ViewVC logotype

Contents of /code/trunk/ChangeLog

Parent Directory Parent Directory | Revision Log Revision Log

Revision 514 - (show annotations)
Mon May 3 12:54:22 2010 UTC (11 years, 4 months ago) by ph10
File size: 185504 byte(s)
Add support for \N.
1 ChangeLog for PCRE
2 ------------------
4 Version 8.10 03 May-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.)
24 Version 8.02 19-Mar-2010
25 ------------------------
27 1. The Unicode data tables have been updated to Unicode 5.2.0.
29 2. Added the option --libs-cpp to pcre-config, but only when C++ support is
30 configured.
32 3. Updated the licensing terms in the pcregexp.pas file, as agreed with the
33 original author of that file, following a query about its status.
35 4. On systems that do not have stdint.h (e.g. Solaris), check for and include
36 inttypes.h instead. This fixes a bug that was introduced by change 8.01/8.
38 5. A pattern such as (?&t)*+(?(DEFINE)(?<t>.)) which has a possessive
39 quantifier applied to a forward-referencing subroutine call, could compile
40 incorrect code or give the error "internal error: previously-checked
41 referenced subpattern not found".
43 6. Both MS Visual Studio and Symbian OS have problems with initializing
44 variables to point to external functions. For these systems, therefore,
45 pcre_malloc etc. are now initialized to local functions that call the
46 relevant global functions.
48 7. There were two entries missing in the vectors called coptable and poptable
49 in pcre_dfa_exec.c. This could lead to memory accesses outsize the vectors.
50 I've fixed the data, and added a kludgy way of testing at compile time that
51 the lengths are correct (equal to the number of opcodes).
53 8. Following on from 7, I added a similar kludge to check the length of the
54 eint vector in pcreposix.c.
56 9. Error texts for pcre_compile() are held as one long string to avoid too
57 much relocation at load time. To find a text, the string is searched,
58 counting zeros. There was no check for running off the end of the string,
59 which could happen if a new error number was added without updating the
60 string.
62 10. \K gave a compile-time error if it appeared in a lookbehind assersion.
64 11. \K was not working if it appeared in an atomic group or in a group that
65 was called as a "subroutine", or in an assertion. Perl 5.11 documents that
66 \K is "not well defined" if used in an assertion. PCRE now accepts it if
67 the assertion is positive, but not if it is negative.
69 12. Change 11 fortuitously reduced the size of the stack frame used in the
70 "match()" function of pcre_exec.c by one pointer. Forthcoming
71 implementation of support for (*MARK) will need an extra pointer on the
72 stack; I have reserved it now, so that the stack frame size does not
73 decrease.
75 13. A pattern such as (?P<L1>(?P<L2>0)|(?P>L2)(?P>L1)) in which the only other
76 item in branch that calls a recursion is a subroutine call - as in the
77 second branch in the above example - was incorrectly given the compile-
78 time error "recursive call could loop indefinitely" because pcre_compile()
79 was not correctly checking the subroutine for matching a non-empty string.
81 14. The checks for overrunning compiling workspace could trigger after an
82 overrun had occurred. This is a "should never occur" error, but it can be
83 triggered by pathological patterns such as hundreds of nested parentheses.
84 The checks now trigger 100 bytes before the end of the workspace.
86 15. Fix typo in configure.ac: "srtoq" should be "strtoq".
89 Version 8.01 19-Jan-2010
90 ------------------------
92 1. If a pattern contained a conditional subpattern with only one branch (in
93 particular, this includes all (*DEFINE) patterns), a call to pcre_study()
94 computed the wrong minimum data length (which is of course zero for such
95 subpatterns). This could cause incorrect "no match" results.
97 2. For patterns such as (?i)a(?-i)b|c where an option setting at the start of
98 the pattern is reset in the first branch, pcre_compile() failed with
99 "internal error: code overflow at offset...". This happened only when
100 the reset was to the original external option setting. (An optimization
101 abstracts leading options settings into an external setting, which was the
102 cause of this.)
104 3. A pattern such as ^(?!a(*SKIP)b) where a negative assertion contained one
105 of the verbs SKIP, PRUNE, or COMMIT, did not work correctly. When the
106 assertion pattern did not match (meaning that the assertion was true), it
107 was incorrectly treated as false if the SKIP had been reached during the
108 matching. This also applied to assertions used as conditions.
110 4. If an item that is not supported by pcre_dfa_exec() was encountered in an
111 assertion subpattern, including such a pattern used as a condition,
112 unpredictable results occurred, instead of the error return
115 5. The C++ GlobalReplace function was not working like Perl for the special
116 situation when an empty string is matched. It now does the fancy magic
117 stuff that is necessary.
119 6. In pcre_internal.h, obsolete includes to setjmp.h and stdarg.h have been
120 removed. (These were left over from very, very early versions of PCRE.)
122 7. Some cosmetic changes to the code to make life easier when compiling it
123 as part of something else:
125 (a) Change DEBUG to PCRE_DEBUG.
127 (b) In pcre_compile(), rename the member of the "branch_chain" structure
128 called "current" as "current_branch", to prevent a collision with the
129 Linux macro when compiled as a kernel module.
131 (c) In pcre_study(), rename the function set_bit() as set_table_bit(), to
132 prevent a collision with the Linux macro when compiled as a kernel
133 module.
135 8. In pcre_compile() there are some checks for integer overflows that used to
136 cast potentially large values to (double). This has been changed to that
137 when building, a check for int64_t is made, and if it is found, it is used
138 instead, thus avoiding the use of floating point arithmetic. (There is no
139 other use of FP in PCRE.) If int64_t is not found, the fallback is to
140 double.
142 9. Added two casts to avoid signed/unsigned warnings from VS Studio Express
143 2005 (difference between two addresses compared to an unsigned value).
145 10. Change the standard AC_CHECK_LIB test for libbz2 in configure.ac to a
146 custom one, because of the following reported problem in Windows:
148 - libbz2 uses the Pascal calling convention (WINAPI) for the functions
149 under Win32.
150 - The standard autoconf AC_CHECK_LIB fails to include "bzlib.h",
151 therefore missing the function definition.
152 - The compiler thus generates a "C" signature for the test function.
153 - The linker fails to find the "C" function.
154 - PCRE fails to configure if asked to do so against libbz2.
156 11. When running libtoolize from libtool-2.2.6b as part of autogen.sh, these
157 messages were output:
159 Consider adding `AC_CONFIG_MACRO_DIR([m4])' to configure.ac and
160 rerunning libtoolize, to keep the correct libtool macros in-tree.
161 Consider adding `-I m4' to ACLOCAL_AMFLAGS in Makefile.am.
163 I have done both of these things.
165 12. Although pcre_dfa_exec() does not use nearly as much stack as pcre_exec()
166 most of the time, it *can* run out if it is given a pattern that contains a
167 runaway infinite recursion. I updated the discussion in the pcrestack man
168 page.
170 13. Now that we have gone to the x.xx style of version numbers, the minor
171 version may start with zero. Using 08 or 09 is a bad idea because users
172 might check the value of PCRE_MINOR in their code, and 08 or 09 may be
173 interpreted as invalid octal numbers. I've updated the previous comment in
174 configure.ac, and also added a check that gives an error if 08 or 09 are
175 used.
177 14. Change 8.00/11 was not quite complete: code had been accidentally omitted,
178 causing partial matching to fail when the end of the subject matched \W
179 in a UTF-8 pattern where \W was quantified with a minimum of 3.
181 15. There were some discrepancies between the declarations in pcre_internal.h
182 of _pcre_is_newline(), _pcre_was_newline(), and _pcre_valid_utf8() and
183 their definitions. The declarations used "const uschar *" and the
184 definitions used USPTR. Even though USPTR is normally defined as "const
185 unsigned char *" (and uschar is typedeffed as "unsigned char"), it was
186 reported that: "This difference in casting confuses some C++ compilers, for
187 example, SunCC recognizes above declarations as different functions and
188 generates broken code for hbpcre." I have changed the declarations to use
189 USPTR.
191 16. GNU libtool is named differently on some systems. The autogen.sh script now
192 tries several variants such as glibtoolize (MacOSX) and libtoolize1x
193 (FreeBSD).
195 17. Applied Craig's patch that fixes an HP aCC compile error in pcre 8.00
196 (strtoXX undefined when compiling pcrecpp.cc). The patch contains this
197 comment: "Figure out how to create a longlong from a string: strtoll and
198 equivalent. It's not enough to call AC_CHECK_FUNCS: hpux has a strtoll, for
199 instance, but it only takes 2 args instead of 3!"
201 18. A subtle bug concerned with back references has been fixed by a change of
202 specification, with a corresponding code fix. A pattern such as
203 ^(xa|=?\1a)+$ which contains a back reference inside the group to which it
204 refers, was giving matches when it shouldn't. For example, xa=xaaa would
205 match that pattern. Interestingly, Perl (at least up to 5.11.3) has the
206 same bug. Such groups have to be quantified to be useful, or contained
207 inside another quantified group. (If there's no repetition, the reference
208 can never match.) The problem arises because, having left the group and
209 moved on to the rest of the pattern, a later failure that backtracks into
210 the group uses the captured value from the final iteration of the group
211 rather than the correct earlier one. I have fixed this in PCRE by forcing
212 any group that contains a reference to itself to be an atomic group; that
213 is, there cannot be any backtracking into it once it has completed. This is
214 similar to recursive and subroutine calls.
217 Version 8.00 19-Oct-09
218 ----------------------
220 1. The table for translating pcre_compile() error codes into POSIX error codes
221 was out-of-date, and there was no check on the pcre_compile() error code
222 being within the table. This could lead to an OK return being given in
223 error.
225 2. Changed the call to open a subject file in pcregrep from fopen(pathname,
226 "r") to fopen(pathname, "rb"), which fixed a problem with some of the tests
227 in a Windows environment.
229 3. The pcregrep --count option prints the count for each file even when it is
230 zero, as does GNU grep. However, pcregrep was also printing all files when
231 --files-with-matches was added. Now, when both options are given, it prints
232 counts only for those files that have at least one match. (GNU grep just
233 prints the file name in this circumstance, but including the count seems
234 more useful - otherwise, why use --count?) Also ensured that the
235 combination -clh just lists non-zero counts, with no names.
237 4. The long form of the pcregrep -F option was incorrectly implemented as
238 --fixed_strings instead of --fixed-strings. This is an incompatible change,
239 but it seems right to fix it, and I didn't think it was worth preserving
240 the old behaviour.
242 5. The command line items --regex=pattern and --regexp=pattern were not
243 recognized by pcregrep, which required --regex pattern or --regexp pattern
244 (with a space rather than an '='). The man page documented the '=' forms,
245 which are compatible with GNU grep; these now work.
247 6. No libpcreposix.pc file was created for pkg-config; there was just
248 libpcre.pc and libpcrecpp.pc. The omission has been rectified.
250 7. Added #ifndef SUPPORT_UCP into the pcre_ucd.c module, to reduce its size
251 when UCP support is not needed, by modifying the Python script that
252 generates it from Unicode data files. This should not matter if the module
253 is correctly used as a library, but I received one complaint about 50K of
254 unwanted data. My guess is that the person linked everything into his
255 program rather than using a library. Anyway, it does no harm.
257 8. A pattern such as /\x{123}{2,2}+/8 was incorrectly compiled; the trigger
258 was a minimum greater than 1 for a wide character in a possessive
259 repetition. The same bug could also affect patterns like /(\x{ff}{0,2})*/8
260 which had an unlimited repeat of a nested, fixed maximum repeat of a wide
261 character. Chaos in the form of incorrect output or a compiling loop could
262 result.
264 9. The restrictions on what a pattern can contain when partial matching is
265 requested for pcre_exec() have been removed. All patterns can now be
266 partially matched by this function. In addition, if there are at least two
267 slots in the offset vector, the offset of the earliest inspected character
268 for the match and the offset of the end of the subject are set in them when
269 PCRE_ERROR_PARTIAL is returned.
271 10. Partial matching has been split into two forms: PCRE_PARTIAL_SOFT, which is
272 synonymous with PCRE_PARTIAL, for backwards compatibility, and
273 PCRE_PARTIAL_HARD, which causes a partial match to supersede a full match,
274 and may be more useful for multi-segment matching.
276 11. Partial matching with pcre_exec() is now more intuitive. A partial match
277 used to be given if ever the end of the subject was reached; now it is
278 given only if matching could not proceed because another character was
279 needed. This makes a difference in some odd cases such as Z(*FAIL) with the
280 string "Z", which now yields "no match" instead of "partial match". In the
281 case of pcre_dfa_exec(), "no match" is given if every matching path for the
282 final character ended with (*FAIL).
284 12. Restarting a match using pcre_dfa_exec() after a partial match did not work
285 if the pattern had a "must contain" character that was already found in the
286 earlier partial match, unless partial matching was again requested. For
287 example, with the pattern /dog.(body)?/, the "must contain" character is
288 "g". If the first part-match was for the string "dog", restarting with
289 "sbody" failed. This bug has been fixed.
291 13. The string returned by pcre_dfa_exec() after a partial match has been
292 changed so that it starts at the first inspected character rather than the
293 first character of the match. This makes a difference only if the pattern
294 starts with a lookbehind assertion or \b or \B (\K is not supported by
295 pcre_dfa_exec()). It's an incompatible change, but it makes the two
296 matching functions compatible, and I think it's the right thing to do.
298 14. Added a pcredemo man page, created automatically from the pcredemo.c file,
299 so that the demonstration program is easily available in environments where
300 PCRE has not been installed from source.
302 15. Arranged to add -DPCRE_STATIC to cflags in libpcre.pc, libpcreposix.cp,
303 libpcrecpp.pc and pcre-config when PCRE is not compiled as a shared
304 library.
306 16. Added REG_UNGREEDY to the pcreposix interface, at the request of a user.
307 It maps to PCRE_UNGREEDY. It is not, of course, POSIX-compatible, but it
308 is not the first non-POSIX option to be added. Clearly some people find
309 these options useful.
311 17. If a caller to the POSIX matching function regexec() passes a non-zero
312 value for nmatch with a NULL value for pmatch, the value of
313 nmatch is forced to zero.
315 18. RunGrepTest did not have a test for the availability of the -u option of
316 the diff command, as RunTest does. It now checks in the same way as
317 RunTest, and also checks for the -b option.
319 19. If an odd number of negated classes containing just a single character
320 interposed, within parentheses, between a forward reference to a named
321 subpattern and the definition of the subpattern, compilation crashed with
322 an internal error, complaining that it could not find the referenced
323 subpattern. An example of a crashing pattern is /(?&A)(([^m])(?<A>))/.
324 [The bug was that it was starting one character too far in when skipping
325 over the character class, thus treating the ] as data rather than
326 terminating the class. This meant it could skip too much.]
328 20. Added PCRE_NOTEMPTY_ATSTART in order to be able to correctly implement the
329 /g option in pcretest when the pattern contains \K, which makes it possible
330 to have an empty string match not at the start, even when the pattern is
331 anchored. Updated pcretest and pcredemo to use this option.
333 21. If the maximum number of capturing subpatterns in a recursion was greater
334 than the maximum at the outer level, the higher number was returned, but
335 with unset values at the outer level. The correct (outer level) value is
336 now given.
338 22. If (*ACCEPT) appeared inside capturing parentheses, previous releases of
339 PCRE did not set those parentheses (unlike Perl). I have now found a way to
340 make it do so. The string so far is captured, making this feature
341 compatible with Perl.
343 23. The tests have been re-organized, adding tests 11 and 12, to make it
344 possible to check the Perl 5.10 features against Perl 5.10.
346 24. Perl 5.10 allows subroutine calls in lookbehinds, as long as the subroutine
347 pattern matches a fixed length string. PCRE did not allow this; now it
348 does. Neither allows recursion.
350 25. I finally figured out how to implement a request to provide the minimum
351 length of subject string that was needed in order to match a given pattern.
352 (It was back references and recursion that I had previously got hung up
353 on.) This code has now been added to pcre_study(); it finds a lower bound
354 to the length of subject needed. It is not necessarily the greatest lower
355 bound, but using it to avoid searching strings that are too short does give
356 some useful speed-ups. The value is available to calling programs via
357 pcre_fullinfo().
359 26. While implementing 25, I discovered to my embarrassment that pcretest had
360 not been passing the result of pcre_study() to pcre_dfa_exec(), so the
361 study optimizations had never been tested with that matching function.
362 Oops. What is worse, even when it was passed study data, there was a bug in
363 pcre_dfa_exec() that meant it never actually used it. Double oops. There
364 were also very few tests of studied patterns with pcre_dfa_exec().
366 27. If (?| is used to create subpatterns with duplicate numbers, they are now
367 allowed to have the same name, even if PCRE_DUPNAMES is not set. However,
368 on the other side of the coin, they are no longer allowed to have different
369 names, because these cannot be distinguished in PCRE, and this has caused
370 confusion. (This is a difference from Perl.)
372 28. When duplicate subpattern names are present (necessarily with different
373 numbers, as required by 27 above), and a test is made by name in a
374 conditional pattern, either for a subpattern having been matched, or for
375 recursion in such a pattern, all the associated numbered subpatterns are
376 tested, and the overall condition is true if the condition is true for any
377 one of them. This is the way Perl works, and is also more like the way
378 testing by number works.
381 Version 7.9 11-Apr-09
382 ---------------------
384 1. When building with support for bzlib/zlib (pcregrep) and/or readline
385 (pcretest), all targets were linked against these libraries. This included
386 libpcre, libpcreposix, and libpcrecpp, even though they do not use these
387 libraries. This caused unwanted dependencies to be created. This problem
388 has been fixed, and now only pcregrep is linked with bzlib/zlib and only
389 pcretest is linked with readline.
391 2. The "typedef int BOOL" in pcre_internal.h that was included inside the
392 "#ifndef FALSE" condition by an earlier change (probably 7.8/18) has been
393 moved outside it again, because FALSE and TRUE are already defined in AIX,
394 but BOOL is not.
396 3. The pcre_config() function was treating the PCRE_MATCH_LIMIT and
397 PCRE_MATCH_LIMIT_RECURSION values as ints, when they should be long ints.
399 4. The pcregrep documentation said spaces were inserted as well as colons (or
400 hyphens) following file names and line numbers when outputting matching
401 lines. This is not true; no spaces are inserted. I have also clarified the
402 wording for the --colour (or --color) option.
404 5. In pcregrep, when --colour was used with -o, the list of matching strings
405 was not coloured; this is different to GNU grep, so I have changed it to be
406 the same.
408 6. When --colo(u)r was used in pcregrep, only the first matching substring in
409 each matching line was coloured. Now it goes on to look for further matches
410 of any of the test patterns, which is the same behaviour as GNU grep.
412 7. A pattern that could match an empty string could cause pcregrep to loop; it
413 doesn't make sense to accept an empty string match in pcregrep, so I have
414 locked it out (using PCRE's PCRE_NOTEMPTY option). By experiment, this
415 seems to be how GNU grep behaves.
417 8. The pattern (?(?=.*b)b|^) was incorrectly compiled as "match must be at
418 start or after a newline", because the conditional assertion was not being
419 correctly handled. The rule now is that both the assertion and what follows
420 in the first alternative must satisfy the test.
422 9. If auto-callout was enabled in a pattern with a conditional group whose
423 condition was an assertion, PCRE could crash during matching, both with
424 pcre_exec() and pcre_dfa_exec().
426 10. The PCRE_DOLLAR_ENDONLY option was not working when pcre_dfa_exec() was
427 used for matching.
429 11. Unicode property support in character classes was not working for
430 characters (bytes) greater than 127 when not in UTF-8 mode.
432 12. Added the -M command line option to pcretest.
434 14. Added the non-standard REG_NOTEMPTY option to the POSIX interface.
436 15. Added the PCRE_NO_START_OPTIMIZE match-time option.
438 16. Added comments and documentation about mis-use of no_arg in the C++
439 wrapper.
441 17. Implemented support for UTF-8 encoding in EBCDIC environments, a patch
442 from Martin Jerabek that uses macro names for all relevant character and
443 string constants.
445 18. Added to pcre_internal.h two configuration checks: (a) If both EBCDIC and
446 SUPPORT_UTF8 are set, give an error; (b) If SUPPORT_UCP is set without
447 SUPPORT_UTF8, define SUPPORT_UTF8. The "configure" script handles both of
448 these, but not everybody uses configure.
450 19. A conditional group that had only one branch was not being correctly
451 recognized as an item that could match an empty string. This meant that an
452 enclosing group might also not be so recognized, causing infinite looping
453 (and probably a segfault) for patterns such as ^"((?(?=[a])[^"])|b)*"$
454 with the subject "ab", where knowledge that the repeated group can match
455 nothing is needed in order to break the loop.
457 20. If a pattern that was compiled with callouts was matched using pcre_dfa_
458 exec(), but without supplying a callout function, matching went wrong.
460 21. If PCRE_ERROR_MATCHLIMIT occurred during a recursion, there was a memory
461 leak if the size of the offset vector was greater than 30. When the vector
462 is smaller, the saved offsets during recursion go onto a local stack
463 vector, but for larger vectors malloc() is used. It was failing to free
464 when the recursion yielded PCRE_ERROR_MATCH_LIMIT (or any other "abnormal"
465 error, in fact).
467 22. There was a missing #ifdef SUPPORT_UTF8 round one of the variables in the
468 heapframe that is used only when UTF-8 support is enabled. This caused no
469 problem, but was untidy.
471 23. Steven Van Ingelgem's patch to CMakeLists.txt to change the name
472 CMAKE_BINARY_DIR to PROJECT_BINARY_DIR so that it works when PCRE is
473 included within another project.
475 24. Steven Van Ingelgem's patches to add more options to the CMake support,
476 slightly modified by me:
478 (a) PCRE_BUILD_TESTS can be set OFF not to build the tests, including
479 not building pcregrep.
481 (b) PCRE_BUILD_PCREGREP can be see OFF not to build pcregrep, but only
482 if PCRE_BUILD_TESTS is also set OFF, because the tests use pcregrep.
484 25. Forward references, both numeric and by name, in patterns that made use of
485 duplicate group numbers, could behave incorrectly or give incorrect errors,
486 because when scanning forward to find the reference group, PCRE was not
487 taking into account the duplicate group numbers. A pattern such as
488 ^X(?3)(a)(?|(b)|(q))(Y) is an example.
490 26. Changed a few more instances of "const unsigned char *" to USPTR, making
491 the feature of a custom pointer more persuasive (as requested by a user).
493 27. Wrapped the definitions of fileno and isatty for Windows, which appear in
494 pcretest.c, inside #ifndefs, because it seems they are sometimes already
495 pre-defined.
497 28. Added support for (*UTF8) at the start of a pattern.
499 29. Arrange for flags added by the "release type" setting in CMake to be shown
500 in the configuration summary.
503 Version 7.8 05-Sep-08
504 ---------------------
506 1. Replaced UCP searching code with optimized version as implemented for Ad
507 Muncher (http://www.admuncher.com/) by Peter Kankowski. This uses a two-
508 stage table and inline lookup instead of a function, giving speed ups of 2
509 to 5 times on some simple patterns that I tested. Permission was given to
510 distribute the MultiStage2.py script that generates the tables (it's not in
511 the tarball, but is in the Subversion repository).
513 2. Updated the Unicode datatables to Unicode 5.1.0. This adds yet more
514 scripts.
516 3. Change 12 for 7.7 introduced a bug in pcre_study() when a pattern contained
517 a group with a zero qualifier. The result of the study could be incorrect,
518 or the function might crash, depending on the pattern.
520 4. Caseless matching was not working for non-ASCII characters in back
521 references. For example, /(\x{de})\1/8i was not matching \x{de}\x{fe}.
522 It now works when Unicode Property Support is available.
524 5. In pcretest, an escape such as \x{de} in the data was always generating
525 a UTF-8 string, even in non-UTF-8 mode. Now it generates a single byte in
526 non-UTF-8 mode. If the value is greater than 255, it gives a warning about
527 truncation.
529 6. Minor bugfix in pcrecpp.cc (change "" == ... to NULL == ...).
531 7. Added two (int) casts to pcregrep when printing the difference of two
532 pointers, in case they are 64-bit values.
534 8. Added comments about Mac OS X stack usage to the pcrestack man page and to
535 test 2 if it fails.
537 9. Added PCRE_CALL_CONVENTION just before the names of all exported functions,
538 and a #define of that name to empty if it is not externally set. This is to
539 allow users of MSVC to set it if necessary.
541 10. The PCRE_EXP_DEFN macro which precedes exported functions was missing from
542 the convenience functions in the pcre_get.c source file.
544 11. An option change at the start of a pattern that had top-level alternatives
545 could cause overwriting and/or a crash. This command provoked a crash in
546 some environments:
548 printf "/(?i)[\xc3\xa9\xc3\xbd]|[\xc3\xa9\xc3\xbdA]/8\n" | pcretest
550 This potential security problem was recorded as CVE-2008-2371.
552 12. For a pattern where the match had to start at the beginning or immediately
553 after a newline (e.g /.*anything/ without the DOTALL flag), pcre_exec() and
554 pcre_dfa_exec() could read past the end of the passed subject if there was
555 no match. To help with detecting such bugs (e.g. with valgrind), I modified
556 pcretest so that it places the subject at the end of its malloc-ed buffer.
558 13. The change to pcretest in 12 above threw up a couple more cases when pcre_
559 exec() might read past the end of the data buffer in UTF-8 mode.
561 14. A similar bug to 7.3/2 existed when the PCRE_FIRSTLINE option was set and
562 the data contained the byte 0x85 as part of a UTF-8 character within its
563 first line. This applied both to normal and DFA matching.
565 15. Lazy qualifiers were not working in some cases in UTF-8 mode. For example,
566 /^[^d]*?$/8 failed to match "abc".
568 16. Added a missing copyright notice to pcrecpp_internal.h.
570 17. Make it more clear in the documentation that values returned from
571 pcre_exec() in ovector are byte offsets, not character counts.
573 18. Tidied a few places to stop certain compilers from issuing warnings.
575 19. Updated the Virtual Pascal + BCC files to compile the latest v7.7, as
576 supplied by Stefan Weber. I made a further small update for 7.8 because
577 there is a change of source arrangements: the pcre_searchfuncs.c module is
578 replaced by pcre_ucd.c.
581 Version 7.7 07-May-08
582 ---------------------
584 1. Applied Craig's patch to sort out a long long problem: "If we can't convert
585 a string to a long long, pretend we don't even have a long long." This is
586 done by checking for the strtoq, strtoll, and _strtoi64 functions.
588 2. Applied Craig's patch to pcrecpp.cc to restore ABI compatibility with
589 pre-7.6 versions, which defined a global no_arg variable instead of putting
590 it in the RE class. (See also #8 below.)
592 3. Remove a line of dead code, identified by coverity and reported by Nuno
593 Lopes.
595 4. Fixed two related pcregrep bugs involving -r with --include or --exclude:
597 (1) The include/exclude patterns were being applied to the whole pathnames
598 of files, instead of just to the final components.
600 (2) If there was more than one level of directory, the subdirectories were
601 skipped unless they satisfied the include/exclude conditions. This is
602 inconsistent with GNU grep (and could even be seen as contrary to the
603 pcregrep specification - which I improved to make it absolutely clear).
604 The action now is always to scan all levels of directory, and just
605 apply the include/exclude patterns to regular files.
607 5. Added the --include_dir and --exclude_dir patterns to pcregrep, and used
608 --exclude_dir in the tests to avoid scanning .svn directories.
610 6. Applied Craig's patch to the QuoteMeta function so that it escapes the
611 NUL character as backslash + 0 rather than backslash + NUL, because PCRE
612 doesn't support NULs in patterns.
614 7. Added some missing "const"s to declarations of static tables in
615 pcre_compile.c and pcre_dfa_exec.c.
617 8. Applied Craig's patch to pcrecpp.cc to fix a problem in OS X that was
618 caused by fix #2 above. (Subsequently also a second patch to fix the
619 first patch. And a third patch - this was a messy problem.)
621 9. Applied Craig's patch to remove the use of push_back().
623 10. Applied Alan Lehotsky's patch to add REG_STARTEND support to the POSIX
624 matching function regexec().
626 11. Added support for the Oniguruma syntax \g<name>, \g<n>, \g'name', \g'n',
627 which, however, unlike Perl's \g{...}, are subroutine calls, not back
628 references. PCRE supports relative numbers with this syntax (I don't think
629 Oniguruma does).
631 12. Previously, a group with a zero repeat such as (...){0} was completely
632 omitted from the compiled regex. However, this means that if the group
633 was called as a subroutine from elsewhere in the pattern, things went wrong
634 (an internal error was given). Such groups are now left in the compiled
635 pattern, with a new opcode that causes them to be skipped at execution
636 time.
638 13. Added the PCRE_JAVASCRIPT_COMPAT option. This makes the following changes
639 to the way PCRE behaves:
641 (a) A lone ] character is dis-allowed (Perl treats it as data).
643 (b) A back reference to an unmatched subpattern matches an empty string
644 (Perl fails the current match path).
646 (c) A data ] in a character class must be notated as \] because if the
647 first data character in a class is ], it defines an empty class. (In
648 Perl it is not possible to have an empty class.) The empty class []
649 never matches; it forces failure and is equivalent to (*FAIL) or (?!).
650 The negative empty class [^] matches any one character, independently
651 of the DOTALL setting.
653 14. A pattern such as /(?2)[]a()b](abc)/ which had a forward reference to a
654 non-existent subpattern following a character class starting with ']' and
655 containing () gave an internal compiling error instead of "reference to
656 non-existent subpattern". Fortunately, when the pattern did exist, the
657 compiled code was correct. (When scanning forwards to check for the
658 existencd of the subpattern, it was treating the data ']' as terminating
659 the class, so got the count wrong. When actually compiling, the reference
660 was subsequently set up correctly.)
662 15. The "always fail" assertion (?!) is optimzed to (*FAIL) by pcre_compile;
663 it was being rejected as not supported by pcre_dfa_exec(), even though
664 other assertions are supported. I have made pcre_dfa_exec() support
665 (*FAIL).
667 16. The implementation of 13c above involved the invention of a new opcode,
668 OP_ALLANY, which is like OP_ANY but doesn't check the /s flag. Since /s
669 cannot be changed at match time, I realized I could make a small
670 improvement to matching performance by compiling OP_ALLANY instead of
671 OP_ANY for "." when DOTALL was set, and then removing the runtime tests
672 on the OP_ANY path.
674 17. Compiling pcretest on Windows with readline support failed without the
675 following two fixes: (1) Make the unistd.h include conditional on
676 HAVE_UNISTD_H; (2) #define isatty and fileno as _isatty and _fileno.
678 18. Changed CMakeLists.txt and cmake/FindReadline.cmake to arrange for the
679 ncurses library to be included for pcretest when ReadLine support is
680 requested, but also to allow for it to be overridden. This patch came from
681 Daniel Bergström.
683 19. There was a typo in the file ucpinternal.h where f0_rangeflag was defined
684 as 0x00f00000 instead of 0x00800000. Luckily, this would not have caused
685 any errors with the current Unicode tables. Thanks to Peter Kankowski for
686 spotting this.
689 Version 7.6 28-Jan-08
690 ---------------------
692 1. A character class containing a very large number of characters with
693 codepoints greater than 255 (in UTF-8 mode, of course) caused a buffer
694 overflow.
696 2. Patch to cut out the "long long" test in pcrecpp_unittest when
697 HAVE_LONG_LONG is not defined.
699 3. Applied Christian Ehrlicher's patch to update the CMake build files to
700 bring them up to date and include new features. This patch includes:
702 - Fixed PH's badly added libz and libbz2 support.
703 - Fixed a problem with static linking.
704 - Added pcredemo. [But later removed - see 7 below.]
705 - Fixed dftables problem and added an option.
706 - Added a number of HAVE_XXX tests, including HAVE_WINDOWS_H and
708 - Added readline support for pcretest.
709 - Added an listing of the option settings after cmake has run.
711 4. A user submitted a patch to Makefile that makes it easy to create
712 "pcre.dll" under mingw when using Configure/Make. I added stuff to
713 Makefile.am that cause it to include this special target, without
714 affecting anything else. Note that the same mingw target plus all
715 the other distribution libraries and programs are now supported
716 when configuring with CMake (see 6 below) instead of with
717 Configure/Make.
719 5. Applied Craig's patch that moves no_arg into the RE class in the C++ code.
720 This is an attempt to solve the reported problem "pcrecpp::no_arg is not
721 exported in the Windows port". It has not yet been confirmed that the patch
722 solves the problem, but it does no harm.
724 6. Applied Sheri's patch to CMakeLists.txt to add NON_STANDARD_LIB_PREFIX and
725 NON_STANDARD_LIB_SUFFIX for dll names built with mingw when configured
726 with CMake, and also correct the comment about stack recursion.
728 7. Remove the automatic building of pcredemo from the ./configure system and
729 from CMakeLists.txt. The whole idea of pcredemo.c is that it is an example
730 of a program that users should build themselves after PCRE is installed, so
731 building it automatically is not really right. What is more, it gave
732 trouble in some build environments.
734 8. Further tidies to CMakeLists.txt from Sheri and Christian.
737 Version 7.5 10-Jan-08
738 ---------------------
740 1. Applied a patch from Craig: "This patch makes it possible to 'ignore'
741 values in parens when parsing an RE using the C++ wrapper."
743 2. Negative specials like \S did not work in character classes in UTF-8 mode.
744 Characters greater than 255 were excluded from the class instead of being
745 included.
747 3. The same bug as (2) above applied to negated POSIX classes such as
748 [:^space:].
750 4. PCRECPP_STATIC was referenced in pcrecpp_internal.h, but nowhere was it
751 defined or documented. It seems to have been a typo for PCRE_STATIC, so
752 I have changed it.
754 5. The construct (?&) was not diagnosed as a syntax error (it referenced the
755 first named subpattern) and a construct such as (?&a) would reference the
756 first named subpattern whose name started with "a" (in other words, the
757 length check was missing). Both these problems are fixed. "Subpattern name
758 expected" is now given for (?&) (a zero-length name), and this patch also
759 makes it give the same error for \k'' (previously it complained that that
760 was a reference to a non-existent subpattern).
762 6. The erroneous patterns (?+-a) and (?-+a) give different error messages;
763 this is right because (?- can be followed by option settings as well as by
764 digits. I have, however, made the messages clearer.
766 7. Patterns such as (?(1)a|b) (a pattern that contains fewer subpatterns
767 than the number used in the conditional) now cause a compile-time error.
768 This is actually not compatible with Perl, which accepts such patterns, but
769 treats the conditional as always being FALSE (as PCRE used to), but it
770 seems to me that giving a diagnostic is better.
772 8. Change "alphameric" to the more common word "alphanumeric" in comments
773 and messages.
775 9. Fix two occurrences of "backslash" in comments that should have been
776 "backspace".
778 10. Remove two redundant lines of code that can never be obeyed (their function
779 was moved elsewhere).
781 11. The program that makes PCRE's Unicode character property table had a bug
782 which caused it to generate incorrect table entries for sequences of
783 characters that have the same character type, but are in different scripts.
784 It amalgamated them into a single range, with the script of the first of
785 them. In other words, some characters were in the wrong script. There were
786 thirteen such cases, affecting characters in the following ranges:
788 U+002b0 - U+002c1
789 U+0060c - U+0060d
790 U+0061e - U+00612
791 U+0064b - U+0065e
792 U+0074d - U+0076d
793 U+01800 - U+01805
794 U+01d00 - U+01d77
795 U+01d9b - U+01dbf
796 U+0200b - U+0200f
797 U+030fc - U+030fe
798 U+03260 - U+0327f
799 U+0fb46 - U+0fbb1
800 U+10450 - U+1049d
802 12. The -o option (show only the matching part of a line) for pcregrep was not
803 compatible with GNU grep in that, if there was more than one match in a
804 line, it showed only the first of them. It now behaves in the same way as
805 GNU grep.
807 13. If the -o and -v options were combined for pcregrep, it printed a blank
808 line for every non-matching line. GNU grep prints nothing, and pcregrep now
809 does the same. The return code can be used to tell if there were any
810 non-matching lines.
812 14. Added --file-offsets and --line-offsets to pcregrep.
814 15. The pattern (?=something)(?R) was not being diagnosed as a potentially
815 infinitely looping recursion. The bug was that positive lookaheads were not
816 being skipped when checking for a possible empty match (negative lookaheads
817 and both kinds of lookbehind were skipped).
819 16. Fixed two typos in the Windows-only code in pcregrep.c, and moved the
820 inclusion of <windows.h> to before rather than after the definition of
821 INVALID_FILE_ATTRIBUTES (patch from David Byron).
823 17. Specifying a possessive quantifier with a specific limit for a Unicode
824 character property caused pcre_compile() to compile bad code, which led at
825 runtime to PCRE_ERROR_INTERNAL (-14). Examples of patterns that caused this
826 are: /\p{Zl}{2,3}+/8 and /\p{Cc}{2}+/8. It was the possessive "+" that
827 caused the error; without that there was no problem.
829 18. Added --enable-pcregrep-libz and --enable-pcregrep-libbz2.
831 19. Added --enable-pcretest-libreadline.
833 20. In pcrecpp.cc, the variable 'count' was incremented twice in
834 RE::GlobalReplace(). As a result, the number of replacements returned was
835 double what it should be. I removed one of the increments, but Craig sent a
836 later patch that removed the other one (the right fix) and added unit tests
837 that check the return values (which was not done before).
839 21. Several CMake things:
841 (1) Arranged that, when cmake is used on Unix, the libraries end up with
842 the names libpcre and libpcreposix, not just pcre and pcreposix.
844 (2) The above change means that pcretest and pcregrep are now correctly
845 linked with the newly-built libraries, not previously installed ones.
849 22. In UTF-8 mode, with newline set to "any", a pattern such as .*a.*=.b.*
850 crashed when matching a string such as a\x{2029}b (note that \x{2029} is a
851 UTF-8 newline character). The key issue is that the pattern starts .*;
852 this means that the match must be either at the beginning, or after a
853 newline. The bug was in the code for advancing after a failed match and
854 checking that the new position followed a newline. It was not taking
855 account of UTF-8 characters correctly.
857 23. PCRE was behaving differently from Perl in the way it recognized POSIX
858 character classes. PCRE was not treating the sequence [:...:] as a
859 character class unless the ... were all letters. Perl, however, seems to
860 allow any characters between [: and :], though of course it rejects as
861 unknown any "names" that contain non-letters, because all the known class
862 names consist only of letters. Thus, Perl gives an error for [[:1234:]],
863 for example, whereas PCRE did not - it did not recognize a POSIX character
864 class. This seemed a bit dangerous, so the code has been changed to be
865 closer to Perl. The behaviour is not identical to Perl, because PCRE will
866 diagnose an unknown class for, for example, [[:l\ower:]] where Perl will
867 treat it as [[:lower:]]. However, PCRE does now give "unknown" errors where
868 Perl does, and where it didn't before.
870 24. Rewrite so as to remove the single use of %n from pcregrep because in some
871 Windows environments %n is disabled by default.
874 Version 7.4 21-Sep-07
875 ---------------------
877 1. Change 7.3/28 was implemented for classes by looking at the bitmap. This
878 means that a class such as [\s] counted as "explicit reference to CR or
879 LF". That isn't really right - the whole point of the change was to try to
880 help when there was an actual mention of one of the two characters. So now
881 the change happens only if \r or \n (or a literal CR or LF) character is
882 encountered.
884 2. The 32-bit options word was also used for 6 internal flags, but the numbers
885 of both had grown to the point where there were only 3 bits left.
886 Fortunately, there was spare space in the data structure, and so I have
887 moved the internal flags into a new 16-bit field to free up more option
888 bits.
890 3. The appearance of (?J) at the start of a pattern set the DUPNAMES option,
891 but did not set the internal JCHANGED flag - either of these is enough to
892 control the way the "get" function works - but the PCRE_INFO_JCHANGED
893 facility is supposed to tell if (?J) was ever used, so now (?J) at the
894 start sets both bits.
896 4. Added options (at build time, compile time, exec time) to change \R from
897 matching any Unicode line ending sequence to just matching CR, LF, or CRLF.
899 5. doc/pcresyntax.html was missing from the distribution.
901 6. Put back the definition of PCRE_ERROR_NULLWSLIMIT, for backward
902 compatibility, even though it is no longer used.
904 7. Added macro for snprintf to pcrecpp_unittest.cc and also for strtoll and
905 strtoull to pcrecpp.cc to select the available functions in WIN32 when the
906 windows.h file is present (where different names are used). [This was
907 reversed later after testing - see 16 below.]
909 8. Changed all #include <config.h> to #include "config.h". There were also
910 some further <pcre.h> cases that I changed to "pcre.h".
912 9. When pcregrep was used with the --colour option, it missed the line ending
913 sequence off the lines that it output.
915 10. It was pointed out to me that arrays of string pointers cause lots of
916 relocations when a shared library is dynamically loaded. A technique of
917 using a single long string with a table of offsets can drastically reduce
918 these. I have refactored PCRE in four places to do this. The result is
919 dramatic:
921 Originally: 290
922 After changing UCP table: 187
923 After changing error message table: 43
924 After changing table of "verbs" 36
925 After changing table of Posix names 22
927 Thanks to the folks working on Gregex for glib for this insight.
929 11. --disable-stack-for-recursion caused compiling to fail unless -enable-
930 unicode-properties was also set.
932 12. Updated the tests so that they work when \R is defaulted to ANYCRLF.
934 13. Added checks for ANY and ANYCRLF to pcrecpp.cc where it previously
935 checked only for CRLF.
937 14. Added casts to pcretest.c to avoid compiler warnings.
939 15. Added Craig's patch to various pcrecpp modules to avoid compiler warnings.
941 16. Added Craig's patch to remove the WINDOWS_H tests, that were not working,
942 and instead check for _strtoi64 explicitly, and avoid the use of snprintf()
943 entirely. This removes changes made in 7 above.
945 17. The CMake files have been updated, and there is now more information about
946 building with CMake in the NON-UNIX-USE document.
949 Version 7.3 28-Aug-07
950 ---------------------
952 1. In the rejigging of the build system that eventually resulted in 7.1, the
953 line "#include <pcre.h>" was included in pcre_internal.h. The use of angle
954 brackets there is not right, since it causes compilers to look for an
955 installed pcre.h, not the version that is in the source that is being
956 compiled (which of course may be different). I have changed it back to:
958 #include "pcre.h"
960 I have a vague recollection that the change was concerned with compiling in
961 different directories, but in the new build system, that is taken care of
962 by the VPATH setting the Makefile.
964 2. The pattern .*$ when run in not-DOTALL UTF-8 mode with newline=any failed
965 when the subject happened to end in the byte 0x85 (e.g. if the last
966 character was \x{1ec5}). *Character* 0x85 is one of the "any" newline
967 characters but of course it shouldn't be taken as a newline when it is part
968 of another character. The bug was that, for an unlimited repeat of . in
969 not-DOTALL UTF-8 mode, PCRE was advancing by bytes rather than by
970 characters when looking for a newline.
972 3. A small performance improvement in the DOTALL UTF-8 mode .* case.
974 4. Debugging: adjusted the names of opcodes for different kinds of parentheses
975 in debug output.
977 5. Arrange to use "%I64d" instead of "%lld" and "%I64u" instead of "%llu" for
978 long printing in the pcrecpp unittest when running under MinGW.
980 6. ESC_K was left out of the EBCDIC table.
982 7. Change 7.0/38 introduced a new limit on the number of nested non-capturing
983 parentheses; I made it 1000, which seemed large enough. Unfortunately, the
984 limit also applies to "virtual nesting" when a pattern is recursive, and in
985 this case 1000 isn't so big. I have been able to remove this limit at the
986 expense of backing off one optimization in certain circumstances. Normally,
987 when pcre_exec() would call its internal match() function recursively and
988 immediately return the result unconditionally, it uses a "tail recursion"
989 feature to save stack. However, when a subpattern that can match an empty
990 string has an unlimited repetition quantifier, it no longer makes this
991 optimization. That gives it a stack frame in which to save the data for
992 checking that an empty string has been matched. Previously this was taken
993 from the 1000-entry workspace that had been reserved. So now there is no
994 explicit limit, but more stack is used.
996 8. Applied Daniel's patches to solve problems with the import/export magic
997 syntax that is required for Windows, and which was going wrong for the
998 pcreposix and pcrecpp parts of the library. These were overlooked when this
999 problem was solved for the main library.
1001 9. There were some crude static tests to avoid integer overflow when computing
1002 the size of patterns that contain repeated groups with explicit upper
1003 limits. As the maximum quantifier is 65535, the maximum group length was
1004 set at 30,000 so that the product of these two numbers did not overflow a
1005 32-bit integer. However, it turns out that people want to use groups that
1006 are longer than 30,000 bytes (though not repeat them that many times).
1007 Change 7.0/17 (the refactoring of the way the pattern size is computed) has
1008 made it possible to implement the integer overflow checks in a much more
1009 dynamic way, which I have now done. The artificial limitation on group
1010 length has been removed - we now have only the limit on the total length of
1011 the compiled pattern, which depends on the LINK_SIZE setting.
1013 10. Fixed a bug in the documentation for get/copy named substring when
1014 duplicate names are permitted. If none of the named substrings are set, the
1015 functions return PCRE_ERROR_NOSUBSTRING (7); the doc said they returned an
1016 empty string.
1018 11. Because Perl interprets \Q...\E at a high level, and ignores orphan \E
1019 instances, patterns such as [\Q\E] or [\E] or even [^\E] cause an error,
1020 because the ] is interpreted as the first data character and the
1021 terminating ] is not found. PCRE has been made compatible with Perl in this
1022 regard. Previously, it interpreted [\Q\E] as an empty class, and [\E] could
1023 cause memory overwriting.
1025 10. Like Perl, PCRE automatically breaks an unlimited repeat after an empty
1026 string has been matched (to stop an infinite loop). It was not recognizing
1027 a conditional subpattern that could match an empty string if that
1028 subpattern was within another subpattern. For example, it looped when
1029 trying to match (((?(1)X|))*) but it was OK with ((?(1)X|)*) where the
1030 condition was not nested. This bug has been fixed.
1032 12. A pattern like \X?\d or \P{L}?\d in non-UTF-8 mode could cause a backtrack
1033 past the start of the subject in the presence of bytes with the top bit
1034 set, for example "\x8aBCD".
1036 13. Added Perl 5.10 experimental backtracking controls (*FAIL), (*F), (*PRUNE),
1037 (*SKIP), (*THEN), (*COMMIT), and (*ACCEPT).
1039 14. Optimized (?!) to (*FAIL).
1041 15. Updated the test for a valid UTF-8 string to conform to the later RFC 3629.
1042 This restricts code points to be within the range 0 to 0x10FFFF, excluding
1043 the "low surrogate" sequence 0xD800 to 0xDFFF. Previously, PCRE allowed the
1044 full range 0 to 0x7FFFFFFF, as defined by RFC 2279. Internally, it still
1045 does: it's just the validity check that is more restrictive.
1047 16. Inserted checks for integer overflows during escape sequence (backslash)
1048 processing, and also fixed erroneous offset values for syntax errors during
1049 backslash processing.
1051 17. Fixed another case of looking too far back in non-UTF-8 mode (cf 12 above)
1052 for patterns like [\PPP\x8a]{1,}\x80 with the subject "A\x80".
1054 18. An unterminated class in a pattern like (?1)\c[ with a "forward reference"
1055 caused an overrun.
1057 19. A pattern like (?:[\PPa*]*){8,} which had an "extended class" (one with
1058 something other than just ASCII characters) inside a group that had an
1059 unlimited repeat caused a loop at compile time (while checking to see
1060 whether the group could match an empty string).
1062 20. Debugging a pattern containing \p or \P could cause a crash. For example,
1063 [\P{Any}] did so. (Error in the code for printing property names.)
1065 21. An orphan \E inside a character class could cause a crash.
1067 22. A repeated capturing bracket such as (A)? could cause a wild memory
1068 reference during compilation.
1070 23. There are several functions in pcre_compile() that scan along a compiled
1071 expression for various reasons (e.g. to see if it's fixed length for look
1072 behind). There were bugs in these functions when a repeated \p or \P was
1073 present in the pattern. These operators have additional parameters compared
1074 with \d, etc, and these were not being taken into account when moving along
1075 the compiled data. Specifically:
1077 (a) A item such as \p{Yi}{3} in a lookbehind was not treated as fixed
1078 length.
1080 (b) An item such as \pL+ within a repeated group could cause crashes or
1081 loops.
1083 (c) A pattern such as \p{Yi}+(\P{Yi}+)(?1) could give an incorrect
1084 "reference to non-existent subpattern" error.
1086 (d) A pattern like (\P{Yi}{2}\277)? could loop at compile time.
1088 24. A repeated \S or \W in UTF-8 mode could give wrong answers when multibyte
1089 characters were involved (for example /\S{2}/8g with "A\x{a3}BC").
1091 25. Using pcregrep in multiline, inverted mode (-Mv) caused it to loop.
1093 26. Patterns such as [\P{Yi}A] which include \p or \P and just one other
1094 character were causing crashes (broken optimization).
1096 27. Patterns such as (\P{Yi}*\277)* (group with possible zero repeat containing
1097 \p or \P) caused a compile-time loop.
1099 28. More problems have arisen in unanchored patterns when CRLF is a valid line
1100 break. For example, the unstudied pattern [\r\n]A does not match the string
1101 "\r\nA" because change 7.0/46 below moves the current point on by two
1102 characters after failing to match at the start. However, the pattern \nA
1103 *does* match, because it doesn't start till \n, and if [\r\n]A is studied,
1104 the same is true. There doesn't seem any very clean way out of this, but
1105 what I have chosen to do makes the common cases work: PCRE now takes note
1106 of whether there can be an explicit match for \r or \n anywhere in the
1107 pattern, and if so, 7.0/46 no longer applies. As part of this change,
1108 there's a new PCRE_INFO_HASCRORLF option for finding out whether a compiled
1109 pattern has explicit CR or LF references.
1111 29. Added (*CR) etc for changing newline setting at start of pattern.
1114 Version 7.2 19-Jun-07
1115 ---------------------
1117 1. If the fr_FR locale cannot be found for test 3, try the "french" locale,
1118 which is apparently normally available under Windows.
1120 2. Re-jig the pcregrep tests with different newline settings in an attempt
1121 to make them independent of the local environment's newline setting.
1123 3. Add code to configure.ac to remove -g from the CFLAGS default settings.
1125 4. Some of the "internals" tests were previously cut out when the link size
1126 was not 2, because the output contained actual offsets. The recent new
1127 "Z" feature of pcretest means that these can be cut out, making the tests
1128 usable with all link sizes.
1130 5. Implemented Stan Switzer's goto replacement for longjmp() when not using
1131 stack recursion. This gives a massive performance boost under BSD, but just
1132 a small improvement under Linux. However, it saves one field in the frame
1133 in all cases.
1135 6. Added more features from the forthcoming Perl 5.10:
1137 (a) (?-n) (where n is a string of digits) is a relative subroutine or
1138 recursion call. It refers to the nth most recently opened parentheses.
1140 (b) (?+n) is also a relative subroutine call; it refers to the nth next
1141 to be opened parentheses.
1143 (c) Conditions that refer to capturing parentheses can be specified
1144 relatively, for example, (?(-2)... or (?(+3)...
1146 (d) \K resets the start of the current match so that everything before
1147 is not part of it.
1149 (e) \k{name} is synonymous with \k<name> and \k'name' (.NET compatible).
1151 (f) \g{name} is another synonym - part of Perl 5.10's unification of
1152 reference syntax.
1154 (g) (?| introduces a group in which the numbering of parentheses in each
1155 alternative starts with the same number.
1157 (h) \h, \H, \v, and \V match horizontal and vertical whitespace.
1159 7. Added two new calls to pcre_fullinfo(): PCRE_INFO_OKPARTIAL and
1162 8. A pattern such as (.*(.)?)* caused pcre_exec() to fail by either not
1163 terminating or by crashing. Diagnosed by Viktor Griph; it was in the code
1164 for detecting groups that can match an empty string.
1166 9. A pattern with a very large number of alternatives (more than several
1167 hundred) was running out of internal workspace during the pre-compile
1168 phase, where pcre_compile() figures out how much memory will be needed. A
1169 bit of new cunning has reduced the workspace needed for groups with
1170 alternatives. The 1000-alternative test pattern now uses 12 bytes of
1171 workspace instead of running out of the 4096 that are available.
1173 10. Inserted some missing (unsigned int) casts to get rid of compiler warnings.
1175 11. Applied patch from Google to remove an optimization that didn't quite work.
1176 The report of the bug said:
1178 pcrecpp::RE("a*").FullMatch("aaa") matches, while
1179 pcrecpp::RE("a*?").FullMatch("aaa") does not, and
1180 pcrecpp::RE("a*?\\z").FullMatch("aaa") does again.
1182 12. If \p or \P was used in non-UTF-8 mode on a character greater than 127
1183 it matched the wrong number of bytes.
1186 Version 7.1 24-Apr-07
1187 ---------------------
1189 1. Applied Bob Rossi and Daniel G's patches to convert the build system to one
1190 that is more "standard", making use of automake and other Autotools. There
1191 is some re-arrangement of the files and adjustment of comments consequent
1192 on this.
1194 2. Part of the patch fixed a problem with the pcregrep tests. The test of -r
1195 for recursive directory scanning broke on some systems because the files
1196 are not scanned in any specific order and on different systems the order
1197 was different. A call to "sort" has been inserted into RunGrepTest for the
1198 approprate test as a short-term fix. In the longer term there may be an
1199 alternative.
1201 3. I had an email from Eric Raymond about problems translating some of PCRE's
1202 man pages to HTML (despite the fact that I distribute HTML pages, some
1203 people do their own conversions for various reasons). The problems
1204 concerned the use of low-level troff macros .br and .in. I have therefore
1205 removed all such uses from the man pages (some were redundant, some could
1206 be replaced by .nf/.fi pairs). The 132html script that I use to generate
1207 HTML has been updated to handle .nf/.fi and to complain if it encounters
1208 .br or .in.
1210 4. Updated comments in configure.ac that get placed in config.h.in and also
1211 arranged for config.h to be included in the distribution, with the name
1212 config.h.generic, for the benefit of those who have to compile without
1213 Autotools (compare pcre.h, which is now distributed as pcre.h.generic).
1215 5. Updated the support (such as it is) for Virtual Pascal, thanks to Stefan
1216 Weber: (1) pcre_internal.h was missing some function renames; (2) updated
1217 makevp.bat for the current PCRE, using the additional files
1218 makevp_c.txt, makevp_l.txt, and pcregexp.pas.
1220 6. A Windows user reported a minor discrepancy with test 2, which turned out
1221 to be caused by a trailing space on an input line that had got lost in his
1222 copy. The trailing space was an accident, so I've just removed it.
1224 7. Add -Wl,-R... flags in pcre-config.in for *BSD* systems, as I'm told
1225 that is needed.
1227 8. Mark ucp_table (in ucptable.h) and ucp_gentype (in pcre_ucp_searchfuncs.c)
1228 as "const" (a) because they are and (b) because it helps the PHP
1229 maintainers who have recently made a script to detect big data structures
1230 in the php code that should be moved to the .rodata section. I remembered
1231 to update Builducptable as well, so it won't revert if ucptable.h is ever
1232 re-created.
1234 9. Added some extra #ifdef SUPPORT_UTF8 conditionals into pcretest.c,
1235 pcre_printint.src, pcre_compile.c, pcre_study.c, and pcre_tables.c, in
1236 order to be able to cut out the UTF-8 tables in the latter when UTF-8
1237 support is not required. This saves 1.5-2K of code, which is important in
1238 some applications.
1240 Later: more #ifdefs are needed in pcre_ord2utf8.c and pcre_valid_utf8.c
1241 so as not to refer to the tables, even though these functions will never be
1242 called when UTF-8 support is disabled. Otherwise there are problems with a
1243 shared library.
1245 10. Fixed two bugs in the emulated memmove() function in pcre_internal.h:
1247 (a) It was defining its arguments as char * instead of void *.
1249 (b) It was assuming that all moves were upwards in memory; this was true
1250 a long time ago when I wrote it, but is no longer the case.
1252 The emulated memove() is provided for those environments that have neither
1253 memmove() nor bcopy(). I didn't think anyone used it these days, but that
1254 is clearly not the case, as these two bugs were recently reported.
1256 11. The script PrepareRelease is now distributed: it calls 132html, CleanTxt,
1257 and Detrail to create the HTML documentation, the .txt form of the man
1258 pages, and it removes trailing spaces from listed files. It also creates
1259 pcre.h.generic and config.h.generic from pcre.h and config.h. In the latter
1260 case, it wraps all the #defines with #ifndefs. This script should be run
1261 before "make dist".
1263 12. Fixed two fairly obscure bugs concerned with quantified caseless matching
1264 with Unicode property support.
1266 (a) For a maximizing quantifier, if the two different cases of the
1267 character were of different lengths in their UTF-8 codings (there are
1268 some cases like this - I found 11), and the matching function had to
1269 back up over a mixture of the two cases, it incorrectly assumed they
1270 were both the same length.
1272 (b) When PCRE was configured to use the heap rather than the stack for
1273 recursion during matching, it was not correctly preserving the data for
1274 the other case of a UTF-8 character when checking ahead for a match
1275 while processing a minimizing repeat. If the check also involved
1276 matching a wide character, but failed, corruption could cause an
1277 erroneous result when trying to check for a repeat of the original
1278 character.
1280 13. Some tidying changes to the testing mechanism:
1282 (a) The RunTest script now detects the internal link size and whether there
1283 is UTF-8 and UCP support by running ./pcretest -C instead of relying on
1284 values substituted by "configure". (The RunGrepTest script already did
1285 this for UTF-8.) The configure.ac script no longer substitutes the
1286 relevant variables.
1288 (b) The debugging options /B and /D in pcretest show the compiled bytecode
1289 with length and offset values. This means that the output is different
1290 for different internal link sizes. Test 2 is skipped for link sizes
1291 other than 2 because of this, bypassing the problem. Unfortunately,
1292 there was also a test in test 3 (the locale tests) that used /B and
1293 failed for link sizes other than 2. Rather than cut the whole test out,
1294 I have added a new /Z option to pcretest that replaces the length and
1295 offset values with spaces. This is now used to make test 3 independent
1296 of link size. (Test 2 will be tidied up later.)
1298 14. If erroroffset was passed as NULL to pcre_compile, it provoked a
1299 segmentation fault instead of returning the appropriate error message.
1301 15. In multiline mode when the newline sequence was set to "any", the pattern
1302 ^$ would give a match between the \r and \n of a subject such as "A\r\nB".
1303 This doesn't seem right; it now treats the CRLF combination as the line
1304 ending, and so does not match in that case. It's only a pattern such as ^$
1305 that would hit this one: something like ^ABC$ would have failed after \r
1306 and then tried again after \r\n.
1308 16. Changed the comparison command for RunGrepTest from "diff -u" to "diff -ub"
1309 in an attempt to make files that differ only in their line terminators
1310 compare equal. This works on Linux.
1312 17. Under certain error circumstances pcregrep might try to free random memory
1313 as it exited. This is now fixed, thanks to valgrind.
1315 19. In pcretest, if the pattern /(?m)^$/g<any> was matched against the string
1316 "abc\r\n\r\n", it found an unwanted second match after the second \r. This
1317 was because its rules for how to advance for /g after matching an empty
1318 string at the end of a line did not allow for this case. They now check for
1319 it specially.
1321 20. pcretest is supposed to handle patterns and data of any length, by
1322 extending its buffers when necessary. It was getting this wrong when the
1323 buffer for a data line had to be extended.
1325 21. Added PCRE_NEWLINE_ANYCRLF which is like ANY, but matches only CR, LF, or
1326 CRLF as a newline sequence.
1328 22. Code for handling Unicode properties in pcre_dfa_exec() wasn't being cut
1329 out by #ifdef SUPPORT_UCP. This did no harm, as it could never be used, but
1330 I have nevertheless tidied it up.
1332 23. Added some casts to kill warnings from HP-UX ia64 compiler.
1334 24. Added a man page for pcre-config.
1337 Version 7.0 19-Dec-06
1338 ---------------------
1340 1. Fixed a signed/unsigned compiler warning in pcre_compile.c, shown up by
1341 moving to gcc 4.1.1.
1343 2. The -S option for pcretest uses setrlimit(); I had omitted to #include
1344 sys/time.h, which is documented as needed for this function. It doesn't
1345 seem to matter on Linux, but it showed up on some releases of OS X.
1347 3. It seems that there are systems where bytes whose values are greater than
1348 127 match isprint() in the "C" locale. The "C" locale should be the
1349 default when a C program starts up. In most systems, only ASCII printing
1350 characters match isprint(). This difference caused the output from pcretest
1351 to vary, making some of the tests fail. I have changed pcretest so that:
1353 (a) When it is outputting text in the compiled version of a pattern, bytes
1354 other than 32-126 are always shown as hex escapes.
1356 (b) When it is outputting text that is a matched part of a subject string,
1357 it does the same, unless a different locale has been set for the match
1358 (using the /L modifier). In this case, it uses isprint() to decide.
1360 4. Fixed a major bug that caused incorrect computation of the amount of memory
1361 required for a compiled pattern when options that changed within the
1362 pattern affected the logic of the preliminary scan that determines the
1363 length. The relevant options are -x, and -i in UTF-8 mode. The result was
1364 that the computed length was too small. The symptoms of this bug were
1365 either the PCRE error "internal error: code overflow" from pcre_compile(),
1366 or a glibc crash with a message such as "pcretest: free(): invalid next
1367 size (fast)". Examples of patterns that provoked this bug (shown in
1368 pcretest format) are:
1370 /(?-x: )/x
1371 /(?x)(?-x: \s*#\s*)/
1372 /((?i)[\x{c0}])/8
1373 /(?i:[\x{c0}])/8
1375 HOWEVER: Change 17 below makes this fix obsolete as the memory computation
1376 is now done differently.
1378 5. Applied patches from Google to: (a) add a QuoteMeta function to the C++
1379 wrapper classes; (b) implement a new function in the C++ scanner that is
1380 more efficient than the old way of doing things because it avoids levels of
1381 recursion in the regex matching; (c) add a paragraph to the documentation
1382 for the FullMatch() function.
1384 6. The escape sequence \n was being treated as whatever was defined as
1385 "newline". Not only was this contrary to the documentation, which states
1386 that \n is character 10 (hex 0A), but it also went horribly wrong when
1387 "newline" was defined as CRLF. This has been fixed.
1389 7. In pcre_dfa_exec.c the value of an unsigned integer (the variable called c)
1390 was being set to -1 for the "end of line" case (supposedly a value that no
1391 character can have). Though this value is never used (the check for end of
1392 line is "zero bytes in current character"), it caused compiler complaints.
1393 I've changed it to 0xffffffff.
1395 8. In pcre_version.c, the version string was being built by a sequence of
1396 C macros that, in the event of PCRE_PRERELEASE being defined as an empty
1397 string (as it is for production releases) called a macro with an empty
1398 argument. The C standard says the result of this is undefined. The gcc
1399 compiler treats it as an empty string (which was what was wanted) but it is
1400 reported that Visual C gives an error. The source has been hacked around to
1401 avoid this problem.
1403 9. On the advice of a Windows user, included <io.h> and <fcntl.h> in Windows
1404 builds of pcretest, and changed the call to _setmode() to use _O_BINARY
1405 instead of 0x8000. Made all the #ifdefs test both _WIN32 and WIN32 (not all
1406 of them did).
1408 10. Originally, pcretest opened its input and output without "b"; then I was
1409 told that "b" was needed in some environments, so it was added for release
1410 5.0 to both the input and output. (It makes no difference on Unix-like
1411 systems.) Later I was told that it is wrong for the input on Windows. I've
1412 now abstracted the modes into two macros, to make it easier to fiddle with
1413 them, and removed "b" from the input mode under Windows.
1415 11. Added pkgconfig support for the C++ wrapper library, libpcrecpp.
1417 12. Added -help and --help to pcretest as an official way of being reminded
1418 of the options.
1420 13. Removed some redundant semicolons after macro calls in pcrecpparg.h.in
1421 and pcrecpp.cc because they annoy compilers at high warning levels.
1423 14. A bit of tidying/refactoring in pcre_exec.c in the main bumpalong loop.
1425 15. Fixed an occurrence of == in configure.ac that should have been = (shell
1426 scripts are not C programs :-) and which was not noticed because it works
1427 on Linux.
1429 16. pcretest is supposed to handle any length of pattern and data line (as one
1430 line or as a continued sequence of lines) by extending its input buffer if
1431 necessary. This feature was broken for very long pattern lines, leading to
1432 a string of junk being passed to pcre_compile() if the pattern was longer
1433 than about 50K.
1435 17. I have done a major re-factoring of the way pcre_compile() computes the
1436 amount of memory needed for a compiled pattern. Previously, there was code
1437 that made a preliminary scan of the pattern in order to do this. That was
1438 OK when PCRE was new, but as the facilities have expanded, it has become
1439 harder and harder to keep it in step with the real compile phase, and there
1440 have been a number of bugs (see for example, 4 above). I have now found a
1441 cunning way of running the real compile function in a "fake" mode that
1442 enables it to compute how much memory it would need, while actually only
1443 ever using a few hundred bytes of working memory and without too many
1444 tests of the mode. This should make future maintenance and development
1445 easier. A side effect of this work is that the limit of 200 on the nesting
1446 depth of parentheses has been removed (though this was never a serious
1447 limitation, I suspect). However, there is a downside: pcre_compile() now
1448 runs more slowly than before (30% or more, depending on the pattern). I
1449 hope this isn't a big issue. There is no effect on runtime performance.
1451 18. Fixed a minor bug in pcretest: if a pattern line was not terminated by a
1452 newline (only possible for the last line of a file) and it was a
1453 pattern that set a locale (followed by /Lsomething), pcretest crashed.
1455 19. Added additional timing features to pcretest. (1) The -tm option now times
1456 matching only, not compiling. (2) Both -t and -tm can be followed, as a
1457 separate command line item, by a number that specifies the number of
1458 repeats to use when timing. The default is 50000; this gives better
1459 precision, but takes uncomfortably long for very large patterns.
1461 20. Extended pcre_study() to be more clever in cases where a branch of a
1462 subpattern has no definite first character. For example, (a*|b*)[cd] would
1463 previously give no result from pcre_study(). Now it recognizes that the
1464 first character must be a, b, c, or d.
1466 21. There was an incorrect error "recursive call could loop indefinitely" if
1467 a subpattern (or the entire pattern) that was being tested for matching an
1468 empty string contained only one non-empty item after a nested subpattern.
1469 For example, the pattern (?>\x{100}*)\d(?R) provoked this error
1470 incorrectly, because the \d was being skipped in the check.
1472 22. The pcretest program now has a new pattern option /B and a command line
1473 option -b, which is equivalent to adding /B to every pattern. This causes
1474 it to show the compiled bytecode, without the additional information that
1475 -d shows. The effect of -d is now the same as -b with -i (and similarly, /D
1476 is the same as /B/I).
1478 23. A new optimization is now able automatically to treat some sequences such
1479 as a*b as a*+b. More specifically, if something simple (such as a character
1480 or a simple class like \d) has an unlimited quantifier, and is followed by
1481 something that cannot possibly match the quantified thing, the quantifier
1482 is automatically "possessified".
1484 24. A recursive reference to a subpattern whose number was greater than 39
1485 went wrong under certain circumstances in UTF-8 mode. This bug could also
1486 have affected the operation of pcre_study().
1488 25. Realized that a little bit of performance could be had by replacing
1489 (c & 0xc0) == 0xc0 with c >= 0xc0 when processing UTF-8 characters.
1491 26. Timing data from pcretest is now shown to 4 decimal places instead of 3.
1493 27. Possessive quantifiers such as a++ were previously implemented by turning
1494 them into atomic groups such as ($>a+). Now they have their own opcodes,
1495 which improves performance. This includes the automatically created ones
1496 from 23 above.
1498 28. A pattern such as (?=(\w+))\1: which simulates an atomic group using a
1499 lookahead was broken if it was not anchored. PCRE was mistakenly expecting
1500 the first matched character to be a colon. This applied both to named and
1501 numbered groups.
1503 29. The ucpinternal.h header file was missing its idempotency #ifdef.
1505 30. I was sent a "project" file called libpcre.a.dev which I understand makes
1506 building PCRE on Windows easier, so I have included it in the distribution.
1508 31. There is now a check in pcretest against a ridiculously large number being
1509 returned by pcre_exec() or pcre_dfa_exec(). If this happens in a /g or /G
1510 loop, the loop is abandoned.
1512 32. Forward references to subpatterns in conditions such as (?(2)...) where
1513 subpattern 2 is defined later cause pcre_compile() to search forwards in
1514 the pattern for the relevant set of parentheses. This search went wrong
1515 when there were unescaped parentheses in a character class, parentheses
1516 escaped with \Q...\E, or parentheses in a #-comment in /x mode.
1518 33. "Subroutine" calls and backreferences were previously restricted to
1519 referencing subpatterns earlier in the regex. This restriction has now
1520 been removed.
1522 34. Added a number of extra features that are going to be in Perl 5.10. On the
1523 whole, these are just syntactic alternatives for features that PCRE had
1524 previously implemented using the Python syntax or my own invention. The
1525 other formats are all retained for compatibility.
1527 (a) Named groups can now be defined as (?<name>...) or (?'name'...) as well
1528 as (?P<name>...). The new forms, as well as being in Perl 5.10, are
1529 also .NET compatible.
1531 (b) A recursion or subroutine call to a named group can now be defined as
1532 (?&name) as well as (?P>name).
1534 (c) A backreference to a named group can now be defined as \k<name> or
1535 \k'name' as well as (?P=name). The new forms, as well as being in Perl
1536 5.10, are also .NET compatible.
1538 (d) A conditional reference to a named group can now use the syntax
1539 (?(<name>) or (?('name') as well as (?(name).
1541 (e) A "conditional group" of the form (?(DEFINE)...) can be used to define
1542 groups (named and numbered) that are never evaluated inline, but can be
1543 called as "subroutines" from elsewhere. In effect, the DEFINE condition
1544 is always false. There may be only one alternative in such a group.
1546 (f) A test for recursion can be given as (?(R1).. or (?(R&name)... as well
1547 as the simple (?(R). The condition is true only if the most recent
1548 recursion is that of the given number or name. It does not search out
1549 through the entire recursion stack.
1551 (g) The escape \gN or \g{N} has been added, where N is a positive or
1552 negative number, specifying an absolute or relative reference.
1554 35. Tidied to get rid of some further signed/unsigned compiler warnings and
1555 some "unreachable code" warnings.
1557 36. Updated the Unicode property tables to Unicode version 5.0.0. Amongst other
1558 things, this adds five new scripts.
1560 37. Perl ignores orphaned \E escapes completely. PCRE now does the same.
1561 There were also incompatibilities regarding the handling of \Q..\E inside
1562 character classes, for example with patterns like [\Qa\E-\Qz\E] where the
1563 hyphen was adjacent to \Q or \E. I hope I've cleared all this up now.
1565 38. Like Perl, PCRE detects when an indefinitely repeated parenthesized group
1566 matches an empty string, and forcibly breaks the loop. There were bugs in
1567 this code in non-simple cases. For a pattern such as ^(a()*)* matched
1568 against aaaa the result was just "a" rather than "aaaa", for example. Two
1569 separate and independent bugs (that affected different cases) have been
1570 fixed.
1572 39. Refactored the code to abolish the use of different opcodes for small
1573 capturing bracket numbers. This is a tidy that I avoided doing when I
1574 removed the limit on the number of capturing brackets for 3.5 back in 2001.
1575 The new approach is not only tidier, it makes it possible to reduce the
1576 memory needed to fix the previous bug (38).
1578 40. Implemented PCRE_NEWLINE_ANY to recognize any of the Unicode newline
1579 sequences (http://unicode.org/unicode/reports/tr18/) as "newline" when
1580 processing dot, circumflex, or dollar metacharacters, or #-comments in /x
1581 mode.
1583 41. Add \R to match any Unicode newline sequence, as suggested in the Unicode
1584 report.
1586 42. Applied patch, originally from Ari Pollak, modified by Google, to allow
1587 copy construction and assignment in the C++ wrapper.
1589 43. Updated pcregrep to support "--newline=any". In the process, I fixed a
1590 couple of bugs that could have given wrong results in the "--newline=crlf"
1591 case.
1593 44. Added a number of casts and did some reorganization of signed/unsigned int
1594 variables following suggestions from Dair Grant. Also renamed the variable
1595 "this" as "item" because it is a C++ keyword.
1597 45. Arranged for dftables to add
1599 #include "pcre_internal.h"
1601 to pcre_chartables.c because without it, gcc 4.x may remove the array
1602 definition from the final binary if PCRE is built into a static library and
1603 dead code stripping is activated.
1605 46. For an unanchored pattern, if a match attempt fails at the start of a
1606 newline sequence, and the newline setting is CRLF or ANY, and the next two
1607 characters are CRLF, advance by two characters instead of one.
1610 Version 6.7 04-Jul-06
1611 ---------------------
1613 1. In order to handle tests when input lines are enormously long, pcretest has
1614 been re-factored so that it automatically extends its buffers when
1615 necessary. The code is crude, but this _is_ just a test program. The
1616 default size has been increased from 32K to 50K.
1618 2. The code in pcre_study() was using the value of the re argument before
1619 testing it for NULL. (Of course, in any sensible call of the function, it
1620 won't be NULL.)
1622 3. The memmove() emulation function in pcre_internal.h, which is used on
1623 systems that lack both memmove() and bcopy() - that is, hardly ever -
1624 was missing a "static" storage class specifier.
1626 4. When UTF-8 mode was not set, PCRE looped when compiling certain patterns
1627 containing an extended class (one that cannot be represented by a bitmap
1628 because it contains high-valued characters or Unicode property items, e.g.
1629 [\pZ]). Almost always one would set UTF-8 mode when processing such a
1630 pattern, but PCRE should not loop if you do not (it no longer does).
1631 [Detail: two cases were found: (a) a repeated subpattern containing an
1632 extended class; (b) a recursive reference to a subpattern that followed a
1633 previous extended class. It wasn't skipping over the extended class
1634 correctly when UTF-8 mode was not set.]
1636 5. A negated single-character class was not being recognized as fixed-length
1637 in lookbehind assertions such as (?<=[^f]), leading to an incorrect
1638 compile error "lookbehind assertion is not fixed length".
1640 6. The RunPerlTest auxiliary script was showing an unexpected difference
1641 between PCRE and Perl for UTF-8 tests. It turns out that it is hard to
1642 write a Perl script that can interpret lines of an input file either as
1643 byte characters or as UTF-8, which is what "perltest" was being required to
1644 do for the non-UTF-8 and UTF-8 tests, respectively. Essentially what you
1645 can't do is switch easily at run time between having the "use utf8;" pragma
1646 or not. In the end, I fudged it by using the RunPerlTest script to insert
1647 "use utf8;" explicitly for the UTF-8 tests.
1649 7. In multiline (/m) mode, PCRE was matching ^ after a terminating newline at
1650 the end of the subject string, contrary to the documentation and to what
1651 Perl does. This was true of both matching functions. Now it matches only at
1652 the start of the subject and immediately after *internal* newlines.
1654 8. A call of pcre_fullinfo() from pcretest to get the option bits was passing
1655 a pointer to an int instead of a pointer to an unsigned long int. This
1656 caused problems on 64-bit systems.
1658 9. Applied a patch from the folks at Google to pcrecpp.cc, to fix "another
1659 instance of the 'standard' template library not being so standard".
1661 10. There was no check on the number of named subpatterns nor the maximum
1662 length of a subpattern name. The product of these values is used to compute
1663 the size of the memory block for a compiled pattern. By supplying a very
1664 long subpattern name and a large number of named subpatterns, the size
1665 computation could be caused to overflow. This is now prevented by limiting
1666 the length of names to 32 characters, and the number of named subpatterns
1667 to 10,000.
1669 11. Subpatterns that are repeated with specific counts have to be replicated in
1670 the compiled pattern. The size of memory for this was computed from the
1671 length of the subpattern and the repeat count. The latter is limited to
1672 65535, but there was no limit on the former, meaning that integer overflow
1673 could in principle occur. The compiled length of a repeated subpattern is
1674 now limited to 30,000 bytes in order to prevent this.
1676 12. Added the optional facility to have named substrings with the same name.
1678 13. Added the ability to use a named substring as a condition, using the
1679 Python syntax: (?(name)yes|no). This overloads (?(R)... and names that
1680 are numbers (not recommended). Forward references are permitted.
1682 14. Added forward references in named backreferences (if you see what I mean).
1684 15. In UTF-8 mode, with the PCRE_DOTALL option set, a quantified dot in the
1685 pattern could run off the end of the subject. For example, the pattern
1686 "(?s)(.{1,5})"8 did this with the subject "ab".
1688 16. If PCRE_DOTALL or PCRE_MULTILINE were set, pcre_dfa_exec() behaved as if
1689 PCRE_CASELESS was set when matching characters that were quantified with ?
1690 or *.
1692 17. A character class other than a single negated character that had a minimum
1693 but no maximum quantifier - for example [ab]{6,} - was not handled
1694 correctly by pce_dfa_exec(). It would match only one character.
1696 18. A valid (though odd) pattern that looked like a POSIX character
1697 class but used an invalid character after [ (for example [[,abc,]]) caused
1698 pcre_compile() to give the error "Failed: internal error: code overflow" or
1699 in some cases to crash with a glibc free() error. This could even happen if
1700 the pattern terminated after [[ but there just happened to be a sequence of
1701 letters, a binary zero, and a closing ] in the memory that followed.
1703 19. Perl's treatment of octal escapes in the range \400 to \777 has changed
1704 over the years. Originally (before any Unicode support), just the bottom 8
1705 bits were taken. Thus, for example, \500 really meant \100. Nowadays the
1706 output from "man perlunicode" includes this:
1708 The regular expression compiler produces polymorphic opcodes. That
1709 is, the pattern adapts to the data and automatically switches to
1710 the Unicode character scheme when presented with Unicode data--or
1711 instead uses a traditional byte scheme when presented with byte
1712 data.
1714 Sadly, a wide octal escape does not cause a switch, and in a string with
1715 no other multibyte characters, these octal escapes are treated as before.
1716 Thus, in Perl, the pattern /\500/ actually matches \100 but the pattern
1717 /\500|\x{1ff}/ matches \500 or \777 because the whole thing is treated as a
1718 Unicode string.
1720 I have not perpetrated such confusion in PCRE. Up till now, it took just
1721 the bottom 8 bits, as in old Perl. I have now made octal escapes with
1722 values greater than \377 illegal in non-UTF-8 mode. In UTF-8 mode they
1723 translate to the appropriate multibyte character.
1725 29. Applied some refactoring to reduce the number of warnings from Microsoft
1726 and Borland compilers. This has included removing the fudge introduced
1727 seven years ago for the OS/2 compiler (see 2.02/2 below) because it caused
1728 a warning about an unused variable.
1730 21. PCRE has not included VT (character 0x0b) in the set of whitespace
1731 characters since release 4.0, because Perl (from release 5.004) does not.
1732 [Or at least, is documented not to: some releases seem to be in conflict
1733 with the documentation.] However, when a pattern was studied with
1734 pcre_study() and all its branches started with \s, PCRE still included VT
1735 as a possible starting character. Of course, this did no harm; it just
1736 caused an unnecessary match attempt.
1738 22. Removed a now-redundant internal flag bit that recorded the fact that case
1739 dependency changed within the pattern. This was once needed for "required
1740 byte" processing, but is no longer used. This recovers a now-scarce options
1741 bit. Also moved the least significant internal flag bit to the most-
1742 significant bit of the word, which was not previously used (hangover from
1743 the days when it was an int rather than a uint) to free up another bit for
1744 the future.
1746 23. Added support for CRLF line endings as well as CR and LF. As well as the
1747 default being selectable at build time, it can now be changed at runtime
1748 via the PCRE_NEWLINE_xxx flags. There are now options for pcregrep to
1749 specify that it is scanning data with non-default line endings.
1751 24. Changed the definition of CXXLINK to make it agree with the definition of
1752 LINK in the Makefile, by replacing LDFLAGS to CXXFLAGS.
1754 25. Applied Ian Taylor's patches to avoid using another stack frame for tail
1755 recursions. This makes a big different to stack usage for some patterns.
1757 26. If a subpattern containing a named recursion or subroutine reference such
1758 as (?P>B) was quantified, for example (xxx(?P>B)){3}, the calculation of
1759 the space required for the compiled pattern went wrong and gave too small a
1760 value. Depending on the environment, this could lead to "Failed: internal
1761 error: code overflow at offset 49" or "glibc detected double free or
1762 corruption" errors.
1764 27. Applied patches from Google (a) to support the new newline modes and (b) to
1765 advance over multibyte UTF-8 characters in GlobalReplace.
1767 28. Change free() to pcre_free() in pcredemo.c. Apparently this makes a
1768 difference for some implementation of PCRE in some Windows version.
1770 29. Added some extra testing facilities to pcretest:
1772 \q<number> in a data line sets the "match limit" value
1773 \Q<number> in a data line sets the "match recursion limt" value
1774 -S <number> sets the stack size, where <number> is in megabytes
1776 The -S option isn't available for Windows.
1779 Version 6.6 06-Feb-06
1780 ---------------------
1782 1. Change 16(a) for 6.5 broke things, because PCRE_DATA_SCOPE was not defined
1783 in pcreposix.h. I have copied the definition from pcre.h.
1785 2. Change 25 for 6.5 broke compilation in a build directory out-of-tree
1786 because pcre.h is no longer a built file.
1788 3. Added Jeff Friedl's additional debugging patches to pcregrep. These are
1789 not normally included in the compiled code.
1792 Version 6.5 01-Feb-06
1793 ---------------------
1795 1. When using the partial match feature with pcre_dfa_exec(), it was not
1796 anchoring the second and subsequent partial matches at the new starting
1797 point. This could lead to incorrect results. For example, with the pattern
1798 /1234/, partially matching against "123" and then "a4" gave a match.
1800 2. Changes to pcregrep:
1802 (a) All non-match returns from pcre_exec() were being treated as failures
1803 to match the line. Now, unless the error is PCRE_ERROR_NOMATCH, an
1804 error message is output. Some extra information is given for the
1806 probably the only errors that are likely to be caused by users (by
1807 specifying a regex that has nested indefinite repeats, for instance).
1808 If there are more than 20 of these errors, pcregrep is abandoned.
1810 (b) A binary zero was treated as data while matching, but terminated the
1811 output line if it was written out. This has been fixed: binary zeroes
1812 are now no different to any other data bytes.
1814 (c) Whichever of the LC_ALL or LC_CTYPE environment variables is set is
1815 used to set a locale for matching. The --locale=xxxx long option has
1816 been added (no short equivalent) to specify a locale explicitly on the
1817 pcregrep command, overriding the environment variables.
1819 (d) When -B was used with -n, some line numbers in the output were one less
1820 than they should have been.
1822 (e) Added the -o (--only-matching) option.
1824 (f) If -A or -C was used with -c (count only), some lines of context were
1825 accidentally printed for the final match.
1827 (g) Added the -H (--with-filename) option.
1829 (h) The combination of options -rh failed to suppress file names for files
1830 that were found from directory arguments.
1832 (i) Added the -D (--devices) and -d (--directories) options.
1834 (j) Added the -F (--fixed-strings) option.
1836 (k) Allow "-" to be used as a file name for -f as well as for a data file.
1838 (l) Added the --colo(u)r option.
1840 (m) Added Jeffrey Friedl's -S testing option, but within #ifdefs so that it
1841 is not present by default.
1843 3. A nasty bug was discovered in the handling of recursive patterns, that is,
1844 items such as (?R) or (?1), when the recursion could match a number of
1845 alternatives. If it matched one of the alternatives, but subsequently,
1846 outside the recursion, there was a failure, the code tried to back up into
1847 the recursion. However, because of the way PCRE is implemented, this is not
1848 possible, and the result was an incorrect result from the match.
1850 In order to prevent this happening, the specification of recursion has
1851 been changed so that all such subpatterns are automatically treated as
1852 atomic groups. Thus, for example, (?R) is treated as if it were (?>(?R)).
1854 4. I had overlooked the fact that, in some locales, there are characters for
1855 which isalpha() is true but neither isupper() nor islower() are true. In
1856 the fr_FR locale, for instance, the \xAA and \xBA characters (ordmasculine
1857 and ordfeminine) are like this. This affected the treatment of \w and \W
1858 when they appeared in character classes, but not when they appeared outside
1859 a character class. The bit map for "word" characters is now created
1860 separately from the results of isalnum() instead of just taking it from the
1861 upper, lower, and digit maps. (Plus the underscore character, of course.)
1863 5. The above bug also affected the handling of POSIX character classes such as
1864 [[:alpha:]] and [[:alnum:]]. These do not have their own bit maps in PCRE's
1865 permanent tables. Instead, the bit maps for such a class were previously
1866 created as the appropriate unions of the upper, lower, and digit bitmaps.
1867 Now they are created by subtraction from the [[:word:]] class, which has
1868 its own bitmap.
1870 6. The [[:blank:]] character class matches horizontal, but not vertical space.
1871 It is created by subtracting the vertical space characters (\x09, \x0a,
1872 \x0b, \x0c) from the [[:space:]] bitmap. Previously, however, the
1873 subtraction was done in the overall bitmap for a character class, meaning
1874 that a class such as [\x0c[:blank:]] was incorrect because \x0c would not
1875 be recognized. This bug has been fixed.
1877 7. Patches from the folks at Google:
1879 (a) pcrecpp.cc: "to handle a corner case that may or may not happen in
1880 real life, but is still worth protecting against".
1882 (b) pcrecpp.cc: "corrects a bug when negative radixes are used with
1883 regular expressions".
1885 (c) pcre_scanner.cc: avoid use of std::count() because not all systems
1886 have it.
1888 (d) Split off pcrecpparg.h from pcrecpp.h and had the former built by
1889 "configure" and the latter not, in order to fix a problem somebody had
1890 with compiling the Arg class on HP-UX.
1892 (e) Improve the error-handling of the C++ wrapper a little bit.
1894 (f) New tests for checking recursion limiting.
1896 8. The pcre_memmove() function, which is used only if the environment does not
1897 have a standard memmove() function (and is therefore rarely compiled),
1898 contained two bugs: (a) use of int instead of size_t, and (b) it was not
1899 returning a result (though PCRE never actually uses the result).
1901 9. In the POSIX regexec() interface, if nmatch is specified as a ridiculously
1902 large number - greater than INT_MAX/(3*sizeof(int)) - REG_ESPACE is
1903 returned instead of calling malloc() with an overflowing number that would
1904 most likely cause subsequent chaos.
1906 10. The debugging option of pcretest was not showing the NO_AUTO_CAPTURE flag.
1908 11. The POSIX flag REG_NOSUB is now supported. When a pattern that was compiled
1909 with this option is matched, the nmatch and pmatch options of regexec() are
1910 ignored.
1912 12. Added REG_UTF8 to the POSIX interface. This is not defined by POSIX, but is
1913 provided in case anyone wants to the the POSIX interface with UTF-8
1914 strings.
1916 13. Added CXXLDFLAGS to the Makefile parameters to provide settings only on the
1917 C++ linking (needed for some HP-UX environments).
1919 14. Avoid compiler warnings in get_ucpname() when compiled without UCP support
1920 (unused parameter) and in the pcre_printint() function (omitted "default"
1921 switch label when the default is to do nothing).
1923 15. Added some code to make it possible, when PCRE is compiled as a C++
1924 library, to replace subject pointers for pcre_exec() with a smart pointer
1925 class, thus making it possible to process discontinuous strings.
1927 16. The two macros PCRE_EXPORT and PCRE_DATA_SCOPE are confusing, and perform
1928 much the same function. They were added by different people who were trying
1929 to make PCRE easy to compile on non-Unix systems. It has been suggested
1930 that PCRE_EXPORT be abolished now that there is more automatic apparatus
1931 for compiling on Windows systems. I have therefore replaced it with
1932 PCRE_DATA_SCOPE. This is set automatically for Windows; if not set it
1933 defaults to "extern" for C or "extern C" for C++, which works fine on
1934 Unix-like systems. It is now possible to override the value of PCRE_DATA_
1935 SCOPE with something explicit in config.h. In addition:
1937 (a) pcreposix.h still had just "extern" instead of either of these macros;
1938 I have replaced it with PCRE_DATA_SCOPE.
1940 (b) Functions such as _pcre_xclass(), which are internal to the library,
1941 but external in the C sense, all had PCRE_EXPORT in their definitions.
1942 This is apparently wrong for the Windows case, so I have removed it.
1943 (It makes no difference on Unix-like systems.)
1945 17. Added a new limit, MATCH_LIMIT_RECURSION, which limits the depth of nesting
1946 of recursive calls to match(). This is different to MATCH_LIMIT because
1947 that limits the total number of calls to match(), not all of which increase
1948 the depth of recursion. Limiting the recursion depth limits the amount of
1949 stack (or heap if NO_RECURSE is set) that is used. The default can be set
1950 when PCRE is compiled, and changed at run time. A patch from Google adds
1951 this functionality to the C++ interface.
1953 18. Changes to the handling of Unicode character properties:
1955 (a) Updated the table to Unicode 4.1.0.
1957 (b) Recognize characters that are not in the table as "Cn" (undefined).
1959 (c) I revised the way the table is implemented to a much improved format
1960 which includes recognition of ranges. It now supports the ranges that
1961 are defined in UnicodeData.txt, and it also amalgamates other
1962 characters into ranges. This has reduced the number of entries in the
1963 table from around 16,000 to around 3,000, thus reducing its size
1964 considerably. I realized I did not need to use a tree structure after
1965 all - a binary chop search is just as efficient. Having reduced the
1966 number of entries, I extended their size from 6 bytes to 8 bytes to
1967 allow for more data.
1969 (d) Added support for Unicode script names via properties such as \p{Han}.
1971 19. In UTF-8 mode, a backslash followed by a non-Ascii character was not
1972 matching that character.
1974 20. When matching a repeated Unicode property with a minimum greater than zero,
1975 (for example \pL{2,}), PCRE could look past the end of the subject if it
1976 reached it while seeking the minimum number of characters. This could
1977 happen only if some of the characters were more than one byte long, because
1978 there is a check for at least the minimum number of bytes.
1980 21. Refactored the implementation of \p and \P so as to be more general, to
1981 allow for more different types of property in future. This has changed the
1982 compiled form incompatibly. Anybody with saved compiled patterns that use
1983 \p or \P will have to recompile them.
1985 22. Added "Any" and "L&" to the supported property types.
1987 23. Recognize \x{...} as a code point specifier, even when not in UTF-8 mode,
1988 but give a compile time error if the value is greater than 0xff.
1990 24. The man pages for pcrepartial, pcreprecompile, and pcre_compile2 were
1991 accidentally not being installed or uninstalled.
1993 25. The pcre.h file was built from pcre.h.in, but the only changes that were
1994 made were to insert the current release number. This seemed silly, because
1995 it made things harder for people building PCRE on systems that don't run
1996 "configure". I have turned pcre.h into a distributed file, no longer built
1997 by "configure", with the version identification directly included. There is
1998 no longer a pcre.h.in file.
2000 However, this change necessitated a change to the pcre-config script as
2001 well. It is built from pcre-config.in, and one of the substitutions was the
2002 release number. I have updated configure.ac so that ./configure now finds
2003 the release number by grepping pcre.h.
2005 26. Added the ability to run the tests under valgrind.
2008 Version 6.4 05-Sep-05
2009 ---------------------
2011 1. Change 6.0/10/(l) to pcregrep introduced a bug that caused separator lines
2012 "--" to be printed when multiple files were scanned, even when none of the
2013 -A, -B, or -C options were used. This is not compatible with Gnu grep, so I
2014 consider it to be a bug, and have restored the previous behaviour.
2016 2. A couple of code tidies to get rid of compiler warnings.
2018 3. The pcretest program used to cheat by referring to symbols in the library
2019 whose names begin with _pcre_. These are internal symbols that are not
2020 really supposed to be visible externally, and in some environments it is
2021 possible to suppress them. The cheating is now confined to including
2022 certain files from the library's source, which is a bit cleaner.
2024 4. Renamed pcre.in as pcre.h.in to go with pcrecpp.h.in; it also makes the
2025 file's purpose clearer.
2027 5. Reorganized pcre_ucp_findchar().
2030 Version 6.3 15-Aug-05
2031 ---------------------
2033 1. The file libpcre.pc.in did not have general read permission in the tarball.
2035 2. There were some problems when building without C++ support:
2037 (a) If C++ support was not built, "make install" and "make test" still
2038 tried to test it.
2040 (b) There were problems when the value of CXX was explicitly set. Some
2041 changes have been made to try to fix these, and ...
2043 (c) --disable-cpp can now be used to explicitly disable C++ support.
2045 (d) The use of @CPP_OBJ@ directly caused a blank line preceded by a
2046 backslash in a target when C++ was disabled. This confuses some
2047 versions of "make", apparently. Using an intermediate variable solves
2048 this. (Same for CPP_LOBJ.)
2050 3. $(LINK_FOR_BUILD) now includes $(CFLAGS_FOR_BUILD) and $(LINK)
2051 (non-Windows) now includes $(CFLAGS) because these flags are sometimes
2052 necessary on certain architectures.
2054 4. Added a setting of -export-symbols-regex to the link command to remove
2055 those symbols that are exported in the C sense, but actually are local
2056 within the library, and not documented. Their names all begin with
2057 "_pcre_". This is not a perfect job, because (a) we have to except some
2058 symbols that pcretest ("illegally") uses, and (b) the facility isn't always
2059 available (and never for static libraries). I have made a note to try to
2060 find a way round (a) in the future.
2063 Version 6.2 01-Aug-05
2064 ---------------------
2066 1. There was no test for integer overflow of quantifier values. A construction
2067 such as {1111111111111111} would give undefined results. What is worse, if
2068 a minimum quantifier for a parenthesized subpattern overflowed and became
2069 negative, the calculation of the memory size went wrong. This could have
2070 led to memory overwriting.
2072 2. Building PCRE using VPATH was broken. Hopefully it is now fixed.
2074 3. Added "b" to the 2nd argument of fopen() in dftables.c, for non-Unix-like
2075 operating environments where this matters.
2077 4. Applied Giuseppe Maxia's patch to add additional features for controlling
2078 PCRE options from within the C++ wrapper.
2080 5. Named capturing subpatterns were not being correctly counted when a pattern
2081 was compiled. This caused two problems: (a) If there were more than 100
2082 such subpatterns, the calculation of the memory needed for the whole
2083 compiled pattern went wrong, leading to an overflow error. (b) Numerical
2084 back references of the form \12, where the number was greater than 9, were
2085 not recognized as back references, even though there were sufficient
2086 previous subpatterns.
2088 6. Two minor patches to pcrecpp.cc in order to allow it to compile on older
2089 versions of gcc, e.g. 2.95.4.
2092 Version 6.1 21-Jun-05
2093 ---------------------
2095 1. There was one reference to the variable "posix" in pcretest.c that was not
2096 surrounded by "#if !defined NOPOSIX".
2098 2. Make it possible to compile pcretest without DFA support, UTF8 support, or
2099 the cross-check on the old pcre_info() function, for the benefit of the
2100 cut-down version of PCRE that is currently imported into Exim.
2102 3. A (silly) pattern starting with (?i)(?-i) caused an internal space
2103 allocation error. I've done the easy fix, which wastes 2 bytes for sensible
2104 patterns that start (?i) but I don't think that matters. The use of (?i) is
2105 just an example; this all applies to the other options as well.
2107 4. Since libtool seems to echo the compile commands it is issuing, the output
2108 from "make" can be reduced a bit by putting "@" in front of each libtool
2109 compile command.
2111 5. Patch from the folks at Google for configure.in to be a bit more thorough
2112 in checking for a suitable C++ installation before trying to compile the
2113 C++ stuff. This should fix a reported problem when a compiler was present,
2114 but no suitable headers.
2116 6. The man pages all had just "PCRE" as their title. I have changed them to
2117 be the relevant file name. I have also arranged that these names are
2118 retained in the file doc/pcre.txt, which is a concatenation in text format
2119 of all the man pages except the little individual ones for each function.
2121 7. The NON-UNIX-USE file had not been updated for the different set of source
2122 files that come with release 6. I also added a few comments about the C++
2123 wrapper.
2126 Version 6.0 07-Jun-05
2127 ---------------------
2129 1. Some minor internal re-organization to help with my DFA experiments.
2131 2. Some missing #ifdef SUPPORT_UCP conditionals in pcretest and printint that
2132 didn't matter for the library itself when fully configured, but did matter
2133 when compiling without UCP support, or within Exim, where the ucp files are
2134 not imported.
2136 3. Refactoring of the library code to split up the various functions into
2137 different source modules. The addition of the new DFA matching code (see
2138 below) to a single monolithic source would have made it really too
2139 unwieldy, quite apart from causing all the code to be include in a
2140 statically linked application, when only some functions are used. This is
2141 relevant even without the DFA addition now that patterns can be compiled in
2142 one application and matched in another.
2144 The downside of splitting up is that there have to be some external
2145 functions and data tables that are used internally in different modules of
2146 the library but which are not part of the API. These have all had their
2147 names changed to start with "_pcre_" so that they are unlikely to clash
2148 with other external names.
2150 4. Added an alternate matching function, pcre_dfa_exec(), which matches using
2151 a different (DFA) algorithm. Although it is slower than the original
2152 function, it does have some advantages for certain types of matching
2153 problem.
2155 5. Upgrades to pcretest in order to test the features of pcre_dfa_exec(),
2156 including restarting after a partial match.
2158 6. A patch for pcregrep that defines INVALID_FILE_ATTRIBUTES if it is not
2159 defined when compiling for Windows was sent to me. I have put it into the
2160 code, though I have no means of testing or verifying it.
2162 7. Added the pcre_refcount() auxiliary function.
2164 8. Added the PCRE_FIRSTLINE option. This constrains an unanchored pattern to
2165 match before or at the first newline in the subject string. In pcretest,
2166 the /f option on a pattern can be used to set this.
2168 9. A repeated \w when used in UTF-8 mode with characters greater than 256
2169 would behave wrongly. This has been present in PCRE since release 4.0.
2171 10. A number of changes to the pcregrep command:
2173 (a) Refactored how -x works; insert ^(...)$ instead of setting
2174 PCRE_ANCHORED and checking the length, in preparation for adding
2175 something similar for -w.
2177 (b) Added the -w (match as a word) option.
2179 (c) Refactored the way lines are read and buffered so as to have more
2180 than one at a time available.
2182 (d) Implemented a pcregrep test script.
2184 (e) Added the -M (multiline match) option. This allows patterns to match
2185 over several lines of the subject. The buffering ensures that at least
2186 8K, or the rest of the document (whichever is the shorter) is available
2187 for matching (and similarly the previous 8K for lookbehind assertions).
2189 (f) Changed the --help output so that it now says
2191 -w, --word-regex(p)
2193 instead of two lines, one with "regex" and the other with "regexp"
2194 because that confused at least one person since the short forms are the
2195 same. (This required a bit of code, as the output is generated
2196 automatically from a table. It wasn't just a text change.)
2198 (g) -- can be used to terminate pcregrep options if the next thing isn't an
2199 option but starts with a hyphen. Could be a pattern or a path name
2200 starting with a hyphen, for instance.
2202 (h) "-" can be given as a file name to represent stdin.
2204 (i) When file names are being printed, "(standard input)" is used for
2205 the standard input, for compatibility with GNU grep. Previously
2206 "<stdin>" was used.
2208 (j) The option --label=xxx can be used to supply a name to be used for
2209 stdin when file names are being printed. There is no short form.
2211 (k) Re-factored the options decoding logic because we are going to add
2212 two more options that take data. Such options can now be given in four
2213 different ways, e.g. "-fname", "-f name", "--file=name", "--file name".
2215 (l) Added the -A, -B, and -C options for requesting that lines of context
2216 around matches be printed.
2218 (m) Added the -L option to print the names of files that do not contain
2219 any matching lines, that is, the complement of -l.
2221 (n) The return code is 2 if any file cannot be opened, but pcregrep does
2222 continue to scan other files.
2224 (o) The -s option was incorrectly implemented. For compatibility with other
2225 greps, it now suppresses the error message for a non-existent or non-
2226 accessible file (but not the return code). There is a new option called
2227 -q that suppresses the output of matching lines, which was what -s was
2228 previously doing.
2230 (p) Added --include and --exclude options to specify files for inclusion
2231 and exclusion when recursing.
2233 11. The Makefile was not using the Autoconf-supported LDFLAGS macro properly.
2234 Hopefully, it now does.
2236 12. Missing cast in pcre_study().
2238 13. Added an "uninstall" target to the makefile.
2240 14. Replaced "extern" in the function prototypes in Makefile.in with
2241 "PCRE_DATA_SCOPE", which defaults to 'extern' or 'extern "C"' in the Unix
2242 world, but is set differently for Windows.
2244 15. Added a second compiling function called pcre_compile2(). The only
2245 difference is that it has an extra argument, which is a pointer to an
2246 integer error code. When there is a compile-time failure, this is set
2247 non-zero, in addition to the error test pointer being set to point to an
2248 error message. The new argument may be NULL if no error number is required
2249 (but then you may as well call pcre_compile(), which is now just a
2250 wrapper). This facility is provided because some applications need a
2251 numeric error indication, but it has also enabled me to tidy up the way
2252 compile-time errors are handled in the POSIX wrapper.
2254 16. Added VPATH=.libs to the makefile; this should help when building with one
2255 prefix path and installing with another. (Or so I'm told by someone who
2256 knows more about this stuff than I do.)
2258 17. Added a new option, REG_DOTALL, to the POSIX function regcomp(). This
2259 passes PCRE_DOTALL to the pcre_compile() function, making the "." character
2260 match everything, including newlines. This is not POSIX-compatible, but
2261 somebody wanted the feature. From pcretest it can be activated by using
2262 both the P and the s flags.
2264 18. AC_PROG_LIBTOOL appeared twice in Makefile.in. Removed one.
2266 19. libpcre.pc was being incorrectly installed as executable.
2268 20. A couple of places in pcretest check for end-of-line by looking for '\n';
2269 it now also looks for '\r' so that it will work unmodified on Windows.
2271 21. Added Google's contributed C++ wrapper to the distribution.
2273 22. Added some untidy missing memory free() calls in pcretest, to keep
2274 Electric Fence happy when testing.
2278 Version 5.0 13-Sep-04
2279 ---------------------
2281 1. Internal change: literal characters are no longer packed up into items
2282 containing multiple characters in a single byte-string. Each character
2283 is now matched using a separate opcode. However, there may be more than one
2284 byte in the character in UTF-8 mode.
2286 2. The pcre_callout_block structure has two new fields: pattern_position and
2287 next_item_length. These contain the offset in the pattern to the next match
2288 item, and its length, respectively.
2290 3. The PCRE_AUTO_CALLOUT option for pcre_compile() requests the automatic
2291 insertion of callouts before each pattern item. Added the /C option to
2292 pcretest to make use of this.
2294 4. On the advice of a Windows user, the lines
2296 #if defined(_WIN32) || defined(WIN32)
2297 _setmode( _fileno( stdout ), 0x8000 );
2298 #endif /* defined(_WIN32) || defined(WIN32) */
2300 have been added to the source of pcretest. This apparently does useful
2301 magic in relation to line terminators.
2303 5. Changed "r" and "w" in the calls to fopen() in pcretest to "rb" and "wb"
2304 for the benefit of those environments where the "b" makes a difference.
2306 6. The icc compiler has the same options as gcc, but "configure" doesn't seem
2307 to know about it. I have put a hack into configure.in that adds in code
2308 to set GCC=yes if CC=icc. This seems to end up at a point in the
2309 generated configure script that is early enough to affect the setting of
2310 compiler options, which is what is needed, but I have no means of testing
2311 whether it really works. (The user who reported this had patched the
2312 generated configure script, which of course I cannot do.)
2314 LATER: After change 22 below (new libtool files), the configure script
2315 seems to know about icc (and also ecc). Therefore, I have commented out
2316 this hack in configure.in.
2318 7. Added support for pkg-config (2 patches were sent in).
2320 8. Negated POSIX character classes that used a combination of internal tables
2321 were completely broken. These were [[:^alpha:]], [[:^alnum:]], and
2322 [[:^ascii]]. Typically, they would match almost any characters. The other
2323 POSIX classes were not broken in this way.
2325 9. Matching the pattern "\b.*?" against "ab cd", starting at offset 1, failed
2326 to find the match, as PCRE was deluded into thinking that the match had to
2327 start at the start point or following a newline. The same bug applied to
2328 patterns with negative forward assertions or any backward assertions
2329 preceding ".*" at the start, unless the pattern required a fixed first
2330 character. This was a failing pattern: "(?!.bcd).*". The bug is now fixed.
2332 10. In UTF-8 mode, when moving forwards in the subject after a failed match
2333 starting at the last subject character, bytes beyond the end of the subject
2334 string were read.
2336 11. Renamed the variable "class" as "classbits" to make life easier for C++
2337 users. (Previously there was a macro definition, but it apparently wasn't
2338 enough.)
2340 12. Added the new field "tables" to the extra data so that tables can be passed
2341 in at exec time, or the internal tables can be re-selected. This allows
2342 a compiled regex to be saved and re-used at a later time by a different
2343 program that might have everything at different addresses.
2345 13. Modified the pcre-config script so that, when run on Solaris, it shows a
2346 -R library as well as a -L library.
2348 14. The debugging options of pcretest (-d on the command line or D on a
2349 pattern) showed incorrect output for anything following an extended class
2350 that contained multibyte characters and which was followed by a quantifier.
2352 15. Added optional support for general category Unicode character properties
2353 via the \p, \P, and \X escapes. Unicode property support implies UTF-8
2354 support. It adds about 90K to the size of the library. The meanings of the
2355 inbuilt class escapes such as \d and \s have NOT been changed.
2357 16. Updated pcredemo.c to include calls to free() to release the memory for the
2358 compiled pattern.
2360 17. The generated file chartables.c was being created in the source directory
2361 instead of in the building directory. This caused the build to fail if the
2362 source directory was different from the building directory, and was
2363 read-only.
2365 18. Added some sample Win commands from Mark Tetrode into the NON-UNIX-USE
2366 file. No doubt somebody will tell me if they don't make sense... Also added
2367 Dan Mooney's comments about building on OpenVMS.
2369 19. Added support for partial matching via the PCRE_PARTIAL option for
2370 pcre_exec() and the \P data escape in pcretest.
2372 20. Extended pcretest with 3 new pattern features:
2374 (i) A pattern option of the form ">rest-of-line" causes pcretest to
2375 write the compiled pattern to the file whose name is "rest-of-line".
2376 This is a straight binary dump of the data, with the saved pointer to
2377 the character tables forced to be NULL. The study data, if any, is
2378 written too. After writing, pcretest reads a new pattern.
2380 (ii) If, instead of a pattern, "<rest-of-line" is given, pcretest reads a
2381 compiled pattern from the given file. There must not be any
2382 occurrences of "<" in the file name (pretty unlikely); if there are,
2383 pcretest will instead treat the initial "<" as a pattern delimiter.
2384 After reading in the pattern, pcretest goes on to read data lines as
2385 usual.
2387 (iii) The F pattern option causes pcretest to flip the bytes in the 32-bit
2388 and 16-bit fields in a compiled pattern, to simulate a pattern that
2389 was compiled on a host of opposite endianness.
2391 21. The pcre-exec() function can now cope with patterns that were compiled on
2392 hosts of opposite endianness, with this restriction:
2394 As for any compiled expression that is saved and used later, the tables
2395 pointer field cannot be preserved; the extra_data field in the arguments
2396 to pcre_exec() should be used to pass in a tables address if a value
2397 other than the default internal tables were used at compile time.
2399 22. Calling pcre_exec() with a negative value of the "ovecsize" parameter is
2400 now diagnosed as an error. Previously, most of the time, a negative number
2401 would have been treated as zero, but if in addition "ovector" was passed as
2402 NULL, a crash could occur.
2404 23. Updated the files ltmain.sh, config.sub, config.guess, and aclocal.m4 with
2405 new versions from the libtool 1.5 distribution (the last one is a copy of
2406 a file called libtool.m4). This seems to have fixed the need to patch
2407 "configure" to support Darwin 1.3 (which I used to do). However, I still
2408 had to patch ltmain.sh to ensure that ${SED} is set (it isn't on my
2409 workstation).
2411 24. Changed the PCRE licence to be the more standard "BSD" licence.
2414 Version 4.5 01-Dec-03
2415 ---------------------
2417 1. There has been some re-arrangement of the code for the match() function so
2418 that it can be compiled in a version that does not call itself recursively.
2419 Instead, it keeps those local variables that need separate instances for
2420 each "recursion" in a frame on the heap, and gets/frees frames whenever it
2421 needs to "recurse". Keeping track of where control must go is done by means
2422 of setjmp/longjmp. The whole thing is implemented by a set of macros that
2423 hide most of the details from the main code, and operates only if
2424 NO_RECURSE is defined while compiling pcre.c. If PCRE is built using the
2425 "configure" mechanism, "--disable-stack-for-recursion" turns on this way of
2426 operating.
2428 To make it easier for callers to provide specially tailored get/free
2429 functions for this usage, two new functions, pcre_stack_malloc, and
2430 pcre_stack_free, are used. They are always called in strict stacking order,
2431 and the size of block requested is always the same.
2433 The PCRE_CONFIG_STACKRECURSE info parameter can be used to find out whether
2434 PCRE has been compiled to use the stack or the heap for recursion. The
2435 -C option of pcretest uses this to show which version is compiled.
2437 A new data escape \S, is added to pcretest; it causes the amounts of store
2438 obtained and freed by both kinds of malloc/free at match time to be added
2439 to the output.
2441 2. Changed the locale test to use "fr_FR" instead of "fr" because that's
2442 what's available on my current Linux desktop machine.
2444 3. When matching a UTF-8 string, the test for a valid string at the start has
2445 been extended. If start_offset is not zero, PCRE now checks that it points
2446 to a byte that is the start of a UTF-8 character. If not, it returns
2447 PCRE_ERROR_BADUTF8_OFFSET (-11). Note: the whole string is still checked;
2448 this is necessary because there may be backward assertions in the pattern.
2449 When matching the same subject several times, it may save resources to use
2450 PCRE_NO_UTF8_CHECK on all but the first call if the string is long.
2452 4. The code for checking the validity of UTF-8 strings has been tightened so
2453 that it rejects (a) strings containing 0xfe or 0xff bytes and (b) strings
2454 containing "overlong sequences".
2456 5. Fixed a bug (appearing twice) that I could not find any way of exploiting!
2457 I had written "if ((digitab[*p++] && chtab_digit) == 0)" where the "&&"
2458 should have been "&", but it just so happened that all the cases this let
2459 through by mistake were picked up later in the function.
2461 6. I had used a variable called "isblank" - this is a C99 function, causing
2462 some compilers to warn. To avoid this, I renamed it (as "blankclass").
2464 7. Cosmetic: (a) only output another newline at the end of pcretest if it is
2465 prompting; (b) run "./pcretest /dev/null" at the start of the test script
2466 so the version is shown; (c) stop "make test" echoing "./RunTest".
2468 8. Added patches from David Burgess to enable PCRE to run on EBCDIC systems.
2470 9. The prototype for memmove() for systems that don't have it was using
2471 size_t, but the inclusion of the header that defines size_t was later. I've
2472 moved the #includes for the C headers earlier to avoid this.
2474 10. Added some adjustments to the code to make it easier to compiler on certain
2475 special systems:
2477 (a) Some "const" qualifiers were missing.
2478 (b) Added the macro EXPORT before all exported functions; by default this
2479 is defined to be empty.
2480 (c) Changed the dftables auxiliary program (that builds chartables.c) so
2481 that it reads its output file name as an argument instead of writing
2482 to the standard output and assuming this can be redirected.
2484 11. In UTF-8 mode, if a recursive reference (e.g. (?1)) followed a character
2485 class containing characters with values greater than 255, PCRE compilation
2486 went into a loop.
2488 12. A recursive reference to a subpattern that was within another subpattern
2489 that had a minimum quantifier of zero caused PCRE to crash. For example,
2490 (x(y(?2))z)? provoked this bug with a subject that got as far as the
2491 recursion. If the recursively-called subpattern itself had a zero repeat,
2492 that was OK.
2494 13. In pcretest, the buffer for reading a data line was set at 30K, but the
2495 buffer into which it was copied (for escape processing) was still set at
2496 1024, so long lines caused crashes.
2498 14. A pattern such as /[ab]{1,3}+/ failed to compile, giving the error
2499 "internal error: code overflow...". This applied to any character class
2500 that was followed by a possessive quantifier.
2502 15. Modified the Makefile to add libpcre.la as a prerequisite for
2503 libpcreposix.la because I was told this is needed for a parallel build to
2504 work.
2506 16. If a pattern that contained .* following optional items at the start was
2507 studied, the wrong optimizing data was generated, leading to matching
2508 errors. For example, studying /[ab]*.*c/ concluded, erroneously, that any
2509 matching string must start with a or b or c. The correct conclusion for
2510 this pattern is that a match can start with any character.
2513 Version 4.4 13-Aug-03
2514 ---------------------
2516 1. In UTF-8 mode, a character class containing characters with values between
2517 127 and 255 was not handled correctly if the compiled pattern was studied.
2518 In fixing this, I have also improved the studying algorithm for such
2519 classes (slightly).
2521 2. Three internal functions had redundant arguments passed to them. Removal
2522 might give a very teeny performance improvement.
2524 3. Documentation bug: the value of the capture_top field in a callout is *one
2525 more than* the number of the hightest numbered captured substring.
2527 4. The Makefile linked pcretest and pcregrep with -lpcre, which could result
2528 in incorrectly linking with a previously installed version. They now link
2529 explicitly with libpcre.la.
2531 5. configure.in no longer needs to recognize Cygwin specially.
2533 6. A problem in pcre.in for Windows platforms is fixed.
2535 7. If a pattern was successfully studied, and the -d (or /D) flag was given to
2536 pcretest, it used to include the size of the study block as part of its
2537 output. Unfortunately, the structure contains a field that has a different
2538 size on different hardware architectures. This meant that the tests that
2539 showed this size failed. As the block is currently always of a fixed size,
2540 this information isn't actually particularly useful in pcretest output, so
2541 I have just removed it.
2543 8. Three pre-processor statements accidentally did not start in column 1.
2544 Sadly, there are *still* compilers around that complain, even though
2545 standard C has not required this for well over a decade. Sigh.
2547 9. In pcretest, the code for checking callouts passed small integers in the
2548 callout_data field, which is a void * field. However, some picky compilers
2549 complained about the casts involved for this on 64-bit systems. Now
2550 pcretest passes the address of the small integer instead, which should get
2551 rid of the warnings.
2553 10. By default, when in UTF-8 mode, PCRE now checks for valid UTF-8 strings at
2554 both compile and run time, and gives an error if an invalid UTF-8 sequence
2555 is found. There is a option for disabling this check in cases where the
2556 string is known to be correct and/or the maximum performance is wanted.
2558 11. In response to a bug report, I changed one line in Makefile.in from
2560 -Wl,--out-implib,.libs/lib@WIN_PREFIX@pcreposix.dll.a \
2561 to
2562 -Wl,--out-implib,.libs/@WIN_PREFIX@libpcreposix.dll.a \
2564 to look similar to other lines, but I have no way of telling whether this
2565 is the right thing to do, as I do not use Windows. No doubt I'll get told
2566 if it's wrong...
2569 Version 4.3 21-May-03
2570 ---------------------
2572 1. Two instances of @WIN_PREFIX@ omitted from the Windows targets in the
2573 Makefile.
2575 2. Some refactoring to improve the quality of the code:
2577 (i) The utf8_table... variables are now declared "const".
2579 (ii) The code for \cx, which used the "case flipping" table to upper case
2580 lower case letters, now just substracts 32. This is ASCII-specific,
2581 but the whole concept of \cx is ASCII-specific, so it seems
2582 reasonable.
2584 (iii) PCRE was using its character types table to recognize decimal and
2585 hexadecimal digits in the pattern. This is silly, because it handles
2586 only 0-9, a-f, and A-F, but the character types table is locale-
2587 specific, which means strange things might happen. A private
2588 table is now used for this - though it costs 256 bytes, a table is
2589 much faster than multiple explicit tests. Of course, the standard
2590 character types table is still used for matching digits in subject
2591 strings against \d.
2593 (iv) Strictly, the identifier ESC_t is reserved by POSIX (all identifiers
2594 ending in _t are). So I've renamed it as ESC_tee.
2596 3. The first argument for regexec() in the POSIX wrapper should have been
2597 defined as "const".
2599 4. Changed pcretest to use malloc() for its buffers so that they can be
2600 Electric Fenced for debugging.
2602 5. There were several places in the code where, in UTF-8 mode, PCRE would try
2603 to read one or more bytes before the start of the subject string. Often this
2604 had no effect on PCRE's behaviour, but in some circumstances it could
2605 provoke a segmentation fault.
2607 6. A lookbehind at the start of a pattern in UTF-8 mode could also cause PCRE
2608 to try to read one or more bytes before the start of the subject string.
2610 7. A lookbehind in a pattern matched in non-UTF-8 mode on a PCRE compiled with
2611 UTF-8 support could misbehave in various ways if the subject string
2612 contained bytes with the 0x80 bit set and the 0x40 bit unset in a lookbehind
2613 area. (PCRE was not checking for the UTF-8 mode flag, and trying to move
2614 back over UTF-8 characters.)
2617 Version 4.2 14-Apr-03
2618 ---------------------
2620 1. Typo "#if SUPPORT_UTF8" instead of "#ifdef SUPPORT_UTF8" fixed.
2622 2. Changes to the building process, supplied by Ronald Landheer-Cieslak
2623 [ON_WINDOWS]: new variable, "#" on non-Windows platforms
2624 [NOT_ON_WINDOWS]: new variable, "#" on Windows platforms
2625 [WIN_PREFIX]: new variable, "cyg" for Cygwin
2626 * Makefile.in: use autoconf substitution for OBJEXT, EXEEXT, BUILD_OBJEXT
2628 Note: automatic setting of the BUILD variables is not yet working
2629 set CPPFLAGS and BUILD_CPPFLAGS (but don't use yet) - should be used at
2630 compile-time but not at link-time
2631 [LINK]: use for linking executables only
2632 make different versions for Windows and non-Windows
2633 [LINKLIB]: new variable, copy of UNIX-style LINK, used for linking
2634 libraries
2635 [LINK_FOR_BUILD]: new variable
2636 [OBJEXT]: use throughout
2637 [EXEEXT]: use throughout
2638 <winshared>: new target
2639 <wininstall>: new target
2640 <dftables.o>: use native compiler
2641 <dftables>: use native linker
2642 <install>: handle Windows platform correctly
2643 <clean>: ditto
2644 <check>: ditto
2645 copy DLL to top builddir before testing
2647 As part of these changes, -no-undefined was removed again. This was reported
2648 to give trouble on HP-UX 11.0, so getting rid of it seems like a good idea
2649 in any case.
2651 3. Some tidies to get rid of compiler warnings:
2653 . In the match_data structure, match_limit was an unsigned long int, whereas
2654 match_call_count was an int. I've made them both unsigned long ints.
2656 . In pcretest the fact that a const uschar * doesn't automatically cast to
2657 a void * provoked a warning.
2659 . Turning on some more compiler warnings threw up some "shadow" variables
2660 and a few more missing casts.
2662 4. If PCRE was complied with UTF-8 support, but called without the PCRE_UTF8
2663 option, a class that contained a single character with a value between 128
2664 and 255 (e.g. /[\xFF]/) caused PCRE to crash.
2666 5. If PCRE was compiled with UTF-8 support, but called without the PCRE_UTF8
2667 option, a class that contained several characters, but with at least one
2668 whose value was between 128 and 255 caused PCRE to crash.
2671 Version 4.1 12-Mar-03
2672 ---------------------
2674 1. Compiling with gcc -pedantic found a couple of places where casts were
2675 needed, and a string in dftables.c that was longer than standard compilers are
2676 required to support.
2678 2. Compiling with Sun's compiler found a few more places where the code could
2679 be tidied up in order to avoid warnings.
2681 3. The variables for cross-compiling were called HOST_CC and HOST_CFLAGS; the
2682 first of these names is deprecated in the latest Autoconf in favour of the name
2683 CC_FOR_BUILD, because "host" is typically used to mean the system on which the
2684 compiled code will be run. I can't find a reference for HOST_CFLAGS, but by
2685 analogy I have changed it to CFLAGS_FOR_BUILD.
2687 4. Added -no-undefined to the linking command in the Makefile, because this is
2688 apparently helpful for Windows. To make it work, also added "-L. -lpcre" to the
2689 linking step for the pcreposix library.
2691 5. PCRE was failing to diagnose the case of two named groups with the same
2692 name.
2694 6. A problem with one of PCRE's optimizations was discovered. PCRE remembers a
2695 literal character that is needed in the subject for a match, and scans along to
2696 ensure that it is present before embarking on the full matching process. This
2697 saves time in cases of nested unlimited repeats that are never going to match.
2698 Problem: the scan can take a lot of time if the subject is very long (e.g.
2699 megabytes), thus penalizing straightforward matches. It is now done only if the
2700 amount of subject to be scanned is less than 1000 bytes.
2702 7. A lesser problem with the same optimization is that it was recording the
2703 first character of an anchored pattern as "needed", thus provoking a search
2704 right along the subject, even when the first match of the pattern was going to
2705 fail. The "needed" character is now not set for anchored patterns, unless it
2706 follows something in the pattern that is of non-fixed length. Thus, it still
2707 fulfils its original purpose of finding quick non-matches in cases of nested
2708 unlimited repeats, but isn't used for simple anchored patterns such as /^abc/.
2711 Version 4.0 17-Feb-03
2712 ---------------------
2714 1. If a comment in an extended regex that started immediately after a meta-item
2715 extended to the end of string, PCRE compiled incorrect data. This could lead to
2716 all kinds of weird effects. Example: /#/ was bad; /()#/ was bad; /a#/ was not.
2718 2. Moved to autoconf 2.53 and libtool 1.4.2.
2720 3. Perl 5.8 no longer needs "use utf8" for doing UTF-8 things. Consequently,
2721 the special perltest8 script is no longer needed - all the tests can be run
2722 from a single perltest script.
2724 4. From 5.004, Perl has not included the VT character (0x0b) in the set defined
2725 by \s. It has now been removed in PCRE. This means it isn't recognized as
2726 whitespace in /x regexes too, which is the same as Perl. Note that the POSIX
2727 class [:space:] *does* include VT, thereby creating a mess.
2729 5. Added the class [:blank:] (a GNU extension from Perl 5.8) to match only
2730 space and tab.
2732 6. Perl 5.005 was a long time ago. It's time to amalgamate the tests that use
2733 its new features into the main test script, reducing the number of scripts.
2735 7. Perl 5.8 has changed the meaning of patterns like /a(?i)b/. Earlier versions
2736 were backward compatible, and made the (?i) apply to the whole pattern, as if
2737 /i were given. Now it behaves more logically, and applies the option setting
2738 only to what follows. PCRE has been changed to follow suit. However, if it
2739 finds options settings right at the start of the pattern, it extracts them into
2740 the global options, as before. Thus, they show up in the info data.
2742 8. Added support for the \Q...\E escape sequence. Characters in between are
2743 treated as literals. This is slightly different from Perl in that $ and @ are
2744 also handled as literals inside the quotes. In Perl, they will cause variable
2745 interpolation. Note the following examples:
2747 Pattern PCRE matches Perl matches
2749 \Qabc$xyz\E abc$xyz abc followed by the contents of $xyz
2750 \Qabc\$xyz\E abc\$xyz abc\$xyz
2751 \Qabc\E\$\Qxyz\E abc$xyz abc$xyz
2753 For compatibility with Perl, \Q...\E sequences are recognized inside character
2754 classes as well as outside them.
2756 9. Re-organized 3 code statements in pcretest to avoid "overflow in
2757 floating-point constant arithmetic" warnings from a Microsoft compiler. Added a
2758 (size_t) cast to one statement in pcretest and one in pcreposix to avoid
2759 signed/unsigned warnings.
2761 10. SunOS4 doesn't have strtoul(). This was used only for unpicking the -o
2762 option for pcretest, so I've replaced it by a simple function that does just
2763 that job.
2765 11. pcregrep was ending with code 0 instead of 2 for the commands "pcregrep" or
2766 "pcregrep -".
2768 12. Added "possessive quantifiers" ?+, *+, ++, and {,}+ which come from Sun's
2769 Java package. This provides some syntactic sugar for simple cases of what my
2770 documentation calls "once-only subpatterns". A pattern such as x*+ is the same
2771 as (?>x*). In other words, if what is inside (?>...) is just a single repeated
2772 item, you can use this simplified notation. Note that only makes sense with
2773 greedy quantifiers. Consequently, the use of the possessive quantifier forces
2774 greediness, whatever the setting of the PCRE_UNGREEDY option.
2776 13. A change of greediness default within a pattern was not taking effect at
2777 the current level for patterns like /(b+(?U)a+)/. It did apply to parenthesized
2778 subpatterns that followed. Patterns like /b+(?U)a+/ worked because the option
2779 was abstracted outside.
2781 14. PCRE now supports the \G assertion. It is true when the current matching
2782 position is at the start point of the match. This differs from \A when the
2783 starting offset is non-zero. Used with the /g option of pcretest (or similar
2784 code), it works in the same way as it does for Perl's /g option. If all
2785 alternatives of a regex begin with \G, the expression is anchored to the start
2786 match position, and the "anchored" flag is set in the compiled expression.
2788 15. Some bugs concerning the handling of certain option changes within patterns
2789 have been fixed. These applied to options other than (?ims). For example,
2790 "a(?x: b c )d" did not match "XabcdY" but did match "Xa b c dY". It should have
2791 been the other way round. Some of this was related to change 7 above.
2793 16. PCRE now gives errors for /[.x.]/ and /[=x=]/ as unsupported POSIX
2794 features, as Perl does. Previously, PCRE gave the warnings only for /[[.x.]]/
2795 and /[[=x=]]/. PCRE now also gives an error for /[:name:]/ because it supports
2796 POSIX classes only within a class (e.g. /[[:alpha:]]/).
2798 17. Added support for Perl's \C escape. This matches one byte, even in UTF8
2799 mode. Unlike ".", it always matches newline, whatever the setting of
2800 PCRE_DOTALL. However, PCRE does not permit \C to appear in lookbehind
2801 assertions. Perl allows it, but it doesn't (in general) work because it can't
2802 calculate the length of the lookbehind. At least, that's the case for Perl
2803 5.8.0 - I've been told they are going to document that it doesn't work in
2804 future.
2806 18. Added an error diagnosis for escapes that PCRE does not support: these are
2807 \L, \l, \N, \P, \p, \U, \u, and \X.
2809 19. Although correctly diagnosing a missing ']' in a character class, PCRE was
2810 reading past the end of the pattern in cases such as /[abcd/.
2812 20. PCRE was getting more memory than necessary for patterns with classes that
2813 contained both POSIX named classes and other characters, e.g. /[[:space:]abc/.
2815 21. Added some code, conditional on #ifdef VPCOMPAT, to make life easier for
2816 compiling PCRE for use with Virtual Pascal.
2818 22. Small fix to the Makefile to make it work properly if the build is done
2819 outside the source tree.
2821 23. Added a new extension: a condition to go with recursion. If a conditional
2822 subpattern starts with (?(R) the "true" branch is used if recursion has
2823 happened, whereas the "false" branch is used only at the top level.
2825 24. When there was a very long string of literal characters (over 255 bytes
2826 without UTF support, over 250 bytes with UTF support), the computation of how
2827 much memory was required could be incorrect, leading to segfaults or other
2828 strange effects.
2830 25. PCRE was incorrectly assuming anchoring (either to start of subject or to
2831 start of line for a non-DOTALL pattern) when a pattern started with (.*) and
2832 there was a subsequent back reference to those brackets. This meant that, for
2833 example, /(.*)\d+\1/ failed to match "abc123bc". Unfortunately, it isn't
2834 possible to check for precisely this case. All we can do is abandon the
2835 optimization if .* occurs inside capturing brackets when there are any back
2836 references whatsoever. (See below for a better fix that came later.)
2838 26. The handling of the optimization for finding the first character of a
2839 non-anchored pattern, and for finding a character that is required later in the
2840 match were failing in some cases. This didn't break the matching; it just
2841 failed to optimize when it could. The way this is done has been re-implemented.
2843 27. Fixed typo in error message for invalid (?R item (it said "(?p").
2845 28. Added a new feature that provides some of the functionality that Perl
2846 provides with (?{...}). The facility is termed a "callout". The way it is done
2847 in PCRE is for the caller to provide an optional function, by setting
2848 pcre_callout to its entry point. Like pcre_malloc and pcre_free, this is a
2849 global variable. By default it is unset, which disables all calling out. To get
2850 the function called, the regex must include (?C) at appropriate points. This
2851 is, in fact, equivalent to (?C0), and any number <= 255 may be given with (?C).
2852 This provides a means of identifying different callout points. When PCRE
2853 reaches such a point in the regex, if pcre_callout has been set, the external
2854 function is called. It is provided with data in a structure called
2855 pcre_callout_block, which is defined in pcre.h. If the function returns 0,
2856 matching continues; if it returns a non-zero value, the match at the current
2857 point fails. However, backtracking will occur if possible. [This was changed
2858 later and other features added - see item 49 below.]
2860 29. pcretest is upgraded to test the callout functionality. It provides a
2861 callout function that displays information. By default, it shows the start of
2862 the match and the current position in the text. There are some new data escapes
2863 to vary what happens:
2865 \C+ in addition, show current contents of captured substrings
2866 \C- do not supply a callout function
2867 \C!n return 1 when callout number n is reached
2868 \C!n!m return 1 when callout number n is reached for the mth time
2870 30. If pcregrep was called with the -l option and just a single file name, it
2871 output "<stdin>" if a match was found, instead of the file name.
2873 31. Improve the efficiency of the POSIX API to PCRE. If the number of capturing
2874 slots is less than POSIX_MALLOC_THRESHOLD, use a block on the stack to pass to
2875 pcre_exec(). This saves a malloc/free per call. The default value of
2876 POSIX_MALLOC_THRESHOLD is 10; it can be changed by --with-posix-malloc-threshold
2877 when configuring.
2879 32. The default maximum size of a compiled pattern is 64K. There have been a
2880 few cases of people hitting this limit. The code now uses macros to handle the
2881 storing of links as offsets within the compiled pattern. It defaults to 2-byte
2882 links, but this can be changed to 3 or 4 bytes by --with-link-size when
2883 configuring. Tests 2 and 5 work only with 2-byte links because they output
2884 debugging information about compiled patterns.
2886 33. Internal code re-arrangements:
2888 (a) Moved the debugging function for printing out a compiled regex into
2889 its own source file (printint.c) and used #include to pull it into
2890 pcretest.c and, when DEBUG is defined, into pcre.c, instead of having two
2891 separate copies.
2893 (b) Defined the list of op-code names for debugging as a macro in
2894 internal.h so that it is next to the definition of the opcodes.
2896 (c) Defined a table of op-code lengths for simpler skipping along compiled
2897 code. This is again a macro in internal.h so that it is next to the
2898 definition of the opcodes.
2900 34. Added support for recursive calls to individual subpatterns, along the
2901 lines of Robin Houston's patch (but implemented somewhat differently).
2903 35. Further mods to the Makefile to help Win32. Also, added code to pcregrep to
2904 allow it to read and process whole directories in Win32. This code was
2905 contributed by Lionel Fourquaux; it has not been tested by me.
2907 36. Added support for named subpatterns. The Python syntax (?P<name>...) is
2908 used to name a group. Names consist of alphanumerics and underscores, and must
2909 be unique. Back references use the syntax (?P=name) and recursive calls use
2910 (?P>name) which is a PCRE extension to the Python extension. Groups still have
2911 numbers. The function pcre_fullinfo() can be used after compilation to extract
2912 a name/number map. There are three relevant calls:
2914 PCRE_INFO_NAMEENTRYSIZE yields the size of each entry in the map
2915 PCRE_INFO_NAMECOUNT yields the number of entries
2916 PCRE_INFO_NAMETABLE yields a pointer to the map.
2918 The map is a vector of fixed-size entries. The size of each entry depends on
2919 the length of the longest name used. The first two bytes of each entry are the
2920 group number, most significant byte first. There follows the corresponding
2921 name, zero terminated. The names are in alphabetical order.
2923 37. Make the maximum literal string in the compiled code 250 for the non-UTF-8
2924 case instead of 255. Making it the same both with and without UTF-8 support
2925 means that the same test output works with both.
2927 38. There was a case of malloc(0) in the POSIX testing code in pcretest. Avoid
2928 calling malloc() with a zero argument.
2930 39. Change 25 above had to resort to a heavy-handed test for the .* anchoring
2931 optimization. I've improved things by keeping a bitmap of backreferences with
2932 numbers 1-31 so that if .* occurs inside capturing brackets that are not in
2933 fact referenced, the optimization can be applied. It is unlikely that a
2934 relevant occurrence of .* (i.e. one which might indicate anchoring or forcing
2935 the match to follow \n) will appear inside brackets with a number greater than
2936 31, but if it does, any back reference > 31 suppresses the optimization.
2938 40. Added a new compile-time option PCRE_NO_AUTO_CAPTURE. This has the effect
2939 of disabling numbered capturing parentheses. Any opening parenthesis that is
2940 not followed by ? behaves as if it were followed by ?: but named parentheses
2941 can still be used for capturing (and they will acquire numbers in the usual
2942 way).
2944 41. Redesigned the return codes from the match() function into yes/no/error so
2945 that errors can be passed back from deep inside the nested calls. A malloc
2946 failure while inside a recursive subpattern call now causes the
2947 PCRE_ERROR_NOMEMORY return instead of quietly going wrong.
2949 42. It is now possible to set a limit on the number of times the match()
2950 function is called in a call to pcre_exec(). This facility makes it possible to
2951 limit the amount of recursion and backtracking, though not in a directly
2952 obvious way, because the match() function is used in a number of different
2953 circumstances. The count starts from zero for each position in the subject
2954 string (for non-anchored patterns). The default limit is, for compatibility, a
2955 large number, namely 10 000 000. You can change this in two ways:
2957 (a) When configuring PCRE before making, you can use --with-match-limit=n
2958 to set a default value for the compiled library.
2960 (b) For each call to pcre_exec(), you can pass a pcre_extra block in which
2961 a different value is set. See 45 below.
2963 If the limit is exceeded, pcre_exec() returns PCRE_ERROR_MATCHLIMIT.
2965 43. Added a new function pcre_config(int, void *) to enable run-time extraction
2966 of things that can be changed at compile time. The first argument specifies
2967 what is wanted and the second points to where the information is to be placed.
2968 The current list of available information is:
2972 The output is an integer that is set to one if UTF-8 support is available;
2973 otherwise it is set to zero.
2977 The output is an integer that it set to the value of the code that is used for
2978 newline. It is either LF (10) or CR (13).
2982 The output is an integer that contains the number of bytes used for internal
2983 linkage in compiled expressions. The value is 2, 3, or 4. See item 32 above.
2987 The output is an integer that contains the threshold above which the POSIX
2988 interface uses malloc() for output vectors. See item 31 above.
2992 The output is an unsigned integer that contains the default limit of the number
2993 of match() calls in a pcre_exec() execution. See 42 above.
2995 44. pcretest has been upgraded by the addition of the -C option. This causes it
2996 to extract all the available output from the new pcre_config() function, and to
2997 output it. The program then exits immediately.
2999 45. A need has arisen to pass over additional data with calls to pcre_exec() in
3000 order to support additional features. One way would have been to define
3001 pcre_exec2() (for example) with extra arguments, but this would not have been
3002 extensible, and would also have required all calls to the original function to
3003 be mapped to the new one. Instead, I have chosen to extend the mechanism that
3004 is used for passing in "extra" data from pcre_study().
3006 The pcre_extra structure is now exposed and defined in pcre.h. It currently
3007 contains the following fields:
3009 flags a bitmap indicating which of the following fields are set
3010 study_data opaque data from pcre_study()
3011 match_limit a way of specifying a limit on match() calls for a specific
3012 call to pcre_exec()
3013 callout_data data for callouts (see 49 below)
3015 The flag bits are also defined in pcre.h, and are
3021 The pcre_study() function now returns one of these new pcre_extra blocks, with
3022 the actual study data pointed to by the study_data field, and the
3023 PCRE_EXTRA_STUDY_DATA flag set. This can be passed directly to pcre_exec() as
3024 before. That is, this change is entirely upwards-compatible and requires no
3025 change to existing code.
3027 If you want to pass in additional data to pcre_exec(), you can either place it
3028 in a pcre_extra block provided by pcre_study(), or create your own pcre_extra
3029 block.
3031 46. pcretest has been extended to test the PCRE_EXTRA_MATCH_LIMIT feature. If a
3032 data string contains the escape sequence \M, pcretest calls pcre_exec() several
3033 times with different match limits, until it finds the minimum value needed for
3034 pcre_exec() to complete. The value is then output. This can be instructive; for
3035 most simple matches the number is quite small, but for pathological cases it
3036 gets very large very quickly.
3038 47. There's a new option for pcre_fullinfo() called PCRE_INFO_STUDYSIZE. It
3039 returns the size of the data block pointed to by the study_data field in a
3040 pcre_extra block, that is, the value that was passed as the argument to
3041 pcre_malloc() when PCRE was getting memory in which to place the information
3042 created by pcre_study(). The fourth argument should point to a size_t variable.
3043 pcretest has been extended so that this information is shown after a successful
3044 pcre_study() call when information about the compiled regex is being displayed.
3046 48. Cosmetic change to Makefile: there's no need to have / after $(DESTDIR)
3047 because what follows is always an absolute path. (Later: it turns out that this
3048 is more than cosmetic for MinGW, because it doesn't like empty path
3049 components.)
3051 49. Some changes have been made to the callout feature (see 28 above):
3053 (i) A callout function now has three choices for what it returns:
3055 0 => success, carry on matching
3056 > 0 => failure at this point, but backtrack if possible
3057 < 0 => serious error, return this value from pcre_exec()
3059 Negative values should normally be chosen from the set of PCRE_ERROR_xxx
3060 values. In particular, returning PCRE_ERROR_NOMATCH forces a standard
3061 "match failed" error. The error number PCRE_ERROR_CALLOUT is reserved for
3062 use by callout functions. It will never be used by PCRE itself.
3064 (ii) The pcre_extra structure (see 45 above) has a void * field called
3065 callout_data, with corresponding flag bit PCRE_EXTRA_CALLOUT_DATA. The
3066 pcre_callout_block structure has a field of the same name. The contents of
3067 the field passed in the pcre_extra structure are passed to the callout
3068 function in the corresponding field in the callout block. This makes it
3069 easier to use the same callout-containing regex from multiple threads. For
3070 testing, the pcretest program has a new data escape
3072 \C*n pass the number n (may be negative) as callout_data
3074 If the callout function in pcretest receives a non-zero value as
3075 callout_data, it returns that value.
3077 50. Makefile wasn't handling CFLAGS properly when compiling dftables. Also,
3078 there were some redundant $(CFLAGS) in commands that are now specified as
3079 $(LINK), which already includes $(CFLAGS).
3081 51. Extensions to UTF-8 support are listed below. These all apply when (a) PCRE
3082 has been compiled with UTF-8 support *and* pcre_compile() has been compiled
3083 with the PCRE_UTF8 flag. Patterns that are compiled without that flag assume
3084 one-byte characters throughout. Note that case-insensitive matching applies
3085 only to characters whose values are less than 256. PCRE doesn't support the
3086 notion of cases for higher-valued characters.
3088 (i) A character class whose characters are all within 0-255 is handled as
3089 a bit map, and the map is inverted for negative classes. Previously, a
3090 character > 255 always failed to match such a class; however it should
3091 match if the class was a negative one (e.g. [^ab]). This has been fixed.
3093 (ii) A negated character class with a single character < 255 is coded as
3094 "not this character" (OP_NOT). This wasn't working properly when the test
3095 character was multibyte, either singly or repeated.
3097 (iii) Repeats of multibyte characters are now handled correctly in UTF-8
3098 mode, for example: \x{100}{2,3}.
3100 (iv) The character escapes \b, \B, \d, \D, \s, \S, \w, and \W (either
3101 singly or repeated) now correctly test multibyte characters. However,
3102 PCRE doesn't recognize any characters with values greater than 255 as
3103 digits, spaces, or word characters. Such characters always match \D, \S,
3104 and \W, and never match \d, \s, or \w.
3106 (v) Classes may now contain characters and character ranges with values
3107 greater than 255. For example: [ab\x{100}-\x{400}].
3109 (vi) pcregrep now has a --utf-8 option (synonym -u) which makes it call
3110 PCRE in UTF-8 mode.
3112 52. The info request value PCRE_INFO_FIRSTCHAR has been renamed
3113 PCRE_INFO_FIRSTBYTE because it is a byte value. However, the old name is
3114 retained for backwards compatibility. (Note that LASTLITERAL is also a byte
3115 value.)
3117 53. The single man page has become too large. I have therefore split it up into
3118 a number of separate man pages. These also give rise to individual HTML pages;
3119 these are now put in a separate directory, and there is an index.html page that
3120 lists them all. Some hyperlinking between the pages has been installed.
3122 54. Added convenience functions for handling named capturing parentheses.
3124 55. Unknown escapes inside character classes (e.g. [\M]) and escapes that
3125 aren't interpreted therein (e.g. [\C]) are literals in Perl. This is now also
3126 true in PCRE, except when the PCRE_EXTENDED option is set, in which case they
3127 are faulted.
3129 56. Introduced HOST_CC and HOST_CFLAGS which can be set in the environment when
3130 calling configure. These values are used when compiling the dftables.c program
3131 which is run to generate the source of the default character tables. They
3132 default to the values of CC and CFLAGS. If you are cross-compiling PCRE,
3133 you will need to set these values.
3135 57. Updated the building process for Windows DLL, as provided by Fred Cox.
3138 Version 3.9 02-Jan-02
3139 ---------------------
3141 1. A bit of extraneous text had somehow crept into the pcregrep documentation.
3143 2. If --disable-static was given, the building process failed when trying to
3144 build pcretest and pcregrep. (For some reason it was using libtool to compile
3145 them, which is not right, as they aren't part of the library.)
3148 Version 3.8 18-Dec-01
3149 ---------------------
3151 1. The experimental UTF-8 code was completely screwed up. It was packing the
3152 bytes in the wrong order. How dumb can you get?
3155 Version 3.7 29-Oct-01
3156 ---------------------
3158 1. In updating pcretest to check change 1 of version 3.6, I screwed up.
3159 This caused pcretest, when used on the test data, to segfault. Unfortunately,
3160 this didn't happen under Solaris 8, where I normally test things.
3162 2. The Makefile had to be changed to make it work on BSD systems, where 'make'
3163 doesn't seem to recognize that ./xxx and xxx are the same file. (This entry
3164 isn't in ChangeLog distributed with 3.7 because I forgot when I hastily made
3165 this fix an hour or so after the initial 3.7 release.)
3168 Version 3.6 23-Oct-01
3169 ---------------------
3171 1. Crashed with /(sens|respons)e and \1ibility/ and "sense and sensibility" if
3172 offsets passed as NULL with zero offset count.
3174 2. The config.guess and config.sub files had not been updated when I moved to
3175 the latest autoconf.
3178 Version 3.5 15-Aug-01
3179 ---------------------
3181 1. Added some missing #if !defined NOPOSIX conditionals in pcretest.c that
3182 had been forgotten.
3184 2. By using declared but undefined structures, we can avoid using "void"
3185 definitions in pcre.h while keeping the internal definitions of the structures
3186 private.
3188 3. The distribution is now built using autoconf 2.50 and libtool 1.4. From a
3189 user point of view, this means that both static and shared libraries are built
3190 by default, but this can be individually controlled. More of the work of
3191 handling this static/shared cases is now inside libtool instead of PCRE's make
3192 file.
3194 4. The pcretest utility is now installed along with pcregrep because it is
3195 useful for users (to test regexs) and by doing this, it automatically gets
3196 relinked by libtool. The documentation has been turned into a man page, so
3197 there are now .1, .txt, and .html versions in /doc.
3199 5. Upgrades to pcregrep:
3200 (i) Added long-form option names like gnu grep.
3201 (ii) Added --help to list all options with an explanatory phrase.
3202 (iii) Added -r, --recursive to recurse into sub-directories.
3203 (iv) Added -f, --file to read patterns from a file.
3205 6. pcre_exec() was referring to its "code" argument before testing that
3206 argument for NULL (and giving an error if it was NULL).
3208 7. Upgraded Makefile.in to allow for compiling in a different directory from
3209 the source directory.
3211 8. Tiny buglet in pcretest: when pcre_fullinfo() was called to retrieve the
3212 options bits, the pointer it was passed was to an int instead of to an unsigned
3213 long int. This mattered only on 64-bit systems.
3215 9. Fixed typo (3.4/1) in pcre.h again. Sigh. I had changed pcre.h (which is
3216 generated) instead of pcre.in, which it its source. Also made the same change
3217 in several of the .c files.
3219 10. A new release of gcc defines printf() as a macro, which broke pcretest
3220 because it had an ifdef in the middle of a string argument for printf(). Fixed
3221 by using separate calls to printf().
3223 11. Added --enable-newline-is-cr and --enable-newline-is-lf to the configure
3224 script, to force use of CR or LF instead of \n in the source. On non-Unix
3225 systems, the value can be set in config.h.
3227 12. The limit of 200 on non-capturing parentheses is a _nesting_ limit, not an
3228 absolute limit. Changed the text of the error message to make this clear, and
3229 likewise updated the man page.
3231 13. The limit of 99 on the number of capturing subpatterns has been removed.
3232 The new limit is 65535, which I hope will not be a "real" limit.
3235 Version 3.4 22-Aug-00
3236 ---------------------
3238 1. Fixed typo in pcre.h: unsigned const char * changed to const unsigned char *.
3240 2. Diagnose condition (?(0) as an error instead of crashing on matching.
3243 Version 3.3 01-Aug-00
3244 ---------------------
3246 1. If an octal character was given, but the value was greater than \377, it
3247 was not getting masked to the least significant bits, as documented. This could
3248 lead to crashes in some systems.
3250 2. Perl 5.6 (if not earlier versions) accepts classes like [a-\d] and treats
3251 the hyphen as a literal. PCRE used to give an error; it now behaves like Perl.
3253 3. Added the functions pcre_free_substring() and pcre_free_substring_list().
3254 These just pass their arguments on to (pcre_free)(), but they are provided
3255 because some uses of PCRE bind it to non-C systems that can call its functions,
3256 but cannot call free() or pcre_free() directly.
3258 4. Add "make test" as a synonym for "make check". Corrected some comments in
3259 the Makefile.
3261 5. Add $(DESTDIR)/ in front of all the paths in the "install" target in the
3262 Makefile.
3264 6. Changed the name of pgrep to pcregrep, because Solaris has introduced a
3265 command called pgrep for grepping around the active processes.
3267 7. Added the beginnings of support for UTF-8 character strings.
3269 8. Arranged for the Makefile to pass over the settings of CC, CFLAGS, and
3270 RANLIB to ./ltconfig so that they are used by libtool. I think these are all
3271 the relevant ones. (AR is not passed because ./ltconfig does its own figuring
3272 out for the ar command.)
3275 Version 3.2 12-May-00
3276 ---------------------
3278 This is purely a bug fixing release.
3280 1. If the pattern /((Z)+|A)*/ was matched agained ZABCDEFG it matched Z instead
3281 of ZA. This was just one example of several cases that could provoke this bug,
3282 which was introduced by change 9 of version 2.00. The code for breaking
3283 infinite loops after an iteration that matches an empty string was't working
3284 correctly.
3286 2. The pcretest program was not imitating Perl correctly for the pattern /a*/g
3287 when matched against abbab (for example). After matching an empty string, it
3288 wasn't forcing anchoring when setting PCRE_NOTEMPTY for the next attempt; this
3289 caused it to match further down the string than it should.
3291 3. The code contained an inclusion of sys/types.h. It isn't clear why this
3292 was there because it doesn't seem to be needed, and it causes trouble on some
3293 systems, as it is not a Standard C header. It has been removed.
3295 4. Made 4 silly changes to the source to avoid stupid compiler warnings that
3296 were reported on the Macintosh. The changes were from
3298 while ((c = *(++ptr)) != 0 && c != '\n');
3299 to
3300 while ((c = *(++ptr)) != 0 && c != '\n') ;
3302 Totally extraordinary, but if that's what it takes...
3304 5. PCRE is being used in one environment where neither memmove() nor bcopy() is
3305 available. Added HAVE_BCOPY and an autoconf test for it; if neither
3306 HAVE_MEMMOVE nor HAVE_BCOPY is set, use a built-in emulation function which
3307 assumes the way PCRE uses memmove() (always moving upwards).
3309 6. PCRE is being used in one environment where strchr() is not available. There
3310 was only one use in pcre.c, and writing it out to avoid strchr() probably gives
3311 faster code anyway.
3314 Version 3.1 09-Feb-00
3315 ---------------------
3317 The only change in this release is the fixing of some bugs in Makefile.in for
3318 the "install" target:
3320 (1) It was failing to install pcreposix.h.
3322 (2) It was overwriting the pcre.3 man page with the pcreposix.3 man page.
3325 Version 3.0 01-Feb-00
3326 ---------------------
3328 1. Add support for the /+ modifier to perltest (to output $` like it does in
3329 pcretest).
3331 2. Add support for the /g modifier to perltest.
3333 3. Fix pcretest so that it behaves even more like Perl for /g when the pattern
3334 matches null strings.
3336 4. Fix perltest so that it doesn't do unwanted things when fed an empty
3337 pattern. Perl treats empty patterns specially - it reuses the most recent
3338 pattern, which is not what we want. Replace // by /(?#)/ in order to avoid this
3339 effect.
3341 5. The POSIX interface was broken in that it was just handing over the POSIX
3342 captured string vector to pcre_exec(), but (since release 2.00) PCRE has
3343 required a bigger vector, with some working space on the end. This means that
3344 the POSIX wrapper now has to get and free some memory, and copy the results.
3346 6. Added some simple autoconf support, placing the test data and the
3347 documentation in separate directories, re-organizing some of the
3348 information files, and making it build pcre-config (a GNU standard). Also added
3349 libtool support for building PCRE as a shared library, which is now the
3350 default.
3352 7. Got rid of the leading zero in the definition of PCRE_MINOR because 08 and
3353 09 are not valid octal constants. Single digits will be used for minor values
3354 less than 10.
3356 8. Defined REG_EXTENDED and REG_NOSUB as zero in the POSIX header, so that
3357 existing programs that set these in the POSIX interface can use PCRE without
3358 modification.
3360 9. Added a new function, pcre_fullinfo() with an extensible interface. It can
3361 return all that pcre_info() returns, plus additional data. The pcre_info()
3362 function is retained for compatibility, but is considered to be obsolete.
3364 10. Added experimental recursion feature (?R) to handle one common case that
3365 Perl 5.6 will be able to do with (?p{...}).
3367 11. Added support for POSIX character classes like [:alpha:], which Perl is
3368 adopting.
3371 Version 2.08 31-Aug-99
3372 ----------------------
3374 1. When startoffset was not zero and the pattern began with ".*", PCRE was not
3375 trying to match at the startoffset position, but instead was moving forward to
3376 the next newline as if a previous match had failed.
3378 2. pcretest was not making use of PCRE_NOTEMPTY when repeating for /g and /G,
3379 and could get into a loop if a null string was matched other than at the start
3380 of the subject.
3382 3. Added definitions of PCRE_MAJOR and PCRE_MINOR to pcre.h so the version can
3383 be distinguished at compile time, and for completeness also added PCRE_DATE.
3385 5. Added Paul Sokolovsky's minor changes to make it easy to compile a Win32 DLL
3386 in GnuWin32 environments.
3389 Version 2.07 29-Jul-99
3390 ----------------------
3392 1. The documentation is now supplied in plain text form and HTML as well as in
3393 the form of man page sources.
3395 2. C++ compilers don't like assigning (void *) values to other pointer types.
3396 In particular this affects malloc(). Although there is no problem in Standard
3397 C, I've put in casts to keep C++ compilers happy.
3399 3. Typo on pcretest.c; a cast of (unsigned char *) in the POSIX regexec() call
3400 should be (const char *).
3402 4. If NOPOSIX is defined, pcretest.c compiles without POSIX support. This may
3403 be useful for non-Unix systems who don't want to bother with the POSIX stuff.
3404 However, I haven't made this a standard facility. The documentation doesn't
3405 mention it, and the Makefile doesn't support it.
3407 5. The Makefile now contains an "install" target, with editable destinations at
3408 the top of the file. The pcretest program is not installed.
3410 6. pgrep -V now gives the PCRE version number and date.
3412 7. Fixed bug: a zero repetition after a literal string (e.g. /abcde{0}/) was
3413 causing the entire string to be ignored, instead of just the last character.
3415 8. If a pattern like /"([^\\"]+|\\.)*"/ is applied in the normal way to a
3416 non-matching string, it can take a very, very long time, even for strings of
3417 quite modest length, because of the nested recursion. PCRE now does better in
3418 some of these cases. It does this by remembering the last required literal
3419 character in the pattern, and pre-searching the subject to ensure it is present
3420 before running the real match. In other words, it applies a heuristic to detect
3421 some types of certain failure quickly, and in the above example, if presented
3422 with a string that has no trailing " it gives "no match" very quickly.
3424 9. A new runtime option PCRE_NOTEMPTY causes null string matches to be ignored;
3425 other alternatives are tried instead.
3428 Version 2.06 09-Jun-99
3429 ----------------------
3431 1. Change pcretest's output for amount of store used to show just the code
3432 space, because the remainder (the data block) varies in size between 32-bit and
3433 64-bit systems.
3435 2. Added an extra argument to pcre_exec() to supply an offset in the subject to
3436 start matching at. This allows lookbehinds to work when searching for multiple
3437 occurrences in a string.
3439 3. Added additional options to pcretest for testing multiple occurrences:
3441 /+ outputs the rest of the string that follows a match
3442 /g loops for multiple occurrences, using the new startoffset argument
3443 /G loops for multiple occurrences by passing an incremented pointer
3445 4. PCRE wasn't doing the "first character" optimization for patterns starting
3446 with \b or \B, though it was doing it for other lookbehind assertions. That is,
3447 it wasn't noticing that a match for a pattern such as /\bxyz/ has to start with
3448 the letter 'x'. On long subject strings, this gives a significant speed-up.
3451 Version 2.05 21-Apr-99
3452 ----------------------
3454 1. Changed the type of magic_number from int to long int so that it works
3455 properly on 16-bit systems.
3457 2. Fixed a bug which caused patterns starting with .* not to work correctly
3458 when the subject string contained newline characters. PCRE was assuming
3459 anchoring for such patterns in all cases, which is not correct because .* will
3460 not pass a newline unless PCRE_DOTALL is set. It now assumes anchoring only if
3461 DOTALL is set at top level; otherwise it knows that patterns starting with .*
3462 must be retried after every newline in the subject.
3465 Version 2.04 18-Feb-99
3466 ----------------------
3468 1. For parenthesized subpatterns with repeats whose minimum was zero, the
3469 computation of the store needed to hold the pattern was incorrect (too large).
3470 If such patterns were nested a few deep, this could multiply and become a real
3471 problem.
3473 2. Added /M option to pcretest to show the memory requirement of a specific
3474 pattern. Made -m a synonym of -s (which does this globally) for compatibility.
3476 3. Subpatterns of the form (regex){n,m} (i.e. limited maximum) were being
3477 compiled in such a way that the backtracking after subsequent failure was
3478 pessimal. Something like (a){0,3} was compiled as (a)?(a)?(a)? instead of
3479 ((a)((a)(a)?)?)? with disastrous performance if the maximum was of any size.
3482 Version 2.03 02-Feb-99
3483 ----------------------
3485 1. Fixed typo and small mistake in man page.
3487 2. Added 4th condition (GPL supersedes if conflict) and created separate
3488 LICENCE file containing the conditions.
3490 3. Updated pcretest so that patterns such as /abc\/def/ work like they do in
3491 Perl, that is the internal \ allows the delimiter to be included in the
3492 pattern. Locked out the use of \ as a delimiter. If \ immediately follows
3493 the final delimiter, add \ to the end of the pattern (to test the error).
3495 4. Added the convenience functions for extracting substrings after a successful
3496 match. Updated pcretest to make it able to test these functions.
3499 Version 2.02 14-Jan-99
3500 ----------------------
3502 1. Initialized the working variables associated with each extraction so that
3503 their saving and restoring doesn't refer to uninitialized store.
3505 2. Put dummy code into study.c in order to trick the optimizer of the IBM C
3506 compiler for OS/2 into generating correct code. Apparently IBM isn't going to
3507 fix the problem.
3509 3. Pcretest: the timing code wasn't using LOOPREPEAT for timing execution
3510 calls, and wasn't printing the correct value for compiling calls. Increased the
3511 default value of LOOPREPEAT, and the number of significant figures in the
3512 times.
3514 4. Changed "/bin/rm" in the Makefile to "-rm" so it works on Windows NT.
3516 5. Renamed "deftables" as "dftables" to get it down to 8 characters, to avoid
3517 a building problem on Windows NT with a FAT file system.
3520 Version 2.01 21-Oct-98
3521 ----------------------
3523 1. Changed the API for pcre_compile() to allow for the provision of a pointer
3524 to character tables built by pcre_maketables() in the current locale. If NULL
3525 is passed, the default tables are used.
3528 Version 2.00 24-Sep-98
3529 ----------------------
3531 1. Since the (>?) facility is in Perl 5.005, don't require PCRE_EXTRA to enable
3532 it any more.
3534 2. Allow quantification of (?>) groups, and make it work correctly.
3536 3. The first character computation wasn't working for (?>) groups.
3538 4. Correct the implementation of \Z (it is permitted to match on the \n at the
3539 end of the subject) and add 5.005's \z, which really does match only at the
3540 very end of the subject.
3542 5. Remove the \X "cut" facility; Perl doesn't have it, and (?> is neater.
3544 6. Remove the ability to specify CASELESS, MULTILINE, DOTALL, and
3545 DOLLAR_END_ONLY at runtime, to make it possible to implement the Perl 5.005
3546 localized options. All options to pcre_study() were also removed.
3548 7. Add other new features from 5.005:
3550 $(?<= positive lookbehind
3551 $(?<! negative lookbehind
3552 (?imsx-imsx) added the unsetting capability
3553 such a setting is global if at outer level; local otherwise
3554 (?imsx-imsx:) non-capturing groups with option setting
3555 (?(cond)re|re) conditional pattern matching
3557 A backreference to itself in a repeated group matches the previous
3558 captured string.
3560 8. General tidying up of studying (both automatic and via "study")
3561 consequential on the addition of new assertions.
3563 9. As in 5.005, unlimited repeated groups that could match an empty substring
3564 are no longer faulted at compile time. Instead, the loop is forcibly broken at
3565 runtime if any iteration does actually match an empty substring.
3567 10. Include the RunTest script in the distribution.
3569 11. Added tests from the Perl 5.005_02 distribution. This showed up a few
3570 discrepancies, some of which were old and were also with respect to 5.004. They
3571 have now been fixed.
3574 Version 1.09 28-Apr-98
3575 ----------------------
3577 1. A negated single character class followed by a quantifier with a minimum
3578 value of one (e.g. [^x]{1,6} ) was not compiled correctly. This could lead to
3579 program crashes, or just wrong answers. This did not apply to negated classes
3580 containing more than one character, or to minima other than one.
3583 Version 1.08 27-Mar-98
3584 ----------------------
3586 1. Add PCRE_UNGREEDY to invert the greediness of quantifiers.
3588 2. Add (?U) and (?X) to set PCRE_UNGREEDY and PCRE_EXTRA respectively. The
3589 latter must appear before anything that relies on it in the pattern.
3592 Version 1.07 16-Feb-98
3593 ----------------------
3595 1. A pattern such as /((a)*)*/ was not being diagnosed as in error (unlimited
3596 repeat of a potentially empty string).
3599 Version 1.06 23-Jan-98
3600 ----------------------
3602 1. Added Markus Oberhumer's little patches for C++.
3604 2. Literal strings longer than 255 characters were broken.
3607 Version 1.05 23-Dec-97
3608 ----------------------
3610 1. Negated character classes containing more than one character were failing if
3611 PCRE_CASELESS was set at run time.
3614 Version 1.04 19-Dec-97
3615 ----------------------
3617 1. Corrected the man page, where some "const" qualifiers had been omitted.
3619 2. Made debugging output print "{0,xxx}" instead of just "{,xxx}" to agree with
3620 input syntax.
3622 3. Fixed memory leak which occurred when a regex with back references was
3623 matched with an offsets vector that wasn't big enough. The temporary memory
3624 that is used in this case wasn't being freed if the match failed.
3626 4. Tidied pcretest to ensure it frees memory that it gets.
3628 5. Temporary memory was being obtained in the case where the passed offsets
3629 vector was exactly big enough.
3631 6. Corrected definition of offsetof() from change 5 below.
3633 7. I had screwed up change 6 below and broken the rules for the use of
3634 setjmp(). Now fixed.
3637 Version 1.03 18-Dec-97
3638 ----------------------
3640 1. A erroneous regex with a missing opening parenthesis was correctly
3641 diagnosed, but PCRE attempted to access brastack[-1], which could cause crashes
3642 on some systems.
3644 2. Replaced offsetof(real_pcre, code) by offsetof(real_pcre, code[0]) because
3645 it was reported that one broken compiler failed on the former because "code" is
3646 also an independent variable.
3648 3. The erroneous regex a[]b caused an array overrun reference.
3650 4. A regex ending with a one-character negative class (e.g. /[^k]$/) did not
3651 fail on data ending with that character. (It was going on too far, and checking
3652 the next character, typically a binary zero.) This was specific to the
3653 optimized code for single-character negative classes.
3655 5. Added a contributed patch from the TIN world which does the following:
3657 + Add an undef for memmove, in case the the system defines a macro for it.
3659 + Add a definition of offsetof(), in case there isn't one. (I don't know
3660 the reason behind this - offsetof() is part of the ANSI standard - but
3661 it does no harm).
3663 + Reduce the ifdef's in pcre.c using macro DPRINTF, thereby eliminating
3664 most of the places where whitespace preceded '#'. I have given up and
3665 allowed the remaining 2 cases to be at the margin.
3667 + Rename some variables in pcre to eliminate shadowing. This seems very
3668 pedantic, but does no harm, of course.
3670 6. Moved the call to setjmp() into its own function, to get rid of warnings
3671 from gcc -Wall, and avoided calling it at all unless PCRE_EXTRA is used.
3673 7. Constructs such as \d{8,} were compiling into the equivalent of
3674 \d{8}\d{0,65527} instead of \d{8}\d* which didn't make much difference to the
3675 outcome, but in this particular case used more store than had been allocated,
3676 which caused the bug to be discovered because it threw up an internal error.
3678 8. The debugging code in both pcre and pcretest for outputting the compiled
3679 form of a regex was going wrong in the case of back references followed by
3680 curly-bracketed repeats.
3683 Version 1.02 12-Dec-97
3684 ----------------------
3686 1. Typos in pcre.3 and comments in the source fixed.
3688 2. Applied a contributed patch to get rid of places where it used to remove
3689 'const' from variables, and fixed some signed/unsigned and uninitialized
3690 variable warnings.
3692 3. Added the "runtest" target to Makefile.
3694 4. Set default compiler flag to -O2 rather than just -O.
3697 Version 1.01 19-Nov-97
3698 ----------------------
3700 1. PCRE was failing to diagnose unlimited repeat of empty string for patterns
3701 like /([ab]*)*/, that is, for classes with more than one character in them.
3703 2. Likewise, it wasn't diagnosing patterns with "once-only" subpatterns, such
3704 as /((?>a*))*/ (a PCRE_EXTRA facility).
3707 Version 1.00 18-Nov-97
3708 ----------------------
3710 1. Added compile-time macros to support systems such as SunOS4 which don't have
3711 memmove() or strerror() but have other things that can be used instead.
3713 2. Arranged that "make clean" removes the executables.
3716 Version 0.99 27-Oct-97
3717 ----------------------
3719 1. Fixed bug in code for optimizing classes with only one character. It was
3720 initializing a 32-byte map regardless, which could cause it to run off the end
3721 of the memory it had got.
3723 2. Added, conditional on PCRE_EXTRA, the proposed (?>REGEX) construction.
3726 Version 0.98 22-Oct-97
3727 ----------------------
3729 1. Fixed bug in code for handling temporary memory usage when there are more
3730 back references than supplied space in the ovector. This could cause segfaults.
3733 Version 0.97 21-Oct-97
3734 ----------------------
3736 1. Added the \X "cut" facility, conditional on PCRE_EXTRA.
3738 2. Optimized negated single characters not to use a bit map.
3740 3. Brought error texts together as macro definitions; clarified some of them;
3741 fixed one that was wrong - it said "range out of order" when it meant "invalid
3742 escape sequence".
3744 4. Changed some char * arguments to const char *.
3746 5. Added PCRE_NOTBOL and PCRE_NOTEOL (from POSIX).
3748 6. Added the POSIX-style API wrapper in pcreposix.a and testing facilities in
3749 pcretest.
3752 Version 0.96 16-Oct-97
3753 ----------------------
3755 1. Added a simple "pgrep" utility to the distribution.
3757 2. Fixed an incompatibility with Perl: "{" is now treated as a normal character
3758 unless it appears in one of the precise forms "{ddd}", "{ddd,}", or "{ddd,ddd}"
3759 where "ddd" means "one or more decimal digits".
3761 3. Fixed serious bug. If a pattern had a back reference, but the call to
3762 pcre_exec() didn't supply a large enough ovector to record the related
3763 identifying subpattern, the match always failed. PCRE now remembers the number
3764 of the largest back reference, and gets some temporary memory in which to save
3765 the offsets during matching if necessary, in order to ensure that
3766 backreferences always work.
3768 4. Increased the compatibility with Perl in a number of ways:
3770 (a) . no longer matches \n by default; an option PCRE_DOTALL is provided
3771 to request this handling. The option can be set at compile or exec time.
3773 (b) $ matches before a terminating newline by default; an option
3774 PCRE_DOLLAR_ENDONLY is provided to override this (but not in multiline
3775 mode). The option can be set at compile or exec time.
3777 (c) The handling of \ followed by a digit other than 0 is now supposed to be
3778 the same as Perl's. If the decimal number it represents is less than 10
3779 or there aren't that many previous left capturing parentheses, an octal
3780 escape is read. Inside a character class, it's always an octal escape,
3781 even if it is a single digit.
3783 (d) An escaped but undefined alphabetic character is taken as a literal,
3784 unless PCRE_EXTRA is set. Currently this just reserves the remaining
3785 escapes.
3787 (e) {0} is now permitted. (The previous item is removed from the compiled
3788 pattern).
3790 5. Changed all the names of code files so that the basic parts are no longer
3791 than 10 characters, and abolished the teeny "globals.c" file.
3793 6. Changed the handling of character classes; they are now done with a 32-byte
3794 bit map always.
3796 7. Added the -d and /D options to pcretest to make it possible to look at the
3797 internals of compilation without having to recompile pcre.
3800 Version 0.95 23-Sep-97
3801 ----------------------
3803 1. Fixed bug in pre-pass concerning escaped "normal" characters such as \x5c or
3804 \x20 at the start of a run of normal characters. These were being treated as
3805 real characters, instead of the source characters being re-checked.
3808 Version 0.94 18-Sep-97
3809 ----------------------
3811 1. The functions are now thread-safe, with the caveat that the global variables
3812 containing pointers to malloc() and free() or alternative functions are the
3813 same for all threads.
3815 2. Get pcre_study() to generate a bitmap of initial characters for non-
3816 anchored patterns when this is possible, and use it if passed to pcre_exec().
3819 Version 0.93 15-Sep-97
3820 ----------------------
3822 1. /(b)|(:+)/ was computing an incorrect first character.
3824 2. Add pcre_study() to the API and the passing of pcre_extra to pcre_exec(),
3825 but not actually doing anything yet.
3827 3. Treat "-" characters in classes that cannot be part of ranges as literals,
3828 as Perl does (e.g. [-az] or [az-]).
3830 4. Set the anchored flag if a branch starts with .* or .*? because that tests
3831 all possible positions.
3833 5. Split up into different modules to avoid including unneeded functions in a
3834 compiled binary. However, compile and exec are still in one module. The "study"
3835 function is split off.
3837 6. The character tables are now in a separate module whose source is generated
3838 by an auxiliary program - but can then be edited by hand if required. There are
3839 now no calls to isalnum(), isspace(), isdigit(), isxdigit(), tolower() or
3840 toupper() in the code.
3842 7. Turn the malloc/free funtions variables into pcre_malloc and pcre_free and
3843 make them global. Abolish the function for setting them, as the caller can now
3844 set them directly.
3847 Version 0.92 11-Sep-97
3848 ----------------------
3850 1. A repeat with a fixed maximum and a minimum of 1 for an ordinary character
3851 (e.g. /a{1,3}/) was broken (I mis-optimized it).
3853 2. Caseless matching was not working in character classes if the characters in
3854 the pattern were in upper case.
3856 3. Make ranges like [W-c] work in the same way as Perl for caseless matching.
3858 4. Make PCRE_ANCHORED public and accept as a compile option.
3860 5. Add an options word to pcre_exec() and accept PCRE_ANCHORED and
3861 PCRE_CASELESS at run time. Add escapes \A and \I to pcretest to cause it to
3862 pass them.
3864 6. Give an error if bad option bits passed at compile or run time.
3866 7. Add PCRE_MULTILINE at compile and exec time, and (?m) as well. Add \M to
3867 pcretest to cause it to pass that flag.
3869 8. Add pcre_info(), to get the number of identifying subpatterns, the stored
3870 options, and the first character, if set.
3872 9. Recognize C+ or C{n,m} where n >= 1 as providing a fixed starting character.
3875 Version 0.91 10-Sep-97
3876 ----------------------
3878 1. PCRE was failing to diagnose unlimited repeats of subpatterns that could
3879 match the empty string as in /(a*)*/. It was looping and ultimately crashing.
3881 2. PCRE was looping on encountering an indefinitely repeated back reference to
3882 a subpattern that had matched an empty string, e.g. /(a|)\1*/. It now does what
3883 Perl does - treats the match as successful.
3885 ****


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

  ViewVC Help
Powered by ViewVC 1.1.5