ViewVC logotype

Contents of /code/trunk/ChangeLog

Parent Directory Parent Directory | Revision Log Revision Log

Revision 515 - (show annotations)
Tue May 4 09:12:25 2010 UTC (5 years, 5 months ago) by ph10
File size: 185861 byte(s)
Avoid warnings about fwrite in pcregrep.
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.)
23 6. When compiling pcregrep with newer versions of gcc which may have
24 FORTIFY_SOURCE set, several warnings "ignoring return value of 'fwrite',
25 declared with attribute warn_unused_result" were given. Just casting the
26 result to (void) does not stop the warnings; a more elaborate fudge is
27 needed. I've used a macro to implement this.
30 Version 8.02 19-Mar-2010
31 ------------------------
33 1. The Unicode data tables have been updated to Unicode 5.2.0.
35 2. Added the option --libs-cpp to pcre-config, but only when C++ support is
36 configured.
38 3. Updated the licensing terms in the pcregexp.pas file, as agreed with the
39 original author of that file, following a query about its status.
41 4. On systems that do not have stdint.h (e.g. Solaris), check for and include
42 inttypes.h instead. This fixes a bug that was introduced by change 8.01/8.
44 5. A pattern such as (?&t)*+(?(DEFINE)(?<t>.)) which has a possessive
45 quantifier applied to a forward-referencing subroutine call, could compile
46 incorrect code or give the error "internal error: previously-checked
47 referenced subpattern not found".
49 6. Both MS Visual Studio and Symbian OS have problems with initializing
50 variables to point to external functions. For these systems, therefore,
51 pcre_malloc etc. are now initialized to local functions that call the
52 relevant global functions.
54 7. There were two entries missing in the vectors called coptable and poptable
55 in pcre_dfa_exec.c. This could lead to memory accesses outsize the vectors.
56 I've fixed the data, and added a kludgy way of testing at compile time that
57 the lengths are correct (equal to the number of opcodes).
59 8. Following on from 7, I added a similar kludge to check the length of the
60 eint vector in pcreposix.c.
62 9. Error texts for pcre_compile() are held as one long string to avoid too
63 much relocation at load time. To find a text, the string is searched,
64 counting zeros. There was no check for running off the end of the string,
65 which could happen if a new error number was added without updating the
66 string.
68 10. \K gave a compile-time error if it appeared in a lookbehind assersion.
70 11. \K was not working if it appeared in an atomic group or in a group that
71 was called as a "subroutine", or in an assertion. Perl 5.11 documents that
72 \K is "not well defined" if used in an assertion. PCRE now accepts it if
73 the assertion is positive, but not if it is negative.
75 12. Change 11 fortuitously reduced the size of the stack frame used in the
76 "match()" function of pcre_exec.c by one pointer. Forthcoming
77 implementation of support for (*MARK) will need an extra pointer on the
78 stack; I have reserved it now, so that the stack frame size does not
79 decrease.
81 13. A pattern such as (?P<L1>(?P<L2>0)|(?P>L2)(?P>L1)) in which the only other
82 item in branch that calls a recursion is a subroutine call - as in the
83 second branch in the above example - was incorrectly given the compile-
84 time error "recursive call could loop indefinitely" because pcre_compile()
85 was not correctly checking the subroutine for matching a non-empty string.
87 14. The checks for overrunning compiling workspace could trigger after an
88 overrun had occurred. This is a "should never occur" error, but it can be
89 triggered by pathological patterns such as hundreds of nested parentheses.
90 The checks now trigger 100 bytes before the end of the workspace.
92 15. Fix typo in configure.ac: "srtoq" should be "strtoq".
95 Version 8.01 19-Jan-2010
96 ------------------------
98 1. If a pattern contained a conditional subpattern with only one branch (in
99 particular, this includes all (*DEFINE) patterns), a call to pcre_study()
100 computed the wrong minimum data length (which is of course zero for such
101 subpatterns). This could cause incorrect "no match" results.
103 2. For patterns such as (?i)a(?-i)b|c where an option setting at the start of
104 the pattern is reset in the first branch, pcre_compile() failed with
105 "internal error: code overflow at offset...". This happened only when
106 the reset was to the original external option setting. (An optimization
107 abstracts leading options settings into an external setting, which was the
108 cause of this.)
110 3. A pattern such as ^(?!a(*SKIP)b) where a negative assertion contained one
111 of the verbs SKIP, PRUNE, or COMMIT, did not work correctly. When the
112 assertion pattern did not match (meaning that the assertion was true), it
113 was incorrectly treated as false if the SKIP had been reached during the
114 matching. This also applied to assertions used as conditions.
116 4. If an item that is not supported by pcre_dfa_exec() was encountered in an
117 assertion subpattern, including such a pattern used as a condition,
118 unpredictable results occurred, instead of the error return
121 5. The C++ GlobalReplace function was not working like Perl for the special
122 situation when an empty string is matched. It now does the fancy magic
123 stuff that is necessary.
125 6. In pcre_internal.h, obsolete includes to setjmp.h and stdarg.h have been
126 removed. (These were left over from very, very early versions of PCRE.)
128 7. Some cosmetic changes to the code to make life easier when compiling it
129 as part of something else:
131 (a) Change DEBUG to PCRE_DEBUG.
133 (b) In pcre_compile(), rename the member of the "branch_chain" structure
134 called "current" as "current_branch", to prevent a collision with the
135 Linux macro when compiled as a kernel module.
137 (c) In pcre_study(), rename the function set_bit() as set_table_bit(), to
138 prevent a collision with the Linux macro when compiled as a kernel
139 module.
141 8. In pcre_compile() there are some checks for integer overflows that used to
142 cast potentially large values to (double). This has been changed to that
143 when building, a check for int64_t is made, and if it is found, it is used
144 instead, thus avoiding the use of floating point arithmetic. (There is no
145 other use of FP in PCRE.) If int64_t is not found, the fallback is to
146 double.
148 9. Added two casts to avoid signed/unsigned warnings from VS Studio Express
149 2005 (difference between two addresses compared to an unsigned value).
151 10. Change the standard AC_CHECK_LIB test for libbz2 in configure.ac to a
152 custom one, because of the following reported problem in Windows:
154 - libbz2 uses the Pascal calling convention (WINAPI) for the functions
155 under Win32.
156 - The standard autoconf AC_CHECK_LIB fails to include "bzlib.h",
157 therefore missing the function definition.
158 - The compiler thus generates a "C" signature for the test function.
159 - The linker fails to find the "C" function.
160 - PCRE fails to configure if asked to do so against libbz2.
162 11. When running libtoolize from libtool-2.2.6b as part of autogen.sh, these
163 messages were output:
165 Consider adding `AC_CONFIG_MACRO_DIR([m4])' to configure.ac and
166 rerunning libtoolize, to keep the correct libtool macros in-tree.
167 Consider adding `-I m4' to ACLOCAL_AMFLAGS in Makefile.am.
169 I have done both of these things.
171 12. Although pcre_dfa_exec() does not use nearly as much stack as pcre_exec()
172 most of the time, it *can* run out if it is given a pattern that contains a
173 runaway infinite recursion. I updated the discussion in the pcrestack man
174 page.
176 13. Now that we have gone to the x.xx style of version numbers, the minor
177 version may start with zero. Using 08 or 09 is a bad idea because users
178 might check the value of PCRE_MINOR in their code, and 08 or 09 may be
179 interpreted as invalid octal numbers. I've updated the previous comment in
180 configure.ac, and also added a check that gives an error if 08 or 09 are
181 used.
183 14. Change 8.00/11 was not quite complete: code had been accidentally omitted,
184 causing partial matching to fail when the end of the subject matched \W
185 in a UTF-8 pattern where \W was quantified with a minimum of 3.
187 15. There were some discrepancies between the declarations in pcre_internal.h
188 of _pcre_is_newline(), _pcre_was_newline(), and _pcre_valid_utf8() and
189 their definitions. The declarations used "const uschar *" and the
190 definitions used USPTR. Even though USPTR is normally defined as "const
191 unsigned char *" (and uschar is typedeffed as "unsigned char"), it was
192 reported that: "This difference in casting confuses some C++ compilers, for
193 example, SunCC recognizes above declarations as different functions and
194 generates broken code for hbpcre." I have changed the declarations to use
195 USPTR.
197 16. GNU libtool is named differently on some systems. The autogen.sh script now
198 tries several variants such as glibtoolize (MacOSX) and libtoolize1x
199 (FreeBSD).
201 17. Applied Craig's patch that fixes an HP aCC compile error in pcre 8.00
202 (strtoXX undefined when compiling pcrecpp.cc). The patch contains this
203 comment: "Figure out how to create a longlong from a string: strtoll and
204 equivalent. It's not enough to call AC_CHECK_FUNCS: hpux has a strtoll, for
205 instance, but it only takes 2 args instead of 3!"
207 18. A subtle bug concerned with back references has been fixed by a change of
208 specification, with a corresponding code fix. A pattern such as
209 ^(xa|=?\1a)+$ which contains a back reference inside the group to which it
210 refers, was giving matches when it shouldn't. For example, xa=xaaa would
211 match that pattern. Interestingly, Perl (at least up to 5.11.3) has the
212 same bug. Such groups have to be quantified to be useful, or contained
213 inside another quantified group. (If there's no repetition, the reference
214 can never match.) The problem arises because, having left the group and
215 moved on to the rest of the pattern, a later failure that backtracks into
216 the group uses the captured value from the final iteration of the group
217 rather than the correct earlier one. I have fixed this in PCRE by forcing
218 any group that contains a reference to itself to be an atomic group; that
219 is, there cannot be any backtracking into it once it has completed. This is
220 similar to recursive and subroutine calls.
223 Version 8.00 19-Oct-09
224 ----------------------
226 1. The table for translating pcre_compile() error codes into POSIX error codes
227 was out-of-date, and there was no check on the pcre_compile() error code
228 being within the table. This could lead to an OK return being given in
229 error.
231 2. Changed the call to open a subject file in pcregrep from fopen(pathname,
232 "r") to fopen(pathname, "rb"), which fixed a problem with some of the tests
233 in a Windows environment.
235 3. The pcregrep --count option prints the count for each file even when it is
236 zero, as does GNU grep. However, pcregrep was also printing all files when
237 --files-with-matches was added. Now, when both options are given, it prints
238 counts only for those files that have at least one match. (GNU grep just
239 prints the file name in this circumstance, but including the count seems
240 more useful - otherwise, why use --count?) Also ensured that the
241 combination -clh just lists non-zero counts, with no names.
243 4. The long form of the pcregrep -F option was incorrectly implemented as
244 --fixed_strings instead of --fixed-strings. This is an incompatible change,
245 but it seems right to fix it, and I didn't think it was worth preserving
246 the old behaviour.
248 5. The command line items --regex=pattern and --regexp=pattern were not
249 recognized by pcregrep, which required --regex pattern or --regexp pattern
250 (with a space rather than an '='). The man page documented the '=' forms,
251 which are compatible with GNU grep; these now work.
253 6. No libpcreposix.pc file was created for pkg-config; there was just
254 libpcre.pc and libpcrecpp.pc. The omission has been rectified.
256 7. Added #ifndef SUPPORT_UCP into the pcre_ucd.c module, to reduce its size
257 when UCP support is not needed, by modifying the Python script that
258 generates it from Unicode data files. This should not matter if the module
259 is correctly used as a library, but I received one complaint about 50K of
260 unwanted data. My guess is that the person linked everything into his
261 program rather than using a library. Anyway, it does no harm.
263 8. A pattern such as /\x{123}{2,2}+/8 was incorrectly compiled; the trigger
264 was a minimum greater than 1 for a wide character in a possessive
265 repetition. The same bug could also affect patterns like /(\x{ff}{0,2})*/8
266 which had an unlimited repeat of a nested, fixed maximum repeat of a wide
267 character. Chaos in the form of incorrect output or a compiling loop could
268 result.
270 9. The restrictions on what a pattern can contain when partial matching is
271 requested for pcre_exec() have been removed. All patterns can now be
272 partially matched by this function. In addition, if there are at least two
273 slots in the offset vector, the offset of the earliest inspected character
274 for the match and the offset of the end of the subject are set in them when
275 PCRE_ERROR_PARTIAL is returned.
277 10. Partial matching has been split into two forms: PCRE_PARTIAL_SOFT, which is
278 synonymous with PCRE_PARTIAL, for backwards compatibility, and
279 PCRE_PARTIAL_HARD, which causes a partial match to supersede a full match,
280 and may be more useful for multi-segment matching.
282 11. Partial matching with pcre_exec() is now more intuitive. A partial match
283 used to be given if ever the end of the subject was reached; now it is
284 given only if matching could not proceed because another character was
285 needed. This makes a difference in some odd cases such as Z(*FAIL) with the
286 string "Z", which now yields "no match" instead of "partial match". In the
287 case of pcre_dfa_exec(), "no match" is given if every matching path for the
288 final character ended with (*FAIL).
290 12. Restarting a match using pcre_dfa_exec() after a partial match did not work
291 if the pattern had a "must contain" character that was already found in the
292 earlier partial match, unless partial matching was again requested. For
293 example, with the pattern /dog.(body)?/, the "must contain" character is
294 "g". If the first part-match was for the string "dog", restarting with
295 "sbody" failed. This bug has been fixed.
297 13. The string returned by pcre_dfa_exec() after a partial match has been
298 changed so that it starts at the first inspected character rather than the
299 first character of the match. This makes a difference only if the pattern
300 starts with a lookbehind assertion or \b or \B (\K is not supported by
301 pcre_dfa_exec()). It's an incompatible change, but it makes the two
302 matching functions compatible, and I think it's the right thing to do.
304 14. Added a pcredemo man page, created automatically from the pcredemo.c file,
305 so that the demonstration program is easily available in environments where
306 PCRE has not been installed from source.
308 15. Arranged to add -DPCRE_STATIC to cflags in libpcre.pc, libpcreposix.cp,
309 libpcrecpp.pc and pcre-config when PCRE is not compiled as a shared
310 library.
312 16. Added REG_UNGREEDY to the pcreposix interface, at the request of a user.
313 It maps to PCRE_UNGREEDY. It is not, of course, POSIX-compatible, but it
314 is not the first non-POSIX option to be added. Clearly some people find
315 these options useful.
317 17. If a caller to the POSIX matching function regexec() passes a non-zero
318 value for nmatch with a NULL value for pmatch, the value of
319 nmatch is forced to zero.
321 18. RunGrepTest did not have a test for the availability of the -u option of
322 the diff command, as RunTest does. It now checks in the same way as
323 RunTest, and also checks for the -b option.
325 19. If an odd number of negated classes containing just a single character
326 interposed, within parentheses, between a forward reference to a named
327 subpattern and the definition of the subpattern, compilation crashed with
328 an internal error, complaining that it could not find the referenced
329 subpattern. An example of a crashing pattern is /(?&A)(([^m])(?<A>))/.
330 [The bug was that it was starting one character too far in when skipping
331 over the character class, thus treating the ] as data rather than
332 terminating the class. This meant it could skip too much.]
334 20. Added PCRE_NOTEMPTY_ATSTART in order to be able to correctly implement the
335 /g option in pcretest when the pattern contains \K, which makes it possible
336 to have an empty string match not at the start, even when the pattern is
337 anchored. Updated pcretest and pcredemo to use this option.
339 21. If the maximum number of capturing subpatterns in a recursion was greater
340 than the maximum at the outer level, the higher number was returned, but
341 with unset values at the outer level. The correct (outer level) value is
342 now given.
344 22. If (*ACCEPT) appeared inside capturing parentheses, previous releases of
345 PCRE did not set those parentheses (unlike Perl). I have now found a way to
346 make it do so. The string so far is captured, making this feature
347 compatible with Perl.
349 23. The tests have been re-organized, adding tests 11 and 12, to make it
350 possible to check the Perl 5.10 features against Perl 5.10.
352 24. Perl 5.10 allows subroutine calls in lookbehinds, as long as the subroutine
353 pattern matches a fixed length string. PCRE did not allow this; now it
354 does. Neither allows recursion.
356 25. I finally figured out how to implement a request to provide the minimum
357 length of subject string that was needed in order to match a given pattern.
358 (It was back references and recursion that I had previously got hung up
359 on.) This code has now been added to pcre_study(); it finds a lower bound
360 to the length of subject needed. It is not necessarily the greatest lower
361 bound, but using it to avoid searching strings that are too short does give
362 some useful speed-ups. The value is available to calling programs via
363 pcre_fullinfo().
365 26. While implementing 25, I discovered to my embarrassment that pcretest had
366 not been passing the result of pcre_study() to pcre_dfa_exec(), so the
367 study optimizations had never been tested with that matching function.
368 Oops. What is worse, even when it was passed study data, there was a bug in
369 pcre_dfa_exec() that meant it never actually used it. Double oops. There
370 were also very few tests of studied patterns with pcre_dfa_exec().
372 27. If (?| is used to create subpatterns with duplicate numbers, they are now
373 allowed to have the same name, even if PCRE_DUPNAMES is not set. However,
374 on the other side of the coin, they are no longer allowed to have different
375 names, because these cannot be distinguished in PCRE, and this has caused
376 confusion. (This is a difference from Perl.)
378 28. When duplicate subpattern names are present (necessarily with different
379 numbers, as required by 27 above), and a test is made by name in a
380 conditional pattern, either for a subpattern having been matched, or for
381 recursion in such a pattern, all the associated numbered subpatterns are
382 tested, and the overall condition is true if the condition is true for any
383 one of them. This is the way Perl works, and is also more like the way
384 testing by number works.
387 Version 7.9 11-Apr-09
388 ---------------------
390 1. When building with support for bzlib/zlib (pcregrep) and/or readline
391 (pcretest), all targets were linked against these libraries. This included
392 libpcre, libpcreposix, and libpcrecpp, even though they do not use these
393 libraries. This caused unwanted dependencies to be created. This problem
394 has been fixed, and now only pcregrep is linked with bzlib/zlib and only
395 pcretest is linked with readline.
397 2. The "typedef int BOOL" in pcre_internal.h that was included inside the
398 "#ifndef FALSE" condition by an earlier change (probably 7.8/18) has been
399 moved outside it again, because FALSE and TRUE are already defined in AIX,
400 but BOOL is not.
402 3. The pcre_config() function was treating the PCRE_MATCH_LIMIT and
403 PCRE_MATCH_LIMIT_RECURSION values as ints, when they should be long ints.
405 4. The pcregrep documentation said spaces were inserted as well as colons (or
406 hyphens) following file names and line numbers when outputting matching
407 lines. This is not true; no spaces are inserted. I have also clarified the
408 wording for the --colour (or --color) option.
410 5. In pcregrep, when --colour was used with -o, the list of matching strings
411 was not coloured; this is different to GNU grep, so I have changed it to be
412 the same.
414 6. When --colo(u)r was used in pcregrep, only the first matching substring in
415 each matching line was coloured. Now it goes on to look for further matches
416 of any of the test patterns, which is the same behaviour as GNU grep.
418 7. A pattern that could match an empty string could cause pcregrep to loop; it
419 doesn't make sense to accept an empty string match in pcregrep, so I have
420 locked it out (using PCRE's PCRE_NOTEMPTY option). By experiment, this
421 seems to be how GNU grep behaves.
423 8. The pattern (?(?=.*b)b|^) was incorrectly compiled as "match must be at
424 start or after a newline", because the conditional assertion was not being
425 correctly handled. The rule now is that both the assertion and what follows
426 in the first alternative must satisfy the test.
428 9. If auto-callout was enabled in a pattern with a conditional group whose
429 condition was an assertion, PCRE could crash during matching, both with
430 pcre_exec() and pcre_dfa_exec().
432 10. The PCRE_DOLLAR_ENDONLY option was not working when pcre_dfa_exec() was
433 used for matching.
435 11. Unicode property support in character classes was not working for
436 characters (bytes) greater than 127 when not in UTF-8 mode.
438 12. Added the -M command line option to pcretest.
440 14. Added the non-standard REG_NOTEMPTY option to the POSIX interface.
442 15. Added the PCRE_NO_START_OPTIMIZE match-time option.
444 16. Added comments and documentation about mis-use of no_arg in the C++
445 wrapper.
447 17. Implemented support for UTF-8 encoding in EBCDIC environments, a patch
448 from Martin Jerabek that uses macro names for all relevant character and
449 string constants.
451 18. Added to pcre_internal.h two configuration checks: (a) If both EBCDIC and
452 SUPPORT_UTF8 are set, give an error; (b) If SUPPORT_UCP is set without
453 SUPPORT_UTF8, define SUPPORT_UTF8. The "configure" script handles both of
454 these, but not everybody uses configure.
456 19. A conditional group that had only one branch was not being correctly
457 recognized as an item that could match an empty string. This meant that an
458 enclosing group might also not be so recognized, causing infinite looping
459 (and probably a segfault) for patterns such as ^"((?(?=[a])[^"])|b)*"$
460 with the subject "ab", where knowledge that the repeated group can match
461 nothing is needed in order to break the loop.
463 20. If a pattern that was compiled with callouts was matched using pcre_dfa_
464 exec(), but without supplying a callout function, matching went wrong.
466 21. If PCRE_ERROR_MATCHLIMIT occurred during a recursion, there was a memory
467 leak if the size of the offset vector was greater than 30. When the vector
468 is smaller, the saved offsets during recursion go onto a local stack
469 vector, but for larger vectors malloc() is used. It was failing to free
470 when the recursion yielded PCRE_ERROR_MATCH_LIMIT (or any other "abnormal"
471 error, in fact).
473 22. There was a missing #ifdef SUPPORT_UTF8 round one of the variables in the
474 heapframe that is used only when UTF-8 support is enabled. This caused no
475 problem, but was untidy.
477 23. Steven Van Ingelgem's patch to CMakeLists.txt to change the name
478 CMAKE_BINARY_DIR to PROJECT_BINARY_DIR so that it works when PCRE is
479 included within another project.
481 24. Steven Van Ingelgem's patches to add more options to the CMake support,
482 slightly modified by me:
484 (a) PCRE_BUILD_TESTS can be set OFF not to build the tests, including
485 not building pcregrep.
487 (b) PCRE_BUILD_PCREGREP can be see OFF not to build pcregrep, but only
488 if PCRE_BUILD_TESTS is also set OFF, because the tests use pcregrep.
490 25. Forward references, both numeric and by name, in patterns that made use of
491 duplicate group numbers, could behave incorrectly or give incorrect errors,
492 because when scanning forward to find the reference group, PCRE was not
493 taking into account the duplicate group numbers. A pattern such as
494 ^X(?3)(a)(?|(b)|(q))(Y) is an example.
496 26. Changed a few more instances of "const unsigned char *" to USPTR, making
497 the feature of a custom pointer more persuasive (as requested by a user).
499 27. Wrapped the definitions of fileno and isatty for Windows, which appear in
500 pcretest.c, inside #ifndefs, because it seems they are sometimes already
501 pre-defined.
503 28. Added support for (*UTF8) at the start of a pattern.
505 29. Arrange for flags added by the "release type" setting in CMake to be shown
506 in the configuration summary.
509 Version 7.8 05-Sep-08
510 ---------------------
512 1. Replaced UCP searching code with optimized version as implemented for Ad
513 Muncher (http://www.admuncher.com/) by Peter Kankowski. This uses a two-
514 stage table and inline lookup instead of a function, giving speed ups of 2
515 to 5 times on some simple patterns that I tested. Permission was given to
516 distribute the MultiStage2.py script that generates the tables (it's not in
517 the tarball, but is in the Subversion repository).
519 2. Updated the Unicode datatables to Unicode 5.1.0. This adds yet more
520 scripts.
522 3. Change 12 for 7.7 introduced a bug in pcre_study() when a pattern contained
523 a group with a zero qualifier. The result of the study could be incorrect,
524 or the function might crash, depending on the pattern.
526 4. Caseless matching was not working for non-ASCII characters in back
527 references. For example, /(\x{de})\1/8i was not matching \x{de}\x{fe}.
528 It now works when Unicode Property Support is available.
530 5. In pcretest, an escape such as \x{de} in the data was always generating
531 a UTF-8 string, even in non-UTF-8 mode. Now it generates a single byte in
532 non-UTF-8 mode. If the value is greater than 255, it gives a warning about
533 truncation.
535 6. Minor bugfix in pcrecpp.cc (change "" == ... to NULL == ...).
537 7. Added two (int) casts to pcregrep when printing the difference of two
538 pointers, in case they are 64-bit values.
540 8. Added comments about Mac OS X stack usage to the pcrestack man page and to
541 test 2 if it fails.
543 9. Added PCRE_CALL_CONVENTION just before the names of all exported functions,
544 and a #define of that name to empty if it is not externally set. This is to
545 allow users of MSVC to set it if necessary.
547 10. The PCRE_EXP_DEFN macro which precedes exported functions was missing from
548 the convenience functions in the pcre_get.c source file.
550 11. An option change at the start of a pattern that had top-level alternatives
551 could cause overwriting and/or a crash. This command provoked a crash in
552 some environments:
554 printf "/(?i)[\xc3\xa9\xc3\xbd]|[\xc3\xa9\xc3\xbdA]/8\n" | pcretest
556 This potential security problem was recorded as CVE-2008-2371.
558 12. For a pattern where the match had to start at the beginning or immediately
559 after a newline (e.g /.*anything/ without the DOTALL flag), pcre_exec() and
560 pcre_dfa_exec() could read past the end of the passed subject if there was
561 no match. To help with detecting such bugs (e.g. with valgrind), I modified
562 pcretest so that it places the subject at the end of its malloc-ed buffer.
564 13. The change to pcretest in 12 above threw up a couple more cases when pcre_
565 exec() might read past the end of the data buffer in UTF-8 mode.
567 14. A similar bug to 7.3/2 existed when the PCRE_FIRSTLINE option was set and
568 the data contained the byte 0x85 as part of a UTF-8 character within its
569 first line. This applied both to normal and DFA matching.
571 15. Lazy qualifiers were not working in some cases in UTF-8 mode. For example,
572 /^[^d]*?$/8 failed to match "abc".
574 16. Added a missing copyright notice to pcrecpp_internal.h.
576 17. Make it more clear in the documentation that values returned from
577 pcre_exec() in ovector are byte offsets, not character counts.
579 18. Tidied a few places to stop certain compilers from issuing warnings.
581 19. Updated the Virtual Pascal + BCC files to compile the latest v7.7, as
582 supplied by Stefan Weber. I made a further small update for 7.8 because
583 there is a change of source arrangements: the pcre_searchfuncs.c module is
584 replaced by pcre_ucd.c.
587 Version 7.7 07-May-08
588 ---------------------
590 1. Applied Craig's patch to sort out a long long problem: "If we can't convert
591 a string to a long long, pretend we don't even have a long long." This is
592 done by checking for the strtoq, strtoll, and _strtoi64 functions.
594 2. Applied Craig's patch to pcrecpp.cc to restore ABI compatibility with
595 pre-7.6 versions, which defined a global no_arg variable instead of putting
596 it in the RE class. (See also #8 below.)
598 3. Remove a line of dead code, identified by coverity and reported by Nuno
599 Lopes.
601 4. Fixed two related pcregrep bugs involving -r with --include or --exclude:
603 (1) The include/exclude patterns were being applied to the whole pathnames
604 of files, instead of just to the final components.
606 (2) If there was more than one level of directory, the subdirectories were
607 skipped unless they satisfied the include/exclude conditions. This is
608 inconsistent with GNU grep (and could even be seen as contrary to the
609 pcregrep specification - which I improved to make it absolutely clear).
610 The action now is always to scan all levels of directory, and just
611 apply the include/exclude patterns to regular files.
613 5. Added the --include_dir and --exclude_dir patterns to pcregrep, and used
614 --exclude_dir in the tests to avoid scanning .svn directories.
616 6. Applied Craig's patch to the QuoteMeta function so that it escapes the
617 NUL character as backslash + 0 rather than backslash + NUL, because PCRE
618 doesn't support NULs in patterns.
620 7. Added some missing "const"s to declarations of static tables in
621 pcre_compile.c and pcre_dfa_exec.c.
623 8. Applied Craig's patch to pcrecpp.cc to fix a problem in OS X that was
624 caused by fix #2 above. (Subsequently also a second patch to fix the
625 first patch. And a third patch - this was a messy problem.)
627 9. Applied Craig's patch to remove the use of push_back().
629 10. Applied Alan Lehotsky's patch to add REG_STARTEND support to the POSIX
630 matching function regexec().
632 11. Added support for the Oniguruma syntax \g<name>, \g<n>, \g'name', \g'n',
633 which, however, unlike Perl's \g{...}, are subroutine calls, not back
634 references. PCRE supports relative numbers with this syntax (I don't think
635 Oniguruma does).
637 12. Previously, a group with a zero repeat such as (...){0} was completely
638 omitted from the compiled regex. However, this means that if the group
639 was called as a subroutine from elsewhere in the pattern, things went wrong
640 (an internal error was given). Such groups are now left in the compiled
641 pattern, with a new opcode that causes them to be skipped at execution
642 time.
644 13. Added the PCRE_JAVASCRIPT_COMPAT option. This makes the following changes
645 to the way PCRE behaves:
647 (a) A lone ] character is dis-allowed (Perl treats it as data).
649 (b) A back reference to an unmatched subpattern matches an empty string
650 (Perl fails the current match path).
652 (c) A data ] in a character class must be notated as \] because if the
653 first data character in a class is ], it defines an empty class. (In
654 Perl it is not possible to have an empty class.) The empty class []
655 never matches; it forces failure and is equivalent to (*FAIL) or (?!).
656 The negative empty class [^] matches any one character, independently
657 of the DOTALL setting.
659 14. A pattern such as /(?2)[]a()b](abc)/ which had a forward reference to a
660 non-existent subpattern following a character class starting with ']' and
661 containing () gave an internal compiling error instead of "reference to
662 non-existent subpattern". Fortunately, when the pattern did exist, the
663 compiled code was correct. (When scanning forwards to check for the
664 existencd of the subpattern, it was treating the data ']' as terminating
665 the class, so got the count wrong. When actually compiling, the reference
666 was subsequently set up correctly.)
668 15. The "always fail" assertion (?!) is optimzed to (*FAIL) by pcre_compile;
669 it was being rejected as not supported by pcre_dfa_exec(), even though
670 other assertions are supported. I have made pcre_dfa_exec() support
671 (*FAIL).
673 16. The implementation of 13c above involved the invention of a new opcode,
674 OP_ALLANY, which is like OP_ANY but doesn't check the /s flag. Since /s
675 cannot be changed at match time, I realized I could make a small
676 improvement to matching performance by compiling OP_ALLANY instead of
677 OP_ANY for "." when DOTALL was set, and then removing the runtime tests
678 on the OP_ANY path.
680 17. Compiling pcretest on Windows with readline support failed without the
681 following two fixes: (1) Make the unistd.h include conditional on
682 HAVE_UNISTD_H; (2) #define isatty and fileno as _isatty and _fileno.
684 18. Changed CMakeLists.txt and cmake/FindReadline.cmake to arrange for the
685 ncurses library to be included for pcretest when ReadLine support is
686 requested, but also to allow for it to be overridden. This patch came from
687 Daniel Bergström.
689 19. There was a typo in the file ucpinternal.h where f0_rangeflag was defined
690 as 0x00f00000 instead of 0x00800000. Luckily, this would not have caused
691 any errors with the current Unicode tables. Thanks to Peter Kankowski for
692 spotting this.
695 Version 7.6 28-Jan-08
696 ---------------------
698 1. A character class containing a very large number of characters with
699 codepoints greater than 255 (in UTF-8 mode, of course) caused a buffer
700 overflow.
702 2. Patch to cut out the "long long" test in pcrecpp_unittest when
703 HAVE_LONG_LONG is not defined.
705 3. Applied Christian Ehrlicher's patch to update the CMake build files to
706 bring them up to date and include new features. This patch includes:
708 - Fixed PH's badly added libz and libbz2 support.
709 - Fixed a problem with static linking.
710 - Added pcredemo. [But later removed - see 7 below.]
711 - Fixed dftables problem and added an option.
712 - Added a number of HAVE_XXX tests, including HAVE_WINDOWS_H and
714 - Added readline support for pcretest.
715 - Added an listing of the option settings after cmake has run.
717 4. A user submitted a patch to Makefile that makes it easy to create
718 "pcre.dll" under mingw when using Configure/Make. I added stuff to
719 Makefile.am that cause it to include this special target, without
720 affecting anything else. Note that the same mingw target plus all
721 the other distribution libraries and programs are now supported
722 when configuring with CMake (see 6 below) instead of with
723 Configure/Make.
725 5. Applied Craig's patch that moves no_arg into the RE class in the C++ code.
726 This is an attempt to solve the reported problem "pcrecpp::no_arg is not
727 exported in the Windows port". It has not yet been confirmed that the patch
728 solves the problem, but it does no harm.
730 6. Applied Sheri's patch to CMakeLists.txt to add NON_STANDARD_LIB_PREFIX and
731 NON_STANDARD_LIB_SUFFIX for dll names built with mingw when configured
732 with CMake, and also correct the comment about stack recursion.
734 7. Remove the automatic building of pcredemo from the ./configure system and
735 from CMakeLists.txt. The whole idea of pcredemo.c is that it is an example
736 of a program that users should build themselves after PCRE is installed, so
737 building it automatically is not really right. What is more, it gave
738 trouble in some build environments.
740 8. Further tidies to CMakeLists.txt from Sheri and Christian.
743 Version 7.5 10-Jan-08
744 ---------------------
746 1. Applied a patch from Craig: "This patch makes it possible to 'ignore'
747 values in parens when parsing an RE using the C++ wrapper."
749 2. Negative specials like \S did not work in character classes in UTF-8 mode.
750 Characters greater than 255 were excluded from the class instead of being
751 included.
753 3. The same bug as (2) above applied to negated POSIX classes such as
754 [:^space:].
756 4. PCRECPP_STATIC was referenced in pcrecpp_internal.h, but nowhere was it
757 defined or documented. It seems to have been a typo for PCRE_STATIC, so
758 I have changed it.
760 5. The construct (?&) was not diagnosed as a syntax error (it referenced the
761 first named subpattern) and a construct such as (?&a) would reference the
762 first named subpattern whose name started with "a" (in other words, the
763 length check was missing). Both these problems are fixed. "Subpattern name
764 expected" is now given for (?&) (a zero-length name), and this patch also
765 makes it give the same error for \k'' (previously it complained that that
766 was a reference to a non-existent subpattern).
768 6. The erroneous patterns (?+-a) and (?-+a) give different error messages;
769 this is right because (?- can be followed by option settings as well as by
770 digits. I have, however, made the messages clearer.
772 7. Patterns such as (?(1)a|b) (a pattern that contains fewer subpatterns
773 than the number used in the conditional) now cause a compile-time error.
774 This is actually not compatible with Perl, which accepts such patterns, but
775 treats the conditional as always being FALSE (as PCRE used to), but it
776 seems to me that giving a diagnostic is better.
778 8. Change "alphameric" to the more common word "alphanumeric" in comments
779 and messages.
781 9. Fix two occurrences of "backslash" in comments that should have been
782 "backspace".
784 10. Remove two redundant lines of code that can never be obeyed (their function
785 was moved elsewhere).
787 11. The program that makes PCRE's Unicode character property table had a bug
788 which caused it to generate incorrect table entries for sequences of
789 characters that have the same character type, but are in different scripts.
790 It amalgamated them into a single range, with the script of the first of
791 them. In other words, some characters were in the wrong script. There were
792 thirteen such cases, affecting characters in the following ranges:
794 U+002b0 - U+002c1
795 U+0060c - U+0060d
796 U+0061e - U+00612
797 U+0064b - U+0065e
798 U+0074d - U+0076d
799 U+01800 - U+01805
800 U+01d00 - U+01d77
801 U+01d9b - U+01dbf
802 U+0200b - U+0200f
803 U+030fc - U+030fe
804 U+03260 - U+0327f
805 U+0fb46 - U+0fbb1
806 U+10450 - U+1049d
808 12. The -o option (show only the matching part of a line) for pcregrep was not
809 compatible with GNU grep in that, if there was more than one match in a
810 line, it showed only the first of them. It now behaves in the same way as
811 GNU grep.
813 13. If the -o and -v options were combined for pcregrep, it printed a blank
814 line for every non-matching line. GNU grep prints nothing, and pcregrep now
815 does the same. The return code can be used to tell if there were any
816 non-matching lines.
818 14. Added --file-offsets and --line-offsets to pcregrep.
820 15. The pattern (?=something)(?R) was not being diagnosed as a potentially
821 infinitely looping recursion. The bug was that positive lookaheads were not
822 being skipped when checking for a possible empty match (negative lookaheads
823 and both kinds of lookbehind were skipped).
825 16. Fixed two typos in the Windows-only code in pcregrep.c, and moved the
826 inclusion of <windows.h> to before rather than after the definition of
827 INVALID_FILE_ATTRIBUTES (patch from David Byron).
829 17. Specifying a possessive quantifier with a specific limit for a Unicode
830 character property caused pcre_compile() to compile bad code, which led at
831 runtime to PCRE_ERROR_INTERNAL (-14). Examples of patterns that caused this
832 are: /\p{Zl}{2,3}+/8 and /\p{Cc}{2}+/8. It was the possessive "+" that
833 caused the error; without that there was no problem.
835 18. Added --enable-pcregrep-libz and --enable-pcregrep-libbz2.
837 19. Added --enable-pcretest-libreadline.
839 20. In pcrecpp.cc, the variable 'count' was incremented twice in
840 RE::GlobalReplace(). As a result, the number of replacements returned was
841 double what it should be. I removed one of the increments, but Craig sent a
842 later patch that removed the other one (the right fix) and added unit tests
843 that check the return values (which was not done before).
845 21. Several CMake things:
847 (1) Arranged that, when cmake is used on Unix, the libraries end up with
848 the names libpcre and libpcreposix, not just pcre and pcreposix.
850 (2) The above change means that pcretest and pcregrep are now correctly
851 linked with the newly-built libraries, not previously installed ones.
855 22. In UTF-8 mode, with newline set to "any", a pattern such as .*a.*=.b.*
856 crashed when matching a string such as a\x{2029}b (note that \x{2029} is a
857 UTF-8 newline character). The key issue is that the pattern starts .*;
858 this means that the match must be either at the beginning, or after a
859 newline. The bug was in the code for advancing after a failed match and
860 checking that the new position followed a newline. It was not taking
861 account of UTF-8 characters correctly.
863 23. PCRE was behaving differently from Perl in the way it recognized POSIX
864 character classes. PCRE was not treating the sequence [:...:] as a
865 character class unless the ... were all letters. Perl, however, seems to
866 allow any characters between [: and :], though of course it rejects as
867 unknown any "names" that contain non-letters, because all the known class
868 names consist only of letters. Thus, Perl gives an error for [[:1234:]],
869 for example, whereas PCRE did not - it did not recognize a POSIX character
870 class. This seemed a bit dangerous, so the code has been changed to be
871 closer to Perl. The behaviour is not identical to Perl, because PCRE will
872 diagnose an unknown class for, for example, [[:l\ower:]] where Perl will
873 treat it as [[:lower:]]. However, PCRE does now give "unknown" errors where
874 Perl does, and where it didn't before.
876 24. Rewrite so as to remove the single use of %n from pcregrep because in some
877 Windows environments %n is disabled by default.
880 Version 7.4 21-Sep-07
881 ---------------------
883 1. Change 7.3/28 was implemented for classes by looking at the bitmap. This
884 means that a class such as [\s] counted as "explicit reference to CR or
885 LF". That isn't really right - the whole point of the change was to try to
886 help when there was an actual mention of one of the two characters. So now
887 the change happens only if \r or \n (or a literal CR or LF) character is
888 encountered.
890 2. The 32-bit options word was also used for 6 internal flags, but the numbers
891 of both had grown to the point where there were only 3 bits left.
892 Fortunately, there was spare space in the data structure, and so I have
893 moved the internal flags into a new 16-bit field to free up more option
894 bits.
896 3. The appearance of (?J) at the start of a pattern set the DUPNAMES option,
897 but did not set the internal JCHANGED flag - either of these is enough to
898 control the way the "get" function works - but the PCRE_INFO_JCHANGED
899 facility is supposed to tell if (?J) was ever used, so now (?J) at the
900 start sets both bits.
902 4. Added options (at build time, compile time, exec time) to change \R from
903 matching any Unicode line ending sequence to just matching CR, LF, or CRLF.
905 5. doc/pcresyntax.html was missing from the distribution.
907 6. Put back the definition of PCRE_ERROR_NULLWSLIMIT, for backward
908 compatibility, even though it is no longer used.
910 7. Added macro for snprintf to pcrecpp_unittest.cc and also for strtoll and
911 strtoull to pcrecpp.cc to select the available functions in WIN32 when the
912 windows.h file is present (where different names are used). [This was
913 reversed later after testing - see 16 below.]
915 8. Changed all #include <config.h> to #include "config.h". There were also
916 some further <pcre.h> cases that I changed to "pcre.h".
918 9. When pcregrep was used with the --colour option, it missed the line ending
919 sequence off the lines that it output.
921 10. It was pointed out to me that arrays of string pointers cause lots of
922 relocations when a shared library is dynamically loaded. A technique of
923 using a single long string with a table of offsets can drastically reduce
924 these. I have refactored PCRE in four places to do this. The result is
925 dramatic:
927 Originally: 290
928 After changing UCP table: 187
929 After changing error message table: 43
930 After changing table of "verbs" 36
931 After changing table of Posix names 22
933 Thanks to the folks working on Gregex for glib for this insight.
935 11. --disable-stack-for-recursion caused compiling to fail unless -enable-
936 unicode-properties was also set.
938 12. Updated the tests so that they work when \R is defaulted to ANYCRLF.
940 13. Added checks for ANY and ANYCRLF to pcrecpp.cc where it previously
941 checked only for CRLF.
943 14. Added casts to pcretest.c to avoid compiler warnings.
945 15. Added Craig's patch to various pcrecpp modules to avoid compiler warnings.
947 16. Added Craig's patch to remove the WINDOWS_H tests, that were not working,
948 and instead check for _strtoi64 explicitly, and avoid the use of snprintf()
949 entirely. This removes changes made in 7 above.
951 17. The CMake files have been updated, and there is now more information about
952 building with CMake in the NON-UNIX-USE document.
955 Version 7.3 28-Aug-07
956 ---------------------
958 1. In the rejigging of the build system that eventually resulted in 7.1, the
959 line "#include <pcre.h>" was included in pcre_internal.h. The use of angle
960 brackets there is not right, since it causes compilers to look for an
961 installed pcre.h, not the version that is in the source that is being
962 compiled (which of course may be different). I have changed it back to:
964 #include "pcre.h"
966 I have a vague recollection that the change was concerned with compiling in
967 different directories, but in the new build system, that is taken care of
968 by the VPATH setting the Makefile.
970 2. The pattern .*$ when run in not-DOTALL UTF-8 mode with newline=any failed
971 when the subject happened to end in the byte 0x85 (e.g. if the last
972 character was \x{1ec5}). *Character* 0x85 is one of the "any" newline
973 characters but of course it shouldn't be taken as a newline when it is part
974 of another character. The bug was that, for an unlimited repeat of . in
975 not-DOTALL UTF-8 mode, PCRE was advancing by bytes rather than by
976 characters when looking for a newline.
978 3. A small performance improvement in the DOTALL UTF-8 mode .* case.
980 4. Debugging: adjusted the names of opcodes for different kinds of parentheses
981 in debug output.
983 5. Arrange to use "%I64d" instead of "%lld" and "%I64u" instead of "%llu" for
984 long printing in the pcrecpp unittest when running under MinGW.
986 6. ESC_K was left out of the EBCDIC table.
988 7. Change 7.0/38 introduced a new limit on the number of nested non-capturing
989 parentheses; I made it 1000, which seemed large enough. Unfortunately, the
990 limit also applies to "virtual nesting" when a pattern is recursive, and in
991 this case 1000 isn't so big. I have been able to remove this limit at the
992 expense of backing off one optimization in certain circumstances. Normally,
993 when pcre_exec() would call its internal match() function recursively and
994 immediately return the result unconditionally, it uses a "tail recursion"
995 feature to save stack. However, when a subpattern that can match an empty
996 string has an unlimited repetition quantifier, it no longer makes this
997 optimization. That gives it a stack frame in which to save the data for
998 checking that an empty string has been matched. Previously this was taken
999 from the 1000-entry workspace that had been reserved. So now there is no
1000 explicit limit, but more stack is used.
1002 8. Applied Daniel's patches to solve problems with the import/export magic
1003 syntax that is required for Windows, and which was going wrong for the
1004 pcreposix and pcrecpp parts of the library. These were overlooked when this
1005 problem was solved for the main library.
1007 9. There were some crude static tests to avoid integer overflow when computing
1008 the size of patterns that contain repeated groups with explicit upper
1009 limits. As the maximum quantifier is 65535, the maximum group length was
1010 set at 30,000 so that the product of these two numbers did not overflow a
1011 32-bit integer. However, it turns out that people want to use groups that
1012 are longer than 30,000 bytes (though not repeat them that many times).
1013 Change 7.0/17 (the refactoring of the way the pattern size is computed) has
1014 made it possible to implement the integer overflow checks in a much more
1015 dynamic way, which I have now done. The artificial limitation on group
1016 length has been removed - we now have only the limit on the total length of
1017 the compiled pattern, which depends on the LINK_SIZE setting.
1019 10. Fixed a bug in the documentation for get/copy named substring when
1020 duplicate names are permitted. If none of the named substrings are set, the
1021 functions return PCRE_ERROR_NOSUBSTRING (7); the doc said they returned an
1022 empty string.
1024 11. Because Perl interprets \Q...\E at a high level, and ignores orphan \E
1025 instances, patterns such as [\Q\E] or [\E] or even [^\E] cause an error,
1026 because the ] is interpreted as the first data character and the
1027 terminating ] is not found. PCRE has been made compatible with Perl in this
1028 regard. Previously, it interpreted [\Q\E] as an empty class, and [\E] could
1029 cause memory overwriting.
1031 10. Like Perl, PCRE automatically breaks an unlimited repeat after an empty
1032 string has been matched (to stop an infinite loop). It was not recognizing
1033 a conditional subpattern that could match an empty string if that
1034 subpattern was within another subpattern. For example, it looped when
1035 trying to match (((?(1)X|))*) but it was OK with ((?(1)X|)*) where the
1036 condition was not nested. This bug has been fixed.
1038 12. A pattern like \X?\d or \P{L}?\d in non-UTF-8 mode could cause a backtrack
1039 past the start of the subject in the presence of bytes with the top bit
1040 set, for example "\x8aBCD".
1042 13. Added Perl 5.10 experimental backtracking controls (*FAIL), (*F), (*PRUNE),
1043 (*SKIP), (*THEN), (*COMMIT), and (*ACCEPT).
1045 14. Optimized (?!) to (*FAIL).
1047 15. Updated the test for a valid UTF-8 string to conform to the later RFC 3629.
1048 This restricts code points to be within the range 0 to 0x10FFFF, excluding
1049 the "low surrogate" sequence 0xD800 to 0xDFFF. Previously, PCRE allowed the
1050 full range 0 to 0x7FFFFFFF, as defined by RFC 2279. Internally, it still
1051 does: it's just the validity check that is more restrictive.
1053 16. Inserted checks for integer overflows during escape sequence (backslash)
1054 processing, and also fixed erroneous offset values for syntax errors during
1055 backslash processing.
1057 17. Fixed another case of looking too far back in non-UTF-8 mode (cf 12 above)
1058 for patterns like [\PPP\x8a]{1,}\x80 with the subject "A\x80".
1060 18. An unterminated class in a pattern like (?1)\c[ with a "forward reference"
1061 caused an overrun.
1063 19. A pattern like (?:[\PPa*]*){8,} which had an "extended class" (one with
1064 something other than just ASCII characters) inside a group that had an
1065 unlimited repeat caused a loop at compile time (while checking to see
1066 whether the group could match an empty string).
1068 20. Debugging a pattern containing \p or \P could cause a crash. For example,
1069 [\P{Any}] did so. (Error in the code for printing property names.)
1071 21. An orphan \E inside a character class could cause a crash.
1073 22. A repeated capturing bracket such as (A)? could cause a wild memory
1074 reference during compilation.
1076 23. There are several functions in pcre_compile() that scan along a compiled
1077 expression for various reasons (e.g. to see if it's fixed length for look
1078 behind). There were bugs in these functions when a repeated \p or \P was
1079 present in the pattern. These operators have additional parameters compared
1080 with \d, etc, and these were not being taken into account when moving along
1081 the compiled data. Specifically:
1083 (a) A item such as \p{Yi}{3} in a lookbehind was not treated as fixed
1084 length.
1086 (b) An item such as \pL+ within a repeated group could cause crashes or
1087 loops.
1089 (c) A pattern such as \p{Yi}+(\P{Yi}+)(?1) could give an incorrect
1090 "reference to non-existent subpattern" error.
1092 (d) A pattern like (\P{Yi}{2}\277)? could loop at compile time.
1094 24. A repeated \S or \W in UTF-8 mode could give wrong answers when multibyte
1095 characters were involved (for example /\S{2}/8g with "A\x{a3}BC").
1097 25. Using pcregrep in multiline, inverted mode (-Mv) caused it to loop.
1099 26. Patterns such as [\P{Yi}A] which include \p or \P and just one other
1100 character were causing crashes (broken optimization).
1102 27. Patterns such as (\P{Yi}*\277)* (group with possible zero repeat containing
1103 \p or \P) caused a compile-time loop.
1105 28. More problems have arisen in unanchored patterns when CRLF is a valid line
1106 break. For example, the unstudied pattern [\r\n]A does not match the string
1107 "\r\nA" because change 7.0/46 below moves the current point on by two
1108 characters after failing to match at the start. However, the pattern \nA
1109 *does* match, because it doesn't start till \n, and if [\r\n]A is studied,
1110 the same is true. There doesn't seem any very clean way out of this, but
1111 what I have chosen to do makes the common cases work: PCRE now takes note
1112 of whether there can be an explicit match for \r or \n anywhere in the
1113 pattern, and if so, 7.0/46 no longer applies. As part of this change,
1114 there's a new PCRE_INFO_HASCRORLF option for finding out whether a compiled
1115 pattern has explicit CR or LF references.
1117 29. Added (*CR) etc for changing newline setting at start of pattern.
1120 Version 7.2 19-Jun-07
1121 ---------------------
1123 1. If the fr_FR locale cannot be found for test 3, try the "french" locale,
1124 which is apparently normally available under Windows.
1126 2. Re-jig the pcregrep tests with different newline settings in an attempt
1127 to make them independent of the local environment's newline setting.
1129 3. Add code to configure.ac to remove -g from the CFLAGS default settings.
1131 4. Some of the "internals" tests were previously cut out when the link size
1132 was not 2, because the output contained actual offsets. The recent new
1133 "Z" feature of pcretest means that these can be cut out, making the tests
1134 usable with all link sizes.
1136 5. Implemented Stan Switzer's goto replacement for longjmp() when not using
1137 stack recursion. This gives a massive performance boost under BSD, but just
1138 a small improvement under Linux. However, it saves one field in the frame
1139 in all cases.
1141 6. Added more features from the forthcoming Perl 5.10:
1143 (a) (?-n) (where n is a string of digits) is a relative subroutine or
1144 recursion call. It refers to the nth most recently opened parentheses.
1146 (b) (?+n) is also a relative subroutine call; it refers to the nth next
1147 to be opened parentheses.
1149 (c) Conditions that refer to capturing parentheses can be specified
1150 relatively, for example, (?(-2)... or (?(+3)...
1152 (d) \K resets the start of the current match so that everything before
1153 is not part of it.
1155 (e) \k{name} is synonymous with \k<name> and \k'name' (.NET compatible).
1157 (f) \g{name} is another synonym - part of Perl 5.10's unification of
1158 reference syntax.
1160 (g) (?| introduces a group in which the numbering of parentheses in each
1161 alternative starts with the same number.
1163 (h) \h, \H, \v, and \V match horizontal and vertical whitespace.
1165 7. Added two new calls to pcre_fullinfo(): PCRE_INFO_OKPARTIAL and
1168 8. A pattern such as (.*(.)?)* caused pcre_exec() to fail by either not
1169 terminating or by crashing. Diagnosed by Viktor Griph; it was in the code
1170 for detecting groups that can match an empty string.
1172 9. A pattern with a very large number of alternatives (more than several
1173 hundred) was running out of internal workspace during the pre-compile
1174 phase, where pcre_compile() figures out how much memory will be needed. A
1175 bit of new cunning has reduced the workspace needed for groups with
1176 alternatives. The 1000-alternative test pattern now uses 12 bytes of
1177 workspace instead of running out of the 4096 that are available.
1179 10. Inserted some missing (unsigned int) casts to get rid of compiler warnings.
1181 11. Applied patch from Google to remove an optimization that didn't quite work.
1182 The report of the bug said:
1184 pcrecpp::RE("a*").FullMatch("aaa") matches, while
1185 pcrecpp::RE("a*?").FullMatch("aaa") does not, and
1186 pcrecpp::RE("a*?\\z").FullMatch("aaa") does again.
1188 12. If \p or \P was used in non-UTF-8 mode on a character greater than 127
1189 it matched the wrong number of bytes.
1192 Version 7.1 24-Apr-07
1193 ---------------------
1195 1. Applied Bob Rossi and Daniel G's patches to convert the build system to one
1196 that is more "standard", making use of automake and other Autotools. There
1197 is some re-arrangement of the files and adjustment of comments consequent
1198 on this.
1200 2. Part of the patch fixed a problem with the pcregrep tests. The test of -r
1201 for recursive directory scanning broke on some systems because the files
1202 are not scanned in any specific order and on different systems the order
1203 was different. A call to "sort" has been inserted into RunGrepTest for the
1204 approprate test as a short-term fix. In the longer term there may be an
1205 alternative.
1207 3. I had an email from Eric Raymond about problems translating some of PCRE's
1208 man pages to HTML (despite the fact that I distribute HTML pages, some
1209 people do their own conversions for various reasons). The problems
1210 concerned the use of low-level troff macros .br and .in. I have therefore
1211 removed all such uses from the man pages (some were redundant, some could
1212 be replaced by .nf/.fi pairs). The 132html script that I use to generate
1213 HTML has been updated to handle .nf/.fi and to complain if it encounters
1214 .br or .in.
1216 4. Updated comments in configure.ac that get placed in config.h.in and also
1217 arranged for config.h to be included in the distribution, with the name
1218 config.h.generic, for the benefit of those who have to compile without
1219 Autotools (compare pcre.h, which is now distributed as pcre.h.generic).
1221 5. Updated the support (such as it is) for Virtual Pascal, thanks to Stefan
1222 Weber: (1) pcre_internal.h was missing some function renames; (2) updated
1223 makevp.bat for the current PCRE, using the additional files
1224 makevp_c.txt, makevp_l.txt, and pcregexp.pas.
1226 6. A Windows user reported a minor discrepancy with test 2, which turned out
1227 to be caused by a trailing space on an input line that had got lost in his
1228 copy. The trailing space was an accident, so I've just removed it.
1230 7. Add -Wl,-R... flags in pcre-config.in for *BSD* systems, as I'm told
1231 that is needed.
1233 8. Mark ucp_table (in ucptable.h) and ucp_gentype (in pcre_ucp_searchfuncs.c)
1234 as "const" (a) because they are and (b) because it helps the PHP
1235 maintainers who have recently made a script to detect big data structures
1236 in the php code that should be moved to the .rodata section. I remembered
1237 to update Builducptable as well, so it won't revert if ucptable.h is ever
1238 re-created.
1240 9. Added some extra #ifdef SUPPORT_UTF8 conditionals into pcretest.c,
1241 pcre_printint.src, pcre_compile.c, pcre_study.c, and pcre_tables.c, in
1242 order to be able to cut out the UTF-8 tables in the latter when UTF-8
1243 support is not required. This saves 1.5-2K of code, which is important in
1244 some applications.
1246 Later: more #ifdefs are needed in pcre_ord2utf8.c and pcre_valid_utf8.c
1247 so as not to refer to the tables, even though these functions will never be
1248 called when UTF-8 support is disabled. Otherwise there are problems with a
1249 shared library.
1251 10. Fixed two bugs in the emulated memmove() function in pcre_internal.h:
1253 (a) It was defining its arguments as char * instead of void *.
1255 (b) It was assuming that all moves were upwards in memory; this was true
1256 a long time ago when I wrote it, but is no longer the case.
1258 The emulated memove() is provided for those environments that have neither
1259 memmove() nor bcopy(). I didn't think anyone used it these days, but that
1260 is clearly not the case, as these two bugs were recently reported.
1262 11. The script PrepareRelease is now distributed: it calls 132html, CleanTxt,
1263 and Detrail to create the HTML documentation, the .txt form of the man
1264 pages, and it removes trailing spaces from listed files. It also creates
1265 pcre.h.generic and config.h.generic from pcre.h and config.h. In the latter
1266 case, it wraps all the #defines with #ifndefs. This script should be run
1267 before "make dist".
1269 12. Fixed two fairly obscure bugs concerned with quantified caseless matching
1270 with Unicode property support.
1272 (a) For a maximizing quantifier, if the two different cases of the
1273 character were of different lengths in their UTF-8 codings (there are
1274 some cases like this - I found 11), and the matching function had to
1275 back up over a mixture of the two cases, it incorrectly assumed they
1276 were both the same length.
1278 (b) When PCRE was configured to use the heap rather than the stack for
1279 recursion during matching, it was not correctly preserving the data for
1280 the other case of a UTF-8 character when checking ahead for a match
1281 while processing a minimizing repeat. If the check also involved
1282 matching a wide character, but failed, corruption could cause an
1283 erroneous result when trying to check for a repeat of the original
1284 character.
1286 13. Some tidying changes to the testing mechanism:
1288 (a) The RunTest script now detects the internal link size and whether there
1289 is UTF-8 and UCP support by running ./pcretest -C instead of relying on
1290 values substituted by "configure". (The RunGrepTest script already did
1291 this for UTF-8.) The configure.ac script no longer substitutes the
1292 relevant variables.
1294 (b) The debugging options /B and /D in pcretest show the compiled bytecode
1295 with length and offset values. This means that the output is different
1296 for different internal link sizes. Test 2 is skipped for link sizes
1297 other than 2 because of this, bypassing the problem. Unfortunately,
1298 there was also a test in test 3 (the locale tests) that used /B and
1299 failed for link sizes other than 2. Rather than cut the whole test out,
1300 I have added a new /Z option to pcretest that replaces the length and
1301 offset values with spaces. This is now used to make test 3 independent
1302 of link size. (Test 2 will be tidied up later.)
1304 14. If erroroffset was passed as NULL to pcre_compile, it provoked a
1305 segmentation fault instead of returning the appropriate error message.
1307 15. In multiline mode when the newline sequence was set to "any", the pattern
1308 ^$ would give a match between the \r and \n of a subject such as "A\r\nB".
1309 This doesn't seem right; it now treats the CRLF combination as the line
1310 ending, and so does not match in that case. It's only a pattern such as ^$
1311 that would hit this one: something like ^ABC$ would have failed after \r
1312 and then tried again after \r\n.
1314 16. Changed the comparison command for RunGrepTest from "diff -u" to "diff -ub"
1315 in an attempt to make files that differ only in their line terminators
1316 compare equal. This works on Linux.
1318 17. Under certain error circumstances pcregrep might try to free random memory
1319 as it exited. This is now fixed, thanks to valgrind.
1321 19. In pcretest, if the pattern /(?m)^$/g<any> was matched against the string
1322 "abc\r\n\r\n", it found an unwanted second match after the second \r. This
1323 was because its rules for how to advance for /g after matching an empty
1324 string at the end of a line did not allow for this case. They now check for
1325 it specially.
1327 20. pcretest is supposed to handle patterns and data of any length, by
1328 extending its buffers when necessary. It was getting this wrong when the
1329 buffer for a data line had to be extended.
1331 21. Added PCRE_NEWLINE_ANYCRLF which is like ANY, but matches only CR, LF, or
1332 CRLF as a newline sequence.
1334 22. Code for handling Unicode properties in pcre_dfa_exec() wasn't being cut
1335 out by #ifdef SUPPORT_UCP. This did no harm, as it could never be used, but
1336 I have nevertheless tidied it up.
1338 23. Added some casts to kill warnings from HP-UX ia64 compiler.
1340 24. Added a man page for pcre-config.
1343 Version 7.0 19-Dec-06
1344 ---------------------
1346 1. Fixed a signed/unsigned compiler warning in pcre_compile.c, shown up by
1347 moving to gcc 4.1.1.
1349 2. The -S option for pcretest uses setrlimit(); I had omitted to #include
1350 sys/time.h, which is documented as needed for this function. It doesn't
1351 seem to matter on Linux, but it showed up on some releases of OS X.
1353 3. It seems that there are systems where bytes whose values are greater than
1354 127 match isprint() in the "C" locale. The "C" locale should be the
1355 default when a C program starts up. In most systems, only ASCII printing
1356 characters match isprint(). This difference caused the output from pcretest
1357 to vary, making some of the tests fail. I have changed pcretest so that:
1359 (a) When it is outputting text in the compiled version of a pattern, bytes
1360 other than 32-126 are always shown as hex escapes.
1362 (b) When it is outputting text that is a matched part of a subject string,
1363 it does the same, unless a different locale has been set for the match
1364 (using the /L modifier). In this case, it uses isprint() to decide.
1366 4. Fixed a major bug that caused incorrect computation of the amount of memory
1367 required for a compiled pattern when options that changed within the
1368 pattern affected the logic of the preliminary scan that determines the
1369 length. The relevant options are -x, and -i in UTF-8 mode. The result was
1370 that the computed length was too small. The symptoms of this bug were
1371 either the PCRE error "internal error: code overflow" from pcre_compile(),
1372 or a glibc crash with a message such as "pcretest: free(): invalid next
1373 size (fast)". Examples of patterns that provoked this bug (shown in
1374 pcretest format) are:
1376 /(?-x: )/x
1377 /(?x)(?-x: \s*#\s*)/
1378 /((?i)[\x{c0}])/8
1379 /(?i:[\x{c0}])/8
1381 HOWEVER: Change 17 below makes this fix obsolete as the memory computation
1382 is now done differently.
1384 5. Applied patches from Google to: (a) add a QuoteMeta function to the C++
1385 wrapper classes; (b) implement a new function in the C++ scanner that is
1386 more efficient than the old way of doing things because it avoids levels of
1387 recursion in the regex matching; (c) add a paragraph to the documentation
1388 for the FullMatch() function.
1390 6. The escape sequence \n was being treated as whatever was defined as
1391 "newline". Not only was this contrary to the documentation, which states
1392 that \n is character 10 (hex 0A), but it also went horribly wrong when
1393 "newline" was defined as CRLF. This has been fixed.
1395 7. In pcre_dfa_exec.c the value of an unsigned integer (the variable called c)
1396 was being set to -1 for the "end of line" case (supposedly a value that no
1397 character can have). Though this value is never used (the check for end of
1398 line is "zero bytes in current character"), it caused compiler complaints.
1399 I've changed it to 0xffffffff.
1401 8. In pcre_version.c, the version string was being built by a sequence of
1402 C macros that, in the event of PCRE_PRERELEASE being defined as an empty
1403 string (as it is for production releases) called a macro with an empty
1404 argument. The C standard says the result of this is undefined. The gcc
1405 compiler treats it as an empty string (which was what was wanted) but it is
1406 reported that Visual C gives an error. The source has been hacked around to
1407 avoid this problem.
1409 9. On the advice of a Windows user, included <io.h> and <fcntl.h> in Windows
1410 builds of pcretest, and changed the call to _setmode() to use _O_BINARY
1411 instead of 0x8000. Made all the #ifdefs test both _WIN32 and WIN32 (not all
1412 of them did).
1414 10. Originally, pcretest opened its input and output without "b"; then I was
1415 told that "b" was needed in some environments, so it was added for release
1416 5.0 to both the input and output. (It makes no difference on Unix-like
1417 systems.) Later I was told that it is wrong for the input on Windows. I've
1418 now abstracted the modes into two macros, to make it easier to fiddle with
1419 them, and removed "b" from the input mode under Windows.
1421 11. Added pkgconfig support for the C++ wrapper library, libpcrecpp.
1423 12. Added -help and --help to pcretest as an official way of being reminded
1424 of the options.
1426 13. Removed some redundant semicolons after macro calls in pcrecpparg.h.in
1427 and pcrecpp.cc because they annoy compilers at high warning levels.
1429 14. A bit of tidying/refactoring in pcre_exec.c in the main bumpalong loop.
1431 15. Fixed an occurrence of == in configure.ac that should have been = (shell
1432 scripts are not C programs :-) and which was not noticed because it works
1433 on Linux.
1435 16. pcretest is supposed to handle any length of pattern and data line (as one
1436 line or as a continued sequence of lines) by extending its input buffer if
1437 necessary. This feature was broken for very long pattern lines, leading to
1438 a string of junk being passed to pcre_compile() if the pattern was longer
1439 than about 50K.
1441 17. I have done a major re-factoring of the way pcre_compile() computes the
1442 amount of memory needed for a compiled pattern. Previously, there was code
1443 that made a preliminary scan of the pattern in order to do this. That was
1444 OK when PCRE was new, but as the facilities have expanded, it has become
1445 harder and harder to keep it in step with the real compile phase, and there
1446 have been a number of bugs (see for example, 4 above). I have now found a
1447 cunning way of running the real compile function in a "fake" mode that
1448 enables it to compute how much memory it would need, while actually only
1449 ever using a few hundred bytes of working memory and without too many
1450 tests of the mode. This should make future maintenance and development
1451 easier. A side effect of this work is that the limit of 200 on the nesting
1452 depth of parentheses has been removed (though this was never a serious
1453 limitation, I suspect). However, there is a downside: pcre_compile() now
1454 runs more slowly than before (30% or more, depending on the pattern). I
1455 hope this isn't a big issue. There is no effect on runtime performance.
1457 18. Fixed a minor bug in pcretest: if a pattern line was not terminated by a
1458 newline (only possible for the last line of a file) and it was a
1459 pattern that set a locale (followed by /Lsomething), pcretest crashed.
1461 19. Added additional timing features to pcretest. (1) The -tm option now times
1462 matching only, not compiling. (2) Both -t and -tm can be followed, as a
1463 separate command line item, by a number that specifies the number of
1464 repeats to use when timing. The default is 50000; this gives better
1465 precision, but takes uncomfortably long for very large patterns.
1467 20. Extended pcre_study() to be more clever in cases where a branch of a
1468 subpattern has no definite first character. For example, (a*|b*)[cd] would
1469 previously give no result from pcre_study(). Now it recognizes that the
1470 first character must be a, b, c, or d.
1472 21. There was an incorrect error "recursive call could loop indefinitely" if
1473 a subpattern (or the entire pattern) that was being tested for matching an
1474 empty string contained only one non-empty item after a nested subpattern.
1475 For example, the pattern (?>\x{100}*)\d(?R) provoked this error
1476 incorrectly, because the \d was being skipped in the check.
1478 22. The pcretest program now has a new pattern option /B and a command line
1479 option -b, which is equivalent to adding /B to every pattern. This causes
1480 it to show the compiled bytecode, without the additional information that
1481 -d shows. The effect of -d is now the same as -b with -i (and similarly, /D
1482 is the same as /B/I).
1484 23. A new optimization is now able automatically to treat some sequences such
1485 as a*b as a*+b. More specifically, if something simple (such as a character
1486 or a simple class like \d) has an unlimited quantifier, and is followed by
1487 something that cannot possibly match the quantified thing, the quantifier
1488 is automatically "possessified".
1490 24. A recursive reference to a subpattern whose number was greater than 39
1491 went wrong under certain circumstances in UTF-8 mode. This bug could also
1492 have affected the operation of pcre_study().
1494 25. Realized that a little bit of performance could be had by replacing
1495 (c & 0xc0) == 0xc0 with c >= 0xc0 when processing UTF-8 characters.
1497 26. Timing data from pcretest is now shown to 4 decimal places instead of 3.
1499 27. Possessive quantifiers such as a++ were previously implemented by turning
1500 them into atomic groups such as ($>a+). Now they have their own opcodes,
1501 which improves performance. This includes the automatically created ones
1502 from 23 above.
1504 28. A pattern such as (?=(\w+))\1: which simulates an atomic group using a
1505 lookahead was broken if it was not anchored. PCRE was mistakenly expecting
1506 the first matched character to be a colon. This applied both to named and
1507 numbered groups.
1509 29. The ucpinternal.h header file was missing its idempotency #ifdef.
1511 30. I was sent a "project" file called libpcre.a.dev which I understand makes
1512 building PCRE on Windows easier, so I have included it in the distribution.
1514 31. There is now a check in pcretest against a ridiculously large number being
1515 returned by pcre_exec() or pcre_dfa_exec(). If this happens in a /g or /G
1516 loop, the loop is abandoned.
1518 32. Forward references to subpatterns in conditions such as (?(2)...) where
1519 subpattern 2 is defined later cause pcre_compile() to search forwards in
1520 the pattern for the relevant set of parentheses. This search went wrong
1521 when there were unescaped parentheses in a character class, parentheses
1522 escaped with \Q...\E, or parentheses in a #-comment in /x mode.
1524 33. "Subroutine" calls and backreferences were previously restricted to
1525 referencing subpatterns earlier in the regex. This restriction has now
1526 been removed.
1528 34. Added a number of extra features that are going to be in Perl 5.10. On the
1529 whole, these are just syntactic alternatives for features that PCRE had
1530 previously implemented using the Python syntax or my own invention. The
1531 other formats are all retained for compatibility.
1533 (a) Named groups can now be defined as (?<name>...) or (?'name'...) as well
1534 as (?P<name>...). The new forms, as well as being in Perl 5.10, are
1535 also .NET compatible.
1537 (b) A recursion or subroutine call to a named group can now be defined as
1538 (?&name) as well as (?P>name).
1540 (c) A backreference to a named group can now be defined as \k<name> or
1541 \k'name' as well as (?P=name). The new forms, as well as being in Perl
1542 5.10, are also .NET compatible.
1544 (d) A conditional reference to a named group can now use the syntax
1545 (?(<name>) or (?('name') as well as (?(name).
1547 (e) A "conditional group" of the form (?(DEFINE)...) can be used to define
1548 groups (named and numbered) that are never evaluated inline, but can be
1549 called as "subroutines" from elsewhere. In effect, the DEFINE condition
1550 is always false. There may be only one alternative in such a group.
1552 (f) A test for recursion can be given as (?(R1).. or (?(R&name)... as well
1553 as the simple (?(R). The condition is true only if the most recent
1554 recursion is that of the given number or name. It does not search out
1555 through the entire recursion stack.
1557 (g) The escape \gN or \g{N} has been added, where N is a positive or
1558 negative number, specifying an absolute or relative reference.
1560 35. Tidied to get rid of some further signed/unsigned compiler warnings and
1561 some "unreachable code" warnings.
1563 36. Updated the Unicode property tables to Unicode version 5.0.0. Amongst other
1564 things, this adds five new scripts.
1566 37. Perl ignores orphaned \E escapes completely. PCRE now does the same.
1567 There were also incompatibilities regarding the handling of \Q..\E inside
1568 character classes, for example with patterns like [\Qa\E-\Qz\E] where the
1569 hyphen was adjacent to \Q or \E. I hope I've cleared all this up now.
1571 38. Like Perl, PCRE detects when an indefinitely repeated parenthesized group
1572 matches an empty string, and forcibly breaks the loop. There were bugs in
1573 this code in non-simple cases. For a pattern such as ^(a()*)* matched
1574 against aaaa the result was just "a" rather than "aaaa", for example. Two
1575 separate and independent bugs (that affected different cases) have been
1576 fixed.
1578 39. Refactored the code to abolish the use of different opcodes for small
1579 capturing bracket numbers. This is a tidy that I avoided doing when I
1580 removed the limit on the number of capturing brackets for 3.5 back in 2001.
1581 The new approach is not only tidier, it makes it possible to reduce the
1582 memory needed to fix the previous bug (38).
1584 40. Implemented PCRE_NEWLINE_ANY to recognize any of the Unicode newline
1585 sequences (http://unicode.org/unicode/reports/tr18/) as "newline" when
1586 processing dot, circumflex, or dollar metacharacters, or #-comments in /x
1587 mode.
1589 41. Add \R to match any Unicode newline sequence, as suggested in the Unicode
1590 report.
1592 42. Applied patch, originally from Ari Pollak, modified by Google, to allow
1593 copy construction and assignment in the C++ wrapper.
1595 43. Updated pcregrep to support "--newline=any". In the process, I fixed a
1596 couple of bugs that could have given wrong results in the "--newline=crlf"
1597 case.
1599 44. Added a number of casts and did some reorganization of signed/unsigned int
1600 variables following suggestions from Dair Grant. Also renamed the variable
1601 "this" as "item" because it is a C++ keyword.
1603 45. Arranged for dftables to add
1605 #include "pcre_internal.h"
1607 to pcre_chartables.c because without it, gcc 4.x may remove the array
1608 definition from the final binary if PCRE is built into a static library and
1609 dead code stripping is activated.
1611 46. For an unanchored pattern, if a match attempt fails at the start of a
1612 newline sequence, and the newline setting is CRLF or ANY, and the next two
1613 characters are CRLF, advance by two characters instead of one.
1616 Version 6.7 04-Jul-06
1617 ---------------------
1619 1. In order to handle tests when input lines are enormously long, pcretest has
1620 been re-factored so that it automatically extends its buffers when
1621 necessary. The code is crude, but this _is_ just a test program. The
1622 default size has been increased from 32K to 50K.
1624 2. The code in pcre_study() was using the value of the re argument before
1625 testing it for NULL. (Of course, in any sensible call of the function, it
1626 won't be NULL.)
1628 3. The memmove() emulation function in pcre_internal.h, which is used on
1629 systems that lack both memmove() and bcopy() - that is, hardly ever -
1630 was missing a "static" storage class specifier.
1632 4. When UTF-8 mode was not set, PCRE looped when compiling certain patterns
1633 containing an extended class (one that cannot be represented by a bitmap
1634 because it contains high-valued characters or Unicode property items, e.g.
1635 [\pZ]). Almost always one would set UTF-8 mode when processing such a
1636 pattern, but PCRE should not loop if you do not (it no longer does).
1637 [Detail: two cases were found: (a) a repeated subpattern containing an
1638 extended class; (b) a recursive reference to a subpattern that followed a
1639 previous extended class. It wasn't skipping over the extended class
1640 correctly when UTF-8 mode was not set.]
1642 5. A negated single-character class was not being recognized as fixed-length
1643 in lookbehind assertions such as (?<=[^f]), leading to an incorrect
1644 compile error "lookbehind assertion is not fixed length".
1646 6. The RunPerlTest auxiliary script was showing an unexpected difference
1647 between PCRE and Perl for UTF-8 tests. It turns out that it is hard to
1648 write a Perl script that can interpret lines of an input file either as
1649 byte characters or as UTF-8, which is what "perltest" was being required to
1650 do for the non-UTF-8 and UTF-8 tests, respectively. Essentially what you
1651 can't do is switch easily at run time between having the "use utf8;" pragma
1652 or not. In the end, I fudged it by using the RunPerlTest script to insert
1653 "use utf8;" explicitly for the UTF-8 tests.
1655 7. In multiline (/m) mode, PCRE was matching ^ after a terminating newline at
1656 the end of the subject string, contrary to the documentation and to what
1657 Perl does. This was true of both matching functions. Now it matches only at
1658 the start of the subject and immediately after *internal* newlines.
1660 8. A call of pcre_fullinfo() from pcretest to get the option bits was passing
1661 a pointer to an int instead of a pointer to an unsigned long int. This
1662 caused problems on 64-bit systems.
1664 9. Applied a patch from the folks at Google to pcrecpp.cc, to fix "another
1665 instance of the 'standard' template library not being so standard".
1667 10. There was no check on the number of named subpatterns nor the maximum
1668 length of a subpattern name. The product of these values is used to compute
1669 the size of the memory block for a compiled pattern. By supplying a very
1670 long subpattern name and a large number of named subpatterns, the size
1671 computation could be caused to overflow. This is now prevented by limiting
1672 the length of names to 32 characters, and the number of named subpatterns
1673 to 10,000.
1675 11. Subpatterns that are repeated with specific counts have to be replicated in
1676 the compiled pattern. The size of memory for this was computed from the
1677 length of the subpattern and the repeat count. The latter is limited to
1678 65535, but there was no limit on the former, meaning that integer overflow
1679 could in principle occur. The compiled length of a repeated subpattern is
1680 now limited to 30,000 bytes in order to prevent this.
1682 12. Added the optional facility to have named substrings with the same name.
1684 13. Added the ability to use a named substring as a condition, using the
1685 Python syntax: (?(name)yes|no). This overloads (?(R)... and names that
1686 are numbers (not recommended). Forward references are permitted.
1688 14. Added forward references in named backreferences (if you see what I mean).
1690 15. In UTF-8 mode, with the PCRE_DOTALL option set, a quantified dot in the
1691 pattern could run off the end of the subject. For example, the pattern
1692 "(?s)(.{1,5})"8 did this with the subject "ab".
1694 16. If PCRE_DOTALL or PCRE_MULTILINE were set, pcre_dfa_exec() behaved as if
1695 PCRE_CASELESS was set when matching characters that were quantified with ?
1696 or *.
1698 17. A character class other than a single negated character that had a minimum
1699 but no maximum quantifier - for example [ab]{6,} - was not handled
1700 correctly by pce_dfa_exec(). It would match only one character.
1702 18. A valid (though odd) pattern that looked like a POSIX character
1703 class but used an invalid character after [ (for example [[,abc,]]) caused
1704 pcre_compile() to give the error "Failed: internal error: code overflow" or
1705 in some cases to crash with a glibc free() error. This could even happen if
1706 the pattern terminated after [[ but there just happened to be a sequence of
1707 letters, a binary zero, and a closing ] in the memory that followed.
1709 19. Perl's treatment of octal escapes in the range \400 to \777 has changed
1710 over the years. Originally (before any Unicode support), just the bottom 8
1711 bits were taken. Thus, for example, \500 really meant \100. Nowadays the
1712 output from "man perlunicode" includes this:
1714 The regular expression compiler produces polymorphic opcodes. That
1715 is, the pattern adapts to the data and automatically switches to
1716 the Unicode character scheme when presented with Unicode data--or
1717 instead uses a traditional byte scheme when presented with byte
1718 data.
1720 Sadly, a wide octal escape does not cause a switch, and in a string with
1721 no other multibyte characters, these octal escapes are treated as before.
1722 Thus, in Perl, the pattern /\500/ actually matches \100 but the pattern
1723 /\500|\x{1ff}/ matches \500 or \777 because the whole thing is treated as a
1724 Unicode string.
1726 I have not perpetrated such confusion in PCRE. Up till now, it took just
1727 the bottom 8 bits, as in old Perl. I have now made octal escapes with
1728 values greater than \377 illegal in non-UTF-8 mode. In UTF-8 mode they
1729 translate to the appropriate multibyte character.
1731 29. Applied some refactoring to reduce the number of warnings from Microsoft
1732 and Borland compilers. This has included removing the fudge introduced
1733 seven years ago for the OS/2 compiler (see 2.02/2 below) because it caused
1734 a warning about an unused variable.
1736 21. PCRE has not included VT (character 0x0b) in the set of whitespace
1737 characters since release 4.0, because Perl (from release 5.004) does not.
1738 [Or at least, is documented not to: some releases seem to be in conflict
1739 with the documentation.] However, when a pattern was studied with
1740 pcre_study() and all its branches started with \s, PCRE still included VT
1741 as a possible starting character. Of course, this did no harm; it just
1742 caused an unnecessary match attempt.
1744 22. Removed a now-redundant internal flag bit that recorded the fact that case
1745 dependency changed within the pattern. This was once needed for "required
1746 byte" processing, but is no longer used. This recovers a now-scarce options
1747 bit. Also moved the least significant internal flag bit to the most-
1748 significant bit of the word, which was not previously used (hangover from
1749 the days when it was an int rather than a uint) to free up another bit for
1750 the future.
1752 23. Added support for CRLF line endings as well as CR and LF. As well as the
1753 default being selectable at build time, it can now be changed at runtime
1754 via the PCRE_NEWLINE_xxx flags. There are now options for pcregrep to
1755 specify that it is scanning data with non-default line endings.
1757 24. Changed the definition of CXXLINK to make it agree with the definition of
1758 LINK in the Makefile, by replacing LDFLAGS to CXXFLAGS.
1760 25. Applied Ian Taylor's patches to avoid using another stack frame for tail
1761 recursions. This makes a big different to stack usage for some patterns.
1763 26. If a subpattern containing a named recursion or subroutine reference such
1764 as (?P>B) was quantified, for example (xxx(?P>B)){3}, the calculation of
1765 the space required for the compiled pattern went wrong and gave too small a
1766 value. Depending on the environment, this could lead to "Failed: internal
1767 error: code overflow at offset 49" or "glibc detected double free or
1768 corruption" errors.
1770 27. Applied patches from Google (a) to support the new newline modes and (b) to
1771 advance over multibyte UTF-8 characters in GlobalReplace.
1773 28. Change free() to pcre_free() in pcredemo.c. Apparently this makes a
1774 difference for some implementation of PCRE in some Windows version.
1776 29. Added some extra testing facilities to pcretest:
1778 \q<number> in a data line sets the "match limit" value
1779 \Q<number> in a data line sets the "match recursion limt" value
1780 -S <number> sets the stack size, where <number> is in megabytes
1782 The -S option isn't available for Windows.
1785 Version 6.6 06-Feb-06
1786 ---------------------
1788 1. Change 16(a) for 6.5 broke things, because PCRE_DATA_SCOPE was not defined
1789 in pcreposix.h. I have copied the definition from pcre.h.
1791 2. Change 25 for 6.5 broke compilation in a build directory out-of-tree
1792 because pcre.h is no longer a built file.
1794 3. Added Jeff Friedl's additional debugging patches to pcregrep. These are
1795 not normally included in the compiled code.
1798 Version 6.5 01-Feb-06
1799 ---------------------
1801 1. When using the partial match feature with pcre_dfa_exec(), it was not
1802 anchoring the second and subsequent partial matches at the new starting
1803 point. This could lead to incorrect results. For example, with the pattern
1804 /1234/, partially matching against "123" and then "a4" gave a match.
1806 2. Changes to pcregrep:
1808 (a) All non-match returns from pcre_exec() were being treated as failures
1809 to match the line. Now, unless the error is PCRE_ERROR_NOMATCH, an
1810 error message is output. Some extra information is given for the
1812 probably the only errors that are likely to be caused by users (by
1813 specifying a regex that has nested indefinite repeats, for instance).
1814 If there are more than 20 of these errors, pcregrep is abandoned.
1816 (b) A binary zero was treated as data while matching, but terminated the
1817 output line if it was written out. This has been fixed: binary zeroes
1818 are now no different to any other data bytes.
1820 (c) Whichever of the LC_ALL or LC_CTYPE environment variables is set is
1821 used to set a locale for matching. The --locale=xxxx long option has
1822 been added (no short equivalent) to specify a locale explicitly on the
1823 pcregrep command, overriding the environment variables.
1825 (d) When -B was used with -n, some line numbers in the output were one less
1826 than they should have been.
1828 (e) Added the -o (--only-matching) option.
1830 (f) If -A or -C was used with -c (count only), some lines of context were
1831 accidentally printed for the final match.
1833 (g) Added the -H (--with-filename) option.
1835 (h) The combination of options -rh failed to suppress file names for files
1836 that were found from directory arguments.
1838 (i) Added the -D (--devices) and -d (--directories) options.
1840 (j) Added the -F (--fixed-strings) option.
1842 (k) Allow "-" to be used as a file name for -f as well as for a data file.
1844 (l) Added the --colo(u)r option.
1846 (m) Added Jeffrey Friedl's -S testing option, but within #ifdefs so that it
1847 is not present by default.
1849 3. A nasty bug was discovered in the handling of recursive patterns, that is,
1850 items such as (?R) or (?1), when the recursion could match a number of
1851 alternatives. If it matched one of the alternatives, but subsequently,
1852 outside the recursion, there was a failure, the code tried to back up into
1853 the recursion. However, because of the way PCRE is implemented, this is not
1854 possible, and the result was an incorrect result from the match.
1856 In order to prevent this happening, the specification of recursion has
1857 been changed so that all such subpatterns are automatically treated as
1858 atomic groups. Thus, for example, (?R) is treated as if it were (?>(?R)).
1860 4. I had overlooked the fact that, in some locales, there are characters for
1861 which isalpha() is true but neither isupper() nor islower() are true. In
1862 the fr_FR locale, for instance, the \xAA and \xBA characters (ordmasculine
1863 and ordfeminine) are like this. This affected the treatment of \w and \W
1864 when they appeared in character classes, but not when they appeared outside
1865 a character class. The bit map for "word" characters is now created
1866 separately from the results of isalnum() instead of just taking it from the
1867 upper, lower, and digit maps. (Plus the underscore character, of course.)
1869 5. The above bug also affected the handling of POSIX character classes such as
1870 [[:alpha:]] and [[:alnum:]]. These do not have their own bit maps in PCRE's
1871 permanent tables. Instead, the bit maps for such a class were previously
1872 created as the appropriate unions of the upper, lower, and digit bitmaps.
1873 Now they are created by subtraction from the [[:word:]] class, which has
1874 its own bitmap.
1876 6. The [[:blank:]] character class matches horizontal, but not vertical space.
1877 It is created by subtracting the vertical space characters (\x09, \x0a,
1878 \x0b, \x0c) from the [[:space:]] bitmap. Previously, however, the
1879 subtraction was done in the overall bitmap for a character class, meaning
1880 that a class such as [\x0c[:blank:]] was incorrect because \x0c would not
1881 be recognized. This bug has been fixed.
1883 7. Patches from the folks at Google:
1885 (a) pcrecpp.cc: "to handle a corner case that may or may not happen in
1886 real life, but is still worth protecting against".
1888 (b) pcrecpp.cc: "corrects a bug when negative radixes are used with
1889 regular expressions".
1891 (c) pcre_scanner.cc: avoid use of std::count() because not all systems
1892 have it.
1894 (d) Split off pcrecpparg.h from pcrecpp.h and had the former built by
1895 "configure" and the latter not, in order to fix a problem somebody had
1896 with compiling the Arg class on HP-UX.
1898 (e) Improve the error-handling of the C++ wrapper a little bit.
1900 (f) New tests for checking recursion limiting.
1902 8. The pcre_memmove() function, which is used only if the environment does not
1903 have a standard memmove() function (and is therefore rarely compiled),
1904 contained two bugs: (a) use of int instead of size_t, and (b) it was not
1905 returning a result (though PCRE never actually uses the result).
1907 9. In the POSIX regexec() interface, if nmatch is specified as a ridiculously
1908 large number - greater than INT_MAX/(3*sizeof(int)) - REG_ESPACE is
1909 returned instead of calling malloc() with an overflowing number that would
1910 most likely cause subsequent chaos.
1912 10. The debugging option of pcretest was not showing the NO_AUTO_CAPTURE flag.
1914 11. The POSIX flag REG_NOSUB is now supported. When a pattern that was compiled
1915 with this option is matched, the nmatch and pmatch options of regexec() are
1916 ignored.
1918 12. Added REG_UTF8 to the POSIX interface. This is not defined by POSIX, but is
1919 provided in case anyone wants to the the POSIX interface with UTF-8
1920 strings.
1922 13. Added CXXLDFLAGS to the Makefile parameters to provide settings only on the
1923 C++ linking (needed for some HP-UX environments).
1925 14. Avoid compiler warnings in get_ucpname() when compiled without UCP support
1926 (unused parameter) and in the pcre_printint() function (omitted "default"
1927 switch label when the default is to do nothing).
1929 15. Added some code to make it possible, when PCRE is compiled as a C++
1930 library, to replace subject pointers for pcre_exec() with a smart pointer
1931 class, thus making it possible to process discontinuous strings.
1933 16. The two macros PCRE_EXPORT and PCRE_DATA_SCOPE are confusing, and perform
1934 much the same function. They were added by different people who were trying
1935 to make PCRE easy to compile on non-Unix systems. It has been suggested
1936 that PCRE_EXPORT be abolished now that there is more automatic apparatus
1937 for compiling on Windows systems. I have therefore replaced it with
1938 PCRE_DATA_SCOPE. This is set automatically for Windows; if not set it
1939 defaults to "extern" for C or "extern C" for C++, which works fine on
1940 Unix-like systems. It is now possible to override the value of PCRE_DATA_
1941 SCOPE with something explicit in config.h. In addition:
1943 (a) pcreposix.h still had just "extern" instead of either of these macros;
1944 I have replaced it with PCRE_DATA_SCOPE.
1946 (b) Functions such as _pcre_xclass(), which are internal to the library,
1947 but external in the C sense, all had PCRE_EXPORT in their definitions.
1948 This is apparently wrong for the Windows case, so I have removed it.
1949 (It makes no difference on Unix-like systems.)
1951 17. Added a new limit, MATCH_LIMIT_RECURSION, which limits the depth of nesting
1952 of recursive calls to match(). This is different to MATCH_LIMIT because
1953 that limits the total number of calls to match(), not all of which increase
1954 the depth of recursion. Limiting the recursion depth limits the amount of
1955 stack (or heap if NO_RECURSE is set) that is used. The default can be set
1956 when PCRE is compiled, and changed at run time. A patch from Google adds
1957 this functionality to the C++ interface.
1959 18. Changes to the handling of Unicode character properties:
1961 (a) Updated the table to Unicode 4.1.0.
1963 (b) Recognize characters that are not in the table as "Cn" (undefined).
1965 (c) I revised the way the table is implemented to a much improved format
1966 which includes recognition of ranges. It now supports the ranges that
1967 are defined in UnicodeData.txt, and it also amalgamates other
1968 characters into ranges. This has reduced the number of entries in the
1969 table from around 16,000 to around 3,000, thus reducing its size
1970 considerably. I realized I did not need to use a tree structure after
1971 all - a binary chop search is just as efficient. Having reduced the
1972 number of entries, I extended their size from 6 bytes to 8 bytes to
1973 allow for more data.
1975 (d) Added support for Unicode script names via properties such as \p{Han}.
1977 19. In UTF-8 mode, a backslash followed by a non-Ascii character was not
1978 matching that character.
1980 20. When matching a repeated Unicode property with a minimum greater than zero,
1981 (for example \pL{2,}), PCRE could look past the end of the subject if it
1982 reached it while seeking the minimum number of characters. This could
1983 happen only if some of the characters were more than one byte long, because
1984 there is a check for at least the minimum number of bytes.
1986 21. Refactored the implementation of \p and \P so as to be more general, to
1987 allow for more different types of property in future. This has changed the
1988 compiled form incompatibly. Anybody with saved compiled patterns that use
1989 \p or \P will have to recompile them.
1991 22. Added "Any" and "L&" to the supported property types.
1993 23. Recognize \x{...} as a code point specifier, even when not in UTF-8 mode,
1994 but give a compile time error if the value is greater than 0xff.
1996 24. The man pages for pcrepartial, pcreprecompile, and pcre_compile2 were
1997 accidentally not being installed or uninstalled.
1999 25. The pcre.h file was built from pcre.h.in, but the only changes that were
2000 made were to insert the current release number. This seemed silly, because
2001 it made things harder for people building PCRE on systems that don't run
2002 "configure". I have turned pcre.h into a distributed file, no longer built
2003 by "configure", with the version identification directly included. There is
2004 no longer a pcre.h.in file.
2006 However, this change necessitated a change to the pcre-config script as
2007 well. It is built from pcre-config.in, and one of the substitutions was the
2008 release number. I have updated configure.ac so that ./configure now finds
2009 the release number by grepping pcre.h.
2011 26. Added the ability to run the tests under valgrind.
2014 Version 6.4 05-Sep-05
2015 ---------------------
2017 1. Change 6.0/10/(l) to pcregrep introduced a bug that caused separator lines
2018 "--" to be printed when multiple files were scanned, even when none of the
2019 -A, -B, or -C options were used. This is not compatible with Gnu grep, so I
2020 consider it to be a bug, and have restored the previous behaviour.
2022 2. A couple of code tidies to get rid of compiler warnings.
2024 3. The pcretest program used to cheat by referring to symbols in the library
2025 whose names begin with _pcre_. These are internal symbols that are not
2026 really supposed to be visible externally, and in some environments it is
2027 possible to suppress them. The cheating is now confined to including
2028 certain files from the library's source, which is a bit cleaner.
2030 4. Renamed pcre.in as pcre.h.in to go with pcrecpp.h.in; it also makes the
2031 file's purpose clearer.
2033 5. Reorganized pcre_ucp_findchar().
2036 Version 6.3 15-Aug-05
2037 ---------------------
2039 1. The file libpcre.pc.in did not have general read permission in the tarball.
2041 2. There were some problems when building without C++ support:
2043 (a) If C++ support was not built, "make install" and "make test" still
2044 tried to test it.
2046 (b) There were problems when the value of CXX was explicitly set. Some
2047 changes have been made to try to fix these, and ...
2049 (c) --disable-cpp can now be used to explicitly disable C++ support.
2051 (d) The use of @CPP_OBJ@ directly caused a blank line preceded by a
2052 backslash in a target when C++ was disabled. This confuses some
2053 versions of "make", apparently. Using an intermediate variable solves
2054 this. (Same for CPP_LOBJ.)
2056 3. $(LINK_FOR_BUILD) now includes $(CFLAGS_FOR_BUILD) and $(LINK)
2057 (non-Windows) now includes $(CFLAGS) because these flags are sometimes
2058 necessary on certain architectures.
2060 4. Added a setting of -export-symbols-regex to the link command to remove
2061 those symbols that are exported in the C sense, but actually are local
2062 within the library, and not documented. Their names all begin with
2063 "_pcre_". This is not a perfect job, because (a) we have to except some
2064 symbols that pcretest ("illegally") uses, and (b) the facility isn't always
2065 available (and never for static libraries). I have made a note to try to
2066 find a way round (a) in the future.
2069 Version 6.2 01-Aug-05
2070 ---------------------
2072 1. There was no test for integer overflow of quantifier values. A construction
2073 such as {1111111111111111} would give undefined results. What is worse, if
2074 a minimum quantifier for a parenthesized subpattern overflowed and became
2075 negative, the calculation of the memory size went wrong. This could have
2076 led to memory overwriting.
2078 2. Building PCRE using VPATH was broken. Hopefully it is now fixed.
2080 3. Added "b" to the 2nd argument of fopen() in dftables.c, for non-Unix-like
2081 operating environments where this matters.
2083 4. Applied Giuseppe Maxia's patch to add additional features for controlling
2084 PCRE options from within the C++ wrapper.
2086 5. Named capturing subpatterns were not being correctly counted when a pattern
2087 was compiled. This caused two problems: (a) If there were more than 100
2088 such subpatterns, the calculation of the memory needed for the whole
2089 compiled pattern went wrong, leading to an overflow error. (b) Numerical
2090 back references of the form \12, where the number was greater than 9, were
2091 not recognized as back references, even though there were sufficient
2092 previous subpatterns.
2094 6. Two minor patches to pcrecpp.cc in order to allow it to compile on older
2095 versions of gcc, e.g. 2.95.4.
2098 Version 6.1 21-Jun-05
2099 ---------------------
2101 1. There was one reference to the variable "posix" in pcretest.c that was not
2102 surrounded by "#if !defined NOPOSIX".
2104 2. Make it possible to compile pcretest without DFA support, UTF8 support, or
2105 the cross-check on the old pcre_info() function, for the benefit of the
2106 cut-down version of PCRE that is currently imported into Exim.
2108 3. A (silly) pattern starting with (?i)(?-i) caused an internal space
2109 allocation error. I've done the easy fix, which wastes 2 bytes for sensible
2110 patterns that start (?i) but I don't think that matters. The use of (?i) is
2111 just an example; this all applies to the other options as well.
2113 4. Since libtool seems to echo the compile commands it is issuing, the output
2114 from "make" can be reduced a bit by putting "@" in front of each libtool
2115 compile command.
2117 5. Patch from the folks at Google for configure.in to be a bit more thorough
2118 in checking for a suitable C++ installation before trying to compile the
2119 C++ stuff. This should fix a reported problem when a compiler was present,
2120 but no suitable headers.
2122 6. The man pages all had just "PCRE" as their title. I have changed them to
2123 be the relevant file name. I have also arranged that these names are
2124 retained in the file doc/pcre.txt, which is a concatenation in text format
2125 of all the man pages except the little individual ones for each function.
2127 7. The NON-UNIX-USE file had not been updated for the different set of source
2128 files that come with release 6. I also added a few comments about the C++
2129 wrapper.
2132 Version 6.0 07-Jun-05
2133 ---------------------
2135 1. Some minor internal re-organization to help with my DFA experiments.
2137 2. Some missing #ifdef SUPPORT_UCP conditionals in pcretest and printint that
2138 didn't matter for the library itself when fully configured, but did matter
2139 when compiling without UCP support, or within Exim, where the ucp files are
2140 not imported.
2142 3. Refactoring of the library code to split up the various functions into
2143 different source modules. The addition of the new DFA matching code (see
2144 below) to a single monolithic source would have made it really too
2145 unwieldy, quite apart from causing all the code to be include in a
2146 statically linked application, when only some functions are used. This is
2147 relevant even without the DFA addition now that patterns can be compiled in
2148 one application and matched in another.
2150 The downside of splitting up is that there have to be some external
2151 functions and data tables that are used internally in different modules of
2152 the library but which are not part of the API. These have all had their
2153 names changed to start with "_pcre_" so that they are unlikely to clash
2154 with other external names.
2156 4. Added an alternate matching function, pcre_dfa_exec(), which matches using
2157 a different (DFA) algorithm. Although it is slower than the original
2158 function, it does have some advantages for certain types of matching
2159 problem.
2161 5. Upgrades to pcretest in order to test the features of pcre_dfa_exec(),
2162 including restarting after a partial match.
2164 6. A patch for pcregrep that defines INVALID_FILE_ATTRIBUTES if it is not
2165 defined when compiling for Windows was sent to me. I have put it into the
2166 code, though I have no means of testing or verifying it.
2168 7. Added the pcre_refcount() auxiliary function.
2170 8. Added the PCRE_FIRSTLINE option. This constrains an unanchored pattern to
2171 match before or at the first newline in the subject string. In pcretest,
2172 the /f option on a pattern can be used to set this.
2174 9. A repeated \w when used in UTF-8 mode with characters greater than 256
2175 would behave wrongly. This has been present in PCRE since release 4.0.
2177 10. A number of changes to the pcregrep command:
2179 (a) Refactored how -x works; insert ^(...)$ instead of setting
2180 PCRE_ANCHORED and checking the length, in preparation for adding
2181 something similar for -w.
2183 (b) Added the -w (match as a word) option.
2185 (c) Refactored the way lines are read and buffered so as to have more
2186 than one at a time available.
2188 (d) Implemented a pcregrep test script.
2190 (e) Added the -M (multiline match) option. This allows patterns to match
2191 over several lines of the subject. The buffering ensures that at least
2192 8K, or the rest of the document (whichever is the shorter) is available
2193 for matching (and similarly the previous 8K for lookbehind assertions).
2195 (f) Changed the --help output so that it now says
2197 -w, --word-regex(p)
2199 instead of two lines, one with "regex" and the other with "regexp"
2200 because that confused at least one person since the short forms are the
2201 same. (This required a bit of code, as the output is generated
2202 automatically from a table. It wasn't just a text change.)
2204 (g) -- can be used to terminate pcregrep options if the next thing isn't an
2205 option but starts with a hyphen. Could be a pattern or a path name
2206 starting with a hyphen, for instance.
2208 (h) "-" can be given as a file name to represent stdin.
2210 (i) When file names are being printed, "(standard input)" is used for
2211 the standard input, for compatibility with GNU grep. Previously
2212 "<stdin>" was used.
2214 (j) The option --label=xxx can be used to supply a name to be used for
2215 stdin when file names are being printed. There is no short form.
2217 (k) Re-factored the options decoding logic because we are going to add
2218 two more options that take data. Such options can now be given in four
2219 different ways, e.g. "-fname", "-f name", "--file=name", "--file name".
2221 (l) Added the -A, -B, and -C options for requesting that lines of context
2222 around matches be printed.
2224 (m) Added the -L option to print the names of files that do not contain
2225 any matching lines, that is, the complement of -l.
2227 (n) The return code is 2 if any file cannot be opened, but pcregrep does
2228 continue to scan other files.
2230 (o) The -s option was incorrectly implemented. For compatibility with other
2231 greps, it now suppresses the error message for a non-existent or non-
2232 accessible file (but not the return code). There is a new option called
2233 -q that suppresses the output of matching lines, which was what -s was
2234 previously doing.
2236 (p) Added --include and --exclude options to specify files for inclusion
2237 and exclusion when recursing.
2239 11. The Makefile was not using the Autoconf-supported LDFLAGS macro properly.
2240 Hopefully, it now does.
2242 12. Missing cast in pcre_study().
2244 13. Added an "uninstall" target to the makefile.
2246 14. Replaced "extern" in the function prototypes in Makefile.in with
2247 "PCRE_DATA_SCOPE", which defaults to 'extern' or 'extern "C"' in the Unix
2248 world, but is set differently for Windows.
2250 15. Added a second compiling function called pcre_compile2(). The only
2251 difference is that it has an extra argument, which is a pointer to an
2252 integer error code. When there is a compile-time failure, this is set
2253 non-zero, in addition to the error test pointer being set to point to an
2254 error message. The new argument may be NULL if no error number is required
2255 (but then you may as well call pcre_compile(), which is now just a
2256 wrapper). This facility is provided because some applications need a
2257 numeric error indication, but it has also enabled me to tidy up the way
2258 compile-time errors are handled in the POSIX wrapper.
2260 16. Added VPATH=.libs to the makefile; this should help when building with one
2261 prefix path and installing with another. (Or so I'm told by someone who
2262 knows more about this stuff than I do.)
2264 17. Added a new option, REG_DOTALL, to the POSIX function regcomp(). This
2265 passes PCRE_DOTALL to the pcre_compile() function, making the "." character
2266 match everything, including newlines. This is not POSIX-compatible, but
2267 somebody wanted the feature. From pcretest it can be activated by using
2268 both the P and the s flags.
2270 18. AC_PROG_LIBTOOL appeared twice in Makefile.in. Removed one.
2272 19. libpcre.pc was being incorrectly installed as executable.
2274 20. A couple of places in pcretest check for end-of-line by looking for '\n';
2275 it now also looks for '\r' so that it will work unmodified on Windows.
2277 21. Added Google's contributed C++ wrapper to the distribution.
2279 22. Added some untidy missing memory free() calls in pcretest, to keep
2280 Electric Fence happy when testing.
2284 Version 5.0 13-Sep-04
2285 ---------------------
2287 1. Internal change: literal characters are no longer packed up into items
2288 containing multiple characters in a single byte-string. Each character
2289 is now matched using a separate opcode. However, there may be more than one
2290 byte in the character in UTF-8 mode.
2292 2. The pcre_callout_block structure has two new fields: pattern_position and
2293 next_item_length. These contain the offset in the pattern to the next match
2294 item, and its length, respectively.
2296 3. The PCRE_AUTO_CALLOUT option for pcre_compile() requests the automatic
2297 insertion of callouts before each pattern item. Added the /C option to
2298 pcretest to make use of this.
2300 4. On the advice of a Windows user, the lines
2302 #if defined(_WIN32) || defined(WIN32)
2303 _setmode( _fileno( stdout ), 0x8000 );
2304 #endif /* defined(_WIN32) || defined(WIN32) */
2306 have been added to the source of pcretest. This apparently does useful
2307 magic in relation to line terminators.
2309 5. Changed "r" and "w" in the calls to fopen() in pcretest to "rb" and "wb"
2310 for the benefit of those environments where the "b" makes a difference.
2312 6. The icc compiler has the same options as gcc, but "configure" doesn't seem
2313 to know about it. I have put a hack into configure.in that adds in code
2314 to set GCC=yes if CC=icc. This seems to end up at a point in the
2315 generated configure script that is early enough to affect the setting of
2316 compiler options, which is what is needed, but I have no means of testing
2317 whether it really works. (The user who reported this had patched the
2318 generated configure script, which of course I cannot do.)
2320 LATER: After change 22 below (new libtool files), the configure script
2321 seems to know about icc (and also ecc). Therefore, I have commented out
2322 this hack in configure.in.
2324 7. Added support for pkg-config (2 patches were sent in).
2326 8. Negated POSIX character classes that used a combination of internal tables
2327 were completely broken. These were [[:^alpha:]], [[:^alnum:]], and
2328 [[:^ascii]]. Typically, they would match almost any characters. The other
2329 POSIX classes were not broken in this way.
2331 9. Matching the pattern "\b.*?" against "ab cd", starting at offset 1, failed
2332 to find the match, as PCRE was deluded into thinking that the match had to
2333 start at the start point or following a newline. The same bug applied to
2334 patterns with negative forward assertions or any backward assertions
2335 preceding ".*" at the start, unless the pattern required a fixed first
2336 character. This was a failing pattern: "(?!.bcd).*". The bug is now fixed.
2338 10. In UTF-8 mode, when moving forwards in the subject after a failed match
2339 starting at the last subject character, bytes beyond the end of the subject
2340 string were read.
2342 11. Renamed the variable "class" as "classbits" to make life easier for C++
2343 users. (Previously there was a macro definition, but it apparently wasn't
2344 enough.)
2346 12. Added the new field "tables" to the extra data so that tables can be passed
2347 in at exec time, or the internal tables can be re-selected. This allows
2348 a compiled regex to be saved and re-used at a later time by a different
2349 program that might have everything at different addresses.
2351 13. Modified the pcre-config script so that, when run on Solaris, it shows a
2352 -R library as well as a -L library.
2354 14. The debugging options of pcretest (-d on the command line or D on a
2355 pattern) showed incorrect output for anything following an extended class
2356 that contained multibyte characters and which was followed by a quantifier.
2358 15. Added optional support for general category Unicode character properties
2359 via the \p, \P, and \X escapes. Unicode property support implies UTF-8
2360 support. It adds about 90K to the size of the library. The meanings of the
2361 inbuilt class escapes such as \d and \s have NOT been changed.
2363 16. Updated pcredemo.c to include calls to free() to release the memory for the
2364 compiled pattern.
2366 17. The generated file chartables.c was being created in the source directory
2367 instead of in the building directory. This caused the build to fail if the
2368 source directory was different from the building directory, and was
2369 read-only.
2371 18. Added some sample Win commands from Mark Tetrode into the NON-UNIX-USE
2372 file. No doubt somebody will tell me if they don't make sense... Also added
2373 Dan Mooney's comments about building on OpenVMS.
2375 19. Added support for partial matching via the PCRE_PARTIAL option for
2376 pcre_exec() and the \P data escape in pcretest.
2378 20. Extended pcretest with 3 new pattern features:
2380 (i) A pattern option of the form ">rest-of-line" causes pcretest to
2381 write the compiled pattern to the file whose name is "rest-of-line".
2382 This is a straight binary dump of the data, with the saved pointer to
2383 the character tables forced to be NULL. The study data, if any, is
2384 written too. After writing, pcretest reads a new pattern.
2386 (ii) If, instead of a pattern, "<rest-of-line" is given, pcretest reads a
2387 compiled pattern from the given file. There must not be any
2388 occurrences of "<" in the file name (pretty unlikely); if there are,
2389 pcretest will instead treat the initial "<" as a pattern delimiter.
2390 After reading in the pattern, pcretest goes on to read data lines as
2391 usual.
2393 (iii) The F pattern option causes pcretest to flip the bytes in the 32-bit
2394 and 16-bit fields in a compiled pattern, to simulate a pattern that
2395 was compiled on a host of opposite endianness.
2397 21. The pcre-exec() function can now cope with patterns that were compiled on
2398 hosts of opposite endianness, with this restriction:
2400 As for any compiled expression that is saved and used later, the tables
2401 pointer field cannot be preserved; the extra_data field in the arguments
2402 to pcre_exec() should be used to pass in a tables address if a value
2403 other than the default internal tables were used at compile time.
2405 22. Calling pcre_exec() with a negative value of the "ovecsize" parameter is
2406 now diagnosed as an error. Previously, most of the time, a negative number
2407 would have been treated as zero, but if in addition "ovector" was passed as
2408 NULL, a crash could occur.
2410 23. Updated the files ltmain.sh, config.sub, config.guess, and aclocal.m4 with
2411 new versions from the libtool 1.5 distribution (the last one is a copy of
2412 a file called libtool.m4). This seems to have fixed the need to patch
2413 "configure" to support Darwin 1.3 (which I used to do). However, I still
2414 had to patch ltmain.sh to ensure that ${SED} is set (it isn't on my
2415 workstation).
2417 24. Changed the PCRE licence to be the more standard "BSD" licence.
2420 Version 4.5 01-Dec-03
2421 ---------------------
2423 1. There has been some re-arrangement of the code for the match() function so
2424 that it can be compiled in a version that does not call itself recursively.
2425 Instead, it keeps those local variables that need separate instances for
2426 each "recursion" in a frame on the heap, and gets/frees frames whenever it
2427 needs to "recurse". Keeping track of where control must go is done by means
2428 of setjmp/longjmp. The whole thing is implemented by a set of macros that
2429 hide most of the details from the main code, and operates only if
2430 NO_RECURSE is defined while compiling pcre.c. If PCRE is built using the
2431 "configure" mechanism, "--disable-stack-for-recursion" turns on this way of
2432 operating.
2434 To make it easier for callers to provide specially tailored get/free
2435 functions for this usage, two new functions, pcre_stack_malloc, and
2436 pcre_stack_free, are used. They are always called in strict stacking order,
2437 and the size of block requested is always the same.
2439 The PCRE_CONFIG_STACKRECURSE info parameter can be used to find out whether
2440 PCRE has been compiled to use the stack or the heap for recursion. The
2441 -C option of pcretest uses this to show which version is compiled.
2443 A new data escape \S, is added to pcretest; it causes the amounts of store
2444 obtained and freed by both kinds of malloc/free at match time to be added
2445 to the output.
2447 2. Changed the locale test to use "fr_FR" instead of "fr" because that's
2448 what's available on my current Linux desktop machine.
2450 3. When matching a UTF-8 string, the test for a valid string at the start has
2451 been extended. If start_offset is not zero, PCRE now checks that it points
2452 to a byte that is the start of a UTF-8 character. If not, it returns
2453 PCRE_ERROR_BADUTF8_OFFSET (-11). Note: the whole string is still checked;
2454 this is necessary because there may be backward assertions in the pattern.
2455 When matching the same subject several times, it may save resources to use
2456 PCRE_NO_UTF8_CHECK on all but the first call if the string is long.
2458 4. The code for checking the validity of UTF-8 strings has been tightened so
2459 that it rejects (a) strings containing 0xfe or 0xff bytes and (b) strings
2460 containing "overlong sequences".
2462 5. Fixed a bug (appearing twice) that I could not find any way of exploiting!
2463 I had written "if ((digitab[*p++] && chtab_digit) == 0)" where the "&&"
2464 should have been "&", but it just so happened that all the cases this let
2465 through by mistake were picked up later in the function.
2467 6. I had used a variable called "isblank" - this is a C99 function, causing
2468 some compilers to warn. To avoid this, I renamed it (as "blankclass").
2470 7. Cosmetic: (a) only output another newline at the end of pcretest if it is
2471 prompting; (b) run "./pcretest /dev/null" at the start of the test script
2472 so the version is shown; (c) stop "make test" echoing "./RunTest".
2474 8. Added patches from David Burgess to enable PCRE to run on EBCDIC systems.
2476 9. The prototype for memmove() for systems that don't have it was using
2477 size_t, but the inclusion of the header that defines size_t was later. I've
2478 moved the #includes for the C headers earlier to avoid this.
2480 10. Added some adjustments to the code to make it easier to compiler on certain
2481 special systems:
2483 (a) Some "const" qualifiers were missing.
2484 (b) Added the macro EXPORT before all exported functions; by default this
2485 is defined to be empty.
2486 (c) Changed the dftables auxiliary program (that builds chartables.c) so
2487 that it reads its output file name as an argument instead of writing
2488 to the standard output and assuming this can be redirected.
2490 11. In UTF-8 mode, if a recursive reference (e.g. (?1)) followed a character
2491 class containing characters with values greater than 255, PCRE compilation
2492 went into a loop.
2494 12. A recursive reference to a subpattern that was within another subpattern
2495 that had a minimum quantifier of zero caused PCRE to crash. For example,
2496 (x(y(?2))z)? provoked this bug with a subject that got as far as the
2497 recursion. If the recursively-called subpattern itself had a zero repeat,
2498 that was OK.
2500 13. In pcretest, the buffer for reading a data line was set at 30K, but the
2501 buffer into which it was copied (for escape processing) was still set at
2502 1024, so long lines caused crashes.
2504 14. A pattern such as /[ab]{1,3}+/ failed to compile, giving the error
2505 "internal error: code overflow...". This applied to any character class
2506 that was followed by a possessive quantifier.
2508 15. Modified the Makefile to add libpcre.la as a prerequisite for
2509 libpcreposix.la because I was told this is needed for a parallel build to
2510 work.
2512 16. If a pattern that contained .* following optional items at the start was
2513 studied, the wrong optimizing data was generated, leading to matching
2514 errors. For example, studying /[ab]*.*c/ concluded, erroneously, that any
2515 matching string must start with a or b or c. The correct conclusion for
2516 this pattern is that a match can start with any character.
2519 Version 4.4 13-Aug-03
2520 ---------------------
2522 1. In UTF-8 mode, a character class containing characters with values between
2523 127 and 255 was not handled correctly if the compiled pattern was studied.
2524 In fixing this, I have also improved the studying algorithm for such
2525 classes (slightly).
2527 2. Three internal functions had redundant arguments passed to them. Removal
2528 might give a very teeny performance improvement.
2530 3. Documentation bug: the value of the capture_top field in a callout is *one
2531 more than* the number of the hightest numbered captured substring.
2533 4. The Makefile linked pcretest and pcregrep with -lpcre, which could result
2534 in incorrectly linking with a previously installed version. They now link
2535 explicitly with libpcre.la.
2537 5. configure.in no longer needs to recognize Cygwin specially.
2539 6. A problem in pcre.in for Windows platforms is fixed.
2541 7. If a pattern was successfully studied, and the -d (or /D) flag was given to
2542 pcretest, it used to include the size of the study block as part of its
2543 output. Unfortunately, the structure contains a field that has a different
2544 size on different hardware architectures. This meant that the tests that
2545 showed this size failed. As the block is currently always of a fixed size,
2546 this information isn't actually particularly useful in pcretest output, so
2547 I have just removed it.
2549 8. Three pre-processor statements accidentally did not start in column 1.
2550 Sadly, there are *still* compilers around that complain, even though
2551 standard C has not required this for well over a decade. Sigh.
2553 9. In pcretest, the code for checking callouts passed small integers in the
2554 callout_data field, which is a void * field. However, some picky compilers
2555 complained about the casts involved for this on 64-bit systems. Now
2556 pcretest passes the address of the small integer instead, which should get
2557 rid of the warnings.
2559 10. By default, when in UTF-8 mode, PCRE now checks for valid UTF-8 strings at
2560 both compile and run time, and gives an error if an invalid UTF-8 sequence
2561 is found. There is a option for disabling this check in cases where the
2562 string is known to be correct and/or the maximum performance is wanted.
2564 11. In response to a bug report, I changed one line in Makefile.in from
2566 -Wl,--out-implib,.libs/lib@WIN_PREFIX@pcreposix.dll.a \
2567 to
2568 -Wl,--out-implib,.libs/@WIN_PREFIX@libpcreposix.dll.a \
2570 to look similar to other lines, but I have no way of telling whether this
2571 is the right thing to do, as I do not use Windows. No doubt I'll get told
2572 if it's wrong...
2575 Version 4.3 21-May-03
2576 ---------------------
2578 1. Two instances of @WIN_PREFIX@ omitted from the Windows targets in the
2579 Makefile.
2581 2. Some refactoring to improve the quality of the code:
2583 (i) The utf8_table... variables are now declared "const".
2585 (ii) The code for \cx, which used the "case flipping" table to upper case
2586 lower case letters, now just substracts 32. This is ASCII-specific,
2587 but the whole concept of \cx is ASCII-specific, so it seems
2588 reasonable.
2590 (iii) PCRE was using its character types table to recognize decimal and
2591 hexadecimal digits in the pattern. This is silly, because it handles
2592 only 0-9, a-f, and A-F, but the character types table is locale-
2593 specific, which means strange things might happen. A private
2594 table is now used for this - though it costs 256 bytes, a table is
2595 much faster than multiple explicit tests. Of course, the standard
2596 character types table is still used for matching digits in subject
2597 strings against \d.
2599 (iv) Strictly, the identifier ESC_t is reserved by POSIX (all identifiers
2600 ending in _t are). So I've renamed it as ESC_tee.
2602 3. The first argument for regexec() in the POSIX wrapper should have been
2603 defined as "const".
2605 4. Changed pcretest to use malloc() for its buffers so that they can be
2606 Electric Fenced for debugging.
2608 5. There were several places in the code where, in UTF-8 mode, PCRE would try
2609 to read one or more bytes before the start of the subject string. Often this
2610 had no effect on PCRE's behaviour, but in some circumstances it could
2611 provoke a segmentation fault.
2613 6. A lookbehind at the start of a pattern in UTF-8 mode could also cause PCRE
2614 to try to read one or more bytes before the start of the subject string.
2616 7. A lookbehind in a pattern matched in non-UTF-8 mode on a PCRE compiled with
2617 UTF-8 support could misbehave in various ways if the subject string
2618 contained bytes with the 0x80 bit set and the 0x40 bit unset in a lookbehind
2619 area. (PCRE was not checking for the UTF-8 mode flag, and trying to move
2620 back over UTF-8 characters.)
2623 Version 4.2 14-Apr-03
2624 ---------------------
2626 1. Typo "#if SUPPORT_UTF8" instead of "#ifdef SUPPORT_UTF8" fixed.
2628 2. Changes to the building process, supplied by Ronald Landheer-Cieslak
2629 [ON_WINDOWS]: new variable, "#" on non-Windows platforms
2630 [NOT_ON_WINDOWS]: new variable, "#" on Windows platforms
2631 [WIN_PREFIX]: new variable, "cyg" for Cygwin
2632 * Makefile.in: use autoconf substitution for OBJEXT, EXEEXT, BUILD_OBJEXT
2634 Note: automatic setting of the BUILD variables is not yet working
2635 set CPPFLAGS and BUILD_CPPFLAGS (but don't use yet) - should be used at
2636 compile-time but not at link-time
2637 [LINK]: use for linking executables only
2638 make different versions for Windows and non-Windows
2639 [LINKLIB]: new variable, copy of UNIX-style LINK, used for linking
2640 libraries
2641 [LINK_FOR_BUILD]: new variable
2642 [OBJEXT]: use throughout
2643 [EXEEXT]: use throughout
2644 <winshared>: new target
2645 <wininstall>: new target
2646 <dftables.o>: use native compiler
2647 <dftables>: use native linker
2648 <install>: handle Windows platform correctly
2649 <clean>: ditto
2650 <check>: ditto
2651 copy DLL to top builddir before testing
2653 As part of these changes, -no-undefined was removed again. This was reported
2654 to give trouble on HP-UX 11.0, so getting rid of it seems like a good idea
2655 in any case.
2657 3. Some tidies to get rid of compiler warnings:
2659 . In the match_data structure, match_limit was an unsigned long int, whereas
2660 match_call_count was an int. I've made them both unsigned long ints.
2662 . In pcretest the fact that a const uschar * doesn't automatically cast to
2663 a void * provoked a warning.
2665 . Turning on some more compiler warnings threw up some "shadow" variables
2666 and a few more missing casts.
2668 4. If PCRE was complied with UTF-8 support, but called without the PCRE_UTF8
2669 option, a class that contained a single character with a value between 128
2670 and 255 (e.g. /[\xFF]/) caused PCRE to crash.
2672 5. If PCRE was compiled with UTF-8 support, but called without the PCRE_UTF8
2673 option, a class that contained several characters, but with at least one
2674 whose value was between 128 and 255 caused PCRE to crash.
2677 Version 4.1 12-Mar-03
2678 ---------------------
2680 1. Compiling with gcc -pedantic found a couple of places where casts were
2681 needed, and a string in dftables.c that was longer than standard compilers are
2682 required to support.
2684 2. Compiling with Sun's compiler found a few more places where the code could
2685 be tidied up in order to avoid warnings.
2687 3. The variables for cross-compiling were called HOST_CC and HOST_CFLAGS; the
2688 first of these names is deprecated in the latest Autoconf in favour of the name
2689 CC_FOR_BUILD, because "host" is typically used to mean the system on which the
2690 compiled code will be run. I can't find a reference for HOST_CFLAGS, but by
2691 analogy I have changed it to CFLAGS_FOR_BUILD.
2693 4. Added -no-undefined to the linking command in the Makefile, because this is
2694 apparently helpful for Windows. To make it work, also added "-L. -lpcre" to the
2695 linking step for the pcreposix library.
2697 5. PCRE was failing to diagnose the case of two named groups with the same
2698 name.
2700 6. A problem with one of PCRE's optimizations was discovered. PCRE remembers a
2701 literal character that is needed in the subject for a match, and scans along to
2702 ensure that it is present before embarking on the full matching process. This
2703 saves time in cases of nested unlimited repeats that are never going to match.
2704 Problem: the scan can take a lot of time if the subject is very long (e.g.
2705 megabytes), thus penalizing straightforward matches. It is now done only if the
2706 amount of subject to be scanned is less than 1000 bytes.
2708 7. A lesser problem with the same optimization is that it was recording the
2709 first character of an anchored pattern as "needed", thus provoking a search
2710 right along the subject, even when the first match of the pattern was going to
2711 fail. The "needed" character is now not set for anchored patterns, unless it
2712 follows something in the pattern that is of non-fixed length. Thus, it still
2713 fulfils its original purpose of finding quick non-matches in cases of nested
2714 unlimited repeats, but isn't used for simple anchored patterns such as /^abc/.
2717 Version 4.0 17-Feb-03
2718 ---------------------
2720 1. If a comment in an extended regex that started immediately after a meta-item
2721 extended to the end of string, PCRE compiled incorrect data. This could lead to
2722 all kinds of weird effects. Example: /#/ was bad; /()#/ was bad; /a#/ was not.
2724 2. Moved to autoconf 2.53 and libtool 1.4.2.
2726 3. Perl 5.8 no longer needs "use utf8" for doing UTF-8 things. Consequently,
2727 the special perltest8 script is no longer needed - all the tests can be run
2728 from a single perltest script.
2730 4. From 5.004, Perl has not included the VT character (0x0b) in the set defined
2731 by \s. It has now been removed in PCRE. This means it isn't recognized as
2732 whitespace in /x regexes too, which is the same as Perl. Note that the POSIX
2733 class [:space:] *does* include VT, thereby creating a mess.
2735 5. Added the class [:blank:] (a GNU extension from Perl 5.8) to match only
2736 space and tab.
2738 6. Perl 5.005 was a long time ago. It's time to amalgamate the tests that use
2739 its new features into the main test script, reducing the number of scripts.
2741 7. Perl 5.8 has changed the meaning of patterns like /a(?i)b/. Earlier versions
2742 were backward compatible, and made the (?i) apply to the whole pattern, as if
2743 /i were given. Now it behaves more logically, and applies the option setting
2744 only to what follows. PCRE has been changed to follow suit. However, if it
2745 finds options settings right at the start of the pattern, it extracts them into
2746 the global options, as before. Thus, they show up in the info data.
2748 8. Added support for the \Q...\E escape sequence. Characters in between are
2749 treated as literals. This is slightly different from Perl in that $ and @ are
2750 also handled as literals inside the quotes. In Perl, they will cause variable
2751 interpolation. Note the following examples:
2753 Pattern PCRE matches Perl matches
2755 \Qabc$xyz\E abc$xyz abc followed by the contents of $xyz
2756 \Qabc\$xyz\E abc\$xyz abc\$xyz
2757 \Qabc\E\$\Qxyz\E abc$xyz abc$xyz
2759 For compatibility with Perl, \Q...\E sequences are recognized inside character
2760 classes as well as outside them.
2762 9. Re-organized 3 code statements in pcretest to avoid "overflow in
2763 floating-point constant arithmetic" warnings from a Microsoft compiler. Added a
2764 (size_t) cast to one statement in pcretest and one in pcreposix to avoid
2765 signed/unsigned warnings.
2767 10. SunOS4 doesn't have strtoul(). This was used only for unpicking the -o
2768 option for pcretest, so I've replaced it by a simple function that does just
2769 that job.
2771 11. pcregrep was ending with code 0 instead of 2 for the commands "pcregrep" or
2772 "pcregrep -".
2774 12. Added "possessive quantifiers" ?+, *+, ++, and {,}+ which come from Sun's
2775 Java package. This provides some syntactic sugar for simple cases of what my
2776 documentation calls "once-only subpatterns". A pattern such as x*+ is the same
2777 as (?>x*). In other words, if what is inside (?>...) is just a single repeated
2778 item, you can use this simplified notation. Note that only makes sense with
2779 greedy quantifiers. Consequently, the use of the possessive quantifier forces
2780 greediness, whatever the setting of the PCRE_UNGREEDY option.
2782 13. A change of greediness default within a pattern was not taking effect at
2783 the current level for patterns like /(b+(?U)a+)/. It did apply to parenthesized
2784 subpatterns that followed. Patterns like /b+(?U)a+/ worked because the option
2785 was abstracted outside.
2787 14. PCRE now supports the \G assertion. It is true when the current matching
2788 position is at the start point of the match. This differs from \A when the
2789 starting offset is non-zero. Used with the /g option of pcretest (or similar
2790 code), it works in the same way as it does for Perl's /g option. If all
2791 alternatives of a regex begin with \G, the expression is anchored to the start
2792 match position, and the "anchored" flag is set in the compiled expression.
2794 15. Some bugs concerning the handling of certain option changes within patterns
2795 have been fixed. These applied to options other than (?ims). For example,
2796 "a(?x: b c )d" did not match "XabcdY" but did match "Xa b c dY". It should have
2797 been the other way round. Some of this was related to change 7 above.
2799 16. PCRE now gives errors for /[.x.]/ and /[=x=]/ as unsupported POSIX
2800 features, as Perl does. Previously, PCRE gave the warnings only for /[[.x.]]/
2801 and /[[=x=]]/. PCRE now also gives an error for /[:name:]/ because it supports
2802 POSIX classes only within a class (e.g. /[[:alpha:]]/).
2804 17. Added support for Perl's \C escape. This matches one byte, even in UTF8
2805 mode. Unlike ".", it always matches newline, whatever the setting of
2806 PCRE_DOTALL. However, PCRE does not permit \C to appear in lookbehind
2807 assertions. Perl allows it, but it doesn't (in general) work because it can't
2808 calculate the length of the lookbehind. At least, that's the case for Perl
2809 5.8.0 - I've been told they are going to document that it doesn't work in
2810 future.
2812 18. Added an error diagnosis for escapes that PCRE does not support: these are
2813 \L, \l, \N, \P, \p, \U, \u, and \X.
2815 19. Although correctly diagnosing a missing ']' in a character class, PCRE was
2816 reading past the end of the pattern in cases such as /[abcd/.
2818 20. PCRE was getting more memory than necessary for patterns with classes that
2819 contained both POSIX named classes and other characters, e.g. /[[:space:]abc/.
2821 21. Added some code, conditional on #ifdef VPCOMPAT, to make life easier for
2822 compiling PCRE for use with Virtual Pascal.
2824 22. Small fix to the Makefile to make it work properly if the build is done
2825 outside the source tree.
2827 23. Added a new extension: a condition to go with recursion. If a conditional
2828 subpattern starts with (?(R) the "true" branch is used if recursion has
2829 happened, whereas the "false" branch is used only at the top level.
2831 24. When there was a very long string of literal characters (over 255 bytes
2832 without UTF support, over 250 bytes with UTF support), the computation of how
2833 much memory was required could be incorrect, leading to segfaults or other
2834 strange effects.
2836 25. PCRE was incorrectly assuming anchoring (either to start of subject or to
2837 start of line for a non-DOTALL pattern) when a pattern started with (.*) and
2838 there was a subsequent back reference to those brackets. This meant that, for
2839 example, /(.*)\d+\1/ failed to match "abc123bc". Unfortunately, it isn't
2840 possible to check for precisely this case. All we can do is abandon the
2841 optimization if .* occurs inside capturing brackets when there are any back
2842 references whatsoever. (See below for a better fix that came later.)
2844 26. The handling of the optimization for finding the first character of a
2845 non-anchored pattern, and for finding a character that is required later in the
2846 match were failing in some cases. This didn't break the matching; it just
2847 failed to optimize when it could. The way this is done has been re-implemented.
2849 27. Fixed typo in error message for invalid (?R item (it said "(?p").
2851 28. Added a new feature that provides some of the functionality that Perl
2852 provides with (?{...}). The facility is termed a "callout". The way it is done
2853 in PCRE is for the caller to provide an optional function, by setting
2854 pcre_callout to its entry point. Like pcre_malloc and pcre_free, this is a
2855 global variable. By default it is unset, which disables all calling out. To get
2856 the function called, the regex must include (?C) at appropriate points. This
2857 is, in fact, equivalent to (?C0), and any number <= 255 may be given with (?C).
2858 This provides a means of identifying different callout points. When PCRE
2859 reaches such a point in the regex, if pcre_callout has been set, the external
2860 function is called. It is provided with data in a structure called
2861 pcre_callout_block, which is defined in pcre.h. If the function returns 0,
2862 matching continues; if it returns a non-zero value, the match at the current
2863 point fails. However, backtracking will occur if possible. [This was changed
2864 later and other features added - see item 49 below.]
2866 29. pcretest is upgraded to test the callout functionality. It provides a
2867 callout function that displays information. By default, it shows the start of
2868 the match and the current position in the text. There are some new data escapes
2869 to vary what happens:
2871 \C+ in addition, show current contents of captured substrings
2872 \C- do not supply a callout function
2873 \C!n return 1 when callout number n is reached
2874 \C!n!m return 1 when callout number n is reached for the mth time
2876 30. If pcregrep was called with the -l option and just a single file name, it
2877 output "<stdin>" if a match was found, instead of the file name.
2879 31. Improve the efficiency of the POSIX API to PCRE. If the number of capturing
2880 slots is less than POSIX_MALLOC_THRESHOLD, use a block on the stack to pass to
2881 pcre_exec(). This saves a malloc/free per call. The default value of
2882 POSIX_MALLOC_THRESHOLD is 10; it can be changed by --with-posix-malloc-threshold
2883 when configuring.
2885 32. The default maximum size of a compiled pattern is 64K. There have been a
2886 few cases of people hitting this limit. The code now uses macros to handle the
2887 storing of links as offsets within the compiled pattern. It defaults to 2-byte
2888 links, but this can be changed to 3 or 4 bytes by --with-link-size when
2889 configuring. Tests 2 and 5 work only with 2-byte links because they output
2890 debugging information about compiled patterns.
2892 33. Internal code re-arrangements:
2894 (a) Moved the debugging function for printing out a compiled regex into
2895 its own source file (printint.c) and used #include to pull it into
2896 pcretest.c and, when DEBUG is defined, into pcre.c, instead of having two
2897 separate copies.
2899 (b) Defined the list of op-code names for debugging as a macro in
2900 internal.h so that it is next to the definition of the opcodes.
2902 (c) Defined a table of op-code lengths for simpler skipping along compiled
2903 code. This is again a macro in internal.h so that it is next to the
2904 definition of the opcodes.
2906 34. Added support for recursive calls to individual subpatterns, along the
2907 lines of Robin Houston's patch (but implemented somewhat differently).
2909 35. Further mods to the Makefile to help Win32. Also, added code to pcregrep to
2910 allow it to read and process whole directories in Win32. This code was
2911 contributed by Lionel Fourquaux; it has not been tested by me.
2913 36. Added support for named subpatterns. The Python syntax (?P<name>...) is
2914 used to name a group. Names consist of alphanumerics and underscores, and must
2915 be unique. Back references use the syntax (?P=name) and recursive calls use
2916 (?P>name) which is a PCRE extension to the Python extension. Groups still have
2917 numbers. The function pcre_fullinfo() can be used after compilation to extract
2918 a name/number map. There are three relevant calls:
2920 PCRE_INFO_NAMEENTRYSIZE yields the size of each entry in the map
2921 PCRE_INFO_NAMECOUNT yields the number of entries
2922 PCRE_INFO_NAMETABLE yields a pointer to the map.
2924 The map is a vector of fixed-size entries. The size of each entry depends on
2925 the length of the longest name used. The first two bytes of each entry are the
2926 group number, most significant byte first. There follows the corresponding
2927 name, zero terminated. The names are in alphabetical order.
2929 37. Make the maximum literal string in the compiled code 250 for the non-UTF-8
2930 case instead of 255. Making it the same both with and without UTF-8 support
2931 means that the same test output works with both.
2933 38. There was a case of malloc(0) in the POSIX testing code in pcretest. Avoid
2934 calling malloc() with a zero argument.
2936 39. Change 25 above had to resort to a heavy-handed test for the .* anchoring
2937 optimization. I've improved things by keeping a bitmap of backreferences with
2938 numbers 1-31 so that if .* occurs inside capturing brackets that are not in
2939 fact referenced, the optimization can be applied. It is unlikely that a
2940 relevant occurrence of .* (i.e. one which might indicate anchoring or forcing
2941 the match to follow \n) will appear inside brackets with a number greater than
2942 31, but if it does, any back reference > 31 suppresses the optimization.
2944 40. Added a new compile-time option PCRE_NO_AUTO_CAPTURE. This has the effect
2945 of disabling numbered capturing parentheses. Any opening parenthesis that is
2946 not followed by ? behaves as if it were followed by ?: but named parentheses
2947 can still be used for capturing (and they will acquire numbers in the usual
2948 way).
2950 41. Redesigned the return codes from the match() function into yes/no/error so
2951 that errors can be passed back from deep inside the nested calls. A malloc
2952 failure while inside a recursive subpattern call now causes the
2953 PCRE_ERROR_NOMEMORY return instead of quietly going wrong.
2955 42. It is now possible to set a limit on the number of times the match()
2956 function is called in a call to pcre_exec(). This facility makes it possible to
2957 limit the amount of recursion and backtracking, though not in a directly
2958 obvious way, because the match() function is used in a number of different
2959 circumstances. The count starts from zero for each position in the subject
2960 string (for non-anchored patterns). The default limit is, for compatibility, a
2961 large number, namely 10 000 000. You can change this in two ways:
2963 (a) When configuring PCRE before making, you can use --with-match-limit=n
2964 to set a default value for the compiled library.
2966 (b) For each call to pcre_exec(), you can pass a pcre_extra block in which
2967 a different value is set. See 45 below.
2969 If the limit is exceeded, pcre_exec() returns PCRE_ERROR_MATCHLIMIT.
2971 43. Added a new function pcre_config(int, void *) to enable run-time extraction
2972 of things that can be changed at compile time. The first argument specifies
2973 what is wanted and the second points to where the information is to be placed.
2974 The current list of available information is:
2978 The output is an integer that is set to one if UTF-8 support is available;
2979 otherwise it is set to zero.
2983 The output is an integer that it set to the value of the code that is used for
2984 newline. It is either LF (10) or CR (13).
2988 The output is an integer that contains the number of bytes used for internal
2989 linkage in compiled expressions. The value is 2, 3, or 4. See item 32 above.
2993 The output is an integer that contains the threshold above which the POSIX
2994 interface uses malloc() for output vectors. See item 31 above.
2998 The output is an unsigned integer that contains the default limit of the number
2999 of match() calls in a pcre_exec() execution. See 42 above.
3001 44. pcretest has been upgraded by the addition of the -C option. This causes it
3002 to extract all the available output from the new pcre_config() function, and to
3003 output it. The program then exits immediately.
3005 45. A need has arisen to pass over additional data with calls to pcre_exec() in
3006 order to support additional features. One way would have been to define
3007 pcre_exec2() (for example) with extra arguments, but this would not have been
3008 extensible, and would also have required all calls to the original function to
3009 be mapped to the new one. Instead, I have chosen to extend the mechanism that
3010 is used for passing in "extra" data from pcre_study().
3012 The pcre_extra structure is now exposed and defined in pcre.h. It currently
3013 contains the following fields:
3015 flags a bitmap indicating which of the following fields are set
3016 study_data opaque data from pcre_study()
3017 match_limit a way of specifying a limit on match() calls for a specific
3018 call to pcre_exec()
3019 callout_data data for callouts (see 49 below)
3021 The flag bits are also defined in pcre.h, and are
3027 The pcre_study() function now returns one of these new pcre_extra blocks, with
3028 the actual study data pointed to by the study_data field, and the
3029 PCRE_EXTRA_STUDY_DATA flag set. This can be passed directly to pcre_exec() as
3030 before. That is, this change is entirely upwards-compatible and requires no
3031 change to existing code.
3033 If you want to pass in additional data to pcre_exec(), you can either place it
3034 in a pcre_extra block provided by pcre_study(), or create your own pcre_extra
3035 block.
3037 46. pcretest has been extended to test the PCRE_EXTRA_MATCH_LIMIT feature. If a
3038 data string contains the escape sequence \M, pcretest calls pcre_exec() several
3039 times with different match limits, until it finds the minimum value needed for
3040 pcre_exec() to complete. The value is then output. This can be instructive; for
3041 most simple matches the number is quite small, but for pathological cases it
3042 gets very large very quickly.
3044 47. There's a new option for pcre_fullinfo() called PCRE_INFO_STUDYSIZE. It
3045 returns the size of the data block pointed to by the study_data field in a
3046 pcre_extra block, that is, the value that was passed as the argument to
3047 pcre_malloc() when PCRE was getting memory in which to place the information
3048 created by pcre_study(). The fourth argument should point to a size_t variable.
3049 pcretest has been extended so that this information is shown after a successful
3050 pcre_study() call when information about the compiled regex is being displayed.
3052 48. Cosmetic change to Makefile: there's no need to have / after $(DESTDIR)
3053 because what follows is always an absolute path. (Later: it turns out that this
3054 is more than cosmetic for MinGW, because it doesn't like empty path
3055 components.)
3057 49. Some changes have been made to the callout feature (see 28 above):
3059 (i) A callout function now has three choices for what it returns:
3061 0 => success, carry on matching
3062 > 0 => failure at this point, but backtrack if possible
3063 < 0 => serious error, return this value from pcre_exec()
3065 Negative values should normally be chosen from the set of PCRE_ERROR_xxx
3066 values. In particular, returning PCRE_ERROR_NOMATCH forces a standard
3067 "match failed" error. The error number PCRE_ERROR_CALLOUT is reserved for
3068 use by callout functions. It will never be used by PCRE itself.
3070 (ii) The pcre_extra structure (see 45 above) has a void * field called
3071 callout_data, with corresponding flag bit PCRE_EXTRA_CALLOUT_DATA. The
3072 pcre_callout_block structure has a field of the same name. The contents of
3073 the field passed in the pcre_extra structure are passed to the callout
3074 function in the corresponding field in the callout block. This makes it
3075 easier to use the same callout-containing regex from multiple threads. For
3076 testing, the pcretest program has a new data escape
3078 \C*n pass the number n (may be negative) as callout_data
3080 If the callout function in pcretest receives a non-zero value as
3081 callout_data, it returns that value.
3083 50. Makefile wasn't handling CFLAGS properly when compiling dftables. Also,
3084 there were some redundant $(CFLAGS) in commands that are now specified as
3085 $(LINK), which already includes $(CFLAGS).
3087 51. Extensions to UTF-8 support are listed below. These all apply when (a) PCRE
3088 has been compiled with UTF-8 support *and* pcre_compile() has been compiled
3089 with the PCRE_UTF8 flag. Patterns that are compiled without that flag assume
3090 one-byte characters throughout. Note that case-insensitive matching applies
3091 only to characters whose values are less than 256. PCRE doesn't support the
3092 notion of cases for higher-valued characters.
3094 (i) A character class whose characters are all within 0-255 is handled as
3095 a bit map, and the map is inverted for negative classes. Previously, a
3096 character > 255 always failed to match such a class; however it should
3097 match if the class was a negative one (e.g. [^ab]). This has been fixed.
3099 (ii) A negated character class with a single character < 255 is coded as
3100 "not this character" (OP_NOT). This wasn't working properly when the test
3101 character was multibyte, either singly or repeated.
3103 (iii) Repeats of multibyte characters are now handled correctly in UTF-8
3104 mode, for example: \x{100}{2,3}.
3106 (iv) The character escapes \b, \B, \d, \D, \s, \S, \w, and \W (either
3107 singly or repeated) now correctly test multibyte characters. However,
3108 PCRE doesn't recognize any characters with values greater than 255 as
3109 digits, spaces, or word characters. Such characters always match \D, \S,
3110 and \W, and never match \d, \s, or \w.
3112 (v) Classes may now contain characters and character ranges with values
3113 greater than 255. For example: [ab\x{100}-\x{400}].
3115 (vi) pcregrep now has a --utf-8 option (synonym -u) which makes it call
3116 PCRE in UTF-8 mode.
3118 52. The info request value PCRE_INFO_FIRSTCHAR has been renamed
3119 PCRE_INFO_FIRSTBYTE because it is a byte value. However, the old name is
3120 retained for backwards compatibility. (Note that LASTLITERAL is also a byte
3121 value.)
3123 53. The single man page has become too large. I have therefore split it up into
3124 a number of separate man pages. These also give rise to individual HTML pages;
3125 these are now put in a separate directory, and there is an index.html page that
3126 lists them all. Some hyperlinking between the pages has been installed.
3128 54. Added convenience functions for handling named capturing parentheses.
3130 55. Unknown escapes inside character classes (e.g. [\M]) and escapes that
3131 aren't interpreted therein (e.g. [\C]) are literals in Perl. This is now also
3132 true in PCRE, except when the PCRE_EXTENDED option is set, in which case they
3133 are faulted.
3135 56. Introduced HOST_CC and HOST_CFLAGS which can be set in the environment when
3136 calling configure. These values are used when compiling the dftables.c program
3137 which is run to generate the source of the default character tables. They
3138 default to the values of CC and CFLAGS. If you are cross-compiling PCRE,
3139 you will need to set these values.
3141 57. Updated the building process for Windows DLL, as provided by Fred Cox.
3144 Version 3.9 02-Jan-02
3145 ---------------------
3147 1. A bit of extraneous text had somehow crept into the pcregrep documentation.
3149 2. If --disable-static was given, the building process failed when trying to
3150 build pcretest and pcregrep. (For some reason it was using libtool to compile
3151 them, which is not right, as they aren't part of the library.)
3154 Version 3.8 18-Dec-01
3155 ---------------------
3157 1. The experimental UTF-8 code was completely screwed up. It was packing the
3158 bytes in the wrong order. How dumb can you get?
3161 Version 3.7 29-Oct-01
3162 ---------------------
3164 1. In updating pcretest to check change 1 of version 3.6, I screwed up.
3165 This caused pcretest, when used on the test data, to segfault. Unfortunately,
3166 this didn't happen under Solaris 8, where I normally test things.
3168 2. The Makefile had to be changed to make it work on BSD systems, where 'make'
3169 doesn't seem to recognize that ./xxx and xxx are the same file. (This entry
3170 isn't in ChangeLog distributed with 3.7 because I forgot when I hastily made
3171 this fix an hour or so after the initial 3.7 release.)
3174 Version 3.6 23-Oct-01
3175 ---------------------
3177 1. Crashed with /(sens|respons)e and \1ibility/ and "sense and sensibility" if
3178 offsets passed as NULL with zero offset count.
3180 2. The config.guess and config.sub files had not been updated when I moved to
3181 the latest autoconf.
3184 Version 3.5 15-Aug-01
3185 ---------------------
3187 1. Added some missing #if !defined NOPOSIX conditionals in pcretest.c that
3188 had been forgotten.
3190 2. By using declared but undefined structures, we can avoid using "void"
3191 definitions in pcre.h while keeping the internal definitions of the structures
3192 private.
3194 3. The distribution is now built using autoconf 2.50 and libtool 1.4. From a
3195 user point of view, this means that both static and shared libraries are built
3196 by default, but this can be individually controlled. More of the work of
3197 handling this static/shared cases is now inside libtool instead of PCRE's make
3198 file.
3200 4. The pcretest utility is now installed along with pcregrep because it is
3201 useful for users (to test regexs) and by doing this, it automatically gets
3202 relinked by libtool. The documentation has been turned into a man page, so
3203 there are now .1, .txt, and .html versions in /doc.
3205 5. Upgrades to pcregrep:
3206 (i) Added long-form option names like gnu grep.
3207 (ii) Added --help to list all options with an explanatory phrase.
3208 (iii) Added -r, --recursive to recurse into sub-directories.
3209 (iv) Added -f, --file to read patterns from a file.
3211 6. pcre_exec() was referring to its "code" argument before testing that
3212 argument for NULL (and giving an error if it was NULL).
3214 7. Upgraded Makefile.in to allow for compiling in a different directory from
3215 the source directory.
3217 8. Tiny buglet in pcretest: when pcre_fullinfo() was called to retrieve the
3218 options bits, the pointer it was passed was to an int instead of to an unsigned
3219 long int. This mattered only on 64-bit systems.
3221 9. Fixed typo (3.4/1) in pcre.h again. Sigh. I had changed pcre.h (which is
3222 generated) instead of pcre.in, which it its source. Also made the same change
3223 in several of the .c files.
3225 10. A new release of gcc defines printf() as a macro, which broke pcretest
3226 because it had an ifdef in the middle of a string argument for printf(). Fixed
3227 by using separate calls to printf().
3229 11. Added --enable-newline-is-cr and --enable-newline-is-lf to the configure
3230 script, to force use of CR or LF instead of \n in the source. On non-Unix
3231 systems, the value can be set in config.h.
3233 12. The limit of 200 on non-capturing parentheses is a _nesting_ limit, not an
3234 absolute limit. Changed the text of the error message to make this clear, and
3235 likewise updated the man page.
3237 13. The limit of 99 on the number of capturing subpatterns has been removed.
3238 The new limit is 65535, which I hope will not be a "real" limit.
3241 Version 3.4 22-Aug-00
3242 ---------------------
3244 1. Fixed typo in pcre.h: unsigned const char * changed to const unsigned char *.
3246 2. Diagnose condition (?(0) as an error instead of crashing on matching.
3249 Version 3.3 01-Aug-00
3250 ---------------------
3252 1. If an octal character was given, but the value was greater than \377, it
3253 was not getting masked to the least significant bits, as documented. This could
3254 lead to crashes in some systems.
3256 2. Perl 5.6 (if not earlier versions) accepts classes like [a-\d] and treats
3257 the hyphen as a literal. PCRE used to give an error; it now behaves like Perl.
3259 3. Added the functions pcre_free_substring() and pcre_free_substring_list().
3260 These just pass their arguments on to (pcre_free)(), but they are provided
3261 because some uses of PCRE bind it to non-C systems that can call its functions,
3262 but cannot call free() or pcre_free() directly.
3264 4. Add "make test" as a synonym for "make check". Corrected some comments in
3265 the Makefile.
3267 5. Add $(DESTDIR)/ in front of all the paths in the "install" target in the
3268 Makefile.
3270 6. Changed the name of pgrep to pcregrep, because Solaris has introduced a
3271 command called pgrep for grepping around the active processes.
3273 7. Added the beginnings of support for UTF-8 character strings.
3275 8. Arranged for the Makefile to pass over the settings of CC, CFLAGS, and
3276 RANLIB to ./ltconfig so that they are used by libtool. I think these are all
3277 the relevant ones. (AR is not passed because ./ltconfig does its own figuring
3278 out for the ar command.)
3281 Version 3.2 12-May-00
3282 ---------------------
3284 This is purely a bug fixing release.
3286 1. If the pattern /((Z)+|A)*/ was matched agained ZABCDEFG it matched Z instead
3287 of ZA. This was just one example of several cases that could provoke this bug,
3288 which was introduced by change 9 of version 2.00. The code for breaking
3289 infinite loops after an iteration that matches an empty string was't working
3290 correctly.
3292 2. The pcretest program was not imitating Perl correctly for the pattern /a*/g
3293 when matched against abbab (for example). After matching an empty string, it
3294 wasn't forcing anchoring when setting PCRE_NOTEMPTY for the next attempt; this
3295 caused it to match further down the string than it should.
3297 3. The code contained an inclusion of sys/types.h. It isn't clear why this
3298 was there because it doesn't seem to be needed, and it causes trouble on some
3299 systems, as it is not a Standard C header. It has been removed.
3301 4. Made 4 silly changes to the source to avoid stupid compiler warnings that
3302 were reported on the Macintosh. The changes were from
3304 while ((c = *(++ptr)) != 0 && c != '\n');
3305 to
3306 while ((c = *(++ptr)) != 0 && c != '\n') ;
3308 Totally extraordinary, but if that's what it takes...
3310 5. PCRE is being used in one environment where neither memmove() nor bcopy() is
3311 available. Added HAVE_BCOPY and an autoconf test for it; if neither
3312 HAVE_MEMMOVE nor HAVE_BCOPY is set, use a built-in emulation function which
3313 assumes the way PCRE uses memmove() (always moving upwards).
3315 6. PCRE is being used in one environment where strchr() is not available. There
3316 was only one use in pcre.c, and writing it out to avoid strchr() probably gives
3317 faster code anyway.
3320 Version 3.1 09-Feb-00
3321 ---------------------
3323 The only change in this release is the fixing of some bugs in Makefile.in for
3324 the "install" target:
3326 (1) It was failing to install pcreposix.h.
3328 (2) It was overwriting the pcre.3 man page with the pcreposix.3 man page.
3331 Version 3.0 01-Feb-00
3332 ---------------------
3334 1. Add support for the /+ modifier to perltest (to output $` like it does in
3335 pcretest).
3337 2. Add support for the /g modifier to perltest.
3339 3. Fix pcretest so that it behaves even more like Perl for /g when the pattern
3340 matches null strings.
3342 4. Fix perltest so that it doesn't do unwanted things when fed an empty
3343 pattern. Perl treats empty patterns specially - it reuses the most recent
3344 pattern, which is not what we want. Replace // by /(?#)/ in order to avoid this
3345 effect.
3347 5. The POSIX interface was broken in that it was just handing over the POSIX
3348 captured string vector to pcre_exec(), but (since release 2.00) PCRE has
3349 required a bigger vector, with some working space on the end. This means that
3350 the POSIX wrapper now has to get and free some memory, and copy the results.
3352 6. Added some simple autoconf support, placing the test data and the
3353 documentation in separate directories, re-organizing some of the
3354 information files, and making it build pcre-config (a GNU standard). Also added
3355 libtool support for building PCRE as a shared library, which is now the
3356 default.
3358 7. Got rid of the leading zero in the definition of PCRE_MINOR because 08 and
3359 09 are not valid octal constants. Single digits will be used for minor values
3360 less than 10.
3362 8. Defined REG_EXTENDED and REG_NOSUB as zero in the POSIX header, so that
3363 existing programs that set these in the POSIX interface can use PCRE without
3364 modification.
3366 9. Added a new function, pcre_fullinfo() with an extensible interface. It can
3367 return all that pcre_info() returns, plus additional data. The pcre_info()
3368 function is retained for compatibility, but is considered to be obsolete.
3370 10. Added experimental recursion feature (?R) to handle one common case that
3371 Perl 5.6 will be able to do with (?p{...}).
3373 11. Added support for POSIX character classes like [:alpha:], which Perl is
3374 adopting.
3377 Version 2.08 31-Aug-99
3378 ----------------------
3380 1. When startoffset was not zero and the pattern began with ".*", PCRE was not
3381 trying to match at the startoffset position, but instead was moving forward to
3382 the next newline as if a previous match had failed.
3384 2. pcretest was not making use of PCRE_NOTEMPTY when repeating for /g and /G,
3385 and could get into a loop if a null string was matched other than at the start
3386 of the subject.
3388 3. Added definitions of PCRE_MAJOR and PCRE_MINOR to pcre.h so the version can
3389 be distinguished at compile time, and for completeness also added PCRE_DATE.
3391 5. Added Paul Sokolovsky's minor changes to make it easy to compile a Win32 DLL
3392 in GnuWin32 environments.
3395 Version 2.07 29-Jul-99
3396 ----------------------
3398 1. The documentation is now supplied in plain text form and HTML as well as in
3399 the form of man page sources.
3401 2. C++ compilers don't like assigning (void *) values to other pointer types.
3402 In particular this affects malloc(). Although there is no problem in Standard
3403 C, I've put in casts to keep C++ compilers happy.
3405 3. Typo on pcretest.c; a cast of (unsigned char *) in the POSIX regexec() call
3406 should be (const char *).
3408 4. If NOPOSIX is defined, pcretest.c compiles without POSIX support. This may
3409 be useful for non-Unix systems who don't want to bother with the POSIX stuff.
3410 However, I haven't made this a standard facility. The documentation doesn't
3411 mention it, and the Makefile doesn't support it.
3413 5. The Makefile now contains an "install" target, with editable destinations at
3414 the top of the file. The pcretest program is not installed.
3416 6. pgrep -V now gives the PCRE version number and date.
3418 7. Fixed bug: a zero repetition after a literal string (e.g. /abcde{0}/) was
3419 causing the entire string to be ignored, instead of just the last character.
3421 8. If a pattern like /"([^\\"]+|\\.)*"/ is applied in the normal way to a
3422 non-matching string, it can take a very, very long time, even for strings of
3423 quite modest length, because of the nested recursion. PCRE now does better in
3424 some of these cases. It does this by remembering the last required literal
3425 character in the pattern, and pre-searching the subject to ensure it is present
3426 before running the real match. In other words, it applies a heuristic to detect
3427 some types of certain failure quickly, and in the above example, if presented
3428 with a string that has no trailing " it gives "no match" very quickly.
3430 9. A new runtime option PCRE_NOTEMPTY causes null string matches to be ignored;
3431 other alternatives are tried instead.
3434 Version 2.06 09-Jun-99
3435 ----------------------
3437 1. Change pcretest's output for amount of store used to show just the code
3438 space, because the remainder (the data block) varies in size between 32-bit and
3439 64-bit systems.
3441 2. Added an extra argument to pcre_exec() to supply an offset in the subject to
3442 start matching at. This allows lookbehinds to work when searching for multiple
3443 occurrences in a string.
3445 3. Added additional options to pcretest for testing multiple occurrences:
3447 /+ outputs the rest of the string that follows a match
3448 /g loops for multiple occurrences, using the new startoffset argument
3449 /G loops for multiple occurrences by passing an incremented pointer
3451 4. PCRE wasn't doing the "first character" optimization for patterns starting
3452 with \b or \B, though it was doing it for other lookbehind assertions. That is,
3453 it wasn't noticing that a match for a pattern such as /\bxyz/ has to start with
3454 the letter 'x'. On long subject strings, this gives a significant speed-up.
3457 Version 2.05 21-Apr-99
3458 ----------------------
3460 1. Changed the type of magic_number from int to long int so that it works
3461 properly on 16-bit systems.
3463 2. Fixed a bug which caused patterns starting with .* not to work correctly
3464 when the subject string contained newline characters. PCRE was assuming
3465 anchoring for such patterns in all cases, which is not correct because .* will
3466 not pass a newline unless PCRE_DOTALL is set. It now assumes anchoring only if
3467 DOTALL is set at top level; otherwise it knows that patterns starting with .*
3468 must be retried after every newline in the subject.
3471 Version 2.04 18-Feb-99
3472 ----------------------
3474 1. For parenthesized subpatterns with repeats whose minimum was zero, the
3475 computation of the store needed to hold the pattern was incorrect (too large).
3476 If such patterns were nested a few deep, this could multiply and become a real
3477 problem.
3479 2. Added /M option to pcretest to show the memory requirement of a specific
3480 pattern. Made -m a synonym of -s (which does this globally) for compatibility.
3482 3. Subpatterns of the form (regex){n,m} (i.e. limited maximum) were being
3483 compiled in such a way that the backtracking after subsequent failure was
3484 pessimal. Something like (a){0,3} was compiled as (a)?(a)?(a)? instead of
3485 ((a)((a)(a)?)?)? with disastrous performance if the maximum was of any size.
3488 Version 2.03 02-Feb-99
3489 ----------------------
3491 1. Fixed typo and small mistake in man page.
3493 2. Added 4th condition (GPL supersedes if conflict) and created separate
3494 LICENCE file containing the conditions.
3496 3. Updated pcretest so that patterns such as /abc\/def/ work like they do in
3497 Perl, that is the internal \ allows the delimiter to be included in the
3498 pattern. Locked out the use of \ as a delimiter. If \ immediately follows
3499 the final delimiter, add \ to the end of the pattern (to test the error).
3501 4. Added the convenience functions for extracting substrings after a successful
3502 match. Updated pcretest to make it able to test these functions.
3505 Version 2.02 14-Jan-99
3506 ----------------------
3508 1. Initialized the working variables associated with each extraction so that
3509 their saving and restoring doesn't refer to uninitialized store.
3511 2. Put dummy code into study.c in order to trick the optimizer of the IBM C
3512 compiler for OS/2 into generating correct code. Apparently IBM isn't going to
3513 fix the problem.
3515 3. Pcretest: the timing code wasn't using LOOPREPEAT for timing execution
3516 calls, and wasn't printing the correct value for compiling calls. Increased the
3517 default value of LOOPREPEAT, and the number of significant figures in the
3518 times.
3520 4. Changed "/bin/rm" in the Makefile to "-rm" so it works on Windows NT.
3522 5. Renamed "deftables" as "dftables" to get it down to 8 characters, to avoid
3523 a building problem on Windows NT with a FAT file system.
3526 Version 2.01 21-Oct-98
3527 ----------------------
3529 1. Changed the API for pcre_compile() to allow for the provision of a pointer
3530 to character tables built by pcre_maketables() in the current locale. If NULL
3531 is passed, the default tables are used.
3534 Version 2.00 24-Sep-98
3535 ----------------------
3537 1. Since the (>?) facility is in Perl 5.005, don't require PCRE_EXTRA to enable
3538 it any more.
3540 2. Allow quantification of (?>) groups, and make it work correctly.
3542 3. The first character computation wasn't working for (?>) groups.
3544 4. Correct the implementation of \Z (it is permitted to match on the \n at the
3545 end of the subject) and add 5.005's \z, which really does match only at the
3546 very end of the subject.
3548 5. Remove the \X "cut" facility; Perl doesn't have it, and (?> is neater.
3550 6. Remove the ability to specify CASELESS, MULTILINE, DOTALL, and
3551 DOLLAR_END_ONLY at runtime, to make it possible to implement the Perl 5.005
3552 localized options. All options to pcre_study() were also removed.
3554 7. Add other new features from 5.005:
3556 $(?<= positive lookbehind
3557 $(?<! negative lookbehind
3558 (?imsx-imsx) added the unsetting capability
3559 such a setting is global if at outer level; local otherwise
3560 (?imsx-imsx:) non-capturing groups with option setting
3561 (?(cond)re|re) conditional pattern matching
3563 A backreference to itself in a repeated group matches the previous
3564 captured string.
3566 8. General tidying up of studying (both automatic and via "study")
3567 consequential on the addition of new assertions.
3569 9. As in 5.005, unlimited repeated groups that could match an empty substring
3570 are no longer faulted at compile time. Instead, the loop is forcibly broken at
3571 runtime if any iteration does actually match an empty substring.
3573 10. Include the RunTest script in the distribution.
3575 11. Added tests from the Perl 5.005_02 distribution. This showed up a few
3576 discrepancies, some of which were old and were also with respect to 5.004. They
3577 have now been fixed.
3580 Version 1.09 28-Apr-98
3581 ----------------------
3583 1. A negated single character class followed by a quantifier with a minimum
3584 value of one (e.g. [^x]{1,6} ) was not compiled correctly. This could lead to
3585 program crashes, or just wrong answers. This did not apply to negated classes
3586 containing more than one character, or to minima other than one.
3589 Version 1.08 27-Mar-98
3590 ----------------------
3592 1. Add PCRE_UNGREEDY to invert the greediness of quantifiers.
3594 2. Add (?U) and (?X) to set PCRE_UNGREEDY and PCRE_EXTRA respectively. The
3595 latter must appear before anything that relies on it in the pattern.
3598 Version 1.07 16-Feb-98
3599 ----------------------
3601 1. A pattern such as /((a)*)*/ was not being diagnosed as in error (unlimited
3602 repeat of a potentially empty string).
3605 Version 1.06 23-Jan-98
3606 ----------------------
3608 1. Added Markus Oberhumer's little patches for C++.
3610 2. Literal strings longer than 255 characters were broken.
3613 Version 1.05 23-Dec-97
3614 ----------------------
3616 1. Negated character classes containing more than one character were failing if
3617 PCRE_CASELESS was set at run time.
3620 Version 1.04 19-Dec-97
3621 ----------------------
3623 1. Corrected the man page, where some "const" qualifiers had been omitted.
3625 2. Made debugging output print "{0,xxx}" instead of just "{,xxx}" to agree with
3626 input syntax.
3628 3. Fixed memory leak which occurred when a regex with back references was
3629 matched with an offsets vector that wasn't big enough. The temporary memory
3630 that is used in this case wasn't being freed if the match failed.
3632 4. Tidied pcretest to ensure it frees memory that it gets.
3634 5. Temporary memory was being obtained in the case where the passed offsets
3635 vector was exactly big enough.
3637 6. Corrected definition of offsetof() from change 5 below.
3639 7. I had screwed up change 6 below and broken the rules for the use of
3640 setjmp(). Now fixed.
3643 Version 1.03 18-Dec-97
3644 ----------------------
3646 1. A erroneous regex with a missing opening parenthesis was correctly
3647 diagnosed, but PCRE attempted to access brastack[-1], which could cause crashes
3648 on some systems.
3650 2. Replaced offsetof(real_pcre, code) by offsetof(real_pcre, code[0]) because
3651 it was reported that one broken compiler failed on the former because "code" is
3652 also an independent variable.
3654 3. The erroneous regex a[]b caused an array overrun reference.
3656 4. A regex ending with a one-character negative class (e.g. /[^k]$/) did not
3657 fail on data ending with that character. (It was going on too far, and checking
3658 the next character, typically a binary zero.) This was specific to the
3659 optimized code for single-character negative classes.
3661 5. Added a contributed patch from the TIN world which does the following:
3663 + Add an undef for memmove, in case the the system defines a macro for it.
3665 + Add a definition of offsetof(), in case there isn't one. (I don't know
3666 the reason behind this - offsetof() is part of the ANSI standard - but
3667 it does no harm).
3669 + Reduce the ifdef's in pcre.c using macro DPRINTF, thereby eliminating
3670 most of the places where whitespace preceded '#'. I have given up and
3671 allowed the remaining 2 cases to be at the margin.
3673 + Rename some variables in pcre to eliminate shadowing. This seems very
3674 pedantic, but does no harm, of course.
3676 6. Moved the call to setjmp() into its own function, to get rid of warnings
3677 from gcc -Wall, and avoided calling it at all unless PCRE_EXTRA is used.
3679 7. Constructs such as \d{8,} were compiling into the equivalent of
3680 \d{8}\d{0,65527} instead of \d{8}\d* which didn't make much difference to the
3681 outcome, but in this particular case used more store than had been allocated,
3682 which caused the bug to be discovered because it threw up an internal error.
3684 8. The debugging code in both pcre and pcretest for outputting the compiled
3685 form of a regex was going wrong in the case of back references followed by
3686 curly-bracketed repeats.
3689 Version 1.02 12-Dec-97
3690 ----------------------
3692 1. Typos in pcre.3 and comments in the source fixed.
3694 2. Applied a contributed patch to get rid of places where it used to remove
3695 'const' from variables, and fixed some signed/unsigned and uninitialized
3696 variable warnings.
3698 3. Added the "runtest" target to Makefile.
3700 4. Set default compiler flag to -O2 rather than just -O.
3703 Version 1.01 19-Nov-97
3704 ----------------------
3706 1. PCRE was failing to diagnose unlimited repeat of empty string for patterns
3707 like /([ab]*)*/, that is, for classes with more than one character in them.
3709 2. Likewise, it wasn't diagnosing patterns with "once-only" subpatterns, such
3710 as /((?>a*))*/ (a PCRE_EXTRA facility).
3713 Version 1.00 18-Nov-97
3714 ----------------------
3716 1. Added compile-time macros to support systems such as SunOS4 which don't have
3717 memmove() or strerror() but have other things that can be used instead.
3719 2. Arranged that "make clean" removes the executables.
3722 Version 0.99 27-Oct-97
3723 ----------------------
3725 1. Fixed bug in code for optimizing classes with only one character. It was
3726 initializing a 32-byte map regardless, which could cause it to run off the end
3727 of the memory it had got.
3729 2. Added, conditional on PCRE_EXTRA, the proposed (?>REGEX) construction.
3732 Version 0.98 22-Oct-97
3733 ----------------------
3735 1. Fixed bug in code for handling temporary memory usage when there are more
3736 back references than supplied space in the ovector. This could cause segfaults.
3739 Version 0.97 21-Oct-97
3740 ----------------------
3742 1. Added the \X "cut" facility, conditional on PCRE_EXTRA.
3744 2. Optimized negated single characters not to use a bit map.
3746 3. Brought error texts together as macro definitions; clarified some of them;
3747 fixed one that was wrong - it said "range out of order" when it meant "invalid
3748 escape sequence".
3750 4. Changed some char * arguments to const char *.
3752 5. Added PCRE_NOTBOL and PCRE_NOTEOL (from POSIX).
3754 6. Added the POSIX-style API wrapper in pcreposix.a and testing facilities in
3755 pcretest.
3758 Version 0.96 16-Oct-97
3759 ----------------------
3761 1. Added a simple "pgrep" utility to the distribution.
3763 2. Fixed an incompatibility with Perl: "{" is now treated as a normal character
3764 unless it appears in one of the precise forms "{ddd}", "{ddd,}", or "{ddd,ddd}"
3765 where "ddd" means "one or more decimal digits".
3767 3. Fixed serious bug. If a pattern had a back reference, but the call to
3768 pcre_exec() didn't supply a large enough ovector to record the related
3769 identifying subpattern, the match always failed. PCRE now remembers the number
3770 of the largest back reference, and gets some temporary memory in which to save
3771 the offsets during matching if necessary, in order to ensure that
3772 backreferences always work.
3774 4. Increased the compatibility with Perl in a number of ways:
3776 (a) . no longer matches \n by default; an option PCRE_DOTALL is provided
3777 to request this handling. The option can be set at compile or exec time.
3779 (b) $ matches before a terminating newline by default; an option
3780 PCRE_DOLLAR_ENDONLY is provided to override this (but not in multiline
3781 mode). The option can be set at compile or exec time.
3783 (c) The handling of \ followed by a digit other than 0 is now supposed to be
3784 the same as Perl's. If the decimal number it represents is less than 10
3785 or there aren't that many previous left capturing parentheses, an octal
3786 escape is read. Inside a character class, it's always an octal escape,
3787 even if it is a single digit.
3789 (d) An escaped but undefined alphabetic character is taken as a literal,
3790 unless PCRE_EXTRA is set. Currently this just reserves the remaining
3791 escapes.
3793 (e) {0} is now permitted. (The previous item is removed from the compiled
3794 pattern).
3796 5. Changed all the names of code files so that the basic parts are no longer
3797 than 10 characters, and abolished the teeny "globals.c" file.
3799 6. Changed the handling of character classes; they are now done with a 32-byte
3800 bit map always.
3802 7. Added the -d and /D options to pcretest to make it possible to look at the
3803 internals of compilation without having to recompile pcre.
3806 Version 0.95 23-Sep-97
3807 ----------------------
3809 1. Fixed bug in pre-pass concerning escaped "normal" characters such as \x5c or
3810 \x20 at the start of a run of normal characters. These were being treated as
3811 real characters, instead of the source characters being re-checked.
3814 Version 0.94 18-Sep-97
3815 ----------------------
3817 1. The functions are now thread-safe, with the caveat that the global variables
3818 containing pointers to malloc() and free() or alternative functions are the
3819 same for all threads.
3821 2. Get pcre_study() to generate a bitmap of initial characters for non-
3822 anchored patterns when this is possible, and use it if passed to pcre_exec().
3825 Version 0.93 15-Sep-97
3826 ----------------------
3828 1. /(b)|(:+)/ was computing an incorrect first character.
3830 2. Add pcre_study() to the API and the passing of pcre_extra to pcre_exec(),
3831 but not actually doing anything yet.
3833 3. Treat "-" characters in classes that cannot be part of ranges as literals,
3834 as Perl does (e.g. [-az] or [az-]).
3836 4. Set the anchored flag if a branch starts with .* or .*? because that tests
3837 all possible positions.
3839 5. Split up into different modules to avoid including unneeded functions in a
3840 compiled binary. However, compile and exec are still in one module. The "study"
3841 function is split off.
3843 6. The character tables are now in a separate module whose source is generated
3844 by an auxiliary program - but can then be edited by hand if required. There are
3845 now no calls to isalnum(), isspace(), isdigit(), isxdigit(), tolower() or
3846 toupper() in the code.
3848 7. Turn the malloc/free funtions variables into pcre_malloc and pcre_free and
3849 make them global. Abolish the function for setting them, as the caller can now
3850 set them directly.
3853 Version 0.92 11-Sep-97
3854 ----------------------
3856 1. A repeat with a fixed maximum and a minimum of 1 for an ordinary character
3857 (e.g. /a{1,3}/) was broken (I mis-optimized it).
3859 2. Caseless matching was not working in character classes if the characters in
3860 the pattern were in upper case.
3862 3. Make ranges like [W-c] work in the same way as Perl for caseless matching.
3864 4. Make PCRE_ANCHORED public and accept as a compile option.
3866 5. Add an options word to pcre_exec() and accept PCRE_ANCHORED and
3867 PCRE_CASELESS at run time. Add escapes \A and \I to pcretest to cause it to
3868 pass them.
3870 6. Give an error if bad option bits passed at compile or run time.
3872 7. Add PCRE_MULTILINE at compile and exec time, and (?m) as well. Add \M to
3873 pcretest to cause it to pass that flag.
3875 8. Add pcre_info(), to get the number of identifying subpatterns, the stored
3876 options, and the first character, if set.
3878 9. Recognize C+ or C{n,m} where n >= 1 as providing a fixed starting character.
3881 Version 0.91 10-Sep-97
3882 ----------------------
3884 1. PCRE was failing to diagnose unlimited repeats of subpatterns that could
3885 match the empty string as in /(a*)*/. It was looping and ultimately crashing.
3887 2. PCRE was looping on encountering an indefinitely repeated back reference to
3888 a subpattern that had matched an empty string, e.g. /(a|)\1*/. It now does what
3889 Perl does - treats the match as successful.
3891 ****


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

  ViewVC Help
Powered by ViewVC 1.1.5