x8aBCD".
1086 13. Added Perl 5.10 experimental backtracking controls (*FAIL), (*F), (*PRUNE),
1087 (*SKIP), (*THEN), (*COMMIT), and (*ACCEPT).
1089 14. Optimized (?!) to (*FAIL).
1091 15. Updated the test for a valid UTF-8 string to conform to the later RFC 3629.
1092 This restricts code points to be within the range 0 to 0x10FFFF, excluding
1093 the "low surrogate" sequence 0xD800 to 0xDFFF. Previously, PCRE allowed the
1094 full range 0 to 0x7FFFFFFF, as defined by RFC 2279. Internally, it still
1095 does: it's just the validity check that is more restrictive.
1097 16. Inserted checks for integer overflows during escape sequence (backslash)
1098 processing, and also fixed erroneous offset values for syntax errors during
1099 backslash processing.
1101 17. Fixed another case of looking too far back in non-UTF-8 mode (cf 12 above)
1102 for patterns like [\PPP\x8a]{1,}\x80 with the subject "A\x80".
1104 18. An unterminated class in a pattern like (?1)\c[ with a "forward reference"
1105 caused an overrun.
1107 19. A pattern like (?:[\PPa*]*){8,} which had an "extended class" (one with
1108 something other than just ASCII characters) inside a group that had an
1109 unlimited repeat caused a loop at compile time (while checking to see
1110 whether the group could match an empty string).
1112 20. Debugging a pattern containing \p or \P could cause a crash. For example,
1113 [\P{Any}] did so. (Error in the code for printing property names.)
1115 21. An orphan \E inside a character class could cause a crash.
1117 22. A repeated capturing bracket such as (A)? could cause a wild memory
1118 reference during compilation.
1120 23. There are several functions in pcre_compile() that scan along a compiled
1121 expression for various reasons (e.g. to see if it's fixed length for look
1122 behind). There were bugs in these functions when a repeated \p or \P was
1123 present in the pattern. These operators have additional parameters compared
1124 with \d, etc, and these were not being taken into account when moving along
1125 the compiled data. Specifically:
1127 (a) A item such as \p{Yi}{3} in a lookbehind was not treated as fixed
1128 length.
1130 (b) An item such as \pL+ within a repeated group could cause crashes or
1131 loops.
1133 (c) A pattern such as \p{Yi}+(\P{Yi}+)(?1) could give an incorrect
1134 "reference to non-existent subpattern" error.
1136 (d) A pattern like (\P{Yi}{2}\277)? could loop at compile time.
1138 24. A repeated \S or \W in UTF-8 mode could give wrong answers when multibyte
1139 characters were involved (for example /\S{2}/8g with "A\x{a3}BC").
1141 25. Using pcregrep in multiline, inverted mode (-Mv) caused it to loop.
1143 26. Patterns such as [\P{Yi}A] which include \p or \P and just one other
1144 character were causing crashes (broken optimization).
1146 27. Patterns such as (\P{Yi}*\277)* (group with possible zero repeat containing
1147 \p or \P) caused a compile-time loop.
1149 28. More problems have arisen in unanchored patterns when CRLF is a valid line
1150 break. For example, the unstudied pattern [\r\n]A does not match the string
1151 "\r\nA" because change 7.0/46 below moves the current point on by two
1152 characters after failing to match at the start. However, the pattern \nA
1153 *does* match, because it doesn't start till \n, and if [\r\n]A is studied,
1154 the same is true. There doesn't seem any very clean way out of this, but
1155 what I have chosen to do makes the common cases work: PCRE now takes note
1156 of whether there can be an explicit match for \r or \n anywhere in the
1157 pattern, and if so, 7.0/46 no longer applies. As part of this change,
1158 there's a new PCRE_INFO_HASCRORLF option for finding out whether a compiled
1159 pattern has explicit CR or LF references.
1161 29. Added (*CR) etc for changing newline setting at start of pattern.
1164 Version 7.2 19-Jun-07
1165 ---------------------
1167 1. If the fr_FR locale cannot be found for test 3, try the "french" locale,
1168 which is apparently normally available under Windows.
1170 2. Re-jig the pcregrep tests with different newline settings in an attempt
1171 to make them independent of the local environment's newline setting.
1173 3. Add code to configure.ac to remove -g from the CFLAGS default settings.
1175 4. Some of the "internals" tests were previously cut out when the link size
1176 was not 2, because the output contained actual offsets. The recent new
1177 "Z" feature of pcretest means that these can be cut out, making the tests
1178 usable with all link sizes.
1180 5. Implemented Stan Switzer's goto replacement for longjmp() when not using
1181 stack recursion. This gives a massive performance boost under BSD, but just
1182 a small improvement under Linux. However, it saves one field in the frame
1183 in all cases.
1185 6. Added more features from the forthcoming Perl 5.10:
1187 (a) (?-n) (where n is a string of digits) is a relative subroutine or
1188 recursion call. It refers to the nth most recently opened parentheses.
1190 (b) (?+n) is also a relative subroutine call; it refers to the nth next
1191 to be opened parentheses.
1193 (c) Conditions that refer to capturing parentheses can be specified
1194 relatively, for example, (?(-2)... or (?(+3)...
1196 (d) \K resets the start of the current match so that everything before
1197 is not part of it.
1199 (e) \k{name} is synonymous with \k<name> and \k'name' (.NET compatible).
1201 (f) \g{name} is another synonym - part of Perl 5.10's unification of
1202 reference syntax.
1204 (g) (?| introduces a group in which the numbering of parentheses in each
1205 alternative starts with the same number.
1207 (h) \h, \H, \v, and \V match horizontal and vertical whitespace.
1209 7. Added two new calls to pcre_fullinfo(): PCRE_INFO_OKPARTIAL and
1212 8. A pattern such as (.*(.)?)* caused pcre_exec() to fail by either not
1213 terminating or by crashing. Diagnosed by Viktor Griph; it was in the code
1214 for detecting groups that can match an empty string.
1216 9. A pattern with a very large number of alternatives (more than several
1217 hundred) was running out of internal workspace during the pre-compile
1218 phase, where pcre_compile() figures out how much memory will be needed. A
1219 bit of new cunning has reduced the workspace needed for groups with
1220 alternatives. The 1000-alternative test pattern now uses 12 bytes of
1221 workspace instead of running out of the 4096 that are available.
1223 10. Inserted some missing (unsigned int) casts to get rid of compiler warnings.
1225 11. Applied patch from Google to remove an optimization that didn't quite work.
1226 The report of the bug said:
1228 pcrecpp::RE("a*").FullMatch("aaa") matches, while
1229 pcrecpp::RE("a*?").FullMatch("aaa") does not, and
1230 pcrecpp::RE("a*?\\z").FullMatch("aaa") does again.
1232 12. If \p or \P was used in non-UTF-8 mode on a character greater than 127
1233 it matched the wrong number of bytes.
1236 Version 7.1 24-Apr-07
1237 ---------------------
1239 1. Applied Bob Rossi and Daniel G's patches to convert the build system to one
1240 that is more "standard", making use of automake and other Autotools. There
1241 is some re-arrangement of the files and adjustment of comments consequent
1242 on this.
1244 2. Part of the patch fixed a problem with the pcregrep tests. The test of -r
1245 for recursive directory scanning broke on some systems because the files
1246 are not scanned in any specific order and on different systems the order
1247 was different. A call to "sort" has been inserted into RunGrepTest for the
1248 approprate test as a short-term fix. In the longer term there may be an
1249 alternative.
1251 3. I had an email from Eric Raymond about problems translating some of PCRE's
1252 man pages to HTML (despite the fact that I distribute HTML pages, some
1253 people do their own conversions for various reasons). The problems
1254 concerned the use of low-level troff macros .br and .in. I have therefore
1255 removed all such uses from the man pages (some were redundant, some could
1256 be replaced by .nf/.fi pairs). The 132html script that I use to generate
1257 HTML has been updated to handle .nf/.fi and to complain if it encounters
1258 .br or .in.
1260 4. Updated comments in configure.ac that get placed in config.h.in and also
1261 arranged for config.h to be included in the distribution, with the name
1262 config.h.generic, for the benefit of those who have to compile without
1263 Autotools (compare pcre.h, which is now distributed as pcre.h.generic).
1265 5. Updated the support (such as it is) for Virtual Pascal, thanks to Stefan
1266 Weber: (1) pcre_internal.h was missing some function renames; (2) updated
1267 makevp.bat for the current PCRE, using the additional files
1268 makevp_c.txt, makevp_l.txt, and pcregexp.pas.
1270 6. A Windows user reported a minor discrepancy with test 2, which turned out
1271 to be caused by a trailing space on an input line that had got lost in his
1272 copy. The trailing space was an accident, so I've just removed it.
1274 7. Add -Wl,-R... flags in pcre-config.in for *BSD* systems, as I'm told
1275 that is needed.
1277 8. Mark ucp_table (in ucptable.h) and ucp_gentype (in pcre_ucp_searchfuncs.c)
1278 as "const" (a) because they are and (b) because it helps the PHP
1279 maintainers who have recently made a script to detect big data structures
1280 in the php code that should be moved to the .rodata section. I remembered
1281 to update Builducptable as well, so it won't revert if ucptable.h is ever
1282 re-created.
1284 9. Added some extra #ifdef SUPPORT_UTF8 conditionals into pcretest.c,
1285 pcre_printint.src, pcre_compile.c, pcre_study.c, and pcre_tables.c, in
1286 order to be able to cut out the UTF-8 tables in the latter when UTF-8
1287 support is not required. This saves 1.5-2K of code, which is important in
1288 some applications.
1290 Later: more #ifdefs are needed in pcre_ord2utf8.c and pcre_valid_utf8.c
1291 so as not to refer to the tables, even though these functions will never be
1292 called when UTF-8 support is disabled. Otherwise there are problems with a
1293 shared library.
1295 10. Fixed two bugs in the emulated memmove() function in pcre_internal.h:
1297 (a) It was defining its arguments as char * instead of void *.
1299 (b) It was assuming that all moves were upwards in memory; this was true
1300 a long time ago when I wrote it, but is no longer the case.
1302 The emulated memove() is provided for those environments that have neither
1303 memmove() nor bcopy(). I didn't think anyone used it these days, but that
1304 is clearly not the case, as these two bugs were recently reported.
1306 11. The script PrepareRelease is now distributed: it calls 132html, CleanTxt,
1307 and Detrail to create the HTML documentation, the .txt form of the man
1308 pages, and it removes trailing spaces from listed files. It also creates
1309 pcre.h.generic and config.h.generic from pcre.h and config.h. In the latter
1310 case, it wraps all the #defines with #ifndefs. This script should be run
1311 before "make dist".
1313 12. Fixed two fairly obscure bugs concerned with quantified caseless matching
1314 with Unicode property support.
1316 (a) For a maximizing quantifier, if the two different cases of the
1317 character were of different lengths in their UTF-8 codings (there are
1318 some cases like this - I found 11), and the matching function had to
1319 back up over a mixture of the two cases, it incorrectly assumed they
1320 were both the same length.
1322 (b) When PCRE was configured to use the heap rather than the stack for
1323 recursion during matching, it was not correctly preserving the data for
1324 the other case of a UTF-8 character when checking ahead for a match
1325 while processing a minimizing repeat. If the check also involved
1326 matching a wide character, but failed, corruption could cause an
1327 erroneous result when trying to check for a repeat of the original
1328 character.
1330 13. Some tidying changes to the testing mechanism:
1332 (a) The RunTest script now detects the internal link size and whether there
1333 is UTF-8 and UCP support by running ./pcretest -C instead of relying on
1334 values substituted by "configure". (The RunGrepTest script already did
1335 this for UTF-8.) The configure.ac script no longer substitutes the
1336 relevant variables.
1338 (b) The debugging options /B and /D in pcretest show the compiled bytecode
1339 with length and offset values. This means that the output is different
1340 for different internal link sizes. Test 2 is skipped for link sizes
1341 other than 2 because of this, bypassing the problem. Unfortunately,
1342 there was also a test in test 3 (the locale tests) that used /B and
1343 failed for link sizes other than 2. Rather than cut the whole test out,
1344 I have added a new /Z option to pcretest that replaces the length and
1345 offset values with spaces. This is now used to make test 3 independent
1346 of link size. (Test 2 will be tidied up later.)
1348 14. If erroroffset was passed as NULL to pcre_compile, it provoked a
1349 segmentation fault instead of returning the appropriate error message.
1351 15. In multiline mode when the newline sequence was set to "any", the pattern
1352 ^$ would give a match between the \r and \n of a subject such as "A\r\nB".
1353 This doesn't seem right; it now treats the CRLF combination as the line
1354 ending, and so does not match in that case. It's only a pattern such as ^$
1355 that would hit this one: something like ^ABC$ would have failed after \r
1356 and then tried again after \r\n.
1358 16. Changed the comparison command for RunGrepTest from "diff -u" to "diff -ub"
1359 in an attempt to make files that differ only in their line terminators
1360 compare equal. This works on Linux.
1362 17. Under certain error circumstances pcregrep might try to free random memory
1363 as it exited. This is now fixed, thanks to valgrind.
1365 19. In pcretest, if the pattern /(?m)^$/g<any> was matched against the string
1366 "abc\r\n\r\n", it found an unwanted second match after the second \r. This
1367 was because its rules for how to advance for /g after matching an empty
1368 string at the end of a line did not allow for this case. They now check for
1369 it specially.
1371 20. pcretest is supposed to handle patterns and data of any length, by
1372 extending its buffers when necessary. It was getting this wrong when the
1373 buffer for a data line had to be extended.
1375 21. Added PCRE_NEWLINE_ANYCRLF which is like ANY, but matches only CR, LF, or
1376 CRLF as a newline sequence.
1378 22. Code for handling Unicode properties in pcre_dfa_exec() wasn't being cut
1379 out by #ifdef SUPPORT_UCP. This did no harm, as it could never be used, but
1380 I have nevertheless tidied it up.
1382 23. Added some casts to kill warnings from HP-UX ia64 compiler.
1384 24. Added a man page for pcre-config.
1387 Version 7.0 19-Dec-06
1388 ---------------------
1390 1. Fixed a signed/unsigned compiler warning in pcre_compile.c, shown up by
1391 moving to gcc 4.1.1.
1393 2. The -S option for pcretest uses setrlimit(); I had omitted to #include
1394 sys/time.h, which is documented as needed for this function. It doesn't
1395 seem to matter on Linux, but it showed up on some releases of OS X.
1397 3. It seems that there are systems where bytes whose values are greater than
1398 127 match isprint() in the "C" locale. The "C" locale should be the
1399 default when a C program starts up. In most systems, only ASCII printing
1400 characters match isprint(). This difference caused the output from pcretest
1401 to vary, making some of the tests fail. I have changed pcretest so that:
1403 (a) When it is outputting text in the compiled version of a pattern, bytes
1404 other than 32-126 are always shown as hex escapes.
1406 (b) When it is outputting text that is a matched part of a subject string,
1407 it does the same, unless a different locale has been set for the match
1408 (using the /L modifier). In this case, it uses isprint() to decide.
1410 4. Fixed a major bug that caused incorrect computation of the amount of memory
1411 required for a compiled pattern when options that changed within the
1412 pattern affected the logic of the preliminary scan that determines the
1413 length. The relevant options are -x, and -i in UTF-8 mode. The result was
1414 that the computed length was too small. The symptoms of this bug were
1415 either the PCRE error "internal error: code overflow" from pcre_compile(),
1416 or a glibc crash with a message such as "pcretest: free(): invalid next
1417 size (fast)". Examples of patterns that provoked this bug (shown in
1418 pcretest format) are:
1420 /(?-x: )/x
1421 /(?x)(?-x: \s*#\s*)/
1422 /((?i)[\x{c0}])/8
1423 /(?i:[\x{c0}])/8
1425 HOWEVER: Change 17 below makes this fix obsolete as the memory computation
1426 is now done differently.
1428 5. Applied patches from Google to: (a) add a QuoteMeta function to the C++
1429 wrapper classes; (b) implement a new function in the C++ scanner that is
1430 more efficient than the old way of doing things because it avoids levels of
1431 recursion in the regex matching; (c) add a paragraph to the documentation
1432 for the FullMatch() function.
1434 6. The escape sequence \n was being treated as whatever was defined as
1435 "newline". Not only was this contrary to the documentation, which states
1436 that \n is character 10 (hex 0A), but it also went horribly wrong when
1437 "newline" was defined as CRLF. This has been fixed.
1439 7. In pcre_dfa_exec.c the value of an unsigned integer (the variable called c)
1440 was being set to -1 for the "end of line" case (supposedly a value that no
1441 character can have). Though this value is never used (the check for end of
1442 line is "zero bytes in current character"), it caused compiler complaints.
1443 I've changed it to 0xffffffff.
1445 8. In pcre_version.c, the version string was being built by a sequence of
1446 C macros that, in the event of PCRE_PRERELEASE being defined as an empty
1447 string (as it is for production releases) called a macro with an empty
1448 argument. The C standard says the result of this is undefined. The gcc
1449 compiler treats it as an empty string (which was what was wanted) but it is
1450 reported that Visual C gives an error. The source has been hacked around to
1451 avoid this problem.
1453 9. On the advice of a Windows user, included <io.h> and <fcntl.h> in Windows
1454 builds of pcretest, and changed the call to _setmode() to use _O_BINARY
1455 instead of 0x8000. Made all the #ifdefs test both _WIN32 and WIN32 (not all
1456 of them did).
1458 10. Originally, pcretest opened its input and output without "b"; then I was
1459 told that "b" was needed in some environments, so it was added for release
1460 5.0 to both the input and output. (It makes no difference on Unix-like
1461 systems.) Later I was told that it is wrong for the input on Windows. I've
1462 now abstracted the modes into two macros, to make it easier to fiddle with
1463 them, and removed "b" from the input mode under Windows.
1465 11. Added pkgconfig support for the C++ wrapper library, libpcrecpp.
1467 12. Added -help and --help to pcretest as an official way of being reminded
1468 of the options.
1470 13. Removed some redundant semicolons after macro calls in pcrecpparg.h.in
1471 and pcrecpp.cc because they annoy compilers at high warning levels.
1473 14. A bit of tidying/refactoring in pcre_exec.c in the main bumpalong loop.
1475 15. Fixed an occurrence of == in configure.ac that should have been = (shell
1476 scripts are not C programs :-) and which was not noticed because it works
1477 on Linux.
1479 16. pcretest is supposed to handle any length of pattern and data line (as one
1480 line or as a continued sequence of lines) by extending its input buffer if
1481 necessary. This feature was broken for very long pattern lines, leading to
1482 a string of junk being passed to pcre_compile() if the pattern was longer
1483 than about 50K.
1485 17. I have done a major re-factoring of the way pcre_compile() computes the
1486 amount of memory needed for a compiled pattern. Previously, there was code
1487 that made a preliminary scan of the pattern in order to do this. That was
1488 OK when PCRE was new, but as the facilities have expanded, it has become
1489 harder and harder to keep it in step with the real compile phase, and there
1490 have been a number of bugs (see for example, 4 above). I have now found a
1491 cunning way of running the real compile function in a "fake" mode that
1492 enables it to compute how much memory it would need, while actually only
1493 ever using a few hundred bytes of working memory and without too many
1494 tests of the mode. This should make future maintenance and development
1495 easier. A side effect of this work is that the limit of 200 on the nesting
1496 depth of parentheses has been removed (though this was never a serious
1497 limitation, I suspect). However, there is a downside: pcre_compile() now
1498 runs more slowly than before (30% or more, depending on the pattern). I
1499 hope this isn't a big issue. There is no effect on runtime performance.
1501 18. Fixed a minor bug in pcretest: if a pattern line was not terminated by a
1502 newline (only possible for the last line of a file) and it was a
1503 pattern that set a locale (followed by /Lsomething), pcretest crashed.
1505 19. Added additional timing features to pcretest. (1) The -tm option now times
1506 matching only, not compiling. (2) Both -t and -tm can be followed, as a
1507 separate command line item, by a number that specifies the number of
1508 repeats to use when timing. The default is 50000; this gives better
1509 precision, but takes uncomfortably long for very large patterns.
1511 20. Extended pcre_study() to be more clever in cases where a branch of a
1512 subpattern has no definite first character. For example, (a*|b*)[cd] would
1513 previously give no result from pcre_study(). Now it recognizes that the
1514 first character must be a, b, c, or d.
1516 21. There was an incorrect error "recursive call could loop indefinitely" if
1517 a subpattern (or the entire pattern) that was being tested for matching an
1518 empty string contained only one non-empty item after a nested subpattern.
1519 For example, the pattern (?>\x{100}*)\d(?R) provoked this error
1520 incorrectly, because the \d was being skipped in the check.
1522 22. The pcretest program now has a new pattern option /B and a command line
1523 option -b, which is equivalent to adding /B to every pattern. This causes
1524 it to show the compiled bytecode, without the additional information that
1525 -d shows. The effect of -d is now the same as -b with -i (and similarly, /D
1526 is the same as /B/I).
1528 23. A new optimization is now able automatically to treat some sequences such
1529 as a*b as a*+b. More specifically, if something simple (such as a character
1530 or a simple class like \d) has an unlimited quantifier, and is followed by
1531 something that cannot possibly match the quantified thing, the quantifier
1532 is automatically "possessified".
1534 24. A recursive reference to a subpattern whose number was greater than 39
1535 went wrong under certain circumstances in UTF-8 mode. This bug could also
1536 have affected the operation of pcre_study().
1538 25. Realized that a little bit of performance could be had by replacing
1539 (c & 0xc0) == 0xc0 with c >= 0xc0 when processing UTF-8 characters.
1541 26. Timing data from pcretest is now shown to 4 decimal places instead of 3.
1543 27. Possessive quantifiers such as a++ were previously implemented by turning
1544 them into atomic groups such as ($>a+). Now they have their own opcodes,
1545 which improves performance. This includes the automatically created ones
1546 from 23 above.
1548 28. A pattern such as (?=(\w+))\1: which simulates an atomic group using a
1549 lookahead was broken if it was not anchored. PCRE was mistakenly expecting
1550 the first matched character to be a colon. This applied both to named and
1551 numbered groups.
1553 29. The ucpinternal.h header file was missing its idempotency #ifdef.
1555 30. I was sent a "project" file called libpcre.a.dev which I understand makes
1556 building PCRE on Windows easier, so I have included it in the distribution.
1558 31. There is now a check in pcretest against a ridiculously large number being
1559 returned by pcre_exec() or pcre_dfa_exec(). If this happens in a /g or /G
1560 loop, the loop is abandoned.
1562 32. Forward references to subpatterns in conditions such as (?(2)...) where
1563 subpattern 2 is defined later cause pcre_compile() to search forwards in
1564 the pattern for the relevant set of parentheses. This search went wrong
1565 when there were unescaped parentheses in a character class, parentheses
1566 escaped with \Q...\E, or parentheses in a #-comment in /x mode.
1568 33. "Subroutine" calls and backreferences were previously restricted to
1569 referencing subpatterns earlier in the regex. This restriction has now
1570 been removed.
1572 34. Added a number of extra features that are going to be in Perl 5.10. On the
1573 whole, these are just syntactic alternatives for features that PCRE had
1574 previously implemented using the Python syntax or my own invention. The
1575 other formats are all retained for compatibility.
1577 (a) Named groups can now be defined as (?<name>...) or (?'name'...) as well
1578 as (?P<name>...). The new forms, as well as being in Perl 5.10, are
1579 also .NET compatible.
1581 (b) A recursion or subroutine call to a named group can now be defined as
1582 (?&name) as well as (?P>name).
1584 (c) A backreference to a named group can now be defined as \k<name> or
1585 \k'name' as well as (?P=name). The new forms, as well as being in Perl
1586 5.10, are also .NET compatible.
1588 (d) A conditional reference to a named group can now use the syntax
1589 (?(<name>) or (?('name') as well as (?(name).
1591 (e) A "conditional group" of the form (?(DEFINE)...) can be used to define
1592 groups (named and numbered) that are never evaluated inline, but can be
1593 called as "subroutines" from elsewhere. In effect, the DEFINE condition
1594 is always false. There may be only one alternative in such a group.
1596 (f) A test for recursion can be given as (?(R1).. or (?(R&name)... as well
1597 as the simple (?(R). The condition is true only if the most recent
1598 recursion is that of the given number or name. It does not search out
1599 through the entire recursion stack.
1601 (g) The escape \gN or \g{N} has been added, where N is a positive or
1602 negative number, specifying an absolute or relative reference.
1604 35. Tidied to get rid of some further signed/unsigned compiler warnings and
1605 some "unreachable code" warnings.
1607 36. Updated the Unicode property tables to Unicode version 5.0.0. Amongst other
1608 things, this adds five new scripts.
1610 37. Perl ignores orphaned \E escapes completely. PCRE now does the same.
1611 There were also incompatibilities regarding the handling of \Q..\E inside
1612 character classes, for example with patterns like [\Qa\E-\Qz\E] where the
1613 hyphen was adjacent to \Q or \E. I hope I've cleared all this up now.
1615 38. Like Perl, PCRE detects when an indefinitely repeated parenthesized group
1616 matches an empty string, and forcibly breaks the loop. There were bugs in
1617 this code in non-simple cases. For a pattern such as ^(a()*)* matched
1618 against aaaa the result was just "a" rather than "aaaa", for example. Two
1619 separate and independent bugs (that affected different cases) have been
1620 fixed.
1622 39. Refactored the code to abolish the use of different opcodes for small
1623 capturing bracket numbers. This is a tidy that I avoided doing when I
1624 removed the limit on the number of capturing brackets for 3.5 back in 2001.
1625 The new approach is not only tidier, it makes it possible to reduce the
1626 memory needed to fix the previous bug (38).
1628 40. Implemented PCRE_NEWLINE_ANY to recognize any of the Unicode newline
1629 sequences (http://unicode.org/unicode/reports/tr18/) as "newline" when
1630 processing dot, circumflex, or dollar metacharacters, or #-comments in /x
1631 mode.
1633 41. Add \R to match any Unicode newline sequence, as suggested in the Unicode
1634 report.
1636 42. Applied patch, originally from Ari Pollak, modified by Google, to allow
1637 copy construction and assignment in the C++ wrapper.
1639 43. Updated pcregrep to support "--newline=any". In the process, I fixed a
1640 couple of bugs that could have given wrong results in the "--newline=crlf"
1641 case.
1643 44. Added a number of casts and did some reorganization of signed/unsigned int
1644 variables following suggestions from Dair Grant. Also renamed the variable
1645 "this" as "item" because it is a C++ keyword.
1647 45. Arranged for dftables to add
1649 #include "pcre_internal.h"
1651 to pcre_chartables.c because without it, gcc 4.x may remove the array
1652 definition from the final binary if PCRE is built into a static library and
1653 dead code stripping is activated.
1655 46. For an unanchored pattern, if a match attempt fails at the start of a
1656 newline sequence, and the newline setting is CRLF or ANY, and the next two
1657 characters are CRLF, advance by two characters instead of one.
1660 Version 6.7 04-Jul-06
1661 ---------------------
1663 1. In order to handle tests when input lines are enormously long, pcretest has
1664 been re-factored so that it automatically extends its buffers when
1665 necessary. The code is crude, but this _is_ just a test program. The
1666 default size has been increased from 32K to 50K.
1668 2. The code in pcre_study() was using the value of the re argument before
1669 testing it for NULL. (Of course, in any sensible call of the function, it
1670 won't be NULL.)
1672 3. The memmove() emulation function in pcre_internal.h, which is used on
1673 systems that lack both memmove() and bcopy() - that is, hardly ever -
1674 was missing a "static" storage class specifier.
1676 4. When UTF-8 mode was not set, PCRE looped when compiling certain patterns
1677 containing an extended class (one that cannot be represented by a bitmap
1678 because it contains high-valued characters or Unicode property items, e.g.
1679 [\pZ]). Almost always one would set UTF-8 mode when processing such a
1680 pattern, but PCRE should not loop if you do not (it no longer does).
1681 [Detail: two cases were found: (a) a repeated subpattern containing an
1682 extended class; (b) a recursive reference to a subpattern that followed a
1683 previous extended class. It wasn't skipping over the extended class
1684 correctly when UTF-8 mode was not set.]
1686 5. A negated single-character class was not being recognized as fixed-length
1687 in lookbehind assertions such as (?<=[^f]), leading to an incorrect
1688 compile error "lookbehind assertion is not fixed length".
1690 6. The RunPerlTest auxiliary script was showing an unexpected difference
1691 between PCRE and Perl for UTF-8 tests. It turns out that it is hard to
1692 write a Perl script that can interpret lines of an input file either as
1693 byte characters or as UTF-8, which is what "perltest" was being required to
1694 do for the non-UTF-8 and UTF-8 tests, respectively. Essentially what you
1695 can't do is switch easily at run time between having the "use utf8;" pragma
1696 or not. In the end, I fudged it by using the RunPerlTest script to insert
1697 "use utf8;" explicitly for the UTF-8 tests.
1699 7. In multiline (/m) mode, PCRE was matching ^ after a terminating newline at
1700 the end of the subject string, contrary to the documentation and to what
1701 Perl does. This was true of both matching functions. Now it matches only at
1702 the start of the subject and immediately after *internal* newlines.
1704 8. A call of pcre_fullinfo() from pcretest to get the option bits was passing
1705 a pointer to an int instead of a pointer to an unsigned long int. This
1706 caused problems on 64-bit systems.
1708 9. Applied a patch from the folks at Google to pcrecpp.cc, to fix "another
1709 instance of the 'standard' template library not being so standard".
1711 10. There was no check on the number of named subpatterns nor the maximum
1712 length of a subpattern name. The product of these values is used to compute
1713 the size of the memory block for a compiled pattern. By supplying a very
1714 long subpattern name and a large number of named subpatterns, the size
1715 computation could be caused to overflow. This is now prevented by limiting
1716 the length of names to 32 characters, and the number of named subpatterns
1717 to 10,000.
1719 11. Subpatterns that are repeated with specific counts have to be replicated in
1720 the compiled pattern. The size of memory for this was computed from the
1721 length of the subpattern and the repeat count. The latter is limited to
1722 65535, but there was no limit on the former, meaning that integer overflow
1723 could in principle occur. The compiled length of a repeated subpattern is
1724 now limited to 30,000 bytes in order to prevent this.
1726 12. Added the optional facility to have named substrings with the same name.
1728 13. Added the ability to use a named substring as a condition, using the
1729 Python syntax: (?(name)yes|no). This overloads (?(R)... and names that
1730 are numbers (not recommended). Forward references are permitted.
1732 14. Added forward references in named backreferences (if you see what I mean).
1734 15. In UTF-8 mode, with the PCRE_DOTALL option set, a quantified dot in the
1735 pattern could run off the end of the subject. For example, the pattern
1736 "(?s)(.{1,5})"8 did this with the subject "ab".
1738 16. If PCRE_DOTALL or PCRE_MULTILINE were set, pcre_dfa_exec() behaved as if
1739 PCRE_CASELESS was set when matching characters that were quantified with ?
1740 or *.
1742 17. A character class other than a single negated character that had a minimum
1743 but no maximum quantifier - for example [ab]{6,} - was not handled
1744 correctly by pce_dfa_exec(). It would match only one character.
1746 18. A valid (though odd) pattern that looked like a POSIX character
1747 class but used an invalid character after [ (for example [[,abc,]]) caused
1748 pcre_compile() to give the error "Failed: internal error: code overflow" or
1749 in some cases to crash with a glibc free() error. This could even happen if
1750 the pattern terminated after [[ but there just happened to be a sequence of
1751 letters, a binary zero, and a closing ] in the memory that followed.
1753 19. Perl's treatment of octal escapes in the range \400 to \777 has changed
1754 over the years. Originally (before any Unicode support), just the bottom 8
1755 bits were taken. Thus, for example, \500 really meant \100. Nowadays the
1756 output from "man perlunicode" includes this:
1758 The regular expression compiler produces polymorphic opcodes. That
1759 is, the pattern adapts to the data and automatically switches to
1760 the Unicode character scheme when presented with Unicode data--or
1761 instead uses a traditional byte scheme when presented with byte
1762 data.
1764 Sadly, a wide octal escape does not cause a switch, and in a string with
1765 no other multibyte characters, these octal escapes are treated as before.
1766 Thus, in Perl, the pattern /\500/ actually matches \100 but the pattern
1767 /\500|\x{1ff}/ matches \500 or \777 because the whole thing is treated as a
1768 Unicode string.
1770 I have not perpetrated such confusion in PCRE. Up till now, it took just
1771 the bottom 8 bits, as in old Perl. I have now made octal escapes with
1772 values greater than \377 illegal in non-UTF-8 mode. In UTF-8 mode they
1773 translate to the appropriate multibyte character.
1775 29. Applied some refactoring to reduce the number of warnings from Microsoft
1776 and Borland compilers. This has included removing the fudge introduced
1777 seven years ago for the OS/2 compiler (see 2.02/2 below) because it caused
1778 a warning about an unused variable.
1780 21. PCRE has not included VT (character 0x0b) in the set of whitespace
1781 characters since release 4.0, because Perl (from release 5.004) does not.
1782 [Or at least, is documented not to: some releases seem to be in conflict
1783 with the documentation.] However, when a pattern was studied with
1784 pcre_study() and all its branches started with \s, PCRE still included VT
1785 as a possible starting character. Of course, this did no harm; it just
1786 caused an unnecessary match attempt.
1788 22. Removed a now-redundant internal flag bit that recorded the fact that case
1789 dependency changed within the pattern. This was once needed for "required
1790 byte" processing, but is no longer used. This recovers a now-scarce options
1791 bit. Also moved the least significant internal flag bit to the most-
1792 significant bit of the word, which was not previously used (hangover from
1793 the days when it was an int rather than a uint) to free up another bit for
1794 the future.
1796 23. Added support for CRLF line endings as well as CR and LF. As well as the
1797 default being selectable at build time, it can now be changed at runtime
1798 via the PCRE_NEWLINE_xxx flags. There are now options for pcregrep to
1799 specify that it is scanning data with non-default line endings.
1801 24. Changed the definition of CXXLINK to make it agree with the definition of
1802 LINK in the Makefile, by replacing LDFLAGS to CXXFLAGS.
1804 25. Applied Ian Taylor's patches to avoid using another stack frame for tail
1805 recursions. This makes a big different to stack usage for some patterns.
1807 26. If a subpattern containing a named recursion or subroutine reference such
1808 as (?P>B) was quantified, for example (xxx(?P>B)){3}, the calculation of
1809 the space required for the compiled pattern went wrong and gave too small a
1810 value. Depending on the environment, this could lead to "Failed: internal
1811 error: code overflow at offset 49" or "glibc detected double free or
1812 corruption" errors.
1814 27. Applied patches from Google (a) to support the new newline modes and (b) to
1815 advance over multibyte UTF-8 characters in GlobalReplace.
1817 28. Change free() to pcre_free() in pcredemo.c. Apparently this makes a
1818 difference for some implementation of PCRE in some Windows version.
1820 29. Added some extra testing facilities to pcretest:
1822 \q<number> in a data line sets the "match limit" value
1823 \Q<number> in a data line sets the "match recursion limt" value
1824 -S <number> sets the stack size, where <number> is in megabytes
1826 The -S option isn't available for Windows.
1829 Version 6.6 06-Feb-06
1830 ---------------------
1832 1. Change 16(a) for 6.5 broke things, because PCRE_DATA_SCOPE was not defined
1833 in pcreposix.h. I have copied the definition from pcre.h.
1835 2. Change 25 for 6.5 broke compilation in a build directory out-of-tree
1836 because pcre.h is no longer a built file.
1838 3. Added Jeff Friedl's additional debugging patches to pcregrep. These are
1839 not normally included in the compiled code.
1842 Version 6.5 01-Feb-06
1843 ---------------------
1845 1. When using the partial match feature with pcre_dfa_exec(), it was not
1846 anchoring the second and subsequent partial matches at the new starting
1847 point. This could lead to incorrect results. For example, with the pattern
1848 /1234/, partially matching against "123" and then "a4" gave a match.
1850 2. Changes to pcregrep:
1852 (a) All non-match returns from pcre_exec() were being treated as failures
1853 to match the line. Now, unless the error is PCRE_ERROR_NOMATCH, an
1854 error message is output. Some extra information is given for the
1856 probably the only errors that are likely to be caused by users (by
1857 specifying a regex that has nested indefinite repeats, for instance).
1858 If there are more than 20 of these errors, pcregrep is abandoned.
1860 (b) A binary zero was treated as data while matching, but terminated the
1861 output line if it was written out. This has been fixed: binary zeroes
1862 are now no different to any other data bytes.
1864 (c) Whichever of the LC_ALL or LC_CTYPE environment variables is set is
1865 used to set a locale for matching. The --locale=xxxx long option has
1866 been added (no short equivalent) to specify a locale explicitly on the
1867 pcregrep command, overriding the environment variables.
1869 (d) When -B was used with -n, some line numbers in the output were one less
1870 than they should have been.
1872 (e) Added the -o (--only-matching) option.
1874 (f) If -A or -C was used with -c (count only), some lines of context were
1875 accidentally printed for the final match.
1877 (g) Added the -H (--with-filename) option.
1879 (h) The combination of options -rh failed to suppress file names for files
1880 that were found from directory arguments.
1882 (i) Added the -D (--devices) and -d (--directories) options.
1884 (j) Added the -F (--fixed-strings) option.
1886 (k) Allow "-" to be used as a file name for -f as well as for a data file.
1888 (l) Added the --colo(u)r option.
1890 (m) Added Jeffrey Friedl's -S testing option, but within #ifdefs so that it
1891 is not present by default.
1893 3. A nasty bug was discovered in the handling of recursive patterns, that is,
1894 items such as (?R) or (?1), when the recursion could match a number of
1895 alternatives. If it matched one of the alternatives, but subsequently,
1896 outside the recursion, there was a failure, the code tried to back up into
1897 the recursion. However, because of the way PCRE is implemented, this is not
1898 possible, and the result was an incorrect result from the match.
1900 In order to prevent this happening, the specification of recursion has
1901 been changed so that all such subpatterns are automatically treated as
1902 atomic groups. Thus, for example, (?R) is treated as if it were (?>(?R)).
1904 4. I had overlooked the fact that, in some locales, there are characters for
1905 which isalpha() is true but neither isupper() nor islower() are true. In
1906 the fr_FR locale, for instance, the \xAA and \xBA characters (ordmasculine
1907 and ordfeminine) are like this. This affected the treatment of \w and \W
1908 when they appeared in character classes, but not when they appeared outside
1909 a character class. The bit map for "word" characters is now created
1910 separately from the results of isalnum() instead of just taking it from the
1911 upper, lower, and digit maps. (Plus the underscore character, of course.)
1913 5. The above bug also affected the handling of POSIX character classes such as
1914 [[:alpha:]] and [[:alnum:]]. These do not have their own bit maps in PCRE's
1915 permanent tables. Instead, the bit maps for such a class were previously
1916 created as the appropriate unions of the upper, lower, and digit bitmaps.
1917 Now they are created by subtraction from the [[:word:]] class, which has
1918 its own bitmap.
1920 6. The [[:blank:]] character class matches horizontal, but not vertical space.
1921 It is created by subtracting the vertical space characters (\x09, \x0a,
1922 \x0b, \x0c) from the [[:space:]] bitmap. Previously, however, the
1923 subtraction was done in the overall bitmap for a character class, meaning
1924 that a class such as [\x0c[:blank:]] was incorrect because \x0c would not
1925 be recognized. This bug has been fixed.
1927 7. Patches from the folks at Google:
1929 (a) pcrecpp.cc: "to handle a corner case that may or may not happen in
1930 real life, but is still worth protecting against".
1932 (b) pcrecpp.cc: "corrects a bug when negative radixes are used with
1933 regular expressions".
1935 (c) pcre_scanner.cc: avoid use of std::count() because not all systems
1936 have it.
1938 (d) Split off pcrecpparg.h from pcrecpp.h and had the former built by
1939 "configure" and the latter not, in order to fix a problem somebody had
1940 with compiling the Arg class on HP-UX.
1942 (e) Improve the error-handling of the C++ wrapper a little bit.
1944 (f) New tests for checking recursion limiting.
1946 8. The pcre_memmove() function, which is used only if the environment does not
1947 have a standard memmove() function (and is therefore rarely compiled),
1948 contained two bugs: (a) use of int instead of size_t, and (b) it was not
1949 returning a result (though PCRE never actually uses the result).
1951 9. In the POSIX regexec() interface, if nmatch is specified as a ridiculously
1952 large number - greater than INT_MAX/(3*sizeof(int)) - REG_ESPACE is
1953 returned instead of calling malloc() with an overflowing number that would
1954 most likely cause subsequent chaos.
1956 10. The debugging option of pcretest was not showing the NO_AUTO_CAPTURE flag.
1958 11. The POSIX flag REG_NOSUB is now supported. When a pattern that was compiled
1959 with this option is matched, the nmatch and pmatch options of regexec() are
1960 ignored.
1962 12. Added REG_UTF8 to the POSIX interface. This is not defined by POSIX, but is
1963 provided in case anyone wants to the the POSIX interface with UTF-8
1964 strings.
1966 13. Added CXXLDFLAGS to the Makefile parameters to provide settings only on the
1967 C++ linking (needed for some HP-UX environments).
1969 14. Avoid compiler warnings in get_ucpname() when compiled without UCP support
1970 (unused parameter) and in the pcre_printint() function (omitted "default"
1971 switch label when the default is to do nothing).
1973 15. Added some code to make it possible, when PCRE is compiled as a C++
1974 library, to replace subject pointers for pcre_exec() with a smart pointer
1975 class, thus making it possible to process discontinuous strings.
1977 16. The two macros PCRE_EXPORT and PCRE_DATA_SCOPE are confusing, and perform
1978 much the same function. They were added by different people who were trying
1979 to make PCRE easy to compile on non-Unix systems. It has been suggested
1980 that PCRE_EXPORT be abolished now that there is more automatic apparatus
1981 for compiling on Windows systems. I have therefore replaced it with
1982 PCRE_DATA_SCOPE. This is set automatically for Windows; if not set it
1983 defaults to "extern" for C or "extern C" for C++, which works fine on
1984 Unix-like systems. It is now possible to override the value of PCRE_DATA_
1985 SCOPE with something explicit in config.h. In addition:
1987 (a) pcreposix.h still had just "extern" instead of either of these macros;
1988 I have replaced it with PCRE_DATA_SCOPE.
1990 (b) Functions such as _pcre_xclass(), which are internal to the library,
1991 but external in the C sense, all had PCRE_EXPORT in their definitions.
1992 This is apparently wrong for the Windows case, so I have removed it.
1993 (It makes no difference on Unix-like systems.)
1995 17. Added a new limit, MATCH_LIMIT_RECURSION, which limits the depth of nesting
1996 of recursive calls to match(). This is different to MATCH_LIMIT because
1997 that limits the total number of calls to match(), not all of which increase
1998 the depth of recursion. Limiting the recursion depth limits the amount of
1999 stack (or heap if NO_RECURSE is set) that is used. The default can be set
2000 when PCRE is compiled, and changed at run time. A patch from Google adds
2001 this functionality to the C++ interface.
2003 18. Changes to the handling of Unicode character properties:
2005 (a) Updated the table to Unicode 4.1.0.
2007 (b) Recognize characters that are not in the table as "Cn" (undefined).
2009 (c) I revised the way the table is implemented to a much improved format
2010 which includes recognition of ranges. It now supports the ranges that
2011 are defined in UnicodeData.txt, and it also amalgamates other
2012 characters into ranges. This has reduced the number of entries in the
2013 table from around 16,000 to around 3,000, thus reducing its size
2014 considerably. I realized I did not need to use a tree structure after
2015 all - a binary chop search is just as efficient. Having reduced the
2016 number of entries, I extended their size from 6 bytes to 8 bytes to
2017 allow for more data.
2019 (d) Added support for Unicode script names via properties such as \p{Han}.
2021 19. In UTF-8 mode, a backslash followed by a non-Ascii character was not
2022 matching that character.
2024 20. When matching a repeated Unicode property with a minimum greater than zero,
2025 (for example \pL{2,}), PCRE could look past the end of the subject if it
2026 reached it while seeking the minimum number of characters. This could
2027 happen only if some of the characters were more than one byte long, because
2028 there is a check for at least the minimum number of bytes.
2030 21. Refactored the implementation of \p and \P so as to be more general, to
2031 allow for more different types of property in future. This has changed the
2032 compiled form incompatibly. Anybody with saved compiled patterns that use
2033 \p or \P will have to recompile them.
2035 22. Added "Any" and "L&" to the supported property types.
2037 23. Recognize \x{...} as a code point specifier, even when not in UTF-8 mode,
2038 but give a compile time error if the value is greater than 0xff.
2040 24. The man pages for pcrepartial, pcreprecompile, and pcre_compile2 were
2041 accidentally not being installed or uninstalled.
2043 25. The pcre.h file was built from pcre.h.in, but the only changes that were
2044 made were to insert the current release number. This seemed silly, because
2045 it made things harder for people building PCRE on systems that don't run
2046 "configure". I have turned pcre.h into a distributed file, no longer built
2047 by "configure", with the version identification directly included. There is
2048 no longer a pcre.h.in file.
2050 However, this change necessitated a change to the pcre-config script as
2051 well. It is built from pcre-config.in, and one of the substitutions was the
2052 release number. I have updated configure.ac so that ./configure now finds
2053 the release number by grepping pcre.h.
2055 26. Added the ability to run the tests under valgrind.
2058 Version 6.4 05-Sep-05
2059 ---------------------
2061 1. Change 6.0/10/(l) to pcregrep introduced a bug that caused separator lines
2062 "--" to be printed when multiple files were scanned, even when none of the
2063 -A, -B, or -C options were used. This is not compatible with Gnu grep, so I
2064 consider it to be a bug, and have restored the previous behaviour.
2066 2. A couple of code tidies to get rid of compiler warnings.
2068 3. The pcretest program used to cheat by referring to symbols in the library
2069 whose names begin with _pcre_. These are internal symbols that are not
2070 really supposed to be visible externally, and in some environments it is
2071 possible to suppress them. The cheating is now confined to including
2072 certain files from the library's source, which is a bit cleaner.
2074 4. Renamed pcre.in as pcre.h.in to go with pcrecpp.h.in; it also makes the
2075 file's purpose clearer.
2077 5. Reorganized pcre_ucp_findchar().
2080 Version 6.3 15-Aug-05
2081 ---------------------
2083 1. The file libpcre.pc.in did not have general read permission in the tarball.
2085 2. There were some problems when building without C++ support:
2087 (a) If C++ support was not built, "make install" and "make test" still
2088 tried to test it.
2090 (b) There were problems when the value of CXX was explicitly set. Some
2091 changes have been made to try to fix these, and ...
2093 (c) --disable-cpp can now be used to explicitly disable C++ support.
2095 (d) The use of @CPP_OBJ@ directly caused a blank line preceded by a
2096 backslash in a target when C++ was disabled. This confuses some
2097 versions of "make", apparently. Using an intermediate variable solves
2098 this. (Same for CPP_LOBJ.)
2100 3. $(LINK_FOR_BUILD) now includes $(CFLAGS_FOR_BUILD) and $(LINK)
2101 (non-Windows) now includes $(CFLAGS) because these flags are sometimes
2102 necessary on certain architectures.
2104 4. Added a setting of -export-symbols-regex to the link command to remove
2105 those symbols that are exported in the C sense, but actually are local
2106 within the library, and not documented. Their names all begin with
2107 "_pcre_". This is not a perfect job, because (a) we have to except some
2108 symbols that pcretest ("illegally") uses, and (b) the facility isn't always
2109 available (and never for static libraries). I have made a note to try to
2110 find a way round (a) in the future.
2113 Version 6.2 01-Aug-05
2114 ---------------------
2116 1. There was no test for integer overflow of quantifier values. A construction
2117 such as {1111111111111111} would give undefined results. What is worse, if
2118 a minimum quantifier for a parenthesized subpattern overflowed and became
2119 negative, the calculation of the memory size went wrong. This could have
2120 led to memory overwriting.
2122 2. Building PCRE using VPATH was broken. Hopefully it is now fixed.
2124 3. Added "b" to the 2nd argument of fopen() in dftables.c, for non-Unix-like
2125 operating environments where this matters.
2127 4. Applied Giuseppe Maxia's patch to add additional features for controlling
2128 PCRE options from within the C++ wrapper.
2130 5. Named capturing subpatterns were not being correctly counted when a pattern
2131 was compiled. This caused two problems: (a) If there were more than 100
2132 such subpatterns, the calculation of the memory needed for the whole
2133 compiled pattern went wrong, leading to an overflow error. (b) Numerical
2134 back references of the form \12, where the number was greater than 9, were
2135 not recognized as back references, even though there were sufficient
2136 previous subpatterns.
2138 6. Two minor patches to pcrecpp.cc in order to allow it to compile on older
2139 versions of gcc, e.g. 2.95.4.
2142 Version 6.1 21-Jun-05
2143 ---------------------
2145 1. There was one reference to the variable "posix" in pcretest.c that was not
2146 surrounded by "#if !defined NOPOSIX".
2148 2. Make it possible to compile pcretest without DFA support, UTF8 support, or
2149 the cross-check on the old pcre_info() function, for the benefit of the
2150 cut-down version of PCRE that is currently imported into Exim.
2152 3. A (silly) pattern starting with (?i)(?-i) caused an internal space
2153 allocation error. I've done the easy fix, which wastes 2 bytes for sensible
2154 patterns that start (?i) but I don't think that matters. The use of (?i) is
2155 just an example; this all applies to the other options as well.
2157 4. Since libtool seems to echo the compile commands it is issuing, the output
2158 from "make" can be reduced a bit by putting "@" in front of each libtool
2159 compile command.
2161 5. Patch from the folks at Google for configure.in to be a bit more thorough
2162 in checking for a suitable C++ installation before trying to compile the
2163 C++ stuff. This should fix a reported problem when a compiler was present,
2164 but no suitable headers.
2166 6. The man pages all had just "PCRE" as their title. I have changed them to
2167 be the relevant file name. I have also arranged that these names are
2168 retained in the file doc/pcre.txt, which is a concatenation in text format
2169 of all the man pages except the little individual ones for each function.
2171 7. The NON-UNIX-USE file had not been updated for the different set of source
2172 files that come with release 6. I also added a few comments about the C++
2173 wrapper.
2176 Version 6.0 07-Jun-05
2177 ---------------------
2179 1. Some minor internal re-organization to help with my DFA experiments.
2181 2. Some missing #ifdef SUPPORT_UCP conditionals in pcretest and printint that
2182 didn't matter for the library itself when fully configured, but did matter
2183 when compiling without UCP support, or within Exim, where the ucp files are
2184 not imported.
2186 3. Refactoring of the library code to split up the various functions into
2187 different source modules. The addition of the new DFA matching code (see
2188 below) to a single monolithic source would have made it really too
2189 unwieldy, quite apart from causing all the code to be include in a
2190 statically linked application, when only some functions are used. This is
2191 relevant even without the DFA addition now that patterns can be compiled in
2192 one application and matched in another.
2194 The downside of splitting up is that there have to be some external
2195 functions and data tables that are used internally in different modules of
2196 the library but which are not part of the API. These have all had their
2197 names changed to start with "_pcre_" so that they are unlikely to clash
2198 with other external names.
2200 4. Added an alternate matching function, pcre_dfa_exec(), which matches using
2201 a different (DFA) algorithm. Although it is slower than the original
2202 function, it does have some advantages for certain types of matching
2203 problem.
2205 5. Upgrades to pcretest in order to test the features of pcre_dfa_exec(),
2206 including restarting after a partial match.
2208 6. A patch for pcregrep that defines INVALID_FILE_ATTRIBUTES if it is not
2209 defined when compiling for Windows was sent to me. I have put it into the
2210 code, though I have no means of testing or verifying it.
2212 7. Added the pcre_refcount() auxiliary function.
2214 8. Added the PCRE_FIRSTLINE option. This constrains an unanchored pattern to
2215 match before or at the first newline in the subject string. In pcretest,
2216 the /f option on a pattern can be used to set this.
2218 9. A repeated \w when used in UTF-8 mode with characters greater than 256
2219 would behave wrongly. This has been present in PCRE since release 4.0.
2221 10. A number of changes to the pcregrep command:
2223 (a) Refactored how -x works; insert ^(...)$ instead of setting
2224 PCRE_ANCHORED and checking the length, in preparation for adding
2225 something similar for -w.
2227 (b) Added the -w (match as a word) option.
2229 (c) Refactored the way lines are read and buffered so as to have more
2230 than one at a time available.
2232 (d) Implemented a pcregrep test script.
2234 (e) Added the -M (multiline match) option. This allows patterns to match
2235 over several lines of the subject. The buffering ensures that at least
2236 8K, or the rest of the document (whichever is the shorter) is available
2237 for matching (and similarly the previous 8K for lookbehind assertions).
2239 (f) Changed the --help output so that it now says
2241 -w, --word-regex(p)
2243 instead of two lines, one with "regex" and the other with "regexp"
2244 because that confused at least one person since the short forms are the
2245 same. (This required a bit of code, as the output is generated
2246 automatically from a table. It wasn't just a text change.)
2248 (g) -- can be used to terminate pcregrep options if the next thing isn't an
2249 option but starts with a hyphen. Could be a pattern or a path name
2250 starting with a hyphen, for instance.
2252 (h) "-" can be given as a file name to represent stdin.
2254 (i) When file names are being printed, "(standard input)" is used for
2255 the standard input, for compatibility with GNU grep. Previously
2256 "<stdin>" was used.
2258 (j) The option --label=xxx can be used to supply a name to be used for
2259 stdin when file names are being printed. There is no short form.
2261 (k) Re-factored the options decoding logic because we are going to add
2262 two more options that take data. Such options can now be given in four
2263 different ways, e.g. "-fname", "-f name", "--file=name", "--file name".
2265 (l) Added the -A, -B, and -C options for requesting that lines of context
2266 around matches be printed.
2268 (m) Added the -L option to print the names of files that do not contain
2269 any matching lines, that is, the complement of -l.
2271 (n) The return code is 2 if any file cannot be opened, but pcregrep does
2272 continue to scan other files.
2274 (o) The -s option was incorrectly implemented. For compatibility with other
2275 greps, it now suppresses the error message for a non-existent or non-
2276 accessible file (but not the return code). There is a new option called
2277 -q that suppresses the output of matching lines, which was what -s was
2278 previously doing.
2280 (p) Added --include and --exclude options to specify files for inclusion
2281 and exclusion when recursing.
2283 11. The Makefile was not using the Autoconf-supported LDFLAGS macro properly.
2284 Hopefully, it now does.
2286 12. Missing cast in pcre_study().
2288 13. Added an "uninstall" target to the makefile.
2290 14. Replaced "extern" in the function prototypes in Makefile.in with
2291 "PCRE_DATA_SCOPE", which defaults to 'extern' or 'extern "C"' in the Unix
2292 world, but is set differently for Windows.
2294 15. Added a second compiling function called pcre_compile2(). The only
2295 difference is that it has an extra argument, which is a pointer to an
2296 integer error code. When there is a compile-time failure, this is set
2297 non-zero, in addition to the error test pointer being set to point to an
2298 error message. The new argument may be NULL if no error number is required
2299 (but then you may as well call pcre_compile(), which is now just a
2300 wrapper). This facility is provided because some applications need a
2301 numeric error indication, but it has also enabled me to tidy up the way
2302 compile-time errors are handled in the POSIX wrapper.
2304 16. Added VPATH=.libs to the makefile; this should help when building with one
2305 prefix path and installing with another. (Or so I'm told by someone who
2306 knows more about this stuff than I do.)
2308 17. Added a new option, REG_DOTALL, to the POSIX function regcomp(). This
2309 passes PCRE_DOTALL to the pcre_compile() function, making the "." character
2310 match everything, including newlines. This is not POSIX-compatible, but
2311 somebody wanted the feature. From pcretest it can be activated by using
2312 both the P and the s flags.
2314 18. AC_PROG_LIBTOOL appeared twice in Makefile.in. Removed one.
2316 19. libpcre.pc was being incorrectly installed as executable.
2318 20. A couple of places in pcretest check for end-of-line by looking for '\n';
2319 it now also looks for '\r' so that it will work unmodified on Windows.
2321 21. Added Google's contributed C++ wrapper to the distribution.
2323 22. Added some untidy missing memory free() calls in pcretest, to keep
2324 Electric Fence happy when testing.
2328 Version 5.0 13-Sep-04
2329 ---------------------
2331 1. Internal change: literal characters are no longer packed up into items
2332 containing multiple characters in a single byte-string. Each character
2333 is now matched using a separate opcode. However, there may be more than one
2334 byte in the character in UTF-8 mode.
2336 2. The pcre_callout_block structure has two new fields: pattern_position and
2337 next_item_length. These contain the offset in the pattern to the next match
2338 item, and its length, respectively.
2340 3. The PCRE_AUTO_CALLOUT option for pcre_compile() requests the automatic
2341 insertion of callouts before each pattern item. Added the /C option to
2342 pcretest to make use of this.
2344 4. On the advice of a Windows user, the lines
2346 #if defined(_WIN32) || defined(WIN32)
2347 _setmode( _fileno( stdout ), 0x8000 );
2348 #endif /* defined(_WIN32) || defined(WIN32) */
2350 have been added to the source of pcretest. This apparently does useful
2351 magic in relation to line terminators.
2353 5. Changed "r" and "w" in the calls to fopen() in pcretest to "rb" and "wb"
2354 for the benefit of those environments where the "b" makes a difference.
2356 6. The icc compiler has the same options as gcc, but "configure" doesn't seem
2357 to know about it. I have put a hack into configure.in that adds in code
2358 to set GCC=yes if CC=icc. This seems to end up at a point in the
2359 generated configure script that is early enough to affect the setting of
2360 compiler options, which is what is needed, but I have no means of testing
2361 whether it really works. (The user who reported this had patched the
2362 generated configure script, which of course I cannot do.)
2364 LATER: After change 22 below (new libtool files), the configure script
2365 seems to know about icc (and also ecc). Therefore, I have commented out
2366 this hack in configure.in.
2368 7. Added support for pkg-config (2 patches were sent in).
2370 8. Negated POSIX character classes that used a combination of internal tables
2371 were completely broken. These were [[:^alpha:]], [[:^alnum:]], and
2372 [[:^ascii]]. Typically, they would match almost any characters. The other
2373 POSIX classes were not broken in this way.
2375 9. Matching the pattern "\b.*?" against "ab cd", starting at offset 1, failed
2376 to find the match, as PCRE was deluded into thinking that the match had to
2377 start at the start point or following a newline. The same bug applied to
2378 patterns with negative forward assertions or any backward assertions
2379 preceding ".*" at the start, unless the pattern required a fixed first
2380 character. This was a failing pattern: "(?!.bcd).*". The bug is now fixed.
2382 10. In UTF-8 mode, when moving forwards in the subject after a failed match
2383 starting at the last subject character, bytes beyond the end of the subject
2384 string were read.
2386 11. Renamed the variable "class" as "classbits" to make life easier for C++
2387 users. (Previously there was a macro definition, but it apparently wasn't
2388 enough.)
2390 12. Added the new field "tables" to the extra data so that tables can be passed
2391 in at exec time, or the internal tables can be re-selected. This allows
2392 a compiled regex to be saved and re-used at a later time by a different
2393 program that might have everything at different addresses.
2395 13. Modified the pcre-config script so that, when run on Solaris, it shows a
2396 -R library as well as a -L library.
2398 14. The debugging options of pcretest (-d on the command line or D on a
2399 pattern) showed incorrect output for anything following an extended class
2400 that contained multibyte characters and which was followed by a quantifier.
2402 15. Added optional support for general category Unicode character properties
2403 via the \p, \P, and \X escapes. Unicode property support implies UTF-8
2404 support. It adds about 90K to the size of the library. The meanings of the
2405 inbuilt class escapes such as \d and \s have NOT been changed.
2407 16. Updated pcredemo.c to include calls to free() to release the memory for the
2408 compiled pattern.
2410 17. The generated file chartables.c was being created in the source directory
2411 instead of in the building directory. This caused the build to fail if the
2412 source directory was different from the building directory, and was
2413 read-only.
2415 18. Added some sample Win commands from Mark Tetrode into the NON-UNIX-USE
2416 file. No doubt somebody will tell me if they don't make sense... Also added
2417 Dan Mooney's comments about building on OpenVMS.
2419 19. Added support for partial matching via the PCRE_PARTIAL option for
2420 pcre_exec() and the \P data escape in pcretest.
2422 20. Extended pcretest with 3 new pattern features:
2424 (i) A pattern option of the form ">rest-of-line" causes pcretest to
2425 write the compiled pattern to the file whose name is "rest-of-line".
2426 This is a straight binary dump of the data, with the saved pointer to
2427 the character tables forced to be NULL. The study data, if any, is
2428 written too. After writing, pcretest reads a new pattern.
2430 (ii) If, instead of a pattern, "<rest-of-line" is given, pcretest reads a
2431 compiled pattern from the given file. There must not be any
2432 occurrences of "<" in the file name (pretty unlikely); if there are,
2433 pcretest will instead treat the initial "<" as a pattern delimiter.
2434 After reading in the pattern, pcretest goes on to read data lines as
2435 usual.
2437 (iii) The F pattern option causes pcretest to flip the bytes in the 32-bit
2438 and 16-bit fields in a compiled pattern, to simulate a pattern that
2439 was compiled on a host of opposite endianness.
2441 21. The pcre-exec() function can now cope with patterns that were compiled on
2442 hosts of opposite endianness, with this restriction:
2444 As for any compiled expression that is saved and used later, the tables
2445 pointer field cannot be preserved; the extra_data field in the arguments
2446 to pcre_exec() should be used to pass in a tables address if a value
2447 other than the default internal tables were used at compile time.
2449 22. Calling pcre_exec() with a negative value of the "ovecsize" parameter is
2450 now diagnosed as an error. Previously, most of the time, a negative number
2451 would have been treated as zero, but if in addition "ovector" was passed as
2452 NULL, a crash could occur.
2454 23. Updated the files ltmain.sh, config.sub, config.guess, and aclocal.m4 with
2455 new versions from the libtool 1.5 distribution (the last one is a copy of
2456 a file called libtool.m4). This seems to have fixed the need to patch
2457 "configure" to support Darwin 1.3 (which I used to do). However, I still
2458 had to patch ltmain.sh to ensure that ${SED} is set (it isn't on my
2459 workstation).
2461 24. Changed the PCRE licence to be the more standard "BSD" licence.
2464 Version 4.5 01-Dec-03
2465 ---------------------
2467 1. There has been some re-arrangement of the code for the match() function so
2468 that it can be compiled in a version that does not call itself recursively.
2469 Instead, it keeps those local variables that need separate instances for
2470 each "recursion" in a frame on the heap, and gets/frees frames whenever it
2471 needs to "recurse". Keeping track of where control must go is done by means
2472 of setjmp/longjmp. The whole thing is implemented by a set of macros that
2473 hide most of the details from the main code, and operates only if
2474 NO_RECURSE is defined while compiling pcre.c. If PCRE is built using the
2475 "configure" mechanism, "--disable-stack-for-recursion" turns on this way of
2476 operating.
2478 To make it easier for callers to provide specially tailored get/free
2479 functions for this usage, two new functions, pcre_stack_malloc, and
2480 pcre_stack_free, are used. They are always called in strict stacking order,
2481 and the size of block requested is always the same.
2483 The PCRE_CONFIG_STACKRECURSE info parameter can be used to find out whether
2484 PCRE has been compiled to use the stack or the heap for recursion. The
2485 -C option of pcretest uses this to show which version is compiled.
2487 A new data escape \S, is added to pcretest; it causes the amounts of store
2488 obtained and freed by both kinds of malloc/free at match time to be added
2489 to the output.
2491 2. Changed the locale test to use "fr_FR" instead of "fr" because that's
2492 what's available on my current Linux desktop machine.
2494 3. When matching a UTF-8 string, the test for a valid string at the start has
2495 been extended. If start_offset is not zero, PCRE now checks that it points
2496 to a byte that is the start of a UTF-8 character. If not, it returns
2497 PCRE_ERROR_BADUTF8_OFFSET (-11). Note: the whole string is still checked;
2498 this is necessary because there may be backward assertions in the pattern.
2499 When matching the same subject several times, it may save resources to use
2500 PCRE_NO_UTF8_CHECK on all but the first call if the string is long.
2502 4. The code for checking the validity of UTF-8 strings has been tightened so
2503 that it rejects (a) strings containing 0xfe or 0xff bytes and (b) strings
2504 containing "overlong sequences".
2506 5. Fixed a bug (appearing twice) that I could not find any way of exploiting!
2507 I had written "if ((digitab[*p++] && chtab_digit) == 0)" where the "&&"
2508 should have been "&", but it just so happened that all the cases this let
2509 through by mistake were picked up later in the function.
2511 6. I had used a variable called "isblank" - this is a C99 function, causing
2512 some compilers to warn. To avoid this, I renamed it (as "blankclass").
2514 7. Cosmetic: (a) only output another newline at the end of pcretest if it is
2515 prompting; (b) run "./pcretest /dev/null" at the start of the test script
2516 so the version is shown; (c) stop "make test" echoing "./RunTest".
2518 8. Added patches from David Burgess to enable PCRE to run on EBCDIC systems.
2520 9. The prototype for memmove() for systems that don't have it was using
2521 size_t, but the inclusion of the header that defines size_t was later. I've
2522 moved the #includes for the C headers earlier to avoid this.
2524 10. Added some adjustments to the code to make it easier to compiler on certain
2525 special systems:
2527 (a) Some "const" qualifiers were missing.
2528 (b) Added the macro EXPORT before all exported functions; by default this
2529 is defined to be empty.
2530 (c) Changed the dftables auxiliary program (that builds chartables.c) so
2531 that it reads its output file name as an argument instead of writing
2532 to the standard output and assuming this can be redirected.
2534 11. In UTF-8 mode, if a recursive reference (e.g. (?1)) followed a character
2535 class containing characters with values greater than 255, PCRE compilation
2536 went into a loop.
2538 12. A recursive reference to a subpattern that was within another subpattern
2539 that had a minimum quantifier of zero caused PCRE to crash. For example,
2540 (x(y(?2))z)? provoked this bug with a subject that got as far as the
2541 recursion. If the recursively-called subpattern itself had a zero repeat,
2542 that was OK.
2544 13. In pcretest, the buffer for reading a data line was set at 30K, but the
2545 buffer into which it was copied (for escape processing) was still set at
2546 1024, so long lines caused crashes.
2548 14. A pattern such as /[ab]{1,3}+/ failed to compile, giving the error
2549 "internal error: code overflow...". This applied to any character class
2550 that was followed by a possessive quantifier.
2552 15. Modified the Makefile to add libpcre.la as a prerequisite for
2553 libpcreposix.la because I was told this is needed for a parallel build to
2554 work.
2556 16. If a pattern that contained .* following optional items at the start was
2557 studied, the wrong optimizing data was generated, leading to matching
2558 errors. For example, studying /[ab]*.*c/ concluded, erroneously, that any
2559 matching string must start with a or b or c. The correct conclusion for
2560 this pattern is that a match can start with any character.
2563 Version 4.4 13-Aug-03
2564 ---------------------
2566 1. In UTF-8 mode, a character class containing characters with values between
2567 127 and 255 was not handled correctly if the compiled pattern was studied.
2568 In fixing this, I have also improved the studying algorithm for such
2569 classes (slightly).
2571 2. Three internal functions had redundant arguments passed to them. Removal
2572 might give a very teeny performance improvement.
2574 3. Documentation bug: the value of the capture_top field in a callout is *one
2575 more than* the number of the hightest numbered captured substring.
2577 4. The Makefile linked pcretest and pcregrep with -lpcre, which could result
2578 in incorrectly linking with a previously installed version. They now link
2579 explicitly with libpcre.la.
2581 5. configure.in no longer needs to recognize Cygwin specially.
2583 6. A problem in pcre.in for Windows platforms is fixed.
2585 7. If a pattern was successfully studied, and the -d (or /D) flag was given to
2586 pcretest, it used to include the size of the study block as part of its
2587 output. Unfortunately, the structure contains a field that has a different
2588 size on different hardware architectures. This meant that the tests that
2589 showed this size failed. As the block is currently always of a fixed size,
2590 this information isn't actually particularly useful in pcretest output, so
2591 I have just removed it.
2593 8. Three pre-processor statements accidentally did not start in column 1.
2594 Sadly, there are *still* compilers around that complain, even though
2595 standard C has not required this for well over a decade. Sigh.
2597 9. In pcretest, the code for checking callouts passed small integers in the
2598 callout_data field, which is a void * field. However, some picky compilers
2599 complained about the casts involved for this on 64-bit systems. Now
2600 pcretest passes the address of the small integer instead, which should get
2601 rid of the warnings.
2603 10. By default, when in UTF-8 mode, PCRE now checks for valid UTF-8 strings at
2604 both compile and run time, and gives an error if an invalid UTF-8 sequence
2605 is found. There is a option for disabling this check in cases where the
2606 string is known to be correct and/or the maximum performance is wanted.
2608 11. In response to a bug report, I changed one line in Makefile.in from
2610 -Wl,--out-implib,.libs/lib@WIN_PREFIX@pcreposix.dll.a \
2611 to
2612 -Wl,--out-implib,.libs/@WIN_PREFIX@libpcreposix.dll.a \
2614 to look similar to other lines, but I have no way of telling whether this
2615 is the right thing to do, as I do not use Windows. No doubt I'll get told
2616 if it's wrong...
2619 Version 4.3 21-May-03
2620 ---------------------
2622 1. Two instances of @WIN_PREFIX@ omitted from the Windows targets in the
2623 Makefile.
2625 2. Some refactoring to improve the quality of the code:
2627 (i) The utf8_table... variables are now declared "const".
2629 (ii) The code for \cx, which used the "case flipping" table to upper case
2630 lower case letters, now just substracts 32. This is ASCII-specific,
2631 but the whole concept of \cx is ASCII-specific, so it seems
2632 reasonable.
2634 (iii) PCRE was using its character types table to recognize decimal and
2635 hexadecimal digits in the pattern. This is silly, because it handles
2636 only 0-9, a-f, and A-F, but the character types table is locale-
2637 specific, which means strange things might happen. A private
2638 table is now used for this - though it costs 256 bytes, a table is
2639 much faster than multiple explicit tests. Of course, the standard
2640 character types table is still used for matching digits in subject
2641 strings against \d.
2643 (iv) Strictly, the identifier ESC_t is reserved by POSIX (all identifiers
2644 ending in _t are). So I've renamed it as ESC_tee.
2646 3. The first argument for regexec() in the POSIX wrapper should have been
2647 defined as "const".
2649 4. Changed pcretest to use malloc() for its buffers so that they can be
2650 Electric Fenced for debugging.
2652 5. There were several places in the code where, in UTF-8 mode, PCRE would try
2653 to read one or more bytes before the start of the subject string. Often this
2654 had no effect on PCRE's behaviour, but in some circumstances it could
2655 provoke a segmentation fault.
2657 6. A lookbehind at the start of a pattern in UTF-8 mode could also cause PCRE
2658 to try to read one or more bytes before the start of the subject string.
2660 7. A lookbehind in a pattern matched in non-UTF-8 mode on a PCRE compiled with
2661 UTF-8 support could misbehave in various ways if the subject string
2662 contained bytes with the 0x80 bit set and the 0x40 bit unset in a lookbehind
2663 area. (PCRE was not checking for the UTF-8 mode flag, and trying to move
2664 back over UTF-8 characters.)
2667 Version 4.2 14-Apr-03
2668 ---------------------
2670 1. Typo "#if SUPPORT_UTF8" instead of "#ifdef SUPPORT_UTF8" fixed.
2672 2. Changes to the building process, supplied by Ronald Landheer-Cieslak
2673 [ON_WINDOWS]: new variable, "#" on non-Windows platforms
2674 [NOT_ON_WINDOWS]: new variable, "#" on Windows platforms
2675 [WIN_PREFIX]: new variable, "cyg" for Cygwin
2676 * Makefile.in: use autoconf substitution for OBJEXT, EXEEXT, BUILD_OBJEXT
2678 Note: automatic setting of the BUILD variables is not yet working
2679 set CPPFLAGS and BUILD_CPPFLAGS (but don't use yet) - should be used at
2680 compile-time but not at link-time
2681 [LINK]: use for linking executables only
2682 make different versions for Windows and non-Windows
2683 [LINKLIB]: new variable, copy of UNIX-style LINK, used for linking
2684 libraries
2685 [LINK_FOR_BUILD]: new variable
2686 [OBJEXT]: use throughout
2687 [EXEEXT]: use throughout
2688 <winshared>: new target
2689 <wininstall>: new target
2690 <dftables.o>: use native compiler
2691 <dftables>: use native linker
2692 <install>: handle Windows platform correctly
2693 <clean>: ditto
2694 <check>: ditto
2695 copy DLL to top builddir before testing
2697 As part of these changes, -no-undefined was removed again. This was reported
2698 to give trouble on HP-UX 11.0, so getting rid of it seems like a good idea
2699 in any case.
2701 3. Some tidies to get rid of compiler warnings:
2703 . In the match_data structure, match_limit was an unsigned long int, whereas
2704 match_call_count was an int. I've made them both unsigned long ints.
2706 . In pcretest the fact that a const uschar * doesn't automatically cast to
2707 a void * provoked a warning.
2709 . Turning on some more compiler warnings threw up some "shadow" variables
2710 and a few more missing casts.
2712 4. If PCRE was complied with UTF-8 support, but called without the PCRE_UTF8
2713 option, a class that contained a single character with a value between 128
2714 and 255 (e.g. /[\xFF]/) caused PCRE to crash.
2716 5. If PCRE was compiled with UTF-8 support, but called without the PCRE_UTF8
2717 option, a class that contained several characters, but with at least one
2718 whose value was between 128 and 255 caused PCRE to crash.
2721 Version 4.1 12-Mar-03
2722 ---------------------
2724 1. Compiling with gcc -pedantic found a couple of places where casts were
2725 needed, and a string in dftables.c that was longer than standard compilers are
2726 required to support.
2728 2. Compiling with Sun's compiler found a few more places where the code could
2729 be tidied up in order to avoid warnings.
2731 3. The variables for cross-compiling were called HOST_CC and HOST_CFLAGS; the
2732 first of these names is deprecated in the latest Autoconf in favour of the name
2733 CC_FOR_BUILD, because "host" is typically used to mean the system on which the
2734 compiled code will be run. I can't find a reference for HOST_CFLAGS, but by
2735 analogy I have changed it to CFLAGS_FOR_BUILD.
2737 4. Added -no-undefined to the linking command in the Makefile, because this is
2738 apparently helpful for Windows. To make it work, also added "-L. -lpcre" to the
2739 linking step for the pcreposix library.
2741 5. PCRE was failing to diagnose the case of two named groups with the same
2742 name.
2744 6. A problem with one of PCRE's optimizations was discovered. PCRE remembers a
2745 literal character that is needed in the subject for a match, and scans along to
2746 ensure that it is present before embarking on the full matching process. This
2747 saves time in cases of nested unlimited repeats that are never going to match.
2748 Problem: the scan can take a lot of time if the subject is very long (e.g.
2749 megabytes), thus penalizing straightforward matches. It is now done only if the
2750 amount of subject to be scanned is less than 1000 bytes.
2752 7. A lesser problem with the same optimization is that it was recording the
2753 first character of an anchored pattern as "needed", thus provoking a search
2754 right along the subject, even when the first match of the pattern was going to
2755 fail. The "needed" character is now not set for anchored patterns, unless it
2756 follows something in the pattern that is of non-fixed length. Thus, it still
2757 fulfils its original purpose of finding quick non-matches in cases of nested
2758 unlimited repeats, but isn't used for simple anchored patterns such as /^abc/.
2761 Version 4.0 17-Feb-03
2762 ---------------------
2764 1. If a comment in an extended regex that started immediately after a meta-item
2765 extended to the end of string, PCRE compiled incorrect data. This could lead to
2766 all kinds of weird effects. Example: /#/ was bad; /()#/ was bad; /a#/ was not.
2768 2. Moved to autoconf 2.53 and libtool 1.4.2.
2770 3. Perl 5.8 no longer needs "use utf8" for doing UTF-8 things. Consequently,
2771 the special perltest8 script is no longer needed - all the tests can be run
2772 from a single perltest script.
2774 4. From 5.004, Perl has not included the VT character (0x0b) in the set defined
2775 by \s. It has now been removed in PCRE. This means it isn't recognized as
2776 whitespace in /x regexes too, which is the same as Perl. Note that the POSIX
2777 class [:space:] *does* include VT, thereby creating a mess.
2779 5. Added the class [:blank:] (a GNU extension from Perl 5.8) to match only
2780 space and tab.
2782 6. Perl 5.005 was a long time ago. It's time to amalgamate the tests that use
2783 its new features into the main test script, reducing the number of scripts.
2785 7. Perl 5.8 has changed the meaning of patterns like /a(?i)b/. Earlier versions
2786 were backward compatible, and made the (?i) apply to the whole pattern, as if
2787 /i were given. Now it behaves more logically, and applies the option setting
2788 only to what follows. PCRE has been changed to follow suit. However, if it
2789 finds options settings right at the start of the pattern, it extracts them into
2790 the global options, as before. Thus, they show up in the info data.
2792 8. Added support for the \Q...\E escape sequence. Characters in between are
2793 treated as literals. This is slightly different from Perl in that $ and @ are
2794 also handled as literals inside the quotes. In Perl, they will cause variable
2795 interpolation. Note the following examples:
2797 Pattern PCRE matches Perl matches
2799 \Qabc$xyz\E abc$xyz abc followed by the contents of $xyz
2800 \Qabc\$xyz\E abc\$xyz abc\$xyz
2801 \Qabc\E\$\Qxyz\E abc$xyz abc$xyz
2803 For compatibility with Perl, \Q...\E sequences are recognized inside character
2804 classes as well as outside them.
2806 9. Re-organized 3 code statements in pcretest to avoid "overflow in
2807 floating-point constant arithmetic" warnings from a Microsoft compiler. Added a
2808 (size_t) cast to one statement in pcretest and one in pcreposix to avoid
2809 signed/unsigned warnings.
2811 10. SunOS4 doesn't have strtoul(). This was used only for unpicking the -o
2812 option for pcretest, so I've replaced it by a simple function that does just
2813 that job.
2815 11. pcregrep was ending with code 0 instead of 2 for the commands "pcregrep" or
2816 "pcregrep -".
2818 12. Added "possessive quantifiers" ?+, *+, ++, and {,}+ which come from Sun's
2819 Java package. This provides some syntactic sugar for simple cases of what my
2820 documentation calls "once-only subpatterns". A pattern such as x*+ is the same
2821 as (?>x*). In other words, if what is inside (?>...) is just a single repeated
2822 item, you can use this simplified notation. Note that only makes sense with
2823 greedy quantifiers. Consequently, the use of the possessive quantifier forces
2824 greediness, whatever the setting of the PCRE_UNGREEDY option.
2826 13. A change of greediness default within a pattern was not taking effect at
2827 the current level for patterns like /(b+(?U)a+)/. It did apply to parenthesized
2828 subpatterns that followed. Patterns like /b+(?U)a+/ worked because the option
2829 was abstracted outside.
2831 14. PCRE now supports the \G assertion. It is true when the current matching
2832 position is at the start point of the match. This differs from \A when the
2833 starting offset is non-zero. Used with the /g option of pcretest (or similar
2834 code), it works in the same way as it does for Perl's /g option. If all
2835 alternatives of a regex begin with \G, the expression is anchored to the start
2836 match position, and the "anchored" flag is set in the compiled expression.
2838 15. Some bugs concerning the handling of certain option changes within patterns
2839 have been fixed. These applied to options other than (?ims). For example,
2840 "a(?x: b c )d" did not match "XabcdY" but did match "Xa b c dY". It should have
2841 been the other way round. Some of this was related to change 7 above.
2843 16. PCRE now gives errors for /[.x.]/ and /[=x=]/ as unsupported POSIX
2844 features, as Perl does. Previously, PCRE gave the warnings only for /[[.x.]]/
2845 and /[[=x=]]/. PCRE now also gives an error for /[:name:]/ because it supports
2846 POSIX classes only within a class (e.g. /[[:alpha:]]/).
2848 17. Added support for Perl's \C escape. This matches one byte, even in UTF8
2849 mode. Unlike ".", it always matches newline, whatever the setting of
2850 PCRE_DOTALL. However, PCRE does not permit \C to appear in lookbehind
2851 assertions. Perl allows it, but it doesn't (in general) work because it can't
2852 calculate the length of the lookbehind. At least, that's the case for Perl
2853 5.8.0 - I've been told they are going to document that it doesn't work in
2854 future.
2856 18. Added an error diagnosis for escapes that PCRE does not support: these are
2857 \L, \l, \N, \P, \p, \U, \u, and \X.
2859 19. Although correctly diagnosing a missing ']' in a character class, PCRE was
2860 reading past the end of the pattern in cases such as /[abcd/.
2862 20. PCRE was getting more memory than necessary for patterns with classes that
2863 contained both POSIX named classes and other characters, e.g. /[[:space:]abc/.
2865 21. Added some code, conditional on #ifdef VPCOMPAT, to make life easier for
2866 compiling PCRE for use with Virtual Pascal.
2868 22. Small fix to the Makefile to make it work properly if the build is done
2869 outside the source tree.
2871 23. Added a new extension: a condition to go with recursion. If a conditional
2872 subpattern starts with (?(R) the "true" branch is used if recursion has
2873 happened, whereas the "false" branch is used only at the top level.
2875 24. When there was a very long string of literal characters (over 255 bytes
2876 without UTF support, over 250 bytes with UTF support), the computation of how
2877 much memory was required could be incorrect, leading to segfaults or other
2878 strange effects.
2880 25. PCRE was incorrectly assuming anchoring (either to start of subject or to
2881 start of line for a non-DOTALL pattern) when a pattern started with (.*) and
2882 there was a subsequent back reference to those brackets. This meant that, for
2883 example, /(.*)\d+\1/ failed to match "abc123bc". Unfortunately, it isn't
2884 possible to check for precisely this case. All we can do is abandon the
2885 optimization if .* occurs inside capturing brackets when there are any back
2886 references whatsoever. (See below for a better fix that came later.)
2888 26. The handling of the optimization for finding the first character of a
2889 non-anchored pattern, and for finding a character that is required later in the
2890 match were failing in some cases. This didn't break the matching; it just
2891 failed to optimize when it could. The way this is done has been re-implemented.
2893 27. Fixed typo in error message for invalid (?R item (it said "(?p").
2895 28. Added a new feature that provides some of the functionality that Perl
2896 provides with (?{...}). The facility is termed a "callout". The way it is done
2897 in PCRE is for the caller to provide an optional function, by setting
2898 pcre_callout to its entry point. Like pcre_malloc and pcre_free, this is a
2899 global variable. By default it is unset, which disables all calling out. To get
2900 the function called, the regex must include (?C) at appropriate points. This
2901 is, in fact, equivalent to (?C0), and any number <= 255 may be given with (?C).
2902 This provides a means of identifying different callout points. When PCRE
2903 reaches such a point in the regex, if pcre_callout has been set, the external
2904 function is called. It is provided with data in a structure called
2905 pcre_callout_block, which is defined in pcre.h. If the function returns 0,
2906 matching continues; if it returns a non-zero value, the match at the current
2907 point fails. However, backtracking will occur if possible. [This was changed
2908 later and other features added - see item 49 below.]
2910 29. pcretest is upgraded to test the callout functionality. It provides a
2911 callout function that displays information. By default, it shows the start of
2912 the match and the current position in the text. There are some new data escapes
2913 to vary what happens:
2915 \C+ in addition, show current contents of captured substrings
2916 \C- do not supply a callout function
2917 \C!n return 1 when callout number n is reached
2918 \C!n!m return 1 when callout number n is reached for the mth time
2920 30. If pcregrep was called with the -l option and just a single file name, it
2921 output "<stdin>" if a match was found, instead of the file name.
2923 31. Improve the efficiency of the POSIX API to PCRE. If the number of capturing
2924 slots is less than POSIX_MALLOC_THRESHOLD, use a block on the stack to pass to
2925 pcre_exec(). This saves a malloc/free per call. The default value of
2926 POSIX_MALLOC_THRESHOLD is 10; it can be changed by --with-posix-malloc-threshold
2927 when configuring.
2929 32. The default maximum size of a compiled pattern is 64K. There have been a
2930 few cases of people hitting this limit. The code now uses macros to handle the
2931 storing of links as offsets within the compiled pattern. It defaults to 2-byte
2932 links, but this can be changed to 3 or 4 bytes by --with-link-size when
2933 configuring. Tests 2 and 5 work only with 2-byte links because they output
2934 debugging information about compiled patterns.
2936 33. Internal code re-arrangements:
2938 (a) Moved the debugging function for printing out a compiled regex into
2939 its own source file (printint.c) and used #include to pull it into
2940 pcretest.c and, when DEBUG is defined, into pcre.c, instead of having two
2941 separate copies.
2943 (b) Defined the list of op-code names for debugging as a macro in
2944 internal.h so that it is next to the definition of the opcodes.
2946 (c) Defined a table of op-code lengths for simpler skipping along compiled
2947 code. This is again a macro in internal.h so that it is next to the
2948 definition of the opcodes.
2950 34. Added support for recursive calls to individual subpatterns, along the
2951 lines of Robin Houston's patch (but implemented somewhat differently).
2953 35. Further mods to the Makefile to help Win32. Also, added code to pcregrep to
2954 allow it to read and process whole directories in Win32. This code was
2955 contributed by Lionel Fourquaux; it has not been tested by me.
2957 36. Added support for named subpatterns. The Python syntax (?P<name>...) is
2958 used to name a group. Names consist of alphanumerics and underscores, and must
2959 be unique. Back references use the syntax (?P=name) and recursive calls use
2960 (?P>name) which is a PCRE extension to the Python extension. Groups still have
2961 numbers. The function pcre_fullinfo() can be used after compilation to extract
2962 a name/number map. There are three relevant calls:
2964 PCRE_INFO_NAMEENTRYSIZE yields the size of each entry in the map
2965 PCRE_INFO_NAMECOUNT yields the number of entries
2966 PCRE_INFO_NAMETABLE yields a pointer to the map.
2968 The map is a vector of fixed-size entries. The size of each entry depends on
2969 the length of the longest name used. The first two bytes of each entry are the
2970 group number, most significant byte first. There follows the corresponding
2971 name, zero terminated. The names are in alphabetical order.
2973 37. Make the maximum literal string in the compiled code 250 for the non-UTF-8
2974 case instead of 255. Making it the same both with and without UTF-8 support
2975 means that the same test output works with both.
2977 38. There was a case of malloc(0) in the POSIX testing code in pcretest. Avoid
2978 calling malloc() with a zero argument.
2980 39. Change 25 above had to resort to a heavy-handed test for the .* anchoring
2981 optimization. I've improved things by keeping a bitmap of backreferences with
2982 numbers 1-31 so that if .* occurs inside capturing brackets that are not in
2983 fact referenced, the optimization can be applied. It is unlikely that a
2984 relevant occurrence of .* (i.e. one which might indicate anchoring or forcing
2985 the match to follow \n) will appear inside brackets with a number greater than
2986 31, but if it does, any back reference > 31 suppresses the optimization.
2988 40. Added a new compile-time option PCRE_NO_AUTO_CAPTURE. This has the effect
2989 of disabling numbered capturing parentheses. Any opening parenthesis that is
2990 not followed by ? behaves as if it were followed by ?: but named parentheses
2991 can still be used for capturing (and they will acquire numbers in the usual
2992 way).
2994 41. Redesigned the return codes from the match() function into yes/no/error so
2995 that errors can be passed back from deep inside the nested calls. A malloc
2996 failure while inside a recursive subpattern call now causes the
2997 PCRE_ERROR_NOMEMORY return instead of quietly going wrong.
2999 42. It is now possible to set a limit on the number of times the match()
3000 function is called in a call to pcre_exec(). This facility makes it possible to
3001 limit the amount of recursion and backtracking, though not in a directly
3002 obvious way, because the match() function is used in a number of different
3003 circumstances. The count starts from zero for each position in the subject
3004 string (for non-anchored patterns). The default limit is, for compatibility, a
3005 large number, namely 10 000 000. You can change this in two ways:
3007 (a) When configuring PCRE before making, you can use --with-match-limit=n
3008 to set a default value for the compiled library.
3010 (b) For each call to pcre_exec(), you can pass a pcre_extra block in which
3011 a different value is set. See 45 below.
3013 If the limit is exceeded, pcre_exec() returns PCRE_ERROR_MATCHLIMIT.
3015 43. Added a new function pcre_config(int, void *) to enable run-time extraction
3016 of things that can be changed at compile time. The first argument specifies
3017 what is wanted and the second points to where the information is to be placed.
3018 The current list of available information is:
3022 The output is an integer that is set to one if UTF-8 support is available;
3023 otherwise it is set to zero.
3027 The output is an integer that it set to the value of the code that is used for
3028 newline. It is either LF (10) or CR (13).
3032 The output is an integer that contains the number of bytes used for internal
3033 linkage in compiled expressions. The value is 2, 3, or 4. See item 32 above.
3037 The output is an integer that contains the threshold above which the POSIX
3038 interface uses malloc() for output vectors. See item 31 above.
3042 The output is an unsigned integer that contains the default limit of the number
3043 of match() calls in a pcre_exec() execution. See 42 above.
3045 44. pcretest has been upgraded by the addition of the -C option. This causes it
3046 to extract all the available output from the new pcre_config() function, and to
3047 output it. The program then exits immediately.
3049 45. A need has arisen to pass over additional data with calls to pcre_exec() in
3050 order to support additional features. One way would have been to define
3051 pcre_exec2() (for example) with extra arguments, but this would not have been
3052 extensible, and would also have required all calls to the original function to
3053 be mapped to the new one. Instead, I have chosen to extend the mechanism that
3054 is used for passing in "extra" data from pcre_study().
3056 The pcre_extra structure is now exposed and defined in pcre.h. It currently
3057 contains the following fields:
3059 flags a bitmap indicating which of the following fields are set
3060 study_data opaque data from pcre_study()
3061 match_limit a way of specifying a limit on match() calls for a specific
3062 call to pcre_exec()
3063 callout_data data for callouts (see 49 below)
3065 The flag bits are also defined in pcre.h, and are
3071 The pcre_study() function now returns one of these new pcre_extra blocks, with
3072 the actual study data pointed to by the study_data field, and the
3073 PCRE_EXTRA_STUDY_DATA flag set. This can be passed directly to pcre_exec() as
3074 before. That is, this change is entirely upwards-compatible and requires no
3075 change to existing code.
3077 If you want to pass in additional data to pcre_exec(), you can either place it
3078 in a pcre_extra block provided by pcre_study(), or create your own pcre_extra
3079 block.
3081 46. pcretest has been extended to test the PCRE_EXTRA_MATCH_LIMIT feature. If a
3082 data string contains the escape sequence \M, pcretest calls pcre_exec() several
3083 times with different match limits, until it finds the minimum value needed for
3084 pcre_exec() to complete. The value is then output. This can be instructive; for
3085 most simple matches the number is quite small, but for pathological cases it
3086 gets very large very quickly.
3088 47. There's a new option for pcre_fullinfo() called PCRE_INFO_STUDYSIZE. It
3089 returns the size of the data block pointed to by the study_data field in a
3090 pcre_extra block, that is, the value that was passed as the argument to
3091 pcre_malloc() when PCRE was getting memory in which to place the information
3092 created by pcre_study(). The fourth argument should point to a size_t variable.
3093 pcretest has been extended so that this information is shown after a successful
3094 pcre_study() call when information about the compiled regex is being displayed.
3096 48. Cosmetic change to Makefile: there's no need to have / after $(DESTDIR)
3097 because what follows is always an absolute path. (Later: it turns out that this
3098 is more than cosmetic for MinGW, because it doesn't like empty path
3099 components.)
3101 49. Some changes have been made to the callout feature (see 28 above):
3103 (i) A callout function now has three choices for what it returns:
3105 0 => success, carry on matching
3106 > 0 => failure at this point, but backtrack if possible
3107 < 0 => serious error, return this value from pcre_exec()
3109 Negative values should normally be chosen from the set of PCRE_ERROR_xxx
3110 values. In particular, returning PCRE_ERROR_NOMATCH forces a standard
3111 "match failed" error. The error number PCRE_ERROR_CALLOUT is reserved for
3112 use by callout functions. It will never be used by PCRE itself.
3114 (ii) The pcre_extra structure (see 45 above) has a void * field called
3115 callout_data, with corresponding flag bit PCRE_EXTRA_CALLOUT_DATA. The
3116 pcre_callout_block structure has a field of the same name. The contents of
3117 the field passed in the pcre_extra structure are passed to the callout
3118 function in the corresponding field in the callout block. This makes it
3119 easier to use the same callout-containing regex from multiple threads. For
3120 testing, the pcretest program has a new data escape
3122 \C*n pass the number n (may be negative) as callout_data
3124 If the callout function in pcretest receives a non-zero value as
3125 callout_data, it returns that value.
3127 50. Makefile wasn't handling CFLAGS properly when compiling dftables. Also,
3128 there were some redundant $(CFLAGS) in commands that are now specified as
3129 $(LINK), which already includes $(CFLAGS).
3131 51. Extensions to UTF-8 support are listed below. These all apply when (a) PCRE
3132 has been compiled with UTF-8 support *and* pcre_compile() has been compiled
3133 with the PCRE_UTF8 flag. Patterns that are compiled without that flag assume
3134 one-byte characters throughout. Note that case-insensitive matching applies
3135 only to characters whose values are less than 256. PCRE doesn't support the
3136 notion of cases for higher-valued characters.
3138 (i) A character class whose characters are all within 0-255 is handled as
3139 a bit map, and the map is inverted for negative classes. Previously, a
3140 character > 255 always failed to match such a class; however it should
3141 match if the class was a negative one (e.g. [^ab]). This has been fixed.
3143 (ii) A negated character class with a single character < 255 is coded as
3144 "not this character" (OP_NOT). This wasn't working properly when the test
3145 character was multibyte, either singly or repeated.
3147 (iii) Repeats of multibyte characters are now handled correctly in UTF-8
3148 mode, for example: \x{100}{2,3}.
3150 (iv) The character escapes \b, \B, \d, \D, \s, \S, \w, and \W (either
3151 singly or repeated) now correctly test multibyte characters. However,
3152 PCRE doesn't recognize any characters with values greater than 255 as
3153 digits, spaces, or word characters. Such characters always match \D, \S,
3154 and \W, and never match \d, \s, or \w.
3156 (v) Classes may now contain characters and character ranges with values
3157 greater than 255. For example: [ab\x{100}-\x{400}].
3159 (vi) pcregrep now has a --utf-8 option (synonym -u) which makes it call
3160 PCRE in UTF-8 mode.
3162 52. The info request value PCRE_INFO_FIRSTCHAR has been renamed
3163 PCRE_INFO_FIRSTBYTE because it is a byte value. However, the old name is
3164 retained for backwards compatibility. (Note that LASTLITERAL is also a byte
3165 value.)
3167 53. The single man page has become too large. I have therefore split it up into
3168 a number of separate man pages. These also give rise to individual HTML pages;
3169 these are now put in a separate directory, and there is an index.html page that
3170 lists them all. Some hyperlinking between the pages has been installed.
3172 54. Added convenience functions for handling named capturing parentheses.
3174 55. Unknown escapes inside character classes (e.g. [\M]) and escapes that
3175 aren't interpreted therein (e.g. [\C]) are literals in Perl. This is now also
3176 true in PCRE, except when the PCRE_EXTENDED option is set, in which case they
3177 are faulted.
3179 56. Introduced HOST_CC and HOST_CFLAGS which can be set in the environment when
3180 calling configure. These values are used when compiling the dftables.c program
3181 which is run to generate the source of the default character tables. They
3182 default to the values of CC and CFLAGS. If you are cross-compiling PCRE,
3183 you will need to set these values.
3185 57. Updated the building process for Windows DLL, as provided by Fred Cox.
3188 Version 3.9 02-Jan-02
3189 ---------------------
3191 1. A bit of extraneous text had somehow crept into the pcregrep documentation.
3193 2. If --disable-static was given, the building process failed when trying to
3194 build pcretest and pcregrep. (For some reason it was using libtool to compile
3195 them, which is not right, as they aren't part of the library.)
3198 Version 3.8 18-Dec-01
3199 ---------------------
3201 1. The experimental UTF-8 code was completely screwed up. It was packing the
3202 bytes in the wrong order. How dumb can you get?
3205 Version 3.7 29-Oct-01
3206 ---------------------
3208 1. In updating pcretest to check change 1 of version 3.6, I screwed up.
3209 This caused pcretest, when used on the test data, to segfault. Unfortunately,
3210 this didn't happen under Solaris 8, where I normally test things.
3212 2. The Makefile had to be changed to make it work on BSD systems, where 'make'
3213 doesn't seem to recognize that ./xxx and xxx are the same file. (This entry
3214 isn't in ChangeLog distributed with 3.7 because I forgot when I hastily made
3215 this fix an hour or so after the initial 3.7 release.)
3218 Version 3.6 23-Oct-01
3219 ---------------------
3221 1. Crashed with /(sens|respons)e and \1ibility/ and "sense and sensibility" if
3222 offsets passed as NULL with zero offset count.
3224 2. The config.guess and config.sub files had not been updated when I moved to
3225 the latest autoconf.
3228 Version 3.5 15-Aug-01
3229 ---------------------
3231 1. Added some missing #if !defined NOPOSIX conditionals in pcretest.c that
3232 had been forgotten.
3234 2. By using declared but undefined structures, we can avoid using "void"
3235 definitions in pcre.h while keeping the internal definitions of the structures
3236 private.
3238 3. The distribution is now built using autoconf 2.50 and libtool 1.4. From a
3239 user point of view, this means that both static and shared libraries are built
3240 by default, but this can be individually controlled. More of the work of
3241 handling this static/shared cases is now inside libtool instead of PCRE's make
3242 file.
3244 4. The pcretest utility is now installed along with pcregrep because it is
3245 useful for users (to test regexs) and by doing this, it automatically gets
3246 relinked by libtool. The documentation has been turned into a man page, so
3247 there are now .1, .txt, and .html versions in /doc.
3249 5. Upgrades to pcregrep:
3250 (i) Added long-form option names like gnu grep.
3251 (ii) Added --help to list all options with an explanatory phrase.
3252 (iii) Added -r, --recursive to recurse into sub-directories.
3253 (iv) Added -f, --file to read patterns from a file.
3255 6. pcre_exec() was referring to its "code" argument before testing that
3256 argument for NULL (and giving an error if it was NULL).
3258 7. Upgraded Makefile.in to allow for compiling in a different directory from
3259 the source directory.
3261 8. Tiny buglet in pcretest: when pcre_fullinfo() was called to retrieve the
3262 options bits, the pointer it was passed was to an int instead of to an unsigned
3263 long int. This mattered only on 64-bit systems.
3265 9. Fixed typo (3.4/1) in pcre.h again. Sigh. I had changed pcre.h (which is
3266 generated) instead of pcre.in, which it its source. Also made the same change
3267 in several of the .c files.
3269 10. A new release of gcc defines printf() as a macro, which broke pcretest
3270 because it had an ifdef in the middle of a string argument for printf(). Fixed
3271 by using separate calls to printf().
3273 11. Added --enable-newline-is-cr and --enable-newline-is-lf to the configure
3274 script, to force use of CR or LF instead of \n in the source. On non-Unix
3275 systems, the value can be set in config.h.
3277 12. The limit of 200 on non-capturing parentheses is a _nesting_ limit, not an
3278 absolute limit. Changed the text of the error message to make this clear, and
3279 likewise updated the man page.
3281 13. The limit of 99 on the number of capturing subpatterns has been removed.
3282 The new limit is 65535, which I hope will not be a "real" limit.
3285 Version 3.4 22-Aug-00
3286 ---------------------
3288 1. Fixed typo in pcre.h: unsigned const char * changed to const unsigned char *.
3290 2. Diagnose condition (?(0) as an error instead of crashing on matching.
3293 Version 3.3 01-Aug-00
3294 ---------------------
3296 1. If an octal character was given, but the value was greater than \377, it
3297 was not getting masked to the least significant bits, as documented. This could
3298 lead to crashes in some systems.
3300 2. Perl 5.6 (if not earlier versions) accepts classes like [a-\d] and treats
3301 the hyphen as a literal. PCRE used to give an error; it now behaves like Perl.
3303 3. Added the functions pcre_free_substring() and pcre_free_substring_list().
3304 These just pass their arguments on to (pcre_free)(), but they are provided
3305 because some uses of PCRE bind it to non-C systems that can call its functions,
3306 but cannot call free() or pcre_free() directly.
3308 4. Add "make test" as a synonym for "make check". Corrected some comments in
3309 the Makefile.
3311 5. Add $(DESTDIR)/ in front of all the paths in the "install" target in the
3312 Makefile.
3314 6. Changed the name of pgrep to pcregrep, because Solaris has introduced a
3315 command called pgrep for grepping around the active processes.
3317 7. Added the beginnings of support for UTF-8 character strings.
3319 8. Arranged for the Makefile to pass over the settings of CC, CFLAGS, and
3320 RANLIB to ./ltconfig so that they are used by libtool. I think these are all
3321 the relevant ones. (AR is not passed because ./ltconfig does its own figuring
3322 out for the ar command.)
3325 Version 3.2 12-May-00
3326 ---------------------
3328 This is purely a bug fixing release.
3330 1. If the pattern /((Z)+|A)*/ was matched agained ZABCDEFG it matched Z instead
3331 of ZA. This was just one example of several cases that could provoke this bug,
3332 which was introduced by change 9 of version 2.00. The code for breaking
3333 infinite loops after an iteration that matches an empty string was't working
3334 correctly.
3336 2. The pcretest program was not imitating Perl correctly for the pattern /a*/g
3337 when matched against abbab (for example). After matching an empty string, it
3338 wasn't forcing anchoring when setting PCRE_NOTEMPTY for the next attempt; this
3339 caused it to match further down the string than it should.
3341 3. The code contained an inclusion of sys/types.h. It isn't clear why this
3342 was there because it doesn't seem to be needed, and it causes trouble on some
3343 systems, as it is not a Standard C header. It has been removed.
3345 4. Made 4 silly changes to the source to avoid stupid compiler warnings that
3346 were reported on the Macintosh. The changes were from
3348 while ((c = *(++ptr)) != 0 && c != '\n');
3349 to
3350 while ((c = *(++ptr)) != 0 && c != '\n') ;
3352 Totally extraordinary, but if that's what it takes...
3354 5. PCRE is being used in one environment where neither memmove() nor bcopy() is
3355 available. Added HAVE_BCOPY and an autoconf test for it; if neither
3356 HAVE_MEMMOVE nor HAVE_BCOPY is set, use a built-in emulation function which
3357 assumes the way PCRE uses memmove() (always moving upwards).
3359 6. PCRE is being used in one environment where strchr() is not available. There
3360 was only one use in pcre.c, and writing it out to avoid strchr() probably gives
3361 faster code anyway.
3364 Version 3.1 09-Feb-00
3365 ---------------------
3367 The only change in this release is the fixing of some bugs in Makefile.in for
3368 the "install" target:
3370 (1) It was failing to install pcreposix.h.
3372 (2) It was overwriting the pcre.3 man page with the pcreposix.3 man page.
3375 Version 3.0 01-Feb-00
3376 ---------------------
3378 1. Add support for the /+ modifier to perltest (to output $` like it does in
3379 pcretest).
3381 2. Add support for the /g modifier to perltest.
3383 3. Fix pcretest so that it behaves even more like Perl for /g when the pattern
3384 matches null strings.
3386 4. Fix perltest so that it doesn't do unwanted things when fed an empty
3387 pattern. Perl treats empty patterns specially - it reuses the most recent
3388 pattern, which is not what we want. Replace // by /(?#)/ in order to avoid this
3389 effect.
3391 5. The POSIX interface was broken in that it was just handing over the POSIX
3392 captured string vector to pcre_exec(), but (since release 2.00) PCRE has
3393 required a bigger vector, with some working space on the end. This means that
3394 the POSIX wrapper now has to get and free some memory, and copy the results.
3396 6. Added some simple autoconf support, placing the test data and the
3397 documentation in separate directories, re-organizing some of the
3398 information files, and making it build pcre-config (a GNU standard). Also added
3399 libtool support for building PCRE as a shared library, which is now the
3400 default.
3402 7. Got rid of the leading zero in the definition of PCRE_MINOR because 08 and
3403 09 are not valid octal constants. Single digits will be used for minor values
3404 less than 10.
3406 8. Defined REG_EXTENDED and REG_NOSUB as zero in the POSIX header, so that
3407 existing programs that set these in the POSIX interface can use PCRE without
3408 modification.
3410 9. Added a new function, pcre_fullinfo() with an extensible interface. It can
3411 return all that pcre_info() returns, plus additional data. The pcre_info()
3412 function is retained for compatibility, but is considered to be obsolete.
3414 10. Added experimental recursion feature (?R) to handle one common case that
3415 Perl 5.6 will be able to do with (?p{...}).
3417 11. Added support for POSIX character classes like [:alpha:], which Perl is
3418 adopting.
3421 Version 2.08 31-Aug-99
3422 ----------------------
3424 1. When startoffset was not zero and the pattern began with ".*", PCRE was not
3425 trying to match at the startoffset position, but instead was moving forward to
3426 the next newline as if a previous match had failed.
3428 2. pcretest was not making use of PCRE_NOTEMPTY when repeating for /g and /G,
3429 and could get into a loop if a null string was matched other than at the start
3430 of the subject.
3432 3. Added definitions of PCRE_MAJOR and PCRE_MINOR to pcre.h so the version can
3433 be distinguished at compile time, and for completeness also added PCRE_DATE.
3435 5. Added Paul Sokolovsky's minor changes to make it easy to compile a Win32 DLL
3436 in GnuWin32 environments.
3439 Version 2.07 29-Jul-99
3440 ----------------------
3442 1. The documentation is now supplied in plain text form and HTML as well as in
3443 the form of man page sources.
3445 2. C++ compilers don't like assigning (void *) values to other pointer types.
3446 In particular this affects malloc(). Although there is no problem in Standard
3447 C, I've put in casts to keep C++ compilers happy.
3449 3. Typo on pcretest.c; a cast of (unsigned char *) in the POSIX regexec() call
3450 should be (const char *).
3452 4. If NOPOSIX is defined, pcretest.c compiles without POSIX support. This may
3453 be useful for non-Unix systems who don't want to bother with the POSIX stuff.
3454 However, I haven't made this a standard facility. The documentation doesn't
3455 mention it, and the Makefile doesn't support it.
3457 5. The Makefile now contains an "install" target, with editable destinations at
3458 the top of the file. The pcretest program is not installed.
3460 6. pgrep -V now gives the PCRE version number and date.
3462 7. Fixed bug: a zero repetition after a literal string (e.g. /abcde{0}/) was
3463 causing the entire string to be ignored, instead of just the last character.
3465 8. If a pattern like /"([^\\"]+|\\.)*"/ is applied in the normal way to a
3466 non-matching string, it can take a very, very long time, even for strings of
3467 quite modest length, because of the nested recursion. PCRE now does better in
3468 some of these cases. It does this by remembering the last required literal
3469 character in the pattern, and pre-searching the subject to ensure it is present
3470 before running the real match. In other words, it applies a heuristic to detect
3471 some types of certain failure quickly, and in the above example, if presented
3472 with a string that has no trailing " it gives "no match" very quickly.
3474 9. A new runtime option PCRE_NOTEMPTY causes null string matches to be ignored;
3475 other alternatives are tried instead.
3478 Version 2.06 09-Jun-99
3479 ----------------------
3481 1. Change pcretest's output for amount of store used to show just the code
3482 space, because the remainder (the data block) varies in size between 32-bit and
3483 64-bit systems.
3485 2. Added an extra argument to pcre_exec() to supply an offset in the subject to
3486 start matching at. This allows lookbehinds to work when searching for multiple
3487 occurrences in a string.
3489 3. Added additional options to pcretest for testing multiple occurrences:
3491 /+ outputs the rest of the string that follows a match
3492 /g loops for multiple occurrences, using the new startoffset argument
3493 /G loops for multiple occurrences by passing an incremented pointer
3495 4. PCRE wasn't doing the "first character" optimization for patterns starting
3496 with \b or \B, though it was doing it for other lookbehind assertions. That is,
3497 it wasn't noticing that a match for a pattern such as /\bxyz/ has to start with
3498 the letter 'x'. On long subject strings, this gives a significant speed-up.
3501 Version 2.05 21-Apr-99
3502 ----------------------
3504 1. Changed the type of magic_number from int to long int so that it works
3505 properly on 16-bit systems.
3507 2. Fixed a bug which caused patterns starting with .* not to work correctly
3508 when the subject string contained newline characters. PCRE was assuming
3509 anchoring for such patterns in all cases, which is not correct because .* will
3510 not pass a newline unless PCRE_DOTALL is set. It now assumes anchoring only if
3511 DOTALL is set at top level; otherwise it knows that patterns starting with .*
3512 must be retried after every newline in the subject.
3515 Version 2.04 18-Feb-99
3516 ----------------------
3518 1. For parenthesized subpatterns with repeats whose minimum was zero, the
3519 computation of the store needed to hold the pattern was incorrect (too large).
3520 If such patterns were nested a few deep, this could multiply and become a real
3521 problem.
3523 2. Added /M option to pcretest to show the memory requirement of a specific
3524 pattern. Made -m a synonym of -s (which does this globally) for compatibility.
3526 3. Subpatterns of the form (regex){n,m} (i.e. limited maximum) were being
3527 compiled in such a way that the backtracking after subsequent failure was
3528 pessimal. Something like (a){0,3} was compiled as (a)?(a)?(a)? instead of
3529 ((a)((a)(a)?)?)? with disastrous performance if the maximum was of any size.
3532 Version 2.03 02-Feb-99
3533 ----------------------
3535 1. Fixed typo and small mistake in man page.
3537 2. Added 4th condition (GPL supersedes if conflict) and created separate
3538 LICENCE file containing the conditions.
3540 3. Updated pcretest so that patterns such as /abc\/def/ work like they do in
3541 Perl, that is the internal \ allows the delimiter to be included in the
3542 pattern. Locked out the use of \ as a delimiter. If \ immediately follows
3543 the final delimiter, add \ to the end of the pattern (to test the error).
3545 4. Added the convenience functions for extracting substrings after a successful
3546 match. Updated pcretest to make it able to test these functions.
3549 Version 2.02 14-Jan-99
3550 ----------------------
3552 1. Initialized the working variables associated with each extraction so that
3553 their saving and restoring doesn't refer to uninitialized store.
3555 2. Put dummy code into study.c in order to trick the optimizer of the IBM C
3556 compiler for OS/2 into generating correct code. Apparently IBM isn't going to
3557 fix the problem.
3559 3. Pcretest: the timing code wasn't using LOOPREPEAT for timing execution
3560 calls, and wasn't printing the correct value for compiling calls. Increased the
3561 default value of LOOPREPEAT, and the number of significant figures in the
3562 times.
3564 4. Changed "/bin/rm" in the Makefile to "-rm" so it works on Windows NT.
3566 5. Renamed "deftables" as "dftables" to get it down to 8 characters, to avoid
3567 a building problem on Windows NT with a FAT file system.
3570 Version 2.01 21-Oct-98
3571 ----------------------
3573 1. Changed the API for pcre_compile() to allow for the provision of a pointer
3574 to character tables built by pcre_maketables() in the current locale. If NULL
3575 is passed, the default tables are used.
3578 Version 2.00 24-Sep-98
3579 ----------------------
3581 1. Since the (>?) facility is in Perl 5.005, don't require PCRE_EXTRA to enable
3582 it any more.
3584 2. Allow quantification of (?>) groups, and make it work correctly.
3586 3. The first character computation wasn't working for (?>) groups.
3588 4. Correct the implementation of \Z (it is permitted to match on the \n at the
3589 end of the subject) and add 5.005's \z, which really does match only at the
3590 very end of the subject.
3592 5. Remove the \X "cut" facility; Perl doesn't have it, and (?> is neater.
3594 6. Remove the ability to specify CASELESS, MULTILINE, DOTALL, and
3595 DOLLAR_END_ONLY at runtime, to make it possible to implement the Perl 5.005
3596 localized options. All options to pcre_study() were also removed.
3598 7. Add other new features from 5.005:
3600 $(?<= positive lookbehind
3601 $(?<! negative lookbehind
3602 (?imsx-imsx) added the unsetting capability
3603 such a setting is global if at outer level; local otherwise
3604 (?imsx-imsx:) non-capturing groups with option setting
3605 (?(cond)re|re) conditional pattern matching
3607 A backreference to itself in a repeated group matches the previous
3608 captured string.
3610 8. General tidying up of studying (both automatic and via "study")
3611 consequential on the addition of new assertions.
3613 9. As in 5.005, unlimited repeated groups that could match an empty substring
3614 are no longer faulted at compile time. Instead, the loop is forcibly broken at
3615 runtime if any iteration does actually match an empty substring.
3617 10. Include the RunTest script in the distribution.
3619 11. Added tests from the Perl 5.005_02 distribution. This showed up a few
3620 discrepancies, some of which were old and were also with respect to 5.004. They
3621 have now been fixed.
3624 Version 1.09 28-Apr-98
3625 ----------------------
3627 1. A negated single character class followed by a quantifier with a minimum
3628 value of one (e.g. [^x]{1,6} ) was not compiled correctly. This could lead to
3629 program crashes, or just wrong answers. This did not apply to negated classes
3630 containing more than one character, or to minima other than one.
3633 Version 1.08 27-Mar-98
3634 ----------------------
3636 1. Add PCRE_UNGREEDY to invert the greediness of quantifiers.
3638 2. Add (?U) and (?X) to set PCRE_UNGREEDY and PCRE_EXTRA respectively. The
3639 latter must appear before anything that relies on it in the pattern.
3642 Version 1.07 16-Feb-98
3643 ----------------------
3645 1. A pattern such as /((a)*)*/ was not being diagnosed as in error (unlimited
3646 repeat of a potentially empty string).
3649 Version 1.06 23-Jan-98
3650 ----------------------
3652 1. Added Markus Oberhumer's little patches for C++.
3654 2. Literal strings longer than 255 characters were broken.
3657 Version 1.05 23-Dec-97
3658 ----------------------
3660 1. Negated character classes containing more than one character were failing if
3661 PCRE_CASELESS was set at run time.
3664 Version 1.04 19-Dec-97
3665 ----------------------
3667 1. Corrected the man page, where some "const" qualifiers had been omitted.
3669 2. Made debugging output print "{0,xxx}" instead of just "{,xxx}" to agree with
3670 input syntax.
3672 3. Fixed memory leak which occurred when a regex with back references was
3673 matched with an offsets vector that wasn't big enough. The temporary memory
3674 that is used in this case wasn't being freed if the match failed.
3676 4. Tidied pcretest to ensure it frees memory that it gets.
3678 5. Temporary memory was being obtained in the case where the passed offsets
3679 vector was exactly big enough.
3681 6. Corrected definition of offsetof() from change 5 below.
3683 7. I had screwed up change 6 below and broken the rules for the use of
3684 setjmp(). Now fixed.
3687 Version 1.03 18-Dec-97
3688 ----------------------
3690 1. A erroneous regex with a missing opening parenthesis was correctly
3691 diagnosed, but PCRE attempted to access brastack[-1], which could cause crashes
3692 on some systems.
3694 2. Replaced offsetof(real_pcre, code) by offsetof(real_pcre, code[0]) because
3695 it was reported that one broken compiler failed on the former because "code" is
3696 also an independent variable.
3698 3. The erroneous regex a[]b caused an array overrun reference.
3700 4. A regex ending with a one-character negative class (e.g. /[^k]$/) did not
3701 fail on data ending with that character. (It was going on too far, and checking
3702 the next character, typically a binary zero.) This was specific to the
3703 optimized code for single-character negative classes.
3705 5. Added a contributed patch from the TIN world which does the following:
3707 + Add an undef for memmove, in case the the system defines a macro for it.
3709 + Add a definition of offsetof(), in case there isn't one. (I don't know
3710 the reason behind this - offsetof() is part of the ANSI standard - but
3711 it does no harm).
3713 + Reduce the ifdef's in pcre.c using macro DPRINTF, thereby eliminating
3714 most of the places where whitespace preceded '#'. I have given up and
3715 allowed the remaining 2 cases to be at the margin.
3717 + Rename some variables in pcre to eliminate shadowing. This seems very
3718 pedantic, but does no harm, of course.
3720 6. Moved the call to setjmp() into its own function, to get rid of warnings
3721 from gcc -Wall, and avoided calling it at all unless PCRE_EXTRA is used.
3723 7. Constructs such as \d{8,} were compiling into the equivalent of
3724 \d{8}\d{0,65527} instead of \d{8}\d* which didn't make much difference to the
3725 outcome, but in this particular case used more store than had been allocated,
3726 which caused the bug to be discovered because it threw up an internal error.
3728 8. The debugging code in both pcre and pcretest for outputting the compiled
3729 form of a regex was going wrong in the case of back references followed by
3730 curly-bracketed repeats.
3733 Version 1.02 12-Dec-97
3734 ----------------------
3736 1. Typos in pcre.3 and comments in the source fixed.
3738 2. Applied a contributed patch to get rid of places where it used to remove
3739 'const' from variables, and fixed some signed/unsigned and uninitialized
3740 variable warnings.
3742 3. Added the "runtest" target to Makefile.
3744 4. Set default compiler flag to -O2 rather than just -O.
3747 Version 1.01 19-Nov-97
3748 ----------------------
3750 1. PCRE was failing to diagnose unlimited repeat of empty string for patterns
3751 like /([ab]*)*/, that is, for classes with more than one character in them.
3753 2. Likewise, it wasn't diagnosing patterns with "once-only" subpatterns, such
3754 as /((?>a*))*/ (a PCRE_EXTRA facility).
3757 Version 1.00 18-Nov-97
3758 ----------------------
3760 1. Added compile-time macros to support systems such as SunOS4 which don't have
3761 memmove() or strerror() but have other things that can be used instead.
3763 2. Arranged that "make clean" removes the executables.
3766 Version 0.99 27-Oct-97
3767 ----------------------
3769 1. Fixed bug in code for optimizing classes with only one character. It was
3770 initializing a 32-byte map regardless, which could cause it to run off the end
3771 of the memory it had got.
3773 2. Added, conditional on PCRE_EXTRA, the proposed (?>REGEX) construction.
3776 Version 0.98 22-Oct-97
3777 ----------------------
3779 1. Fixed bug in code for handling temporary memory usage when there are more
3780 back references than supplied space in the ovector. This could cause segfaults.
3783 Version 0.97 21-Oct-97
3784 ----------------------
3786 1. Added the \X "cut" facility, conditional on PCRE_EXTRA.
3788 2. Optimized negated single characters not to use a bit map.
3790 3. Brought error texts together as macro definitions; clarified some of them;
3791 fixed one that was wrong - it said "range out of order" when it meant "invalid
3792 escape sequence".
3794 4. Changed some char * arguments to const char *.
3796 5. Added PCRE_NOTBOL and PCRE_NOTEOL (from POSIX).
3798 6. Added the POSIX-style API wrapper in pcreposix.a and testing facilities in
3799 pcretest.
3802 Version 0.96 16-Oct-97
3803 ----------------------
3805 1. Added a simple "pgrep" utility to the distribution.
3807 2. Fixed an incompatibility with Perl: "{" is now treated as a normal character
3808 unless it appears in one of the precise forms "{ddd}", "{ddd,}", or "{ddd,ddd}"
3809 where "ddd" means "one or more decimal digits".
3811 3. Fixed serious bug. If a pattern had a back reference, but the call to
3812 pcre_exec() didn't supply a large enough ovector to record the related
3813 identifying subpattern, the match always failed. PCRE now remembers the number
3814 of the largest back reference, and gets some temporary memory in which to save
3815 the offsets during matching if necessary, in order to ensure that
3816 backreferences always work.
3818 4. Increased the compatibility with Perl in a number of ways:
3820 (a) . no longer matches \n by default; an option PCRE_DOTALL is provided
3821 to request this handling. The option can be set at compile or exec time.
3823 (b) $ matches before a terminating newline by default; an option
3824 PCRE_DOLLAR_ENDONLY is provided to override this (but not in multiline
3825 mode). The option can be set at compile or exec time.
3827 (c) The handling of \ followed by a digit other than 0 is now supposed to be
3828 the same as Perl's. If the decimal number it represents is less than 10
3829 or there aren't that many previous left capturing parentheses, an octal
3830 escape is read. Inside a character class, it's always an octal escape,
3831 even if it is a single digit.
3833 (d) An escaped but undefined alphabetic character is taken as a literal,
3834 unless PCRE_EXTRA is set. Currently this just reserves the remaining
3835 escapes.
3837 (e) {0} is now permitted. (The previous item is removed from the compiled
3838 pattern).
3840 5. Changed all the names of code files so that the basic parts are no longer
3841 than 10 characters, and abolished the teeny "globals.c" file.
3843 6. Changed the handling of character classes; they are now done with a 32-byte
3844 bit map always.
3846 7. Added the -d and /D options to pcretest to make it possible to look at the
3847 internals of compilation without having to recompile pcre.
3850 Version 0.95 23-Sep-97
3851 ----------------------
3853 1. Fixed bug in pre-pass concerning escaped "normal" characters such as \x5c or
3854 \x20 at the start of a run of normal characters. These were being treated as
3855 real characters, instead of the source characters being re-checked.
3858 Version 0.94 18-Sep-97
3859 ----------------------
3861 1. The functions are now thread-safe, with the caveat that the global variables
3862 containing pointers to malloc() and free() or alternative functions are the
3863 same for all threads.
3865 2. Get pcre_study() to generate a bitmap of initial characters for non-
3866 anchored patterns when this is possible, and use it if passed to pcre_exec().
3869 Version 0.93 15-Sep-97
3870 ----------------------
3872 1. /(b)|(:+)/ was computing an incorrect first character.
3874 2. Add pcre_study() to the API and the passing of pcre_extra to pcre_exec(),
3875 but not actually doing anything yet.
3877 3. Treat "-" characters in classes that cannot be part of ranges as literals,
3878 as Perl does (e.g. [-az] or [az-]).
3880 4. Set the anchored flag if a branch starts with .* or .*? because that tests
3881 all possible positions.
3883 5. Split up into different modules to avoid including unneeded functions in a
3884 compiled binary. However, compile and exec are still in one module. The "study"
3885 function is split off.
3887 6. The character tables are now in a separate module whose source is generated
3888 by an auxiliary program - but can then be edited by hand if required. There are
3889 now no calls to isalnum(), isspace(), isdigit(), isxdigit(), tolower() or
3890 toupper() in the code.
3892 7. Turn the malloc/free funtions variables into pcre_malloc and pcre_free and
3893 make them global. Abolish the function for setting them, as the caller can now
3894 set them directly.
3897 Version 0.92 11-Sep-97
3898 ----------------------
3900 1. A repeat with a fixed maximum and a minimum of 1 for an ordinary character
3901 (e.g. /a{1,3}/) was broken (I mis-optimized it).
3903 2. Caseless matching was not working in character classes if the characters in
3904 the pattern were in upper case.
3906 3. Make ranges like [W-c] work in the same way as Perl for caseless matching.
3908 4. Make PCRE_ANCHORED public and accept as a compile option.
3910 5. Add an options word to pcre_exec() and accept PCRE_ANCHORED and
3911 PCRE_CASELESS at run time. Add escapes \A and \I to pcretest to cause it to
3912 pass them.
3914 6. Give an error if bad option bits passed at compile or run time.
3916 7. Add PCRE_MULTILINE at compile and exec time, and (?m) as well. Add \M to
3917 pcretest to cause it to pass that flag.
3919 8. Add pcre_info(), to get the number of identifying subpatterns, the stored
3920 options, and the first character, if set.
3922 9. Recognize C+ or C{n,m} where n >= 1 as providing a fixed starting character.
3925 Version 0.91 10-Sep-97
3926 ----------------------
3928 1. PCRE was failing to diagnose unlimited repeats of subpatterns that could
3929 match the empty string as in /(a*)*/. It was looping and ultimately crashing.
3931 2. PCRE was looping on encountering an indefinitely repeated back reference to
3932 a subpattern that had matched an empty string, e.g. /(a|)\1*/. It now does what
3933 Perl does - treats the match as successful.
3935 ****


Name Value
svn:eol-style native
svn:keywords "Author Date Id Revision Url"

  ViewVC Help
Powered by ViewVC 1.1.5