ViewVC logotype

Contents of /code/trunk/ChangeLog

Parent Directory Parent Directory | Revision Log Revision Log

Revision 516 - (show annotations)
Tue May 4 15:51:35 2010 UTC (11 years, 5 months ago) by ph10
File size: 185927 byte(s)
Avoid compiler warning in pcretest.
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.
29 7. Minor change to pcretest.c to avoid a compiler warning.
32 Version 8.02 19-Mar-2010
33 ------------------------
35 1. The Unicode data tables have been updated to Unicode 5.2.0.
37 2. Added the option --libs-cpp to pcre-config, but only when C++ support is
38 configured.
40 3. Updated the licensing terms in the pcregexp.pas file, as agreed with the
41 original author of that file, following a query about its status.
43 4. On systems that do not have stdint.h (e.g. Solaris), check for and include
44 inttypes.h instead. This fixes a bug that was introduced by change 8.01/8.
46 5. A pattern such as (?&t)*+(?(DEFINE)(?<t>.)) which has a possessive
47 quantifier applied to a forward-referencing subroutine call, could compile
48 incorrect code or give the error "internal error: previously-checked
49 referenced subpattern not found".
51 6. Both MS Visual Studio and Symbian OS have problems with initializing
52 variables to point to external functions. For these systems, therefore,
53 pcre_malloc etc. are now initialized to local functions that call the
54 relevant global functions.
56 7. There were two entries missing in the vectors called coptable and poptable
57 in pcre_dfa_exec.c. This could lead to memory accesses outsize the vectors.
58 I've fixed the data, and added a kludgy way of testing at compile time that
59 the lengths are correct (equal to the number of opcodes).
61 8. Following on from 7, I added a similar kludge to check the length of the
62 eint vector in pcreposix.c.
64 9. Error texts for pcre_compile() are held as one long string to avoid too
65 much relocation at load time. To find a text, the string is searched,
66 counting zeros. There was no check for running off the end of the string,
67 which could happen if a new error number was added without updating the
68 string.
70 10. \K gave a compile-time error if it appeared in a lookbehind assersion.
72 11. \K was not working if it appeared in an atomic group or in a group that
73 was called as a "subroutine", or in an assertion. Perl 5.11 documents that
74 \K is "not well defined" if used in an assertion. PCRE now accepts it if
75 the assertion is positive, but not if it is negative.
77 12. Change 11 fortuitously reduced the size of the stack frame used in the
78 "match()" function of pcre_exec.c by one pointer. Forthcoming
79 implementation of support for (*MARK) will need an extra pointer on the
80 stack; I have reserved it now, so that the stack frame size does not
81 decrease.
83 13. A pattern such as (?P<L1>(?P<L2>0)|(?P>L2)(?P>L1)) in which the only other
84 item in branch that calls a recursion is a subroutine call - as in the
85 second branch in the above example - was incorrectly given the compile-
86 time error "recursive call could loop indefinitely" because pcre_compile()
87 was not correctly checking the subroutine for matching a non-empty string.
89 14. The checks for overrunning compiling workspace could trigger after an
90 overrun had occurred. This is a "should never occur" error, but it can be
91 triggered by pathological patterns such as hundreds of nested parentheses.
92 The checks now trigger 100 bytes before the end of the workspace.
94 15. Fix typo in configure.ac: "srtoq" should be "strtoq".
97 Version 8.01 19-Jan-2010
98 ------------------------
100 1. If a pattern contained a conditional subpattern with only one branch (in
101 particular, this includes all (*DEFINE) patterns), a call to pcre_study()
102 computed the wrong minimum data length (which is of course zero for such
103 subpatterns). This could cause incorrect "no match" results.
105 2. For patterns such as (?i)a(?-i)b|c where an option setting at the start of
106 the pattern is reset in the first branch, pcre_compile() failed with
107 "internal error: code overflow at offset...". This happened only when
108 the reset was to the original external option setting. (An optimization
109 abstracts leading options settings into an external setting, which was the
110 cause of this.)
112 3. A pattern such as ^(?!a(*SKIP)b) where a negative assertion contained one
113 of the verbs SKIP, PRUNE, or COMMIT, did not work correctly. When the
114 assertion pattern did not match (meaning that the assertion was true), it
115 was incorrectly treated as false if the SKIP had been reached during the
116 matching. This also applied to assertions used as conditions.
118 4. If an item that is not supported by pcre_dfa_exec() was encountered in an
119 assertion subpattern, including such a pattern used as a condition,
120 unpredictable results occurred, instead of the error return
123 5. The C++ GlobalReplace function was not working like Perl for the special
124 situation when an empty string is matched. It now does the fancy magic
125 stuff that is necessary.
127 6. In pcre_internal.h, obsolete includes to setjmp.h and stdarg.h have been
128 removed. (These were left over from very, very early versions of PCRE.)
130 7. Some cosmetic changes to the code to make life easier when compiling it
131 as part of something else:
133 (a) Change DEBUG to PCRE_DEBUG.
135 (b) In pcre_compile(), rename the member of the "branch_chain" structure
136 called "current" as "current_branch", to prevent a collision with the
137 Linux macro when compiled as a kernel module.
139 (c) In pcre_study(), rename the function set_bit() as set_table_bit(), to
140 prevent a collision with the Linux macro when compiled as a kernel
141 module.
143 8. In pcre_compile() there are some checks for integer overflows that used to
144 cast potentially large values to (double). This has been changed to that
145 when building, a check for int64_t is made, and if it is found, it is used
146 instead, thus avoiding the use of floating point arithmetic. (There is no
147 other use of FP in PCRE.) If int64_t is not found, the fallback is to
148 double.
150 9. Added two casts to avoid signed/unsigned warnings from VS Studio Express
151 2005 (difference between two addresses compared to an unsigned value).
153 10. Change the standard AC_CHECK_LIB test for libbz2 in configure.ac to a
154 custom one, because of the following reported problem in Windows:
156 - libbz2 uses the Pascal calling convention (WINAPI) for the functions
157 under Win32.
158 - The standard autoconf AC_CHECK_LIB fails to include "bzlib.h",
159 therefore missing the function definition.
160 - The compiler thus generates a "C" signature for the test function.
161 - The linker fails to find the "C" function.
162 - PCRE fails to configure if asked to do so against libbz2.
164 11. When running libtoolize from libtool-2.2.6b as part of autogen.sh, these
165 messages were output:
167 Consider adding `AC_CONFIG_MACRO_DIR([m4])' to configure.ac and
168 rerunning libtoolize, to keep the correct libtool macros in-tree.
169 Consider adding `-I m4' to ACLOCAL_AMFLAGS in Makefile.am.
171 I have done both of these things.
173 12. Although pcre_dfa_exec() does not use nearly as much stack as pcre_exec()
174 most of the time, it *can* run out if it is given a pattern that contains a
175 runaway infinite recursion. I updated the discussion in the pcrestack man
176 page.
178 13. Now that we have gone to the x.xx style of version numbers, the minor
179 version may start with zero. Using 08 or 09 is a bad idea because users
180 might check the value of PCRE_MINOR in their code, and 08 or 09 may be
181 interpreted as invalid octal numbers. I've updated the previous comment in
182 configure.ac, and also added a check that gives an error if 08 or 09 are
183 used.
185 14. Change 8.00/11 was not quite complete: code had been accidentally omitted,
186 causing partial matching to fail when the end of the subject matched \W
187 in a UTF-8 pattern where \W was quantified with a minimum of 3.
189 15. There were some discrepancies between the declarations in pcre_internal.h
190 of _pcre_is_newline(), _pcre_was_newline(), and _pcre_valid_utf8() and
191 their definitions. The declarations used "const uschar *" and the
192 definitions used USPTR. Even though USPTR is normally defined as "const
193 unsigned char *" (and uschar is typedeffed as "unsigned char"), it was
194 reported that: "This difference in casting confuses some C++ compilers, for
195 example, SunCC recognizes above declarations as different functions and
196 generates broken code for hbpcre." I have changed the declarations to use
197 USPTR.
199 16. GNU libtool is named differently on some systems. The autogen.sh script now
200 tries several variants such as glibtoolize (MacOSX) and libtoolize1x
201 (FreeBSD).
203 17. Applied Craig's patch that fixes an HP aCC compile error in pcre 8.00
204 (strtoXX undefined when compiling pcrecpp.cc). The patch contains this
205 comment: "Figure out how to create a longlong from a string: strtoll and
206 equivalent. It's not enough to call AC_CHECK_FUNCS: hpux has a strtoll, for
207 instance, but it only takes 2 args instead of 3!"
209 18. A subtle bug concerned with back references has been fixed by a change of
210 specification, with a corresponding code fix. A pattern such as
211 ^(xa|=?\1a)+$ which contains a back reference inside the group to which it
212 refers, was giving matches when it shouldn't. For example, xa=xaaa would
213 match that pattern. Interestingly, Perl (at least up to 5.11.3) has the
214 same bug. Such groups have to be quantified to be useful, or contained
215 inside another quantified group. (If there's no repetition, the reference
216 can never match.) The problem arises because, having left the group and
217 moved on to the rest of the pattern, a later failure that backtracks into
218 the group uses the captured value from the final iteration of the group
219 rather than the correct earlier one. I have fixed this in PCRE by forcing
220 any group that contains a reference to itself to be an atomic group; that
221 is, there cannot be any backtracking into it once it has completed. This is
222 similar to recursive and subroutine calls.
225 Version 8.00 19-Oct-09
226 ----------------------
228 1. The table for translating pcre_compile() error codes into POSIX error codes
229 was out-of-date, and there was no check on the pcre_compile() error code
230 being within the table. This could lead to an OK return being given in
231 error.
233 2. Changed the call to open a subject file in pcregrep from fopen(pathname,
234 "r") to fopen(pathname, "rb"), which fixed a problem with some of the tests
235 in a Windows environment.
237 3. The pcregrep --count option prints the count for each file even when it is
238 zero, as does GNU grep. However, pcregrep was also printing all files when
239 --files-with-matches was added. Now, when both options are given, it prints
240 counts only for those files that have at least one match. (GNU grep just
241 prints the file name in this circumstance, but including the count seems
242 more useful - otherwise, why use --count?) Also ensured that the
243 combination -clh just lists non-zero counts, with no names.
245 4. The long form of the pcregrep -F option was incorrectly implemented as
246 --fixed_strings instead of --fixed-strings. This is an incompatible change,
247 but it seems right to fix it, and I didn't think it was worth preserving
248 the old behaviour.
250 5. The command line items --regex=pattern and --regexp=pattern were not
251 recognized by pcregrep, which required --regex pattern or --regexp pattern
252 (with a space rather than an '='). The man page documented the '=' forms,
253 which are compatible with GNU grep; these now work.
255 6. No libpcreposix.pc file was created for pkg-config; there was just
256 libpcre.pc and libpcrecpp.pc. The omission has been rectified.
258 7. Added #ifndef SUPPORT_UCP into the pcre_ucd.c module, to reduce its size
259 when UCP support is not needed, by modifying the Python script that
260 generates it from Unicode data files. This should not matter if the module
261 is correctly used as a library, but I received one complaint about 50K of
262 unwanted data. My guess is that the person linked everything into his
263 program rather than using a library. Anyway, it does no harm.
265 8. A pattern such as /\x{123}{2,2}+/8 was incorrectly compiled; the trigger
266 was a minimum greater than 1 for a wide character in a possessive
267 repetition. The same bug could also affect patterns like /(\x{ff}{0,2})*/8
268 which had an unlimited repeat of a nested, fixed maximum repeat of a wide
269 character. Chaos in the form of incorrect output or a compiling loop could
270 result.
272 9. The restrictions on what a pattern can contain when partial matching is
273 requested for pcre_exec() have been removed. All patterns can now be
274 partially matched by this function. In addition, if there are at least two
275 slots in the offset vector, the offset of the earliest inspected character
276 for the match and the offset of the end of the subject are set in them when
277 PCRE_ERROR_PARTIAL is returned.
279 10. Partial matching has been split into two forms: PCRE_PARTIAL_SOFT, which is
280 synonymous with PCRE_PARTIAL, for backwards compatibility, and
281 PCRE_PARTIAL_HARD, which causes a partial match to supersede a full match,
282 and may be more useful for multi-segment matching.
284 11. Partial matching with pcre_exec() is now more intuitive. A partial match
285 used to be given if ever the end of the subject was reached; now it is
286 given only if matching could not proceed because another character was
287 needed. This makes a difference in some odd cases such as Z(*FAIL) with the
288 string "Z", which now yields "no match" instead of "partial match". In the
289 case of pcre_dfa_exec(), "no match" is given if every matching path for the
290 final character ended with (*FAIL).
292 12. Restarting a match using pcre_dfa_exec() after a partial match did not work
293 if the pattern had a "must contain" character that was already found in the
294 earlier partial match, unless partial matching was again requested. For
295 example, with the pattern /dog.(body)?/, the "must contain" character is
296 "g". If the first part-match was for the string "dog", restarting with
297 "sbody" failed. This bug has been fixed.
299 13. The string returned by pcre_dfa_exec() after a partial match has been
300 changed so that it starts at the first inspected character rather than the
301 first character of the match. This makes a difference only if the pattern
302 starts with a lookbehind assertion or \b or \B (\K is not supported by
303 pcre_dfa_exec()). It's an incompatible change, but it makes the two
304 matching functions compatible, and I think it's the right thing to do.
306 14. Added a pcredemo man page, created automatically from the pcredemo.c file,
307 so that the demonstration program is easily available in environments where
308 PCRE has not been installed from source.
310 15. Arranged to add -DPCRE_STATIC to cflags in libpcre.pc, libpcreposix.cp,
311 libpcrecpp.pc and pcre-config when PCRE is not compiled as a shared
312 library.
314 16. Added REG_UNGREEDY to the pcreposix interface, at the request of a user.
315 It maps to PCRE_UNGREEDY. It is not, of course, POSIX-compatible, but it
316 is not the first non-POSIX option to be added. Clearly some people find
317 these options useful.
319 17. If a caller to the POSIX matching function regexec() passes a non-zero
320 value for nmatch with a NULL value for pmatch, the value of
321 nmatch is forced to zero.
323 18. RunGrepTest did not have a test for the availability of the -u option of
324 the diff command, as RunTest does. It now checks in the same way as
325 RunTest, and also checks for the -b option.
327 19. If an odd number of negated classes containing just a single character
328 interposed, within parentheses, between a forward reference to a named
329 subpattern and the definition of the subpattern, compilation crashed with
330 an internal error, complaining that it could not find the referenced
331 subpattern. An example of a crashing pattern is /(?&A)(([^m])(?<A>))/.
332 [The bug was that it was starting one character too far in when skipping
333 over the character class, thus treating the ] as data rather than
334 terminating the class. This meant it could skip too much.]
336 20. Added PCRE_NOTEMPTY_ATSTART in order to be able to correctly implement the
337 /g option in pcretest when the pattern contains \K, which makes it possible
338 to have an empty string match not at the start, even when the pattern is
339 anchored. Updated pcretest and pcredemo to use this option.
341 21. If the maximum number of capturing subpatterns in a recursion was greater
342 than the maximum at the outer level, the higher number was returned, but
343 with unset values at the outer level. The correct (outer level) value is
344 now given.
346 22. If (*ACCEPT) appeared inside capturing parentheses, previous releases of
347 PCRE did not set those parentheses (unlike Perl). I have now found a way to
348 make it do so. The string so far is captured, making this feature
349 compatible with Perl.
351 23. The tests have been re-organized, adding tests 11 and 12, to make it
352 possible to check the Perl 5.10 features against Perl 5.10.
354 24. Perl 5.10 allows subroutine calls in lookbehinds, as long as the subroutine
355 pattern matches a fixed length string. PCRE did not allow this; now it
356 does. Neither allows recursion.
358 25. I finally figured out how to implement a request to provide the minimum
359 length of subject string that was needed in order to match a given pattern.
360 (It was back references and recursion that I had previously got hung up
361 on.) This code has now been added to pcre_study(); it finds a lower bound
362 to the length of subject needed. It is not necessarily the greatest lower
363 bound, but using it to avoid searching strings that are too short does give
364 some useful speed-ups. The value is available to calling programs via
365 pcre_fullinfo().
367 26. While implementing 25, I discovered to my embarrassment that pcretest had
368 not been passing the result of pcre_study() to pcre_dfa_exec(), so the
369 study optimizations had never been tested with that matching function.
370 Oops. What is worse, even when it was passed study data, there was a bug in
371 pcre_dfa_exec() that meant it never actually used it. Double oops. There
372 were also very few tests of studied patterns with pcre_dfa_exec().
374 27. If (?| is used to create subpatterns with duplicate numbers, they are now
375 allowed to have the same name, even if PCRE_DUPNAMES is not set. However,
376 on the other side of the coin, they are no longer allowed to have different
377 names, because these cannot be distinguished in PCRE, and this has caused
378 confusion. (This is a difference from Perl.)
380 28. When duplicate subpattern names are present (necessarily with different
381 numbers, as required by 27 above), and a test is made by name in a
382 conditional pattern, either for a subpattern having been matched, or for
383 recursion in such a pattern, all the associated numbered subpatterns are
384 tested, and the overall condition is true if the condition is true for any
385 one of them. This is the way Perl works, and is also more like the way
386 testing by number works.
389 Version 7.9 11-Apr-09
390 ---------------------
392 1. When building with support for bzlib/zlib (pcregrep) and/or readline
393 (pcretest), all targets were linked against these libraries. This included
394 libpcre, libpcreposix, and libpcrecpp, even though they do not use these
395 libraries. This caused unwanted dependencies to be created. This problem
396 has been fixed, and now only pcregrep is linked with bzlib/zlib and only
397 pcretest is linked with readline.
399 2. The "typedef int BOOL" in pcre_internal.h that was included inside the
400 "#ifndef FALSE" condition by an earlier change (probably 7.8/18) has been
401 moved outside it again, because FALSE and TRUE are already defined in AIX,
402 but BOOL is not.
404 3. The pcre_config() function was treating the PCRE_MATCH_LIMIT and
405 PCRE_MATCH_LIMIT_RECURSION values as ints, when they should be long ints.
407 4. The pcregrep documentation said spaces were inserted as well as colons (or
408 hyphens) following file names and line numbers when outputting matching
409 lines. This is not true; no spaces are inserted. I have also clarified the
410 wording for the --colour (or --color) option.
412 5. In pcregrep, when --colour was used with -o, the list of matching strings
413 was not coloured; this is different to GNU grep, so I have changed it to be
414 the same.
416 6. When --colo(u)r was used in pcregrep, only the first matching substring in
417 each matching line was coloured. Now it goes on to look for further matches
418 of any of the test patterns, which is the same behaviour as GNU grep.
420 7. A pattern that could match an empty string could cause pcregrep to loop; it
421 doesn't make sense to accept an empty string match in pcregrep, so I have
422 locked it out (using PCRE's PCRE_NOTEMPTY option). By experiment, this
423 seems to be how GNU grep behaves.
425 8. The pattern (?(?=.*b)b|^) was incorrectly compiled as "match must be at
426 start or after a newline", because the conditional assertion was not being
427 correctly handled. The rule now is that both the assertion and what follows
428 in the first alternative must satisfy the test.
430 9. If auto-callout was enabled in a pattern with a conditional group whose
431 condition was an assertion, PCRE could crash during matching, both with
432 pcre_exec() and pcre_dfa_exec().
434 10. The PCRE_DOLLAR_ENDONLY option was not working when pcre_dfa_exec() was
435 used for matching.
437 11. Unicode property support in character classes was not working for
438 characters (bytes) greater than 127 when not in UTF-8 mode.
440 12. Added the -M command line option to pcretest.
442 14. Added the non-standard REG_NOTEMPTY option to the POSIX interface.
444 15. Added the PCRE_NO_START_OPTIMIZE match-time option.
446 16. Added comments and documentation about mis-use of no_arg in the C++
447 wrapper.
449 17. Implemented support for UTF-8 encoding in EBCDIC environments, a patch
450 from Martin Jerabek that uses macro names for all relevant character and
451 string constants.
453 18. Added to pcre_internal.h two configuration checks: (a) If both EBCDIC and
454 SUPPORT_UTF8 are set, give an error; (b) If SUPPORT_UCP is set without
455 SUPPORT_UTF8, define SUPPORT_UTF8. The "configure" script handles both of
456 these, but not everybody uses configure.
458 19. A conditional group that had only one branch was not being correctly
459 recognized as an item that could match an empty string. This meant that an
460 enclosing group might also not be so recognized, causing infinite looping
461 (and probably a segfault) for patterns such as ^"((?(?=[a])[^"])|b)*"$
462 with the subject "ab", where knowledge that the repeated group can match
463 nothing is needed in order to break the loop.
465 20. If a pattern that was compiled with callouts was matched using pcre_dfa_
466 exec(), but without supplying a callout function, matching went wrong.
468 21. If PCRE_ERROR_MATCHLIMIT occurred during a recursion, there was a memory
469 leak if the size of the offset vector was greater than 30. When the vector
470 is smaller, the saved offsets during recursion go onto a local stack
471 vector, but for larger vectors malloc() is used. It was failing to free
472 when the recursion yielded PCRE_ERROR_MATCH_LIMIT (or any other "abnormal"
473 error, in fact).
475 22. There was a missing #ifdef SUPPORT_UTF8 round one of the variables in the
476 heapframe that is used only when UTF-8 support is enabled. This caused no
477 problem, but was untidy.
479 23. Steven Van Ingelgem's patch to CMakeLists.txt to change the name
480 CMAKE_BINARY_DIR to PROJECT_BINARY_DIR so that it works when PCRE is
481 included within another project.
483 24. Steven Van Ingelgem's patches to add more options to the CMake support,
484 slightly modified by me:
486 (a) PCRE_BUILD_TESTS can be set OFF not to build the tests, including
487 not building pcregrep.
489 (b) PCRE_BUILD_PCREGREP can be see OFF not to build pcregrep, but only
490 if PCRE_BUILD_TESTS is also set OFF, because the tests use pcregrep.
492 25. Forward references, both numeric and by name, in patterns that made use of
493 duplicate group numbers, could behave incorrectly or give incorrect errors,
494 because when scanning forward to find the reference group, PCRE was not
495 taking into account the duplicate group numbers. A pattern such as
496 ^X(?3)(a)(?|(b)|(q))(Y) is an example.
498 26. Changed a few more instances of "const unsigned char *" to USPTR, making
499 the feature of a custom pointer more persuasive (as requested by a user).
501 27. Wrapped the definitions of fileno and isatty for Windows, which appear in
502 pcretest.c, inside #ifndefs, because it seems they are sometimes already
503 pre-defined.
505 28. Added support for (*UTF8) at the start of a pattern.
507 29. Arrange for flags added by the "release type" setting in CMake to be shown
508 in the configuration summary.
511 Version 7.8 05-Sep-08
512 ---------------------
514 1. Replaced UCP searching code with optimized version as implemented for Ad
515 Muncher (http://www.admuncher.com/) by Peter Kankowski. This uses a two-
516 stage table and inline lookup instead of a function, giving speed ups of 2
517 to 5 times on some simple patterns that I tested. Permission was given to
518 distribute the MultiStage2.py script that generates the tables (it's not in
519 the tarball, but is in the Subversion repository).
521 2. Updated the Unicode datatables to Unicode 5.1.0. This adds yet more
522 scripts.
524 3. Change 12 for 7.7 introduced a bug in pcre_study() when a pattern contained
525 a group with a zero qualifier. The result of the study could be incorrect,
526 or the function might crash, depending on the pattern.
528 4. Caseless matching was not working for non-ASCII characters in back
529 references. For example, /(\x{de})\1/8i was not matching \x{de}\x{fe}.
530 It now works when Unicode Property Support is available.
532 5. In pcretest, an escape such as \x{de} in the data was always generating
533 a UTF-8 string, even in non-UTF-8 mode. Now it generates a single byte in
534 non-UTF-8 mode. If the value is greater than 255, it gives a warning about
535 truncation.
537 6. Minor bugfix in pcrecpp.cc (change "" == ... to NULL == ...).
539 7. Added two (int) casts to pcregrep when printing the difference of two
540 pointers, in case they are 64-bit values.
542 8. Added comments about Mac OS X stack usage to the pcrestack man page and to
543 test 2 if it fails.
545 9. Added PCRE_CALL_CONVENTION just before the names of all exported functions,
546 and a #define of that name to empty if it is not externally set. This is to
547 allow users of MSVC to set it if necessary.
549 10. The PCRE_EXP_DEFN macro which precedes exported functions was missing from
550 the convenience functions in the pcre_get.c source file.
552 11. An option change at the start of a pattern that had top-level alternatives
553 could cause overwriting and/or a crash. This command provoked a crash in
554 some environments:
556 printf "/(?i)[\xc3\xa9\xc3\xbd]|[\xc3\xa9\xc3\xbdA]/8\n" | pcretest
558 This potential security problem was recorded as CVE-2008-2371.
560 12. For a pattern where the match had to start at the beginning or immediately
561 after a newline (e.g /.*anything/ without the DOTALL flag), pcre_exec() and
562 pcre_dfa_exec() could read past the end of the passed subject if there was
563 no match. To help with detecting such bugs (e.g. with valgrind), I modified
564 pcretest so that it places the subject at the end of its malloc-ed buffer.
566 13. The change to pcretest in 12 above threw up a couple more cases when pcre_
567 exec() might read past the end of the data buffer in UTF-8 mode.
569 14. A similar bug to 7.3/2 existed when the PCRE_FIRSTLINE option was set and
570 the data contained the byte 0x85 as part of a UTF-8 character within its
571 first line. This applied both to normal and DFA matching.
573 15. Lazy qualifiers were not working in some cases in UTF-8 mode. For example,
574 /^[^d]*?$/8 failed to match "abc".
576 16. Added a missing copyright notice to pcrecpp_internal.h.
578 17. Make it more clear in the documentation that values returned from
579 pcre_exec() in ovector are byte offsets, not character counts.
581 18. Tidied a few places to stop certain compilers from issuing warnings.
583 19. Updated the Virtual Pascal + BCC files to compile the latest v7.7, as
584 supplied by Stefan Weber. I made a further small update for 7.8 because
585 there is a change of source arrangements: the pcre_searchfuncs.c module is
586 replaced by pcre_ucd.c.
589 Version 7.7 07-May-08
590 ---------------------
592 1. Applied Craig's patch to sort out a long long problem: "If we can't convert
593 a string to a long long, pretend we don't even have a long long." This is
594 done by checking for the strtoq, strtoll, and _strtoi64 functions.
596 2. Applied Craig's patch to pcrecpp.cc to restore ABI compatibility with
597 pre-7.6 versions, which defined a global no_arg variable instead of putting
598 it in the RE class. (See also #8 below.)
600 3. Remove a line of dead code, identified by coverity and reported by Nuno
601 Lopes.
603 4. Fixed two related pcregrep bugs involving -r with --include or --exclude:
605 (1) The include/exclude patterns were being applied to the whole pathnames
606 of files, instead of just to the final components.
608 (2) If there was more than one level of directory, the subdirectories were
609 skipped unless they satisfied the include/exclude conditions. This is
610 inconsistent with GNU grep (and could even be seen as contrary to the
611 pcregrep specification - which I improved to make it absolutely clear).
612 The action now is always to scan all levels of directory, and just
613 apply the include/exclude patterns to regular files.
615 5. Added the --include_dir and --exclude_dir patterns to pcregrep, and used
616 --exclude_dir in the tests to avoid scanning .svn directories.
618 6. Applied Craig's patch to the QuoteMeta function so that it escapes the
619 NUL character as backslash + 0 rather than backslash + NUL, because PCRE
620 doesn't support NULs in patterns.
622 7. Added some missing "const"s to declarations of static tables in
623 pcre_compile.c and pcre_dfa_exec.c.
625 8. Applied Craig's patch to pcrecpp.cc to fix a problem in OS X that was
626 caused by fix #2 above. (Subsequently also a second patch to fix the
627 first patch. And a third patch - this was a messy problem.)
629 9. Applied Craig's patch to remove the use of push_back().
631 10. Applied Alan Lehotsky's patch to add REG_STARTEND support to the POSIX
632 matching function regexec().
634 11. Added support for the Oniguruma syntax \g<name>, \g<n>, \g'name', \g'n',
635 which, however, unlike Perl's \g{...}, are subroutine calls, not back
636 references. PCRE supports relative numbers with this syntax (I don't think
637 Oniguruma does).
639 12. Previously, a group with a zero repeat such as (...){0} was completely
640 omitted from the compiled regex. However, this means that if the group
641 was called as a subroutine from elsewhere in the pattern, things went wrong
642 (an internal error was given). Such groups are now left in the compiled
643 pattern, with a new opcode that causes them to be skipped at execution
644 time.
646 13. Added the PCRE_JAVASCRIPT_COMPAT option. This makes the following changes
647 to the way PCRE behaves:
649 (a) A lone ] character is dis-allowed (Perl treats it as data).
651 (b) A back reference to an unmatched subpattern matches an empty string
652 (Perl fails the current match path).
654 (c) A data ] in a character class must be notated as \] because if the
655 first data character in a class is ], it defines an empty class. (In
656 Perl it is not possible to have an empty class.) The empty class []
657 never matches; it forces failure and is equivalent to (*FAIL) or (?!).
658 The negative empty class [^] matches any one character, independently
659 of the DOTALL setting.
661 14. A pattern such as /(?2)[]a()b](abc)/ which had a forward reference to a
662 non-existent subpattern following a character class starting with ']' and
663 containing () gave an internal compiling error instead of "reference to
664 non-existent subpattern". Fortunately, when the pattern did exist, the
665 compiled code was correct. (When scanning forwards to check for the
666 existencd of the subpattern, it was treating the data ']' as terminating
667 the class, so got the count wrong. When actually compiling, the reference
668 was subsequently set up correctly.)
670 15. The "always fail" assertion (?!) is optimzed to (*FAIL) by pcre_compile;
671 it was being rejected as not supported by pcre_dfa_exec(), even though
672 other assertions are supported. I have made pcre_dfa_exec() support
673 (*FAIL).
675 16. The implementation of 13c above involved the invention of a new opcode,
676 OP_ALLANY, which is like OP_ANY but doesn't check the /s flag. Since /s
677 cannot be changed at match time, I realized I could make a small
678 improvement to matching performance by compiling OP_ALLANY instead of
679 OP_ANY for "." when DOTALL was set, and then removing the runtime tests
680 on the OP_ANY path.
682 17. Compiling pcretest on Windows with readline support failed without the
683 following two fixes: (1) Make the unistd.h include conditional on
684 HAVE_UNISTD_H; (2) #define isatty and fileno as _isatty and _fileno.
686 18. Changed CMakeLists.txt and cmake/FindReadline.cmake to arrange for the
687 ncurses library to be included for pcretest when ReadLine support is
688 requested, but also to allow for it to be overridden. This patch came from
689 Daniel Bergström.
691 19. There was a typo in the file ucpinternal.h where f0_rangeflag was defined
692 as 0x00f00000 instead of 0x00800000. Luckily, this would not have caused
693 any errors with the current Unicode tables. Thanks to Peter Kankowski for
694 spotting this.
697 Version 7.6 28-Jan-08
698 ---------------------
700 1. A character class containing a very large number of characters with
701 codepoints greater than 255 (in UTF-8 mode, of course) caused a buffer
702 overflow.
704 2. Patch to cut out the "long long" test in pcrecpp_unittest when
705 HAVE_LONG_LONG is not defined.
707 3. Applied Christian Ehrlicher's patch to update the CMake build files to
708 bring them up to date and include new features. This patch includes:
710 - Fixed PH's badly added libz and libbz2 support.
711 - Fixed a problem with static linking.
712 - Added pcredemo. [But later removed - see 7 below.]
713 - Fixed dftables problem and added an option.
714 - Added a number of HAVE_XXX tests, including HAVE_WINDOWS_H and
716 - Added readline support for pcretest.
717 - Added an listing of the option settings after cmake has run.
719 4. A user submitted a patch to Makefile that makes it easy to create
720 "pcre.dll" under mingw when using Configure/Make. I added stuff to
721 Makefile.am that cause it to include this special target, without
722 affecting anything else. Note that the same mingw target plus all
723 the other distribution libraries and programs are now supported
724 when configuring with CMake (see 6 below) instead of with
725 Configure/Make.
727 5. Applied Craig's patch that moves no_arg into the RE class in the C++ code.
728 This is an attempt to solve the reported problem "pcrecpp::no_arg is not
729 exported in the Windows port". It has not yet been confirmed that the patch
730 solves the problem, but it does no harm.
732 6. Applied Sheri's patch to CMakeLists.txt to add NON_STANDARD_LIB_PREFIX and
733 NON_STANDARD_LIB_SUFFIX for dll names built with mingw when configured
734 with CMake, and also correct the comment about stack recursion.
736 7. Remove the automatic building of pcredemo from the ./configure system and
737 from CMakeLists.txt. The whole idea of pcredemo.c is that it is an example
738 of a program that users should build themselves after PCRE is installed, so
739 building it automatically is not really right. What is more, it gave
740 trouble in some build environments.
742 8. Further tidies to CMakeLists.txt from Sheri and Christian.
745 Version 7.5 10-Jan-08
746 ---------------------
748 1. Applied a patch from Craig: "This patch makes it possible to 'ignore'
749 values in parens when parsing an RE using the C++ wrapper."
751 2. Negative specials like \S did not work in character classes in UTF-8 mode.
752 Characters greater than 255 were excluded from the class instead of being
753 included.
755 3. The same bug as (2) above applied to negated POSIX classes such as
756 [:^space:].
758 4. PCRECPP_STATIC was referenced in pcrecpp_internal.h, but nowhere was it
759 defined or documented. It seems to have been a typo for PCRE_STATIC, so
760 I have changed it.
762 5. The construct (?&) was not diagnosed as a syntax error (it referenced the
763 first named subpattern) and a construct such as (?&a) would reference the
764 first named subpattern whose name started with "a" (in other words, the
765 length check was missing). Both these problems are fixed. "Subpattern name
766 expected" is now given for (?&) (a zero-length name), and this patch also
767 makes it give the same error for \k'' (previously it complained that that
768 was a reference to a non-existent subpattern).
770 6. The erroneous patterns (?+-a) and (?-+a) give different error messages;
771 this is right because (?- can be followed by option settings as well as by
772 digits. I have, however, made the messages clearer.
774 7. Patterns such as (?(1)a|b) (a pattern that contains fewer subpatterns
775 than the number used in the conditional) now cause a compile-time error.
776 This is actually not compatible with Perl, which accepts such patterns, but
777 treats the conditional as always being FALSE (as PCRE used to), but it
778 seems to me that giving a diagnostic is better.
780 8. Change "alphameric" to the more common word "alphanumeric" in comments
781 and messages.
783 9. Fix two occurrences of "backslash" in comments that should have been
784 "backspace".
786 10. Remove two redundant lines of code that can never be obeyed (their function
787 was moved elsewhere).
789 11. The program that makes PCRE's Unicode character property table had a bug
790 which caused it to generate incorrect table entries for sequences of
791 characters that have the same character type, but are in different scripts.
792 It amalgamated them into a single range, with the script of the first of
793 them. In other words, some characters were in the wrong script. There were
794 thirteen such cases, affecting characters in the following ranges:
796 U+002b0 - U+002c1
797 U+0060c - U+0060d
798 U+0061e - U+00612
799 U+0064b - U+0065e
800 U+0074d - U+0076d
801 U+01800 - U+01805
802 U+01d00 - U+01d77
803 U+01d9b - U+01dbf
804 U+0200b - U+0200f
805 U+030fc - U+030fe
806 U+03260 - U+0327f
807 U+0fb46 - U+0fbb1
808 U+10450 - U+1049d
810 12. The -o option (show only the matching part of a line) for pcregrep was not
811 compatible with GNU grep in that, if there was more than one match in a
812 line, it showed only the first of them. It now behaves in the same way as
813 GNU grep.
815 13. If the -o and -v options were combined for pcregrep, it printed a blank
816 line for every non-matching line. GNU grep prints nothing, and pcregrep now
817 does the same. The return code can be used to tell if there were any
818 non-matching lines.
820 14. Added --file-offsets and --line-offsets to pcregrep.
822 15. The pattern (?=something)(?R) was not being diagnosed as a potentially
823 infinitely looping recursion. The bug was that positive lookaheads were not
824 being skipped when checking for a possible empty match (negative lookaheads
825 and both kinds of lookbehind were skipped).
827 16. Fixed two typos in the Windows-only code in pcregrep.c, and moved the
828 inclusion of <windows.h> to before rather than after the definition of
829 INVALID_FILE_ATTRIBUTES (patch from David Byron).
831 17. Specifying a possessive quantifier with a specific limit for a Unicode
832 character property caused pcre_compile() to compile bad code, which led at
833 runtime to PCRE_ERROR_INTERNAL (-14). Examples of patterns that caused this
834 are: /\p{Zl}{2,3}+/8 and /\p{Cc}{2}+/8. It was the possessive "+" that
835 caused the error; without that there was no problem.
837 18. Added --enable-pcregrep-libz and --enable-pcregrep-libbz2.
839 19. Added --enable-pcretest-libreadline.
841 20. In pcrecpp.cc, the variable 'count' was incremented twice in
842 RE::GlobalReplace(). As a result, the number of replacements returned was
843 double what it should be. I removed one of the increments, but Craig sent a
844 later patch that removed the other one (the right fix) and added unit tests
845 that check the return values (which was not done before).
847 21. Several CMake things:
849 (1) Arranged that, when cmake is used on Unix, the libraries end up with
850 the names libpcre and libpcreposix, not just pcre and pcreposix.
852 (2) The above change means that pcretest and pcregrep are now correctly
853 linked with the newly-built libraries, not previously installed ones.
857 22. In UTF-8 mode, with newline set to "any", a pattern such as .*a.*=.b.*
858 crashed when matching a string such as a\x{2029}b (note that \x{2029} is a
859 UTF-8 newline character). The key issue is that the pattern starts .*;
860 this means that the match must be either at the beginning, or after a
861 newline. The bug was in the code for advancing after a failed match and
862 checking that the new position followed a newline. It was not taking
863 account of UTF-8 characters correctly.
865 23. PCRE was behaving differently from Perl in the way it recognized POSIX
866 character classes. PCRE was not treating the sequence [:...:] as a
867 character class unless the ... were all letters. Perl, however, seems to
868 allow any characters between [: and :], though of course it rejects as
869 unknown any "names" that contain non-letters, because all the known class
870 names consist only of letters. Thus, Perl gives an error for [[:1234:]],
871 for example, whereas PCRE did not - it did not recognize a POSIX character
872 class. This seemed a bit dangerous, so the code has been changed to be
873 closer to Perl. The behaviour is not identical to Perl, because PCRE will
874 diagnose an unknown class for, for example, [[:l\ower:]] where Perl will
875 treat it as [[:lower:]]. However, PCRE does now give "unknown" errors where
876 Perl does, and where it didn't before.
878 24. Rewrite so as to remove the single use of %n from pcregrep because in some
879 Windows environments %n is disabled by default.
882 Version 7.4 21-Sep-07
883 ---------------------
885 1. Change 7.3/28 was implemented for classes by looking at the bitmap. This
886 means that a class such as [\s] counted as "explicit reference to CR or
887 LF". That isn't really right - the whole point of the change was to try to
888 help when there was an actual mention of one of the two characters. So now
889 the change happens only if \r or \n (or a literal CR or LF) character is
890 encountered.
892 2. The 32-bit options word was also used for 6 internal flags, but the numbers
893 of both had grown to the point where there were only 3 bits left.
894 Fortunately, there was spare space in the data structure, and so I have
895 moved the internal flags into a new 16-bit field to free up more option
896 bits.
898 3. The appearance of (?J) at the start of a pattern set the DUPNAMES option,
899 but did not set the internal JCHANGED flag - either of these is enough to
900 control the way the "get" function works - but the PCRE_INFO_JCHANGED
901 facility is supposed to tell if (?J) was ever used, so now (?J) at the
902 start sets both bits.
904 4. Added options (at build time, compile time, exec time) to change \R from
905 matching any Unicode line ending sequence to just matching CR, LF, or CRLF.
907 5. doc/pcresyntax.html was missing from the distribution.
909 6. Put back the definition of PCRE_ERROR_NULLWSLIMIT, for backward
910 compatibility, even though it is no longer used.
912 7. Added macro for snprintf to pcrecpp_unittest.cc and also for strtoll and
913 strtoull to pcrecpp.cc to select the available functions in WIN32 when the
914 windows.h file is present (where different names are used). [This was
915 reversed later after testing - see 16 below.]
917 8. Changed all #include <config.h> to #include "config.h". There were also
918 some further <pcre.h> cases that I changed to "pcre.h".
920 9. When pcregrep was used with the --colour option, it missed the line ending
921 sequence off the lines that it output.
923 10. It was pointed out to me that arrays of string pointers cause lots of
924 relocations when a shared library is dynamically loaded. A technique of
925 using a single long string with a table of offsets can drastically reduce
926 these. I have refactored PCRE in four places to do this. The result is
927 dramatic:
929 Originally: 290
930 After changing UCP table: 187
931 After changing error message table: 43
932 After changing table of "verbs" 36
933 After changing table of Posix names 22
935 Thanks to the folks working on Gregex for glib for this insight.
937 11. --disable-stack-for-recursion caused compiling to fail unless -enable-
938 unicode-properties was also set.
940 12. Updated the tests so that they work when \R is defaulted to ANYCRLF.
942 13. Added checks for ANY and ANYCRLF to pcrecpp.cc where it previously
943 checked only for CRLF.
945 14. Added casts to pcretest.c to avoid compiler warnings.
947 15. Added Craig's patch to various pcrecpp modules to avoid compiler warnings.
949 16. Added Craig's patch to remove the WINDOWS_H tests, that were not working,
950 and instead check for _strtoi64 explicitly, and avoid the use of snprintf()
951 entirely. This removes changes made in 7 above.
953 17. The CMake files have been updated, and there is now more information about
954 building with CMake in the NON-UNIX-USE document.
957 Version 7.3 28-Aug-07
958 ---------------------
960 1. In the rejigging of the build system that eventually resulted in 7.1, the
961 line "#include <pcre.h>" was included in pcre_internal.h. The use of angle
962 brackets there is not right, since it causes compilers to look for an
963 installed pcre.h, not the version that is in the source that is being
964 compiled (which of course may be different). I have changed it back to:
966 #include "pcre.h"
968 I have a vague recollection that the change was concerned with compiling in
969 different directories, but in the new build system, that is taken care of
970 by the VPATH setting the Makefile.
972 2. The pattern .*$ when run in not-DOTALL UTF-8 mode with newline=any failed
973 when the subject happened to end in the byte 0x85 (e.g. if the last
974 character was \x{1ec5}). *Character* 0x85 is one of the "any" newline
975 characters but of course it shouldn't be taken as a newline when it is part
976 of another character. The bug was that, for an unlimited repeat of . in
977 not-DOTALL UTF-8 mode, PCRE was advancing by bytes rather than by
978 characters when looking for a newline.
980 3. A small performance improvement in the DOTALL UTF-8 mode .* case.
982 4. Debugging: adjusted the names of opcodes for different kinds of parentheses
983 in debug output.
985 5. Arrange to use "%I64d" instead of "%lld" and "%I64u" instead of "%llu" for
986 long printing in the pcrecpp unittest when running under MinGW.
988 6. ESC_K was left out of the EBCDIC table.
990 7. Change 7.0/38 introduced a new limit on the number of nested non-capturing
991 parentheses; I made it 1000, which seemed large enough. Unfortunately, the
992 limit also applies to "virtual nesting" when a pattern is recursive, and in
993 this case 1000 isn't so big. I have been able to remove this limit at the
994 expense of backing off one optimization in certain circumstances. Normally,
995 when pcre_exec() would call its internal match() function recursively and
996 immediately return the result unconditionally, it uses a "tail recursion"
997 feature to save stack. However, when a subpattern that can match an empty
998 string has an unlimited repetition quantifier, it no longer makes this
999 optimization. That gives it a stack frame in which to save the data for
1000 checking that an empty string has been matched. Previously this was taken
1001 from the 1000-entry workspace that had been reserved. So now there is no
1002 explicit limit, but more stack is used.
1004 8. Applied Daniel's patches to solve problems with the import/export magic
1005 syntax that is required for Windows, and which was going wrong for the
1006 pcreposix and pcrecpp parts of the library. These were overlooked when this
1007 problem was solved for the main library.
1009 9. There were some crude static tests to avoid integer overflow when computing
1010 the size of patterns that contain repeated groups with explicit upper
1011 limits. As the maximum quantifier is 65535, the maximum group length was
1012 set at 30,000 so that the product of these two numbers did not overflow a
1013 32-bit integer. However, it turns out that people want to use groups that
1014 are longer than 30,000 bytes (though not repeat them that many times).
1015 Change 7.0/17 (the refactoring of the way the pattern size is computed) has
1016 made it possible to implement the integer overflow checks in a much more
1017 dynamic way, which I have now done. The artificial limitation on group
1018 length has been removed - we now have only the limit on the total length of
1019 the compiled pattern, which depends on the LINK_SIZE setting.
1021 10. Fixed a bug in the documentation for get/copy named substring when
1022 duplicate names are permitted. If none of the named substrings are set, the
1023 functions return PCRE_ERROR_NOSUBSTRING (7); the doc said they returned an
1024 empty string.
1026 11. Because Perl interprets \Q...\E at a high level, and ignores orphan \E
1027 instances, patterns such as [\Q\E] or [\E] or even [^\E] cause an error,
1028 because the ] is interpreted as the first data character and the
1029 terminating ] is not found. PCRE has been made compatible with Perl in this
1030 regard. Previously, it interpreted [\Q\E] as an empty class, and [\E] could
1031 cause memory overwriting.
1033 10. Like Perl, PCRE automatically breaks an unlimited repeat after an empty
1034 string has been matched (to stop an infinite loop). It was not recognizing
1035 a conditional subpattern that could match an empty string if that
1036 subpattern was within another subpattern. For example, it looped when
1037 trying to match (((?(1)X|))*) but it was OK with ((?(1)X|)*) where the
1038 condition was not nested. This bug has been fixed.
1040 12. A pattern like \X?\d or \P{L}?\d in non-UTF-8 mode could cause a backtrack
1041 past the start of the subject in the presence of bytes with the top bit
1042 set, for example "\x8aBCD".
1044 13. Added Perl 5.10 experimental backtracking controls (*FAIL), (*F), (*PRUNE),
1045 (*SKIP), (*THEN), (*COMMIT), and (*ACCEPT).
1047 14. Optimized (?!) to (*FAIL).
1049 15. Updated the test for a valid UTF-8 string to conform to the later RFC 3629.
1050 This restricts code points to be within the range 0 to 0x10FFFF, excluding
1051 the "low surrogate" sequence 0xD800 to 0xDFFF. Previously, PCRE allowed the
1052 full range 0 to 0x7FFFFFFF, as defined by RFC 2279. Internally, it still
1053 does: it's just the validity check that is more restrictive.
1055 16. Inserted checks for integer overflows during escape sequence (backslash)
1056 processing, and also fixed erroneous offset values for syntax errors during
1057 backslash processing.
1059 17. Fixed another case of looking too far back in non-UTF-8 mode (cf 12 above)
1060 for patterns like [\PPP\x8a]{1,}\x80 with the subject "A\x80".
1062 18. An unterminated class in a pattern like (?1)\c[ with a "forward reference"
1063 caused an overrun.
1065 19. A pattern like (?:[\PPa*]*){8,} which had an "extended class" (one with
1066 something other than just ASCII characters) inside a group that had an
1067 unlimited repeat caused a loop at compile time (while checking to see
1068 whether the group could match an empty string).
1070 20. Debugging a pattern containing \p or \P could cause a crash. For example,
1071 [\P{Any}] did so. (Error in the code for printing property names.)
1073 21. An orphan \E inside a character class could cause a crash.
1075 22. A repeated capturing bracket such as (A)? could cause a wild memory
1076 reference during compilation.
1078 23. There are several functions in pcre_compile() that scan along a compiled
1079 expression for various reasons (e.g. to see if it's fixed length for look
1080 behind). There were bugs in these functions when a repeated \p or \P was
1081 present in the pattern. These operators have additional parameters compared
1082 with \d, etc, and these were not being taken into account when moving along
1083 the compiled data. Specifically:
1085 (a) A item such as \p{Yi}{3} in a lookbehind was not treated as fixed
1086 length.
1088 (b) An item such as \pL+ within a repeated group could cause crashes or
1089 loops.
1091 (c) A pattern such as \p{Yi}+(\P{Yi}+)(?1) could give an incorrect
1092 "reference to non-existent subpattern" error.
1094 (d) A pattern like (\P{Yi}{2}\277)? could loop at compile time.
1096 24. A repeated \S or \W in UTF-8 mode could give wrong answers when multibyte
1097 characters were involved (for example /\S{2}/8g with "A\x{a3}BC").
1099 25. Using pcregrep in multiline, inverted mode (-Mv) caused it to loop.
1101 26. Patterns such as [\P{Yi}A] which include \p or \P and just one other
1102 character were causing crashes (broken optimization).
1104 27. Patterns such as (\P{Yi}*\277)* (group with possible zero repeat containing
1105 \p or \P) caused a compile-time loop.
1107 28. More problems have arisen in unanchored patterns when CRLF is a valid line
1108 break. For example, the unstudied pattern [\r\n]A does not match the string
1109 "\r\nA" because change 7.0/46 below moves the current point on by two
1110 characters after failing to match at the start. However, the pattern \nA
1111 *does* match, because it doesn't start till \n, and if [\r\n]A is studied,
1112 the same is true. There doesn't seem any very clean way out of this, but
1113 what I have chosen to do makes the common cases work: PCRE now takes note
1114 of whether there can be an explicit match for \r or \n anywhere in the
1115 pattern, and if so, 7.0/46 no longer applies. As part of this change,
1116 there's a new PCRE_INFO_HASCRORLF option for finding out whether a compiled
1117 pattern has explicit CR or LF references.
1119 29. Added (*CR) etc for changing newline setting at start of pattern.
1122 Version 7.2 19-Jun-07
1123 ---------------------
1125 1. If the fr_FR locale cannot be found for test 3, try the "french" locale,
1126 which is apparently normally available under Windows.
1128 2. Re-jig the pcregrep tests with different newline settings in an attempt
1129 to make them independent of the local environment's newline setting.
1131 3. Add code to configure.ac to remove -g from the CFLAGS default settings.
1133 4. Some of the "internals" tests were previously cut out when the link size
1134 was not 2, because the output contained actual offsets. The recent new
1135 "Z" feature of pcretest means that these can be cut out, making the tests
1136 usable with all link sizes.
1138 5. Implemented Stan Switzer's goto replacement for longjmp() when not using
1139 stack recursion. This gives a massive performance boost under BSD, but just
1140 a small improvement under Linux. However, it saves one field in the frame
1141 in all cases.
1143 6. Added more features from the forthcoming Perl 5.10:
1145 (a) (?-n) (where n is a string of digits) is a relative subroutine or
1146 recursion call. It refers to the nth most recently opened parentheses.
1148 (b) (?+n) is also a relative subroutine call; it refers to the nth next
1149 to be opened parentheses.
1151 (c) Conditions that refer to capturing parentheses can be specified
1152 relatively, for example, (?(-2)... or (?(+3)...
1154 (d) \K resets the start of the current match so that everything before
1155 is not part of it.
1157 (e) \k{name} is synonymous with \k<name> and \k'name' (.NET compatible).
1159 (f) \g{name} is another synonym - part of Perl 5.10's unification of
1160 reference syntax.
1162 (g) (?| introduces a group in which the numbering of parentheses in each
1163 alternative starts with the same number.
1165 (h) \h, \H, \v, and \V match horizontal and vertical whitespace.
1167 7. Added two new calls to pcre_fullinfo(): PCRE_INFO_OKPARTIAL and
1170 8. A pattern such as (.*(.)?)* caused pcre_exec() to fail by either not
1171 terminating or by crashing. Diagnosed by Viktor Griph; it was in the code
1172 for detecting groups that can match an empty string.
1174 9. A pattern with a very large number of alternatives (more than several
1175 hundred) was running out of internal workspace during the pre-compile
1176 phase, where pcre_compile() figures out how much memory will be needed. A
1177 bit of new cunning has reduced the workspace needed for groups with
1178 alternatives. The 1000-alternative test pattern now uses 12 bytes of
1179 workspace instead of running out of the 4096 that are available.
1181 10. Inserted some missing (unsigned int) casts to get rid of compiler warnings.
1183 11. Applied patch from Google to remove an optimization that didn't quite work.
1184 The report of the bug said:
1186 pcrecpp::RE("a*").FullMatch("aaa") matches, while
1187 pcrecpp::RE("a*?").FullMatch("aaa") does not, and
1188 pcrecpp::RE("a*?\\z").FullMatch("aaa") does again.
1190 12. If \p or \P was used in non-UTF-8 mode on a character greater than 127
1191 it matched the wrong number of bytes.
1194 Version 7.1 24-Apr-07
1195 ---------------------
1197 1. Applied Bob Rossi and Daniel G's patches to convert the build system to one
1198 that is more "standard", making use of automake and other Autotools. There
1199 is some re-arrangement of the files and adjustment of comments consequent
1200 on this.
1202 2. Part of the patch fixed a problem with the pcregrep tests. The test of -r
1203 for recursive directory scanning broke on some systems because the files
1204 are not scanned in any specific order and on different systems the order
1205 was different. A call to "sort" has been inserted into RunGrepTest for the
1206 approprate test as a short-term fix. In the longer term there may be an
1207 alternative.
1209 3. I had an email from Eric Raymond about problems translating some of PCRE's
1210 man pages to HTML (despite the fact that I distribute HTML pages, some
1211 people do their own conversions for various reasons). The problems
1212 concerned the use of low-level troff macros .br and .in. I have therefore
1213 removed all such uses from the man pages (some were redundant, some could
1214 be replaced by .nf/.fi pairs). The 132html script that I use to generate
1215 HTML has been updated to handle .nf/.fi and to complain if it encounters
1216 .br or .in.
1218 4. Updated comments in configure.ac that get placed in config.h.in and also
1219 arranged for config.h to be included in the distribution, with the name
1220 config.h.generic, for the benefit of those who have to compile without
1221 Autotools (compare pcre.h, which is now distributed as pcre.h.generic).
1223 5. Updated the support (such as it is) for Virtual Pascal, thanks to Stefan
1224 Weber: (1) pcre_internal.h was missing some function renames; (2) updated
1225 makevp.bat for the current PCRE, using the additional files
1226 makevp_c.txt, makevp_l.txt, and pcregexp.pas.
1228 6. A Windows user reported a minor discrepancy with test 2, which turned out
1229 to be caused by a trailing space on an input line that had got lost in his
1230 copy. The trailing space was an accident, so I've just removed it.
1232 7. Add -Wl,-R... flags in pcre-config.in for *BSD* systems, as I'm told
1233 that is needed.
1235 8. Mark ucp_table (in ucptable.h) and ucp_gentype (in pcre_ucp_searchfuncs.c)
1236 as "const" (a) because they are and (b) because it helps the PHP
1237 maintainers who have recently made a script to detect big data structures
1238 in the php code that should be moved to the .rodata section. I remembered
1239 to update Builducptable as well, so it won't revert if ucptable.h is ever
1240 re-created.
1242 9. Added some extra #ifdef SUPPORT_UTF8 conditionals into pcretest.c,
1243 pcre_printint.src, pcre_compile.c, pcre_study.c, and pcre_tables.c, in
1244 order to be able to cut out the UTF-8 tables in the latter when UTF-8
1245 support is not required. This saves 1.5-2K of code, which is important in
1246 some applications.
1248 Later: more #ifdefs are needed in pcre_ord2utf8.c and pcre_valid_utf8.c
1249 so as not to refer to the tables, even though these functions will never be
1250 called when UTF-8 support is disabled. Otherwise there are problems with a
1251 shared library.
1253 10. Fixed two bugs in the emulated memmove() function in pcre_internal.h:
1255 (a) It was defining its arguments as char * instead of void *.
1257 (b) It was assuming that all moves were upwards in memory; this was true
1258 a long time ago when I wrote it, but is no longer the case.
1260 The emulated memove() is provided for those environments that have neither
1261 memmove() nor bcopy(). I didn't think anyone used it these days, but that
1262 is clearly not the case, as these two bugs were recently reported.
1264 11. The script PrepareRelease is now distributed: it calls 132html, CleanTxt,
1265 and Detrail to create the HTML documentation, the .txt form of the man
1266 pages, and it removes trailing spaces from listed files. It also creates
1267 pcre.h.generic and config.h.generic from pcre.h and config.h. In the latter
1268 case, it wraps all the #defines with #ifndefs. This script should be run
1269 before "make dist".
1271 12. Fixed two fairly obscure bugs concerned with quantified caseless matching
1272 with Unicode property support.
1274 (a) For a maximizing quantifier, if the two different cases of the
1275 character were of different lengths in their UTF-8 codings (there are
1276 some cases like this - I found 11), and the matching function had to
1277 back up over a mixture of the two cases, it incorrectly assumed they
1278 were both the same length.
1280 (b) When PCRE was configured to use the heap rather than the stack for
1281 recursion during matching, it was not correctly preserving the data for
1282 the other case of a UTF-8 character when checking ahead for a match
1283 while processing a minimizing repeat. If the check also involved
1284 matching a wide character, but failed, corruption could cause an
1285 erroneous result when trying to check for a repeat of the original
1286 character.
1288 13. Some tidying changes to the testing mechanism:
1290 (a) The RunTest script now detects the internal link size and whether there
1291 is UTF-8 and UCP support by running ./pcretest -C instead of relying on
1292 values substituted by "configure". (The RunGrepTest script already did
1293 this for UTF-8.) The configure.ac script no longer substitutes the
1294 relevant variables.
1296 (b) The debugging options /B and /D in pcretest show the compiled bytecode
1297 with length and offset values. This means that the output is different
1298 for different internal link sizes. Test 2 is skipped for link sizes
1299 other than 2 because of this, bypassing the problem. Unfortunately,
1300 there was also a test in test 3 (the locale tests) that used /B and
1301 failed for link sizes other than 2. Rather than cut the whole test out,
1302 I have added a new /Z option to pcretest that replaces the length and
1303 offset values with spaces. This is now used to make test 3 independent
1304 of link size. (Test 2 will be tidied up later.)
1306 14. If erroroffset was passed as NULL to pcre_compile, it provoked a
1307 segmentation fault instead of returning the appropriate error message.
1309 15. In multiline mode when the newline sequence was set to "any", the pattern
1310 ^$ would give a match between the \r and \n of a subject such as "A\r\nB".
1311 This doesn't seem right; it now treats the CRLF combination as the line
1312 ending, and so does not match in that case. It's only a pattern such as ^$
1313 that would hit this one: something like ^ABC$ would have failed after \r
1314 and then tried again after \r\n.
1316 16. Changed the comparison command for RunGrepTest from "diff -u" to "diff -ub"
1317 in an attempt to make files that differ only in their line terminators
1318 compare equal. This works on Linux.
1320 17. Under certain error circumstances pcregrep might try to free random memory
1321 as it exited. This is now fixed, thanks to valgrind.
1323 19. In pcretest, if the pattern /(?m)^$/g<any> was matched against the string
1324 "abc\r\n\r\n", it found an unwanted second match after the second \r. This
1325 was because its rules for how to advance for /g after matching an empty
1326 string at the end of a line did not allow for this case. They now check for
1327 it specially.
1329 20. pcretest is supposed to handle patterns and data of any length, by
1330 extending its buffers when necessary. It was getting this wrong when the
1331 buffer for a data line had to be extended.
1333 21. Added PCRE_NEWLINE_ANYCRLF which is like ANY, but matches only CR, LF, or
1334 CRLF as a newline sequence.
1336 22. Code for handling Unicode properties in pcre_dfa_exec() wasn't being cut
1337 out by #ifdef SUPPORT_UCP. This did no harm, as it could never be used, but
1338 I have nevertheless tidied it up.
1340 23. Added some casts to kill warnings from HP-UX ia64 compiler.
1342 24. Added a man page for pcre-config.
1345 Version 7.0 19-Dec-06
1346 ---------------------
1348 1. Fixed a signed/unsigned compiler warning in pcre_compile.c, shown up by
1349 moving to gcc 4.1.1.
1351 2. The -S option for pcretest uses setrlimit(); I had omitted to #include
1352 sys/time.h, which is documented as needed for this function. It doesn't
1353 seem to matter on Linux, but it showed up on some releases of OS X.
1355 3. It seems that there are systems where bytes whose values are greater than
1356 127 match isprint() in the "C" locale. The "C" locale should be the
1357 default when a C program starts up. In most systems, only ASCII printing
1358 characters match isprint(). This difference caused the output from pcretest
1359 to vary, making some of the tests fail. I have changed pcretest so that:
1361 (a) When it is outputting text in the compiled version of a pattern, bytes
1362 other than 32-126 are always shown as hex escapes.
1364 (b) When it is outputting text that is a matched part of a subject string,
1365 it does the same, unless a different locale has been set for the match
1366 (using the /L modifier). In this case, it uses isprint() to decide.
1368 4. Fixed a major bug that caused incorrect computation of the amount of memory
1369 required for a compiled pattern when options that changed within the
1370 pattern affected the logic of the preliminary scan that determines the
1371 length. The relevant options are -x, and -i in UTF-8 mode. The result was
1372 that the computed length was too small. The symptoms of this bug were
1373 either the PCRE error "internal error: code overflow" from pcre_compile(),
1374 or a glibc crash with a message such as "pcretest: free(): invalid next
1375 size (fast)". Examples of patterns that provoked this bug (shown in
1376 pcretest format) are:
1378 /(?-x: )/x
1379 /(?x)(?-x: \s*#\s*)/
1380 /((?i)[\x{c0}])/8
1381 /(?i:[\x{c0}])/8
1383 HOWEVER: Change 17 below makes this fix obsolete as the memory computation
1384 is now done differently.
1386 5. Applied patches from Google to: (a) add a QuoteMeta function to the C++
1387 wrapper classes; (b) implement a new function in the C++ scanner that is
1388 more efficient than the old way of doing things because it avoids levels of
1389 recursion in the regex matching; (c) add a paragraph to the documentation
1390 for the FullMatch() function.
1392 6. The escape sequence \n was being treated as whatever was defined as
1393 "newline". Not only was this contrary to the documentation, which states
1394 that \n is character 10 (hex 0A), but it also went horribly wrong when
1395 "newline" was defined as CRLF. This has been fixed.
1397 7. In pcre_dfa_exec.c the value of an unsigned integer (the variable called c)
1398 was being set to -1 for the "end of line" case (supposedly a value that no
1399 character can have). Though this value is never used (the check for end of
1400 line is "zero bytes in current character"), it caused compiler complaints.
1401 I've changed it to 0xffffffff.
1403 8. In pcre_version.c, the version string was being built by a sequence of
1404 C macros that, in the event of PCRE_PRERELEASE being defined as an empty
1405 string (as it is for production releases) called a macro with an empty
1406 argument. The C standard says the result of this is undefined. The gcc
1407 compiler treats it as an empty string (which was what was wanted) but it is
1408 reported that Visual C gives an error. The source has been hacked around to
1409 avoid this problem.
1411 9. On the advice of a Windows user, included <io.h> and <fcntl.h> in Windows
1412 builds of pcretest, and changed the call to _setmode() to use _O_BINARY
1413 instead of 0x8000. Made all the #ifdefs test both _WIN32 and WIN32 (not all
1414 of them did).
1416 10. Originally, pcretest opened its input and output without "b"; then I was
1417 told that "b" was needed in some environments, so it was added for release
1418 5.0 to both the input and output. (It makes no difference on Unix-like
1419 systems.) Later I was told that it is wrong for the input on Windows. I've
1420 now abstracted the modes into two macros, to make it easier to fiddle with
1421 them, and removed "b" from the input mode under Windows.
1423 11. Added pkgconfig support for the C++ wrapper library, libpcrecpp.
1425 12. Added -help and --help to pcretest as an official way of being reminded
1426 of the options.
1428 13. Removed some redundant semicolons after macro calls in pcrecpparg.h.in
1429 and pcrecpp.cc because they annoy compilers at high warning levels.
1431 14. A bit of tidying/refactoring in pcre_exec.c in the main bumpalong loop.
1433 15. Fixed an occurrence of == in configure.ac that should have been = (shell
1434 scripts are not C programs :-) and which was not noticed because it works
1435 on Linux.
1437 16. pcretest is supposed to handle any length of pattern and data line (as one
1438 line or as a continued sequence of lines) by extending its input buffer if
1439 necessary. This feature was broken for very long pattern lines, leading to
1440 a string of junk being passed to pcre_compile() if the pattern was longer
1441 than about 50K.
1443 17. I have done a major re-factoring of the way pcre_compile() computes the
1444 amount of memory needed for a compiled pattern. Previously, there was code
1445 that made a preliminary scan of the pattern in order to do this. That was
1446 OK when PCRE was new, but as the facilities have expanded, it has become
1447 harder and harder to keep it in step with the real compile phase, and there
1448 have been a number of bugs (see for example, 4 above). I have now found a
1449 cunning way of running the real compile function in a "fake" mode that
1450 enables it to compute how much memory it would need, while actually only
1451 ever using a few hundred bytes of working memory and without too many
1452 tests of the mode. This should make future maintenance and development
1453 easier. A side effect of this work is that the limit of 200 on the nesting
1454 depth of parentheses has been removed (though this was never a serious
1455 limitation, I suspect). However, there is a downside: pcre_compile() now
1456 runs more slowly than before (30% or more, depending on the pattern). I
1457 hope this isn't a big issue. There is no effect on runtime performance.
1459 18. Fixed a minor bug in pcretest: if a pattern line was not terminated by a
1460 newline (only possible for the last line of a file) and it was a
1461 pattern that set a locale (followed by /Lsomething), pcretest crashed.
1463 19. Added additional timing features to pcretest. (1) The -tm option now times
1464 matching only, not compiling. (2) Both -t and -tm can be followed, as a
1465 separate command line item, by a number that specifies the number of
1466 repeats to use when timing. The default is 50000; this gives better
1467 precision, but takes uncomfortably long for very large patterns.
1469 20. Extended pcre_study() to be more clever in cases where a branch of a
1470 subpattern has no definite first character. For example, (a*|b*)[cd] would
1471 previously give no result from pcre_study(). Now it recognizes that the
1472 first character must be a, b, c, or d.
1474 21. There was an incorrect error "recursive call could loop indefinitely" if
1475 a subpattern (or the entire pattern) that was being tested for matching an
1476 empty string contained only one non-empty item after a nested subpattern.
1477 For example, the pattern (?>\x{100}*)\d(?R) provoked this error
1478 incorrectly, because the \d was being skipped in the check.
1480 22. The pcretest program now has a new pattern option /B and a command line
1481 option -b, which is equivalent to adding /B to every pattern. This causes
1482 it to show the compiled bytecode, without the additional information that
1483 -d shows. The effect of -d is now the same as -b with -i (and similarly, /D
1484 is the same as /B/I).
1486 23. A new optimization is now able automatically to treat some sequences such
1487 as a*b as a*+b. More specifically, if something simple (such as a character
1488 or a simple class like \d) has an unlimited quantifier, and is followed by
1489 something that cannot possibly match the quantified thing, the quantifier
1490 is automatically "possessified".
1492 24. A recursive reference to a subpattern whose number was greater than 39
1493 went wrong under certain circumstances in UTF-8 mode. This bug could also
1494 have affected the operation of pcre_study().
1496 25. Realized that a little bit of performance could be had by replacing
1497 (c & 0xc0) == 0xc0 with c >= 0xc0 when processing UTF-8 characters.
1499 26. Timing data from pcretest is now shown to 4 decimal places instead of 3.
1501 27. Possessive quantifiers such as a++ were previously implemented by turning
1502 them into atomic groups such as ($>a+). Now they have their own opcodes,
1503 which improves performance. This includes the automatically created ones
1504 from 23 above.
1506 28. A pattern such as (?=(\w+))\1: which simulates an atomic group using a
1507 lookahead was broken if it was not anchored. PCRE was mistakenly expecting
1508 the first matched character to be a colon. This applied both to named and
1509 numbered groups.
1511 29. The ucpinternal.h header file was missing its idempotency #ifdef.
1513 30. I was sent a "project" file called libpcre.a.dev which I understand makes
1514 building PCRE on Windows easier, so I have included it in the distribution.
1516 31. There is now a check in pcretest against a ridiculously large number being
1517 returned by pcre_exec() or pcre_dfa_exec(). If this happens in a /g or /G
1518 loop, the loop is abandoned.
1520 32. Forward references to subpatterns in conditions such as (?(2)...) where
1521 subpattern 2 is defined later cause pcre_compile() to search forwards in
1522 the pattern for the relevant set of parentheses. This search went wrong
1523 when there were unescaped parentheses in a character class, parentheses
1524 escaped with \Q...\E, or parentheses in a #-comment in /x mode.
1526 33. "Subroutine" calls and backreferences were previously restricted to
1527 referencing subpatterns earlier in the regex. This restriction has now
1528 been removed.
1530 34. Added a number of extra features that are going to be in Perl 5.10. On the
1531 whole, these are just syntactic alternatives for features that PCRE had
1532 previously implemented using the Python syntax or my own invention. The
1533 other formats are all retained for compatibility.
1535 (a) Named groups can now be defined as (?<name>...) or (?'name'...) as well
1536 as (?P<name>...). The new forms, as well as being in Perl 5.10, are
1537 also .NET compatible.
1539 (b) A recursion or subroutine call to a named group can now be defined as
1540 (?&name) as well as (?P>name).
1542 (c) A backreference to a named group can now be defined as \k<name> or
1543 \k'name' as well as (?P=name). The new forms, as well as being in Perl
1544 5.10, are also .NET compatible.
1546 (d) A conditional reference to a named group can now use the syntax
1547 (?(<name>) or (?('name') as well as (?(name).
1549 (e) A "conditional group" of the form (?(DEFINE)...) can be used to define
1550 groups (named and numbered) that are never evaluated inline, but can be
1551 called as "subroutines" from elsewhere. In effect, the DEFINE condition
1552 is always false. There may be only one alternative in such a group.
1554 (f) A test for recursion can be given as (?(R1).. or (?(R&name)... as well
1555 as the simple (?(R). The condition is true only if the most recent
1556 recursion is that of the given number or name. It does not search out
1557 through the entire recursion stack.
1559 (g) The escape \gN or \g{N} has been added, where N is a positive or
1560 negative number, specifying an absolute or relative reference.
1562 35. Tidied to get rid of some further signed/unsigned compiler warnings and
1563 some "unreachable code" warnings.
1565 36. Updated the Unicode property tables to Unicode version 5.0.0. Amongst other
1566 things, this adds five new scripts.
1568 37. Perl ignores orphaned \E escapes completely. PCRE now does the same.
1569 There were also incompatibilities regarding the handling of \Q..\E inside
1570 character classes, for example with patterns like [\Qa\E-\Qz\E] where the
1571 hyphen was adjacent to \Q or \E. I hope I've cleared all this up now.
1573 38. Like Perl, PCRE detects when an indefinitely repeated parenthesized group
1574 matches an empty string, and forcibly breaks the loop. There were bugs in
1575 this code in non-simple cases. For a pattern such as ^(a()*)* matched
1576 against aaaa the result was just "a" rather than "aaaa", for example. Two
1577 separate and independent bugs (that affected different cases) have been
1578 fixed.
1580 39. Refactored the code to abolish the use of different opcodes for small
1581 capturing bracket numbers. This is a tidy that I avoided doing when I
1582 removed the limit on the number of capturing brackets for 3.5 back in 2001.
1583 The new approach is not only tidier, it makes it possible to reduce the
1584 memory needed to fix the previous bug (38).
1586 40. Implemented PCRE_NEWLINE_ANY to recognize any of the Unicode newline
1587 sequences (http://unicode.org/unicode/reports/tr18/) as "newline" when
1588 processing dot, circumflex, or dollar metacharacters, or #-comments in /x
1589 mode.
1591 41. Add \R to match any Unicode newline sequence, as suggested in the Unicode
1592 report.
1594 42. Applied patch, originally from Ari Pollak, modified by Google, to allow
1595 copy construction and assignment in the C++ wrapper.
1597 43. Updated pcregrep to support "--newline=any". In the process, I fixed a
1598 couple of bugs that could have given wrong results in the "--newline=crlf"
1599 case.
1601 44. Added a number of casts and did some reorganization of signed/unsigned int
1602 variables following suggestions from Dair Grant. Also renamed the variable
1603 "this" as "item" because it is a C++ keyword.
1605 45. Arranged for dftables to add
1607 #include "pcre_internal.h"
1609 to pcre_chartables.c because without it, gcc 4.x may remove the array
1610 definition from the final binary if PCRE is built into a static library and
1611 dead code stripping is activated.
1613 46. For an unanchored pattern, if a match attempt fails at the start of a
1614 newline sequence, and the newline setting is CRLF or ANY, and the next two
1615 characters are CRLF, advance by two characters instead of one.
1618 Version 6.7 04-Jul-06
1619 ---------------------
1621 1. In order to handle tests when input lines are enormously long, pcretest has
1622 been re-factored so that it automatically extends its buffers when
1623 necessary. The code is crude, but this _is_ just a test program. The
1624 default size has been increased from 32K to 50K.
1626 2. The code in pcre_study() was using the value of the re argument before
1627 testing it for NULL. (Of course, in any sensible call of the function, it
1628 won't be NULL.)
1630 3. The memmove() emulation function in pcre_internal.h, which is used on
1631 systems that lack both memmove() and bcopy() - that is, hardly ever -
1632 was missing a "static" storage class specifier.
1634 4. When UTF-8 mode was not set, PCRE looped when compiling certain patterns
1635 containing an extended class (one that cannot be represented by a bitmap
1636 because it contains high-valued characters or Unicode property items, e.g.
1637 [\pZ]). Almost always one would set UTF-8 mode when processing such a
1638 pattern, but PCRE should not loop if you do not (it no longer does).
1639 [Detail: two cases were found: (a) a repeated subpattern containing an
1640 extended class; (b) a recursive reference to a subpattern that followed a
1641 previous extended class. It wasn't skipping over the extended class
1642 correctly when UTF-8 mode was not set.]
1644 5. A negated single-character class was not being recognized as fixed-length
1645 in lookbehind assertions such as (?<=[^f]), leading to an incorrect
1646 compile error "lookbehind assertion is not fixed length".
1648 6. The RunPerlTest auxiliary script was showing an unexpected difference
1649 between PCRE and Perl for UTF-8 tests. It turns out that it is hard to
1650 write a Perl script that can interpret lines of an input file either as
1651 byte characters or as UTF-8, which is what "perltest" was being required to
1652 do for the non-UTF-8 and UTF-8 tests, respectively. Essentially what you
1653 can't do is switch easily at run time between having the "use utf8;" pragma
1654 or not. In the end, I fudged it by using the RunPerlTest script to insert
1655 "use utf8;" explicitly for the UTF-8 tests.
1657 7. In multiline (/m) mode, PCRE was matching ^ after a terminating newline at
1658 the end of the subject string, contrary to the documentation and to what
1659 Perl does. This was true of both matching functions. Now it matches only at
1660 the start of the subject and immediately after *internal* newlines.
1662 8. A call of pcre_fullinfo() from pcretest to get the option bits was passing
1663 a pointer to an int instead of a pointer to an unsigned long int. This
1664 caused problems on 64-bit systems.
1666 9. Applied a patch from the folks at Google to pcrecpp.cc, to fix "another
1667 instance of the 'standard' template library not being so standard".
1669 10. There was no check on the number of named subpatterns nor the maximum
1670 length of a subpattern name. The product of these values is used to compute
1671 the size of the memory block for a compiled pattern. By supplying a very
1672 long subpattern name and a large number of named subpatterns, the size
1673 computation could be caused to overflow. This is now prevented by limiting
1674 the length of names to 32 characters, and the number of named subpatterns
1675 to 10,000.
1677 11. Subpatterns that are repeated with specific counts have to be replicated in
1678 the compiled pattern. The size of memory for this was computed from the
1679 length of the subpattern and the repeat count. The latter is limited to
1680 65535, but there was no limit on the former, meaning that integer overflow
1681 could in principle occur. The compiled length of a repeated subpattern is
1682 now limited to 30,000 bytes in order to prevent this.
1684 12. Added the optional facility to have named substrings with the same name.
1686 13. Added the ability to use a named substring as a condition, using the
1687 Python syntax: (?(name)yes|no). This overloads (?(R)... and names that
1688 are numbers (not recommended). Forward references are permitted.
1690 14. Added forward references in named backreferences (if you see what I mean).
1692 15. In UTF-8 mode, with the PCRE_DOTALL option set, a quantified dot in the
1693 pattern could run off the end of the subject. For example, the pattern
1694 "(?s)(.{1,5})"8 did this with the subject "ab".
1696 16. If PCRE_DOTALL or PCRE_MULTILINE were set, pcre_dfa_exec() behaved as if
1697 PCRE_CASELESS was set when matching characters that were quantified with ?
1698 or *.
1700 17. A character class other than a single negated character that had a minimum
1701 but no maximum quantifier - for example [ab]{6,} - was not handled
1702 correctly by pce_dfa_exec(). It would match only one character.
1704 18. A valid (though odd) pattern that looked like a POSIX character
1705 class but used an invalid character after [ (for example [[,abc,]]) caused
1706 pcre_compile() to give the error "Failed: internal error: code overflow" or
1707 in some cases to crash with a glibc free() error. This could even happen if
1708 the pattern terminated after [[ but there just happened to be a sequence of
1709 letters, a binary zero, and a closing ] in the memory that followed.
1711 19. Perl's treatment of octal escapes in the range \400 to \777 has changed
1712 over the years. Originally (before any Unicode support), just the bottom 8
1713 bits were taken. Thus, for example, \500 really meant \100. Nowadays the
1714 output from "man perlunicode" includes this:
1716 The regular expression compiler produces polymorphic opcodes. That
1717 is, the pattern adapts to the data and automatically switches to
1718 the Unicode character scheme when presented with Unicode data--or
1719 instead uses a traditional byte scheme when presented with byte
1720 data.
1722 Sadly, a wide octal escape does not cause a switch, and in a string with
1723 no other multibyte characters, these octal escapes are treated as before.
1724 Thus, in Perl, the pattern /\500/ actually matches \100 but the pattern
1725 /\500|\x{1ff}/ matches \500 or \777 because the whole thing is treated as a
1726 Unicode string.
1728 I have not perpetrated such confusion in PCRE. Up till now, it took just
1729 the bottom 8 bits, as in old Perl. I have now made octal escapes with
1730 values greater than \377 illegal in non-UTF-8 mode. In UTF-8 mode they
1731 translate to the appropriate multibyte character.
1733 29. Applied some refactoring to reduce the number of warnings from Microsoft
1734 and Borland compilers. This has included removing the fudge introduced
1735 seven years ago for the OS/2 compiler (see 2.02/2 below) because it caused
1736 a warning about an unused variable.
1738 21. PCRE has not included VT (character 0x0b) in the set of whitespace
1739 characters since release 4.0, because Perl (from release 5.004) does not.
1740 [Or at least, is documented not to: some releases seem to be in conflict
1741 with the documentation.] However, when a pattern was studied with
1742 pcre_study() and all its branches started with \s, PCRE still included VT
1743 as a possible starting character. Of course, this did no harm; it just
1744 caused an unnecessary match attempt.
1746 22. Removed a now-redundant internal flag bit that recorded the fact that case
1747 dependency changed within the pattern. This was once needed for "required
1748 byte" processing, but is no longer used. This recovers a now-scarce options
1749 bit. Also moved the least significant internal flag bit to the most-
1750 significant bit of the word, which was not previously used (hangover from
1751 the days when it was an int rather than a uint) to free up another bit for
1752 the future.
1754 23. Added support for CRLF line endings as well as CR and LF. As well as the
1755 default being selectable at build time, it can now be changed at runtime
1756 via the PCRE_NEWLINE_xxx flags. There are now options for pcregrep to
1757 specify that it is scanning data with non-default line endings.
1759 24. Changed the definition of CXXLINK to make it agree with the definition of
1760 LINK in the Makefile, by replacing LDFLAGS to CXXFLAGS.
1762 25. Applied Ian Taylor's patches to avoid using another stack frame for tail
1763 recursions. This makes a big different to stack usage for some patterns.
1765 26. If a subpattern containing a named recursion or subroutine reference such
1766 as (?P>B) was quantified, for example (xxx(?P>B)){3}, the calculation of
1767 the space required for the compiled pattern went wrong and gave too small a
1768 value. Depending on the environment, this could lead to "Failed: internal
1769 error: code overflow at offset 49" or "glibc detected double free or
1770 corruption" errors.
1772 27. Applied patches from Google (a) to support the new newline modes and (b) to
1773 advance over multibyte UTF-8 characters in GlobalReplace.
1775 28. Change free() to pcre_free() in pcredemo.c. Apparently this makes a
1776 difference for some implementation of PCRE in some Windows version.
1778 29. Added some extra testing facilities to pcretest:
1780 \q<number> in a data line sets the "match limit" value
1781 \Q<number> in a data line sets the "match recursion limt" value
1782 -S <number> sets the stack size, where <number> is in megabytes
1784 The -S option isn't available for Windows.
1787 Version 6.6 06-Feb-06
1788 ---------------------
1790 1. Change 16(a) for 6.5 broke things, because PCRE_DATA_SCOPE was not defined
1791 in pcreposix.h. I have copied the definition from pcre.h.
1793 2. Change 25 for 6.5 broke compilation in a build directory out-of-tree
1794 because pcre.h is no longer a built file.
1796 3. Added Jeff Friedl's additional debugging patches to pcregrep. These are
1797 not normally included in the compiled code.
1800 Version 6.5 01-Feb-06
1801 ---------------------
1803 1. When using the partial match feature with pcre_dfa_exec(), it was not
1804 anchoring the second and subsequent partial matches at the new starting
1805 point. This could lead to incorrect results. For example, with the pattern
1806 /1234/, partially matching against "123" and then "a4" gave a match.
1808 2. Changes to pcregrep:
1810 (a) All non-match returns from pcre_exec() were being treated as failures
1811 to match the line. Now, unless the error is PCRE_ERROR_NOMATCH, an
1812 error message is output. Some extra information is given for the
1814 probably the only errors that are likely to be caused by users (by
1815 specifying a regex that has nested indefinite repeats, for instance).
1816 If there are more than 20 of these errors, pcregrep is abandoned.
1818 (b) A binary zero was treated as data while matching, but terminated the
1819 output line if it was written out. This has been fixed: binary zeroes
1820 are now no different to any other data bytes.
1822 (c) Whichever of the LC_ALL or LC_CTYPE environment variables is set is
1823 used to set a locale for matching. The --locale=xxxx long option has
1824 been added (no short equivalent) to specify a locale explicitly on the
1825 pcregrep command, overriding the environment variables.
1827 (d) When -B was used with -n, some line numbers in the output were one less
1828 than they should have been.
1830 (e) Added the -o (--only-matching) option.
1832 (f) If -A or -C was used with -c (count only), some lines of context were
1833 accidentally printed for the final match.
1835 (g) Added the -H (--with-filename) option.
1837 (h) The combination of options -rh failed to suppress file names for files
1838 that were found from directory arguments.
1840 (i) Added the -D (--devices) and -d (--directories) options.
1842 (j) Added the -F (--fixed-strings) option.
1844 (k) Allow "-" to be used as a file name for -f as well as for a data file.
1846 (l) Added the --colo(u)r option.
1848 (m) Added Jeffrey Friedl's -S testing option, but within #ifdefs so that it
1849 is not present by default.
1851 3. A nasty bug was discovered in the handling of recursive patterns, that is,
1852 items such as (?R) or (?1), when the recursion could match a number of
1853 alternatives. If it matched one of the alternatives, but subsequently,
1854 outside the recursion, there was a failure, the code tried to back up into
1855 the recursion. However, because of the way PCRE is implemented, this is not
1856 possible, and the result was an incorrect result from the match.
1858 In order to prevent this happening, the specification of recursion has
1859 been changed so that all such subpatterns are automatically treated as
1860 atomic groups. Thus, for example, (?R) is treated as if it were (?>(?R)).
1862 4. I had overlooked the fact that, in some locales, there are characters for
1863 which isalpha() is true but neither isupper() nor islower() are true. In
1864 the fr_FR locale, for instance, the \xAA and \xBA characters (ordmasculine
1865 and ordfeminine) are like this. This affected the treatment of \w and \W
1866 when they appeared in character classes, but not when they appeared outside
1867 a character class. The bit map for "word" characters is now created
1868 separately from the results of isalnum() instead of just taking it from the
1869 upper, lower, and digit maps. (Plus the underscore character, of course.)
1871 5. The above bug also affected the handling of POSIX character classes such as
1872 [[:alpha:]] and [[:alnum:]]. These do not have their own bit maps in PCRE's
1873 permanent tables. Instead, the bit maps for such a class were previously
1874 created as the appropriate unions of the upper, lower, and digit bitmaps.
1875 Now they are created by subtraction from the [[:word:]] class, which has
1876 its own bitmap.
1878 6. The [[:blank:]] character class matches horizontal, but not vertical space.
1879 It is created by subtracting the vertical space characters (\x09, \x0a,
1880 \x0b, \x0c) from the [[:space:]] bitmap. Previously, however, the
1881 subtraction was done in the overall bitmap for a character class, meaning
1882 that a class such as [\x0c[:blank:]] was incorrect because \x0c would not
1883 be recognized. This bug has been fixed.
1885 7. Patches from the folks at Google:
1887 (a) pcrecpp.cc: "to handle a corner case that may or may not happen in
1888 real life, but is still worth protecting against".
1890 (b) pcrecpp.cc: "corrects a bug when negative radixes are used with
1891 regular expressions".
1893 (c) pcre_scanner.cc: avoid use of std::count() because not all systems
1894 have it.
1896 (d) Split off pcrecpparg.h from pcrecpp.h and had the former built by
1897 "configure" and the latter not, in order to fix a problem somebody had
1898 with compiling the Arg class on HP-UX.
1900 (e) Improve the error-handling of the C++ wrapper a little bit.
1902 (f) New tests for checking recursion limiting.
1904 8. The pcre_memmove() function, which is used only if the environment does not
1905 have a standard memmove() function (and is therefore rarely compiled),
1906 contained two bugs: (a) use of int instead of size_t, and (b) it was not
1907 returning a result (though PCRE never actually uses the result).
1909 9. In the POSIX regexec() interface, if nmatch is specified as a ridiculously
1910 large number - greater than INT_MAX/(3*sizeof(int)) - REG_ESPACE is
1911 returned instead of calling malloc() with an overflowing number that would
1912 most likely cause subsequent chaos.
1914 10. The debugging option of pcretest was not showing the NO_AUTO_CAPTURE flag.
1916 11. The POSIX flag REG_NOSUB is now supported. When a pattern that was compiled
1917 with this option is matched, the nmatch and pmatch options of regexec() are
1918 ignored.
1920 12. Added REG_UTF8 to the POSIX interface. This is not defined by POSIX, but is
1921 provided in case anyone wants to the the POSIX interface with UTF-8
1922 strings.
1924 13. Added CXXLDFLAGS to the Makefile parameters to provide settings only on the
1925 C++ linking (needed for some HP-UX environments).
1927 14. Avoid compiler warnings in get_ucpname() when compiled without UCP support
1928 (unused parameter) and in the pcre_printint() function (omitted "default"
1929 switch label when the default is to do nothing).
1931 15. Added some code to make it possible, when PCRE is compiled as a C++
1932 library, to replace subject pointers for pcre_exec() with a smart pointer
1933 class, thus making it possible to process discontinuous strings.
1935 16. The two macros PCRE_EXPORT and PCRE_DATA_SCOPE are confusing, and perform
1936 much the same function. They were added by different people who were trying
1937 to make PCRE easy to compile on non-Unix systems. It has been suggested
1938 that PCRE_EXPORT be abolished now that there is more automatic apparatus
1939 for compiling on Windows systems. I have therefore replaced it with
1940 PCRE_DATA_SCOPE. This is set automatically for Windows; if not set it
1941 defaults to "extern" for C or "extern C" for C++, which works fine on
1942 Unix-like systems. It is now possible to override the value of PCRE_DATA_
1943 SCOPE with something explicit in config.h. In addition:
1945 (a) pcreposix.h still had just "extern" instead of either of these macros;
1946 I have replaced it with PCRE_DATA_SCOPE.
1948 (b) Functions such as _pcre_xclass(), which are internal to the library,
1949 but external in the C sense, all had PCRE_EXPORT in their definitions.
1950 This is apparently wrong for the Windows case, so I have removed it.
1951 (It makes no difference on Unix-like systems.)
1953 17. Added a new limit, MATCH_LIMIT_RECURSION, which limits the depth of nesting
1954 of recursive calls to match(). This is different to MATCH_LIMIT because
1955 that limits the total number of calls to match(), not all of which increase
1956 the depth of recursion. Limiting the recursion depth limits the amount of
1957 stack (or heap if NO_RECURSE is set) that is used. The default can be set
1958 when PCRE is compiled, and changed at run time. A patch from Google adds
1959 this functionality to the C++ interface.
1961 18. Changes to the handling of Unicode character properties:
1963 (a) Updated the table to Unicode 4.1.0.
1965 (b) Recognize characters that are not in the table as "Cn" (undefined).
1967 (c) I revised the way the table is implemented to a much improved format
1968 which includes recognition of ranges. It now supports the ranges that
1969 are defined in UnicodeData.txt, and it also amalgamates other
1970 characters into ranges. This has reduced the number of entries in the
1971 table from around 16,000 to around 3,000, thus reducing its size
1972 considerably. I realized I did not need to use a tree structure after
1973 all - a binary chop search is just as efficient. Having reduced the
1974 number of entries, I extended their size from 6 bytes to 8 bytes to
1975 allow for more data.
1977 (d) Added support for Unicode script names via properties such as \p{Han}.
1979 19. In UTF-8 mode, a backslash followed by a non-Ascii character was not
1980 matching that character.
1982 20. When matching a repeated Unicode property with a minimum greater than zero,
1983 (for example \pL{2,}), PCRE could look past the end of the subject if it
1984 reached it while seeking the minimum number of characters. This could
1985 happen only if some of the characters were more than one byte long, because
1986 there is a check for at least the minimum number of bytes.
1988 21. Refactored the implementation of \p and \P so as to be more general, to
1989 allow for more different types of property in future. This has changed the
1990 compiled form incompatibly. Anybody with saved compiled patterns that use
1991 \p or \P will have to recompile them.
1993 22. Added "Any" and "L&" to the supported property types.
1995 23. Recognize \x{...} as a code point specifier, even when not in UTF-8 mode,
1996 but give a compile time error if the value is greater than 0xff.
1998 24. The man pages for pcrepartial, pcreprecompile, and pcre_compile2 were
1999 accidentally not being installed or uninstalled.
2001 25. The pcre.h file was built from pcre.h.in, but the only changes that were
2002 made were to insert the current release number. This seemed silly, because
2003 it made things harder for people building PCRE on systems that don't run
2004 "configure". I have turned pcre.h into a distributed file, no longer built
2005 by "configure", with the version identification directly included. There is
2006 no longer a pcre.h.in file.
2008 However, this change necessitated a change to the pcre-config script as
2009 well. It is built from pcre-config.in, and one of the substitutions was the
2010 release number. I have updated configure.ac so that ./configure now finds
2011 the release number by grepping pcre.h.
2013 26. Added the ability to run the tests under valgrind.
2016 Version 6.4 05-Sep-05
2017 ---------------------
2019 1. Change 6.0/10/(l) to pcregrep introduced a bug that caused separator lines
2020 "--" to be printed when multiple files were scanned, even when none of the
2021 -A, -B, or -C options were used. This is not compatible with Gnu grep, so I
2022 consider it to be a bug, and have restored the previous behaviour.
2024 2. A couple of code tidies to get rid of compiler warnings.
2026 3. The pcretest program used to cheat by referring to symbols in the library
2027 whose names begin with _pcre_. These are internal symbols that are not
2028 really supposed to be visible externally, and in some environments it is
2029 possible to suppress them. The cheating is now confined to including
2030 certain files from the library's source, which is a bit cleaner.
2032 4. Renamed pcre.in as pcre.h.in to go with pcrecpp.h.in; it also makes the
2033 file's purpose clearer.
2035 5. Reorganized pcre_ucp_findchar().
2038 Version 6.3 15-Aug-05
2039 ---------------------
2041 1. The file libpcre.pc.in did not have general read permission in the tarball.
2043 2. There were some problems when building without C++ support:
2045 (a) If C++ support was not built, "make install" and "make test" still
2046 tried to test it.
2048 (b) There were problems when the value of CXX was explicitly set. Some
2049 changes have been made to try to fix these, and ...
2051 (c) --disable-cpp can now be used to explicitly disable C++ support.
2053 (d) The use of @CPP_OBJ@ directly caused a blank line preceded by a
2054 backslash in a target when C++ was disabled. This confuses some
2055 versions of "make", apparently. Using an intermediate variable solves
2056 this. (Same for CPP_LOBJ.)
2058 3. $(LINK_FOR_BUILD) now includes $(CFLAGS_FOR_BUILD) and $(LINK)
2059 (non-Windows) now includes $(CFLAGS) because these flags are sometimes
2060 necessary on certain architectures.
2062 4. Added a setting of -export-symbols-regex to the link command to remove
2063 those symbols that are exported in the C sense, but actually are local
2064 within the library, and not documented. Their names all begin with
2065 "_pcre_". This is not a perfect job, because (a) we have to except some
2066 symbols that pcretest ("illegally") uses, and (b) the facility isn't always
2067 available (and never for static libraries). I have made a note to try to
2068 find a way round (a) in the future.
2071 Version 6.2 01-Aug-05
2072 ---------------------
2074 1. There was no test for integer overflow of quantifier values. A construction
2075 such as {1111111111111111} would give undefined results. What is worse, if
2076 a minimum quantifier for a parenthesized subpattern overflowed and became
2077 negative, the calculation of the memory size went wrong. This could have
2078 led to memory overwriting.
2080 2. Building PCRE using VPATH was broken. Hopefully it is now fixed.
2082 3. Added "b" to the 2nd argument of fopen() in dftables.c, for non-Unix-like
2083 operating environments where this matters.
2085 4. Applied Giuseppe Maxia's patch to add additional features for controlling
2086 PCRE options from within the C++ wrapper.
2088 5. Named capturing subpatterns were not being correctly counted when a pattern
2089 was compiled. This caused two problems: (a) If there were more than 100
2090 such subpatterns, the calculation of the memory needed for the whole
2091 compiled pattern went wrong, leading to an overflow error. (b) Numerical
2092 back references of the form \12, where the number was greater than 9, were
2093 not recognized as back references, even though there were sufficient
2094 previous subpatterns.
2096 6. Two minor patches to pcrecpp.cc in order to allow it to compile on older
2097 versions of gcc, e.g. 2.95.4.
2100 Version 6.1 21-Jun-05
2101 ---------------------
2103 1. There was one reference to the variable "posix" in pcretest.c that was not
2104 surrounded by "#if !defined NOPOSIX".
2106 2. Make it possible to compile pcretest without DFA support, UTF8 support, or
2107 the cross-check on the old pcre_info() function, for the benefit of the
2108 cut-down version of PCRE that is currently imported into Exim.
2110 3. A (silly) pattern starting with (?i)(?-i) caused an internal space
2111 allocation error. I've done the easy fix, which wastes 2 bytes for sensible
2112 patterns that start (?i) but I don't think that matters. The use of (?i) is
2113 just an example; this all applies to the other options as well.
2115 4. Since libtool seems to echo the compile commands it is issuing, the output
2116 from "make" can be reduced a bit by putting "@" in front of each libtool
2117 compile command.
2119 5. Patch from the folks at Google for configure.in to be a bit more thorough
2120 in checking for a suitable C++ installation before trying to compile the
2121 C++ stuff. This should fix a reported problem when a compiler was present,
2122 but no suitable headers.
2124 6. The man pages all had just "PCRE" as their title. I have changed them to
2125 be the relevant file name. I have also arranged that these names are
2126 retained in the file doc/pcre.txt, which is a concatenation in text format
2127 of all the man pages except the little individual ones for each function.
2129 7. The NON-UNIX-USE file had not been updated for the different set of source
2130 files that come with release 6. I also added a few comments about the C++
2131 wrapper.
2134 Version 6.0 07-Jun-05
2135 ---------------------
2137 1. Some minor internal re-organization to help with my DFA experiments.
2139 2. Some missing #ifdef SUPPORT_UCP conditionals in pcretest and printint that
2140 didn't matter for the library itself when fully configured, but did matter
2141 when compiling without UCP support, or within Exim, where the ucp files are
2142 not imported.
2144 3. Refactoring of the library code to split up the various functions into
2145 different source modules. The addition of the new DFA matching code (see
2146 below) to a single monolithic source would have made it really too
2147 unwieldy, quite apart from causing all the code to be include in a
2148 statically linked application, when only some functions are used. This is
2149 relevant even without the DFA addition now that patterns can be compiled in
2150 one application and matched in another.
2152 The downside of splitting up is that there have to be some external
2153 functions and data tables that are used internally in different modules of
2154 the library but which are not part of the API. These have all had their
2155 names changed to start with "_pcre_" so that they are unlikely to clash
2156 with other external names.
2158 4. Added an alternate matching function, pcre_dfa_exec(), which matches using
2159 a different (DFA) algorithm. Although it is slower than the original
2160 function, it does have some advantages for certain types of matching
2161 problem.
2163 5. Upgrades to pcretest in order to test the features of pcre_dfa_exec(),
2164 including restarting after a partial match.
2166 6. A patch for pcregrep that defines INVALID_FILE_ATTRIBUTES if it is not
2167 defined when compiling for Windows was sent to me. I have put it into the
2168 code, though I have no means of testing or verifying it.
2170 7. Added the pcre_refcount() auxiliary function.
2172 8. Added the PCRE_FIRSTLINE option. This constrains an unanchored pattern to
2173 match before or at the first newline in the subject string. In pcretest,
2174 the /f option on a pattern can be used to set this.
2176 9. A repeated \w when used in UTF-8 mode with characters greater than 256
2177 would behave wrongly. This has been present in PCRE since release 4.0.
2179 10. A number of changes to the pcregrep command:
2181 (a) Refactored how -x works; insert ^(...)$ instead of setting
2182 PCRE_ANCHORED and checking the length, in preparation for adding
2183 something similar for -w.
2185 (b) Added the -w (match as a word) option.
2187 (c) Refactored the way lines are read and buffered so as to have more
2188 than one at a time available.
2190 (d) Implemented a pcregrep test script.
2192 (e) Added the -M (multiline match) option. This allows patterns to match
2193 over several lines of the subject. The buffering ensures that at least
2194 8K, or the rest of the document (whichever is the shorter) is available
2195 for matching (and similarly the previous 8K for lookbehind assertions).
2197 (f) Changed the --help output so that it now says
2199 -w, --word-regex(p)
2201 instead of two lines, one with "regex" and the other with "regexp"
2202 because that confused at least one person since the short forms are the
2203 same. (This required a bit of code, as the output is generated
2204 automatically from a table. It wasn't just a text change.)
2206 (g) -- can be used to terminate pcregrep options if the next thing isn't an
2207 option but starts with a hyphen. Could be a pattern or a path name
2208 starting with a hyphen, for instance.
2210 (h) "-" can be given as a file name to represent stdin.
2212 (i) When file names are being printed, "(standard input)" is used for
2213 the standard input, for compatibility with GNU grep. Previously
2214 "<stdin>" was used.
2216 (j) The option --label=xxx can be used to supply a name to be used for
2217 stdin when file names are being printed. There is no short form.
2219 (k) Re-factored the options decoding logic because we are going to add
2220 two more options that take data. Such options can now be given in four
2221 different ways, e.g. "-fname", "-f name", "--file=name", "--file name".
2223 (l) Added the -A, -B, and -C options for requesting that lines of context
2224 around matches be printed.
2226 (m) Added the -L option to print the names of files that do not contain
2227 any matching lines, that is, the complement of -l.
2229 (n) The return code is 2 if any file cannot be opened, but pcregrep does
2230 continue to scan other files.
2232 (o) The -s option was incorrectly implemented. For compatibility with other
2233 greps, it now suppresses the error message for a non-existent or non-
2234 accessible file (but not the return code). There is a new option called
2235 -q that suppresses the output of matching lines, which was what -s was
2236 previously doing.
2238 (p) Added --include and --exclude options to specify files for inclusion
2239 and exclusion when recursing.
2241 11. The Makefile was not using the Autoconf-supported LDFLAGS macro properly.
2242 Hopefully, it now does.
2244 12. Missing cast in pcre_study().
2246 13. Added an "uninstall" target to the makefile.
2248 14. Replaced "extern" in the function prototypes in Makefile.in with
2249 "PCRE_DATA_SCOPE", which defaults to 'extern' or 'extern "C"' in the Unix
2250 world, but is set differently for Windows.
2252 15. Added a second compiling function called pcre_compile2(). The only
2253 difference is that it has an extra argument, which is a pointer to an
2254 integer error code. When there is a compile-time failure, this is set
2255 non-zero, in addition to the error test pointer being set to point to an
2256 error message. The new argument may be NULL if no error number is required
2257 (but then you may as well call pcre_compile(), which is now just a
2258 wrapper). This facility is provided because some applications need a
2259 numeric error indication, but it has also enabled me to tidy up the way
2260 compile-time errors are handled in the POSIX wrapper.
2262 16. Added VPATH=.libs to the makefile; this should help when building with one
2263 prefix path and installing with another. (Or so I'm told by someone who
2264 knows more about this stuff than I do.)
2266 17. Added a new option, REG_DOTALL, to the POSIX function regcomp(). This
2267 passes PCRE_DOTALL to the pcre_compile() function, making the "." character
2268 match everything, including newlines. This is not POSIX-compatible, but
2269 somebody wanted the feature. From pcretest it can be activated by using
2270 both the P and the s flags.
2272 18. AC_PROG_LIBTOOL appeared twice in Makefile.in. Removed one.
2274 19. libpcre.pc was being incorrectly installed as executable.
2276 20. A couple of places in pcretest check for end-of-line by looking for '\n';
2277 it now also looks for '\r' so that it will work unmodified on Windows.
2279 21. Added Google's contributed C++ wrapper to the distribution.
2281 22. Added some untidy missing memory free() calls in pcretest, to keep
2282 Electric Fence happy when testing.
2286 Version 5.0 13-Sep-04
2287 ---------------------
2289 1. Internal change: literal characters are no longer packed up into items
2290 containing multiple characters in a single byte-string. Each character
2291 is now matched using a separate opcode. However, there may be more than one
2292 byte in the character in UTF-8 mode.
2294 2. The pcre_callout_block structure has two new fields: pattern_position and
2295 next_item_length. These contain the offset in the pattern to the next match
2296 item, and its length, respectively.
2298 3. The PCRE_AUTO_CALLOUT option for pcre_compile() requests the automatic
2299 insertion of callouts before each pattern item. Added the /C option to
2300 pcretest to make use of this.
2302 4. On the advice of a Windows user, the lines
2304 #if defined(_WIN32) || defined(WIN32)
2305 _setmode( _fileno( stdout ), 0x8000 );
2306 #endif /* defined(_WIN32) || defined(WIN32) */
2308 have been added to the source of pcretest. This apparently does useful
2309 magic in relation to line terminators.
2311 5. Changed "r" and "w" in the calls to fopen() in pcretest to "rb" and "wb"
2312 for the benefit of those environments where the "b" makes a difference.
2314 6. The icc compiler has the same options as gcc, but "configure" doesn't seem
2315 to know about it. I have put a hack into configure.in that adds in code
2316 to set GCC=yes if CC=icc. This seems to end up at a point in the
2317 generated configure script that is early enough to affect the setting of
2318 compiler options, which is what is needed, but I have no means of testing
2319 whether it really works. (The user who reported this had patched the
2320 generated configure script, which of course I cannot do.)
2322 LATER: After change 22 below (new libtool files), the configure script
2323 seems to know about icc (and also ecc). Therefore, I have commented out
2324 this hack in configure.in.
2326 7. Added support for pkg-config (2 patches were sent in).
2328 8. Negated POSIX character classes that used a combination of internal tables
2329 were completely broken. These were [[:^alpha:]], [[:^alnum:]], and
2330 [[:^ascii]]. Typically, they would match almost any characters. The other
2331 POSIX classes were not broken in this way.
2333 9. Matching the pattern "\b.*?" against "ab cd", starting at offset 1, failed
2334 to find the match, as PCRE was deluded into thinking that the match had to
2335 start at the start point or following a newline. The same bug applied to
2336 patterns with negative forward assertions or any backward assertions
2337 preceding ".*" at the start, unless the pattern required a fixed first
2338 character. This was a failing pattern: "(?!.bcd).*". The bug is now fixed.
2340 10. In UTF-8 mode, when moving forwards in the subject after a failed match
2341 starting at the last subject character, bytes beyond the end of the subject
2342 string were read.
2344 11. Renamed the variable "class" as "classbits" to make life easier for C++
2345 users. (Previously there was a macro definition, but it apparently wasn't
2346 enough.)
2348 12. Added the new field "tables" to the extra data so that tables can be passed
2349 in at exec time, or the internal tables can be re-selected. This allows
2350 a compiled regex to be saved and re-used at a later time by a different
2351 program that might have everything at different addresses.
2353 13. Modified the pcre-config script so that, when run on Solaris, it shows a
2354 -R library as well as a -L library.
2356 14. The debugging options of pcretest (-d on the command line or D on a
2357 pattern) showed incorrect output for anything following an extended class
2358 that contained multibyte characters and which was followed by a quantifier.
2360 15. Added optional support for general category Unicode character properties
2361 via the \p, \P, and \X escapes. Unicode property support implies UTF-8
2362 support. It adds about 90K to the size of the library. The meanings of the
2363 inbuilt class escapes such as \d and \s have NOT been changed.
2365 16. Updated pcredemo.c to include calls to free() to release the memory for the
2366 compiled pattern.
2368 17. The generated file chartables.c was being created in the source directory
2369 instead of in the building directory. This caused the build to fail if the
2370 source directory was different from the building directory, and was
2371 read-only.
2373 18. Added some sample Win commands from Mark Tetrode into the NON-UNIX-USE
2374 file. No doubt somebody will tell me if they don't make sense... Also added
2375 Dan Mooney's comments about building on OpenVMS.
2377 19. Added support for partial matching via the PCRE_PARTIAL option for
2378 pcre_exec() and the \P data escape in pcretest.
2380 20. Extended pcretest with 3 new pattern features:
2382 (i) A pattern option of the form ">rest-of-line" causes pcretest to
2383 write the compiled pattern to the file whose name is "rest-of-line".
2384 This is a straight binary dump of the data, with the saved pointer to
2385 the character tables forced to be NULL. The study data, if any, is
2386 written too. After writing, pcretest reads a new pattern.
2388 (ii) If, instead of a pattern, "<rest-of-line" is given, pcretest reads a
2389 compiled pattern from the given file. There must not be any
2390 occurrences of "<" in the file name (pretty unlikely); if there are,
2391 pcretest will instead treat the initial "<" as a pattern delimiter.
2392 After reading in the pattern, pcretest goes on to read data lines as
2393 usual.
2395 (iii) The F pattern option causes pcretest to flip the bytes in the 32-bit
2396 and 16-bit fields in a compiled pattern, to simulate a pattern that
2397 was compiled on a host of opposite endianness.
2399 21. The pcre-exec() function can now cope with patterns that were compiled on
2400 hosts of opposite endianness, with this restriction:
2402 As for any compiled expression that is saved and used later, the tables
2403 pointer field cannot be preserved; the extra_data field in the arguments
2404 to pcre_exec() should be used to pass in a tables address if a value
2405 other than the default internal tables were used at compile time.
2407 22. Calling pcre_exec() with a negative value of the "ovecsize" parameter is
2408 now diagnosed as an error. Previously, most of the time, a negative number
2409 would have been treated as zero, but if in addition "ovector" was passed as
2410 NULL, a crash could occur.
2412 23. Updated the files ltmain.sh, config.sub, config.guess, and aclocal.m4 with
2413 new versions from the libtool 1.5 distribution (the last one is a copy of
2414 a file called libtool.m4). This seems to have fixed the need to patch
2415 "configure" to support Darwin 1.3 (which I used to do). However, I still
2416 had to patch ltmain.sh to ensure that ${SED} is set (it isn't on my
2417 workstation).
2419 24. Changed the PCRE licence to be the more standard "BSD" licence.
2422 Version 4.5 01-Dec-03
2423 ---------------------
2425 1. There has been some re-arrangement of the code for the match() function so
2426 that it can be compiled in a version that does not call itself recursively.
2427 Instead, it keeps those local variables that need separate instances for
2428 each "recursion" in a frame on the heap, and gets/frees frames whenever it
2429 needs to "recurse". Keeping track of where control must go is done by means
2430 of setjmp/longjmp. The whole thing is implemented by a set of macros that
2431 hide most of the details from the main code, and operates only if
2432 NO_RECURSE is defined while compiling pcre.c. If PCRE is built using the
2433 "configure" mechanism, "--disable-stack-for-recursion" turns on this way of
2434 operating.
2436 To make it easier for callers to provide specially tailored get/free
2437 functions for this usage, two new functions, pcre_stack_malloc, and
2438 pcre_stack_free, are used. They are always called in strict stacking order,
2439 and the size of block requested is always the same.
2441 The PCRE_CONFIG_STACKRECURSE info parameter can be used to find out whether
2442 PCRE has been compiled to use the stack or the heap for recursion. The
2443 -C option of pcretest uses this to show which version is compiled.
2445 A new data escape \S, is added to pcretest; it causes the amounts of store
2446 obtained and freed by both kinds of malloc/free at match time to be added
2447 to the output.
2449 2. Changed the locale test to use "fr_FR" instead of "fr" because that's
2450 what's available on my current Linux desktop machine.
2452 3. When matching a UTF-8 string, the test for a valid string at the start has
2453 been extended. If start_offset is not zero, PCRE now checks that it points
2454 to a byte that is the start of a UTF-8 character. If not, it returns
2455 PCRE_ERROR_BADUTF8_OFFSET (-11). Note: the whole string is still checked;
2456 this is necessary because there may be backward assertions in the pattern.
2457 When matching the same subject several times, it may save resources to use
2458 PCRE_NO_UTF8_CHECK on all but the first call if the string is long.
2460 4. The code for checking the validity of UTF-8 strings has been tightened so
2461 that it rejects (a) strings containing 0xfe or 0xff bytes and (b) strings
2462 containing "overlong sequences".
2464 5. Fixed a bug (appearing twice) that I could not find any way of exploiting!
2465 I had written "if ((digitab[*p++] && chtab_digit) == 0)" where the "&&"
2466 should have been "&", but it just so happened that all the cases this let
2467 through by mistake were picked up later in the function.
2469 6. I had used a variable called "isblank" - this is a C99 function, causing
2470 some compilers to warn. To avoid this, I renamed it (as "blankclass").
2472 7. Cosmetic: (a) only output another newline at the end of pcretest if it is
2473 prompting; (b) run "./pcretest /dev/null" at the start of the test script
2474 so the version is shown; (c) stop "make test" echoing "./RunTest".
2476 8. Added patches from David Burgess to enable PCRE to run on EBCDIC systems.
2478 9. The prototype for memmove() for systems that don't have it was using
2479 size_t, but the inclusion of the header that defines size_t was later. I've
2480 moved the #includes for the C headers earlier to avoid this.
2482 10. Added some adjustments to the code to make it easier to compiler on certain
2483 special systems:
2485 (a) Some "const" qualifiers were missing.
2486 (b) Added the macro EXPORT before all exported functions; by default this
2487 is defined to be empty.
2488 (c) Changed the dftables auxiliary program (that builds chartables.c) so
2489 that it reads its output file name as an argument instead of writing
2490 to the standard output and assuming this can be redirected.
2492 11. In UTF-8 mode, if a recursive reference (e.g. (?1)) followed a character
2493 class containing characters with values greater than 255, PCRE compilation
2494 went into a loop.
2496 12. A recursive reference to a subpattern that was within another subpattern
2497 that had a minimum quantifier of zero caused PCRE to crash. For example,
2498 (x(y(?2))z)? provoked this bug with a subject that got as far as the
2499 recursion. If the recursively-called subpattern itself had a zero repeat,
2500 that was OK.
2502 13. In pcretest, the buffer for reading a data line was set at 30K, but the
2503 buffer into which it was copied (for escape processing) was still set at
2504 1024, so long lines caused crashes.
2506 14. A pattern such as /[ab]{1,3}+/ failed to compile, giving the error
2507 "internal error: code overflow...". This applied to any character class
2508 that was followed by a possessive quantifier.
2510 15. Modified the Makefile to add libpcre.la as a prerequisite for
2511 libpcreposix.la because I was told this is needed for a parallel build to
2512 work.
2514 16. If a pattern that contained .* following optional items at the start was
2515 studied, the wrong optimizing data was generated, leading to matching
2516 errors. For example, studying /[ab]*.*c/ concluded, erroneously, that any
2517 matching string must start with a or b or c. The correct conclusion for
2518 this pattern is that a match can start with any character.
2521 Version 4.4 13-Aug-03
2522 ---------------------
2524 1. In UTF-8 mode, a character class containing characters with values between
2525 127 and 255 was not handled correctly if the compiled pattern was studied.
2526 In fixing this, I have also improved the studying algorithm for such
2527 classes (slightly).
2529 2. Three internal functions had redundant arguments passed to them. Removal
2530 might give a very teeny performance improvement.
2532 3. Documentation bug: the value of the capture_top field in a callout is *one
2533 more than* the number of the hightest numbered captured substring.
2535 4. The Makefile linked pcretest and pcregrep with -lpcre, which could result
2536 in incorrectly linking with a previously installed version. They now link
2537 explicitly with libpcre.la.
2539 5. configure.in no longer needs to recognize Cygwin specially.
2541 6. A problem in pcre.in for Windows platforms is fixed.
2543 7. If a pattern was successfully studied, and the -d (or /D) flag was given to
2544 pcretest, it used to include the size of the study block as part of its
2545 output. Unfortunately, the structure contains a field that has a different
2546 size on different hardware architectures. This meant that the tests that
2547 showed this size failed. As the block is currently always of a fixed size,
2548 this information isn't actually particularly useful in pcretest output, so
2549 I have just removed it.
2551 8. Three pre-processor statements accidentally did not start in column 1.
2552 Sadly, there are *still* compilers around that complain, even though
2553 standard C has not required this for well over a decade. Sigh.
2555 9. In pcretest, the code for checking callouts passed small integers in the
2556 callout_data field, which is a void * field. However, some picky compilers
2557 complained about the casts involved for this on 64-bit systems. Now
2558 pcretest passes the address of the small integer instead, which should get
2559 rid of the warnings.
2561 10. By default, when in UTF-8 mode, PCRE now checks for valid UTF-8 strings at
2562 both compile and run time, and gives an error if an invalid UTF-8 sequence
2563 is found. There is a option for disabling this check in cases where the
2564 string is known to be correct and/or the maximum performance is wanted.
2566 11. In response to a bug report, I changed one line in Makefile.in from
2568 -Wl,--out-implib,.libs/lib@WIN_PREFIX@pcreposix.dll.a \
2569 to
2570 -Wl,--out-implib,.libs/@WIN_PREFIX@libpcreposix.dll.a \
2572 to look similar to other lines, but I have no way of telling whether this
2573 is the right thing to do, as I do not use Windows. No doubt I'll get told
2574 if it's wrong...
2577 Version 4.3 21-May-03
2578 ---------------------
2580 1. Two instances of @WIN_PREFIX@ omitted from the Windows targets in the
2581 Makefile.
2583 2. Some refactoring to improve the quality of the code:
2585 (i) The utf8_table... variables are now declared "const".
2587 (ii) The code for \cx, which used the "case flipping" table to upper case
2588 lower case letters, now just substracts 32. This is ASCII-specific,
2589 but the whole concept of \cx is ASCII-specific, so it seems
2590 reasonable.
2592 (iii) PCRE was using its character types table to recognize decimal and
2593 hexadecimal digits in the pattern. This is silly, because it handles
2594 only 0-9, a-f, and A-F, but the character types table is locale-
2595 specific, which means strange things might happen. A private
2596 table is now used for this - though it costs 256 bytes, a table is
2597 much faster than multiple explicit tests. Of course, the standard
2598 character types table is still used for matching digits in subject
2599 strings against \d.
2601 (iv) Strictly, the identifier ESC_t is reserved by POSIX (all identifiers
2602 ending in _t are). So I've renamed it as ESC_tee.
2604 3. The first argument for regexec() in the POSIX wrapper should have been
2605 defined as "const".
2607 4. Changed pcretest to use malloc() for its buffers so that they can be
2608 Electric Fenced for debugging.
2610 5. There were several places in the code where, in UTF-8 mode, PCRE would try
2611 to read one or more bytes before the start of the subject string. Often this
2612 had no effect on PCRE's behaviour, but in some circumstances it could
2613 provoke a segmentation fault.
2615 6. A lookbehind at the start of a pattern in UTF-8 mode could also cause PCRE
2616 to try to read one or more bytes before the start of the subject string.
2618 7. A lookbehind in a pattern matched in non-UTF-8 mode on a PCRE compiled with
2619 UTF-8 support could misbehave in various ways if the subject string
2620 contained bytes with the 0x80 bit set and the 0x40 bit unset in a lookbehind
2621 area. (PCRE was not checking for the UTF-8 mode flag, and trying to move
2622 back over UTF-8 characters.)
2625 Version 4.2 14-Apr-03
2626 ---------------------
2628 1. Typo "#if SUPPORT_UTF8" instead of "#ifdef SUPPORT_UTF8" fixed.
2630 2. Changes to the building process, supplied by Ronald Landheer-Cieslak
2631 [ON_WINDOWS]: new variable, "#" on non-Windows platforms
2632 [NOT_ON_WINDOWS]: new variable, "#" on Windows platforms
2633 [WIN_PREFIX]: new variable, "cyg" for Cygwin
2634 * Makefile.in: use autoconf substitution for OBJEXT, EXEEXT, BUILD_OBJEXT
2636 Note: automatic setting of the BUILD variables is not yet working
2637 set CPPFLAGS and BUILD_CPPFLAGS (but don't use yet) - should be used at
2638 compile-time but not at link-time
2639 [LINK]: use for linking executables only
2640 make different versions for Windows and non-Windows
2641 [LINKLIB]: new variable, copy of UNIX-style LINK, used for linking
2642 libraries
2643 [LINK_FOR_BUILD]: new variable
2644 [OBJEXT]: use throughout
2645 [EXEEXT]: use throughout
2646 <winshared>: new target
2647 <wininstall>: new target
2648 <dftables.o>: use native compiler
2649 <dftables>: use native linker
2650 <install>: handle Windows platform correctly
2651 <clean>: ditto
2652 <check>: ditto
2653 copy DLL to top builddir before testing
2655 As part of these changes, -no-undefined was removed again. This was reported
2656 to give trouble on HP-UX 11.0, so getting rid of it seems like a good idea
2657 in any case.
2659 3. Some tidies to get rid of compiler warnings:
2661 . In the match_data structure, match_limit was an unsigned long int, whereas
2662 match_call_count was an int. I've made them both unsigned long ints.
2664 . In pcretest the fact that a const uschar * doesn't automatically cast to
2665 a void * provoked a warning.
2667 . Turning on some more compiler warnings threw up some "shadow" variables
2668 and a few more missing casts.
2670 4. If PCRE was complied with UTF-8 support, but called without the PCRE_UTF8
2671 option, a class that contained a single character with a value between 128
2672 and 255 (e.g. /[\xFF]/) caused PCRE to crash.
2674 5. If PCRE was compiled with UTF-8 support, but called without the PCRE_UTF8
2675 option, a class that contained several characters, but with at least one
2676 whose value was between 128 and 255 caused PCRE to crash.
2679 Version 4.1 12-Mar-03
2680 ---------------------
2682 1. Compiling with gcc -pedantic found a couple of places where casts were
2683 needed, and a string in dftables.c that was longer than standard compilers are
2684 required to support.
2686 2. Compiling with Sun's compiler found a few more places where the code could
2687 be tidied up in order to avoid warnings.
2689 3. The variables for cross-compiling were called HOST_CC and HOST_CFLAGS; the
2690 first of these names is deprecated in the latest Autoconf in favour of the name
2691 CC_FOR_BUILD, because "host" is typically used to mean the system on which the
2692 compiled code will be run. I can't find a reference for HOST_CFLAGS, but by
2693 analogy I have changed it to CFLAGS_FOR_BUILD.
2695 4. Added -no-undefined to the linking command in the Makefile, because this is
2696 apparently helpful for Windows. To make it work, also added "-L. -lpcre" to the
2697 linking step for the pcreposix library.
2699 5. PCRE was failing to diagnose the case of two named groups with the same
2700 name.
2702 6. A problem with one of PCRE's optimizations was discovered. PCRE remembers a
2703 literal character that is needed in the subject for a match, and scans along to
2704 ensure that it is present before embarking on the full matching process. This
2705 saves time in cases of nested unlimited repeats that are never going to match.
2706 Problem: the scan can take a lot of time if the subject is very long (e.g.
2707 megabytes), thus penalizing straightforward matches. It is now done only if the
2708 amount of subject to be scanned is less than 1000 bytes.
2710 7. A lesser problem with the same optimization is that it was recording the
2711 first character of an anchored pattern as "needed", thus provoking a search
2712 right along the subject, even when the first match of the pattern was going to
2713 fail. The "needed" character is now not set for anchored patterns, unless it
2714 follows something in the pattern that is of non-fixed length. Thus, it still
2715 fulfils its original purpose of finding quick non-matches in cases of nested
2716 unlimited repeats, but isn't used for simple anchored patterns such as /^abc/.
2719 Version 4.0 17-Feb-03
2720 ---------------------
2722 1. If a comment in an extended regex that started immediately after a meta-item
2723 extended to the end of string, PCRE compiled incorrect data. This could lead to
2724 all kinds of weird effects. Example: /#/ was bad; /()#/ was bad; /a#/ was not.
2726 2. Moved to autoconf 2.53 and libtool 1.4.2.
2728 3. Perl 5.8 no longer needs "use utf8" for doing UTF-8 things. Consequently,
2729 the special perltest8 script is no longer needed - all the tests can be run
2730 from a single perltest script.
2732 4. From 5.004, Perl has not included the VT character (0x0b) in the set defined
2733 by \s. It has now been removed in PCRE. This means it isn't recognized as
2734 whitespace in /x regexes too, which is the same as Perl. Note that the POSIX
2735 class [:space:] *does* include VT, thereby creating a mess.
2737 5. Added the class [:blank:] (a GNU extension from Perl 5.8) to match only
2738 space and tab.
2740 6. Perl 5.005 was a long time ago. It's time to amalgamate the tests that use
2741 its new features into the main test script, reducing the number of scripts.
2743 7. Perl 5.8 has changed the meaning of patterns like /a(?i)b/. Earlier versions
2744 were backward compatible, and made the (?i) apply to the whole pattern, as if
2745 /i were given. Now it behaves more logically, and applies the option setting
2746 only to what follows. PCRE has been changed to follow suit. However, if it
2747 finds options settings right at the start of the pattern, it extracts them into
2748 the global options, as before. Thus, they show up in the info data.
2750 8. Added support for the \Q...\E escape sequence. Characters in between are
2751 treated as literals. This is slightly different from Perl in that $ and @ are
2752 also handled as literals inside the quotes. In Perl, they will cause variable
2753 interpolation. Note the following examples:
2755 Pattern PCRE matches Perl matches
2757 \Qabc$xyz\E abc$xyz abc followed by the contents of $xyz
2758 \Qabc\$xyz\E abc\$xyz abc\$xyz
2759 \Qabc\E\$\Qxyz\E abc$xyz abc$xyz
2761 For compatibility with Perl, \Q...\E sequences are recognized inside character
2762 classes as well as outside them.
2764 9. Re-organized 3 code statements in pcretest to avoid "overflow in
2765 floating-point constant arithmetic" warnings from a Microsoft compiler. Added a
2766 (size_t) cast to one statement in pcretest and one in pcreposix to avoid
2767 signed/unsigned warnings.
2769 10. SunOS4 doesn't have strtoul(). This was used only for unpicking the -o
2770 option for pcretest, so I've replaced it by a simple function that does just
2771 that job.
2773 11. pcregrep was ending with code 0 instead of 2 for the commands "pcregrep" or
2774 "pcregrep -".
2776 12. Added "possessive quantifiers" ?+, *+, ++, and {,}+ which come from Sun's
2777 Java package. This provides some syntactic sugar for simple cases of what my
2778 documentation calls "once-only subpatterns". A pattern such as x*+ is the same
2779 as (?>x*). In other words, if what is inside (?>...) is just a single repeated
2780 item, you can use this simplified notation. Note that only makes sense with
2781 greedy quantifiers. Consequently, the use of the possessive quantifier forces
2782 greediness, whatever the setting of the PCRE_UNGREEDY option.
2784 13. A change of greediness default within a pattern was not taking effect at
2785 the current level for patterns like /(b+(?U)a+)/. It did apply to parenthesized
2786 subpatterns that followed. Patterns like /b+(?U)a+/ worked because the option
2787 was abstracted outside.
2789 14. PCRE now supports the \G assertion. It is true when the current matching
2790 position is at the start point of the match. This differs from \A when the
2791 starting offset is non-zero. Used with the /g option of pcretest (or similar
2792 code), it works in the same way as it does for Perl's /g option. If all
2793 alternatives of a regex begin with \G, the expression is anchored to the start
2794 match position, and the "anchored" flag is set in the compiled expression.
2796 15. Some bugs concerning the handling of certain option changes within patterns
2797 have been fixed. These applied to options other than (?ims). For example,
2798 "a(?x: b c )d" did not match "XabcdY" but did match "Xa b c dY". It should have
2799 been the other way round. Some of this was related to change 7 above.
2801 16. PCRE now gives errors for /[.x.]/ and /[=x=]/ as unsupported POSIX
2802 features, as Perl does. Previously, PCRE gave the warnings only for /[[.x.]]/
2803 and /[[=x=]]/. PCRE now also gives an error for /[:name:]/ because it supports
2804 POSIX classes only within a class (e.g. /[[:alpha:]]/).
2806 17. Added support for Perl's \C escape. This matches one byte, even in UTF8
2807 mode. Unlike ".", it always matches newline, whatever the setting of
2808 PCRE_DOTALL. However, PCRE does not permit \C to appear in lookbehind
2809 assertions. Perl allows it, but it doesn't (in general) work because it can't
2810 calculate the length of the lookbehind. At least, that's the case for Perl
2811 5.8.0 - I've been told they are going to document that it doesn't work in
2812 future.
2814 18. Added an error diagnosis for escapes that PCRE does not support: these are
2815 \L, \l, \N, \P, \p, \U, \u, and \X.
2817 19. Although correctly diagnosing a missing ']' in a character class, PCRE was
2818 reading past the end of the pattern in cases such as /[abcd/.
2820 20. PCRE was getting more memory than necessary for patterns with classes that
2821 contained both POSIX named classes and other characters, e.g. /[[:space:]abc/.
2823 21. Added some code, conditional on #ifdef VPCOMPAT, to make life easier for
2824 compiling PCRE for use with Virtual Pascal.
2826 22. Small fix to the Makefile to make it work properly if the build is done
2827 outside the source tree.
2829 23. Added a new extension: a condition to go with recursion. If a conditional
2830 subpattern starts with (?(R) the "true" branch is used if recursion has
2831 happened, whereas the "false" branch is used only at the top level.
2833 24. When there was a very long string of literal characters (over 255 bytes
2834 without UTF support, over 250 bytes with UTF support), the computation of how
2835 much memory was required could be incorrect, leading to segfaults or other
2836 strange effects.
2838 25. PCRE was incorrectly assuming anchoring (either to start of subject or to
2839 start of line for a non-DOTALL pattern) when a pattern started with (.*) and
2840 there was a subsequent back reference to those brackets. This meant that, for
2841 example, /(.*)\d+\1/ failed to match "abc123bc". Unfortunately, it isn't
2842 possible to check for precisely this case. All we can do is abandon the
2843 optimization if .* occurs inside capturing brackets when there are any back
2844 references whatsoever. (See below for a better fix that came later.)
2846 26. The handling of the optimization for finding the first character of a
2847 non-anchored pattern, and for finding a character that is required later in the
2848 match were failing in some cases. This didn't break the matching; it just
2849 failed to optimize when it could. The way this is done has been re-implemented.
2851 27. Fixed typo in error message for invalid (?R item (it said "(?p").
2853 28. Added a new feature that provides some of the functionality that Perl
2854 provides with (?{...}). The facility is termed a "callout". The way it is done
2855 in PCRE is for the caller to provide an optional function, by setting
2856 pcre_callout to its entry point. Like pcre_malloc and pcre_free, this is a
2857 global variable. By default it is unset, which disables all calling out. To get
2858 the function called, the regex must include (?C) at appropriate points. This
2859 is, in fact, equivalent to (?C0), and any number <= 255 may be given with (?C).
2860 This provides a means of identifying different callout points. When PCRE
2861 reaches such a point in the regex, if pcre_callout has been set, the external
2862 function is called. It is provided with data in a structure called
2863 pcre_callout_block, which is defined in pcre.h. If the function returns 0,
2864 matching continues; if it returns a non-zero value, the match at the current
2865 point fails. However, backtracking will occur if possible. [This was changed
2866 later and other features added - see item 49 below.]
2868 29. pcretest is upgraded to test the callout functionality. It provides a
2869 callout function that displays information. By default, it shows the start of
2870 the match and the current position in the text. There are some new data escapes
2871 to vary what happens:
2873 \C+ in addition, show current contents of captured substrings
2874 \C- do not supply a callout function
2875 \C!n return 1 when callout number n is reached
2876 \C!n!m return 1 when callout number n is reached for the mth time
2878 30. If pcregrep was called with the -l option and just a single file name, it
2879 output "<stdin>" if a match was found, instead of the file name.
2881 31. Improve the efficiency of the POSIX API to PCRE. If the number of capturing
2882 slots is less than POSIX_MALLOC_THRESHOLD, use a block on the stack to pass to
2883 pcre_exec(). This saves a malloc/free per call. The default value of
2884 POSIX_MALLOC_THRESHOLD is 10; it can be changed by --with-posix-malloc-threshold
2885 when configuring.
2887 32. The default maximum size of a compiled pattern is 64K. There have been a
2888 few cases of people hitting this limit. The code now uses macros to handle the
2889 storing of links as offsets within the compiled pattern. It defaults to 2-byte
2890 links, but this can be changed to 3 or 4 bytes by --with-link-size when
2891 configuring. Tests 2 and 5 work only with 2-byte links because they output
2892 debugging information about compiled patterns.
2894 33. Internal code re-arrangements:
2896 (a) Moved the debugging function for printing out a compiled regex into
2897 its own source file (printint.c) and used #include to pull it into
2898 pcretest.c and, when DEBUG is defined, into pcre.c, instead of having two
2899 separate copies.
2901 (b) Defined the list of op-code names for debugging as a macro in
2902 internal.h so that it is next to the definition of the opcodes.
2904 (c) Defined a table of op-code lengths for simpler skipping along compiled
2905 code. This is again a macro in internal.h so that it is next to the
2906 definition of the opcodes.
2908 34. Added support for recursive calls to individual subpatterns, along the
2909 lines of Robin Houston's patch (but implemented somewhat differently).
2911 35. Further mods to the Makefile to help Win32. Also, added code to pcregrep to
2912 allow it to read and process whole directories in Win32. This code was
2913 contributed by Lionel Fourquaux; it has not been tested by me.
2915 36. Added support for named subpatterns. The Python syntax (?P<name>...) is
2916 used to name a group. Names consist of alphanumerics and underscores, and must
2917 be unique. Back references use the syntax (?P=name) and recursive calls use
2918 (?P>name) which is a PCRE extension to the Python extension. Groups still have
2919 numbers. The function pcre_fullinfo() can be used after compilation to extract
2920 a name/number map. There are three relevant calls:
2922 PCRE_INFO_NAMEENTRYSIZE yields the size of each entry in the map
2923 PCRE_INFO_NAMECOUNT yields the number of entries
2924 PCRE_INFO_NAMETABLE yields a pointer to the map.
2926 The map is a vector of fixed-size entries. The size of each entry depends on
2927 the length of the longest name used. The first two bytes of each entry are the
2928 group number, most significant byte first. There follows the corresponding
2929 name, zero terminated. The names are in alphabetical order.
2931 37. Make the maximum literal string in the compiled code 250 for the non-UTF-8
2932 case instead of 255. Making it the same both with and without UTF-8 support
2933 means that the same test output works with both.
2935 38. There was a case of malloc(0) in the POSIX testing code in pcretest. Avoid
2936 calling malloc() with a zero argument.
2938 39. Change 25 above had to resort to a heavy-handed test for the .* anchoring
2939 optimization. I've improved things by keeping a bitmap of backreferences with
2940 numbers 1-31 so that if .* occurs inside capturing brackets that are not in
2941 fact referenced, the optimization can be applied. It is unlikely that a
2942 relevant occurrence of .* (i.e. one which might indicate anchoring or forcing
2943 the match to follow \n) will appear inside brackets with a number greater than
2944 31, but if it does, any back reference > 31 suppresses the optimization.
2946 40. Added a new compile-time option PCRE_NO_AUTO_CAPTURE. This has the effect
2947 of disabling numbered capturing parentheses. Any opening parenthesis that is
2948 not followed by ? behaves as if it were followed by ?: but named parentheses
2949 can still be used for capturing (and they will acquire numbers in the usual
2950 way).
2952 41. Redesigned the return codes from the match() function into yes/no/error so
2953 that errors can be passed back from deep inside the nested calls. A malloc
2954 failure while inside a recursive subpattern call now causes the
2955 PCRE_ERROR_NOMEMORY return instead of quietly going wrong.
2957 42. It is now possible to set a limit on the number of times the match()
2958 function is called in a call to pcre_exec(). This facility makes it possible to
2959 limit the amount of recursion and backtracking, though not in a directly
2960 obvious way, because the match() function is used in a number of different
2961 circumstances. The count starts from zero for each position in the subject
2962 string (for non-anchored patterns). The default limit is, for compatibility, a
2963 large number, namely 10 000 000. You can change this in two ways:
2965 (a) When configuring PCRE before making, you can use --with-match-limit=n
2966 to set a default value for the compiled library.
2968 (b) For each call to pcre_exec(), you can pass a pcre_extra block in which
2969 a different value is set. See 45 below.
2971 If the limit is exceeded, pcre_exec() returns PCRE_ERROR_MATCHLIMIT.
2973 43. Added a new function pcre_config(int, void *) to enable run-time extraction
2974 of things that can be changed at compile time. The first argument specifies
2975 what is wanted and the second points to where the information is to be placed.
2976 The current list of available information is:
2980 The output is an integer that is set to one if UTF-8 support is available;
2981 otherwise it is set to zero.
2985 The output is an integer that it set to the value of the code that is used for
2986 newline. It is either LF (10) or CR (13).
2990 The output is an integer that contains the number of bytes used for internal
2991 linkage in compiled expressions. The value is 2, 3, or 4. See item 32 above.
2995 The output is an integer that contains the threshold above which the POSIX
2996 interface uses malloc() for output vectors. See item 31 above.
3000 The output is an unsigned integer that contains the default limit of the number
3001 of match() calls in a pcre_exec() execution. See 42 above.
3003 44. pcretest has been upgraded by the addition of the -C option. This causes it
3004 to extract all the available output from the new pcre_config() function, and to
3005 output it. The program then exits immediately.
3007 45. A need has arisen to pass over additional data with calls to pcre_exec() in
3008 order to support additional features. One way would have been to define
3009 pcre_exec2() (for example) with extra arguments, but this would not have been
3010 extensible, and would also have required all calls to the original function to
3011 be mapped to the new one. Instead, I have chosen to extend the mechanism that
3012 is used for passing in "extra" data from pcre_study().
3014 The pcre_extra structure is now exposed and defined in pcre.h. It currently
3015 contains the following fields:
3017 flags a bitmap indicating which of the following fields are set
3018 study_data opaque data from pcre_study()
3019 match_limit a way of specifying a limit on match() calls for a specific
3020 call to pcre_exec()
3021 callout_data data for callouts (see 49 below)
3023 The flag bits are also defined in pcre.h, and are
3029 The pcre_study() function now returns one of these new pcre_extra blocks, with
3030 the actual study data pointed to by the study_data field, and the
3031 PCRE_EXTRA_STUDY_DATA flag set. This can be passed directly to pcre_exec() as
3032 before. That is, this change is entirely upwards-compatible and requires no
3033 change to existing code.
3035 If you want to pass in additional data to pcre_exec(), you can either place it
3036 in a pcre_extra block provided by pcre_study(), or create your own pcre_extra
3037 block.
3039 46. pcretest has been extended to test the PCRE_EXTRA_MATCH_LIMIT feature. If a
3040 data string contains the escape sequence \M, pcretest calls pcre_exec() several
3041 times with different match limits, until it finds the minimum value needed for
3042 pcre_exec() to complete. The value is then output. This can be instructive; for
3043 most simple matches the number is quite small, but for pathological cases it
3044 gets very large very quickly.
3046 47. There's a new option for pcre_fullinfo() called PCRE_INFO_STUDYSIZE. It
3047 returns the size of the data block pointed to by the study_data field in a
3048 pcre_extra block, that is, the value that was passed as the argument to
3049 pcre_malloc() when PCRE was getting memory in which to place the information
3050 created by pcre_study(). The fourth argument should point to a size_t variable.
3051 pcretest has been extended so that this information is shown after a successful
3052 pcre_study() call when information about the compiled regex is being displayed.
3054 48. Cosmetic change to Makefile: there's no need to have / after $(DESTDIR)
3055 because what follows is always an absolute path. (Later: it turns out that this
3056 is more than cosmetic for MinGW, because it doesn't like empty path
3057 components.)
3059 49. Some changes have been made to the callout feature (see 28 above):
3061 (i) A callout function now has three choices for what it returns:
3063 0 => success, carry on matching
3064 > 0 => failure at this point, but backtrack if possible
3065 < 0 => serious error, return this value from pcre_exec()
3067 Negative values should normally be chosen from the set of PCRE_ERROR_xxx
3068 values. In particular, returning PCRE_ERROR_NOMATCH forces a standard
3069 "match failed" error. The error number PCRE_ERROR_CALLOUT is reserved for
3070 use by callout functions. It will never be used by PCRE itself.
3072 (ii) The pcre_extra structure (see 45 above) has a void * field called
3073 callout_data, with corresponding flag bit PCRE_EXTRA_CALLOUT_DATA. The
3074 pcre_callout_block structure has a field of the same name. The contents of
3075 the field passed in the pcre_extra structure are passed to the callout
3076 function in the corresponding field in the callout block. This makes it
3077 easier to use the same callout-containing regex from multiple threads. For
3078 testing, the pcretest program has a new data escape
3080 \C*n pass the number n (may be negative) as callout_data
3082 If the callout function in pcretest receives a non-zero value as
3083 callout_data, it returns that value.
3085 50. Makefile wasn't handling CFLAGS properly when compiling dftables. Also,
3086 there were some redundant $(CFLAGS) in commands that are now specified as
3087 $(LINK), which already includes $(CFLAGS).
3089 51. Extensions to UTF-8 support are listed below. These all apply when (a) PCRE
3090 has been compiled with UTF-8 support *and* pcre_compile() has been compiled
3091 with the PCRE_UTF8 flag. Patterns that are compiled without that flag assume
3092 one-byte characters throughout. Note that case-insensitive matching applies
3093 only to characters whose values are less than 256. PCRE doesn't support the
3094 notion of cases for higher-valued characters.
3096 (i) A character class whose characters are all within 0-255 is handled as
3097 a bit map, and the map is inverted for negative classes. Previously, a
3098 character > 255 always failed to match such a class; however it should
3099 match if the class was a negative one (e.g. [^ab]). This has been fixed.
3101 (ii) A negated character class with a single character < 255 is coded as
3102 "not this character" (OP_NOT). This wasn't working properly when the test
3103 character was multibyte, either singly or repeated.
3105 (iii) Repeats of multibyte characters are now handled correctly in UTF-8
3106 mode, for example: \x{100}{2,3}.
3108 (iv) The character escapes \b, \B, \d, \D, \s, \S, \w, and \W (either
3109 singly or repeated) now correctly test multibyte characters. However,
3110 PCRE doesn't recognize any characters with values greater than 255 as
3111 digits, spaces, or word characters. Such characters always match \D, \S,
3112 and \W, and never match \d, \s, or \w.
3114 (v) Classes may now contain characters and character ranges with values
3115 greater than 255. For example: [ab\x{100}-\x{400}].
3117 (vi) pcregrep now has a --utf-8 option (synonym -u) which makes it call
3118 PCRE in UTF-8 mode.
3120 52. The info request value PCRE_INFO_FIRSTCHAR has been renamed
3121 PCRE_INFO_FIRSTBYTE because it is a byte value. However, the old name is
3122 retained for backwards compatibility. (Note that LASTLITERAL is also a byte
3123 value.)
3125 53. The single man page has become too large. I have therefore split it up into
3126 a number of separate man pages. These also give rise to individual HTML pages;
3127 these are now put in a separate directory, and there is an index.html page that
3128 lists them all. Some hyperlinking between the pages has been installed.
3130 54. Added convenience functions for handling named capturing parentheses.
3132 55. Unknown escapes inside character classes (e.g. [\M]) and escapes that
3133 aren't interpreted therein (e.g. [\C]) are literals in Perl. This is now also
3134 true in PCRE, except when the PCRE_EXTENDED option is set, in which case they
3135 are faulted.
3137 56. Introduced HOST_CC and HOST_CFLAGS which can be set in the environment when
3138 calling configure. These values are used when compiling the dftables.c program
3139 which is run to generate the source of the default character tables. They
3140 default to the values of CC and CFLAGS. If you are cross-compiling PCRE,
3141 you will need to set these values.
3143 57. Updated the building process for Windows DLL, as provided by Fred Cox.
3146 Version 3.9 02-Jan-02
3147 ---------------------
3149 1. A bit of extraneous text had somehow crept into the pcregrep documentation.
3151 2. If --disable-static was given, the building process failed when trying to
3152 build pcretest and pcregrep. (For some reason it was using libtool to compile
3153 them, which is not right, as they aren't part of the library.)
3156 Version 3.8 18-Dec-01
3157 ---------------------
3159 1. The experimental UTF-8 code was completely screwed up. It was packing the
3160 bytes in the wrong order. How dumb can you get?
3163 Version 3.7 29-Oct-01
3164 ---------------------
3166 1. In updating pcretest to check change 1 of version 3.6, I screwed up.
3167 This caused pcretest, when used on the test data, to segfault. Unfortunately,
3168 this didn't happen under Solaris 8, where I normally test things.
3170 2. The Makefile had to be changed to make it work on BSD systems, where 'make'
3171 doesn't seem to recognize that ./xxx and xxx are the same file. (This entry
3172 isn't in ChangeLog distributed with 3.7 because I forgot when I hastily made
3173 this fix an hour or so after the initial 3.7 release.)
3176 Version 3.6 23-Oct-01
3177 ---------------------
3179 1. Crashed with /(sens|respons)e and \1ibility/ and "sense and sensibility" if
3180 offsets passed as NULL with zero offset count.
3182 2. The config.guess and config.sub files had not been updated when I moved to
3183 the latest autoconf.
3186 Version 3.5 15-Aug-01
3187 ---------------------
3189 1. Added some missing #if !defined NOPOSIX conditionals in pcretest.c that
3190 had been forgotten.
3192 2. By using declared but undefined structures, we can avoid using "void"
3193 definitions in pcre.h while keeping the internal definitions of the structures
3194 private.
3196 3. The distribution is now built using autoconf 2.50 and libtool 1.4. From a
3197 user point of view, this means that both static and shared libraries are built
3198 by default, but this can be individually controlled. More of the work of
3199 handling this static/shared cases is now inside libtool instead of PCRE's make
3200 file.
3202 4. The pcretest utility is now installed along with pcregrep because it is
3203 useful for users (to test regexs) and by doing this, it automatically gets
3204 relinked by libtool. The documentation has been turned into a man page, so
3205 there are now .1, .txt, and .html versions in /doc.
3207 5. Upgrades to pcregrep:
3208 (i) Added long-form option names like gnu grep.
3209 (ii) Added --help to list all options with an explanatory phrase.
3210 (iii) Added -r, --recursive to recurse into sub-directories.
3211 (iv) Added -f, --file to read patterns from a file.
3213 6. pcre_exec() was referring to its "code" argument before testing that
3214 argument for NULL (and giving an error if it was NULL).
3216 7. Upgraded Makefile.in to allow for compiling in a different directory from
3217 the source directory.
3219 8. Tiny buglet in pcretest: when pcre_fullinfo() was called to retrieve the
3220 options bits, the pointer it was passed was to an int instead of to an unsigned
3221 long int. This mattered only on 64-bit systems.
3223 9. Fixed typo (3.4/1) in pcre.h again. Sigh. I had changed pcre.h (which is
3224 generated) instead of pcre.in, which it its source. Also made the same change
3225 in several of the .c files.
3227 10. A new release of gcc defines printf() as a macro, which broke pcretest
3228 because it had an ifdef in the middle of a string argument for printf(). Fixed
3229 by using separate calls to printf().
3231 11. Added --enable-newline-is-cr and --enable-newline-is-lf to the configure
3232 script, to force use of CR or LF instead of \n in the source. On non-Unix
3233 systems, the value can be set in config.h.
3235 12. The limit of 200 on non-capturing parentheses is a _nesting_ limit, not an
3236 absolute limit. Changed the text of the error message to make this clear, and
3237 likewise updated the man page.
3239 13. The limit of 99 on the number of capturing subpatterns has been removed.
3240 The new limit is 65535, which I hope will not be a "real" limit.
3243 Version 3.4 22-Aug-00
3244 ---------------------
3246 1. Fixed typo in pcre.h: unsigned const char * changed to const unsigned char *.
3248 2. Diagnose condition (?(0) as an error instead of crashing on matching.
3251 Version 3.3 01-Aug-00
3252 ---------------------
3254 1. If an octal character was given, but the value was greater than \377, it
3255 was not getting masked to the least significant bits, as documented. This could
3256 lead to crashes in some systems.
3258 2. Perl 5.6 (if not earlier versions) accepts classes like [a-\d] and treats
3259 the hyphen as a literal. PCRE used to give an error; it now behaves like Perl.
3261 3. Added the functions pcre_free_substring() and pcre_free_substring_list().
3262 These just pass their arguments on to (pcre_free)(), but they are provided
3263 because some uses of PCRE bind it to non-C systems that can call its functions,
3264 but cannot call free() or pcre_free() directly.
3266 4. Add "make test" as a synonym for "make check". Corrected some comments in
3267 the Makefile.
3269 5. Add $(DESTDIR)/ in front of all the paths in the "install" target in the
3270 Makefile.
3272 6. Changed the name of pgrep to pcregrep, because Solaris has introduced a
3273 command called pgrep for grepping around the active processes.
3275 7. Added the beginnings of support for UTF-8 character strings.
3277 8. Arranged for the Makefile to pass over the settings of CC, CFLAGS, and
3278 RANLIB to ./ltconfig so that they are used by libtool. I think these are all
3279 the relevant ones. (AR is not passed because ./ltconfig does its own figuring
3280 out for the ar command.)
3283 Version 3.2 12-May-00
3284 ---------------------
3286 This is purely a bug fixing release.
3288 1. If the pattern /((Z)+|A)*/ was matched agained ZABCDEFG it matched Z instead
3289 of ZA. This was just one example of several cases that could provoke this bug,
3290 which was introduced by change 9 of version 2.00. The code for breaking
3291 infinite loops after an iteration that matches an empty string was't working
3292 correctly.
3294 2. The pcretest program was not imitating Perl correctly for the pattern /a*/g
3295 when matched against abbab (for example). After matching an empty string, it
3296 wasn't forcing anchoring when setting PCRE_NOTEMPTY for the next attempt; this
3297 caused it to match further down the string than it should.
3299 3. The code contained an inclusion of sys/types.h. It isn't clear why this
3300 was there because it doesn't seem to be needed, and it causes trouble on some
3301 systems, as it is not a Standard C header. It has been removed.
3303 4. Made 4 silly changes to the source to avoid stupid compiler warnings that
3304 were reported on the Macintosh. The changes were from
3306 while ((c = *(++ptr)) != 0 && c != '\n');
3307 to
3308 while ((c = *(++ptr)) != 0 && c != '\n') ;
3310 Totally extraordinary, but if that's what it takes...
3312 5. PCRE is being used in one environment where neither memmove() nor bcopy() is
3313 available. Added HAVE_BCOPY and an autoconf test for it; if neither
3314 HAVE_MEMMOVE nor HAVE_BCOPY is set, use a built-in emulation function which
3315 assumes the way PCRE uses memmove() (always moving upwards).
3317 6. PCRE is being used in one environment where strchr() is not available. There
3318 was only one use in pcre.c, and writing it out to avoid strchr() probably gives
3319 faster code anyway.
3322 Version 3.1 09-Feb-00
3323 ---------------------
3325 The only change in this release is the fixing of some bugs in Makefile.in for
3326 the "install" target:
3328 (1) It was failing to install pcreposix.h.
3330 (2) It was overwriting the pcre.3 man page with the pcreposix.3 man page.
3333 Version 3.0 01-Feb-00
3334 ---------------------
3336 1. Add support for the /+ modifier to perltest (to output $` like it does in
3337 pcretest).
3339 2. Add support for the /g modifier to perltest.
3341 3. Fix pcretest so that it behaves even more like Perl for /g when the pattern
3342 matches null strings.
3344 4. Fix perltest so that it doesn't do unwanted things when fed an empty
3345 pattern. Perl treats empty patterns specially - it reuses the most recent
3346 pattern, which is not what we want. Replace // by /(?#)/ in order to avoid this
3347 effect.
3349 5. The POSIX interface was broken in that it was just handing over the POSIX
3350 captured string vector to pcre_exec(), but (since release 2.00) PCRE has
3351 required a bigger vector, with some working space on the end. This means that
3352 the POSIX wrapper now has to get and free some memory, and copy the results.
3354 6. Added some simple autoconf support, placing the test data and the
3355 documentation in separate directories, re-organizing some of the
3356 information files, and making it build pcre-config (a GNU standard). Also added
3357 libtool support for building PCRE as a shared library, which is now the
3358 default.
3360 7. Got rid of the leading zero in the definition of PCRE_MINOR because 08 and
3361 09 are not valid octal constants. Single digits will be used for minor values
3362 less than 10.
3364 8. Defined REG_EXTENDED and REG_NOSUB as zero in the POSIX header, so that
3365 existing programs that set these in the POSIX interface can use PCRE without
3366 modification.
3368 9. Added a new function, pcre_fullinfo() with an extensible interface. It can
3369 return all that pcre_info() returns, plus additional data. The pcre_info()
3370 function is retained for compatibility, but is considered to be obsolete.
3372 10. Added experimental recursion feature (?R) to handle one common case that
3373 Perl 5.6 will be able to do with (?p{...}).
3375 11. Added support for POSIX character classes like [:alpha:], which Perl is
3376 adopting.
3379 Version 2.08 31-Aug-99
3380 ----------------------
3382 1. When startoffset was not zero and the pattern began with ".*", PCRE was not
3383 trying to match at the startoffset position, but instead was moving forward to
3384 the next newline as if a previous match had failed.
3386 2. pcretest was not making use of PCRE_NOTEMPTY when repeating for /g and /G,
3387 and could get into a loop if a null string was matched other than at the start
3388 of the subject.
3390 3. Added definitions of PCRE_MAJOR and PCRE_MINOR to pcre.h so the version can
3391 be distinguished at compile time, and for completeness also added PCRE_DATE.
3393 5. Added Paul Sokolovsky's minor changes to make it easy to compile a Win32 DLL
3394 in GnuWin32 environments.
3397 Version 2.07 29-Jul-99
3398 ----------------------
3400 1. The documentation is now supplied in plain text form and HTML as well as in
3401 the form of man page sources.
3403 2. C++ compilers don't like assigning (void *) values to other pointer types.
3404 In particular this affects malloc(). Although there is no problem in Standard
3405 C, I've put in casts to keep C++ compilers happy.
3407 3. Typo on pcretest.c; a cast of (unsigned char *) in the POSIX regexec() call
3408 should be (const char *).
3410 4. If NOPOSIX is defined, pcretest.c compiles without POSIX support. This may
3411 be useful for non-Unix systems who don't want to bother with the POSIX stuff.
3412 However, I haven't made this a standard facility. The documentation doesn't
3413 mention it, and the Makefile doesn't support it.
3415 5. The Makefile now contains an "install" target, with editable destinations at
3416 the top of the file. The pcretest program is not installed.
3418 6. pgrep -V now gives the PCRE version number and date.
3420 7. Fixed bug: a zero repetition after a literal string (e.g. /abcde{0}/) was
3421 causing the entire string to be ignored, instead of just the last character.
3423 8. If a pattern like /"([^\\"]+|\\.)*"/ is applied in the normal way to a
3424 non-matching string, it can take a very, very long time, even for strings of
3425 quite modest length, because of the nested recursion. PCRE now does better in
3426 some of these cases. It does this by remembering the last required literal
3427 character in the pattern, and pre-searching the subject to ensure it is present
3428 before running the real match. In other words, it applies a heuristic to detect
3429 some types of certain failure quickly, and in the above example, if presented
3430 with a string that has no trailing " it gives "no match" very quickly.
3432 9. A new runtime option PCRE_NOTEMPTY causes null string matches to be ignored;
3433 other alternatives are tried instead.
3436 Version 2.06 09-Jun-99
3437 ----------------------
3439 1. Change pcretest's output for amount of store used to show just the code
3440 space, because the remainder (the data block) varies in size between 32-bit and
3441 64-bit systems.
3443 2. Added an extra argument to pcre_exec() to supply an offset in the subject to
3444 start matching at. This allows lookbehinds to work when searching for multiple
3445 occurrences in a string.
3447 3. Added additional options to pcretest for testing multiple occurrences:
3449 /+ outputs the rest of the string that follows a match
3450 /g loops for multiple occurrences, using the new startoffset argument
3451 /G loops for multiple occurrences by passing an incremented pointer
3453 4. PCRE wasn't doing the "first character" optimization for patterns starting
3454 with \b or \B, though it was doing it for other lookbehind assertions. That is,
3455 it wasn't noticing that a match for a pattern such as /\bxyz/ has to start with
3456 the letter 'x'. On long subject strings, this gives a significant speed-up.
3459 Version 2.05 21-Apr-99
3460 ----------------------
3462 1. Changed the type of magic_number from int to long int so that it works
3463 properly on 16-bit systems.
3465 2. Fixed a bug which caused patterns starting with .* not to work correctly
3466 when the subject string contained newline characters. PCRE was assuming
3467 anchoring for such patterns in all cases, which is not correct because .* will
3468 not pass a newline unless PCRE_DOTALL is set. It now assumes anchoring only if
3469 DOTALL is set at top level; otherwise it knows that patterns starting with .*
3470 must be retried after every newline in the subject.
3473 Version 2.04 18-Feb-99
3474 ----------------------
3476 1. For parenthesized subpatterns with repeats whose minimum was zero, the
3477 computation of the store needed to hold the pattern was incorrect (too large).
3478 If such patterns were nested a few deep, this could multiply and become a real
3479 problem.
3481 2. Added /M option to pcretest to show the memory requirement of a specific
3482 pattern. Made -m a synonym of -s (which does this globally) for compatibility.
3484 3. Subpatterns of the form (regex){n,m} (i.e. limited maximum) were being
3485 compiled in such a way that the backtracking after subsequent failure was
3486 pessimal. Something like (a){0,3} was compiled as (a)?(a)?(a)? instead of
3487 ((a)((a)(a)?)?)? with disastrous performance if the maximum was of any size.
3490 Version 2.03 02-Feb-99
3491 ----------------------
3493 1. Fixed typo and small mistake in man page.
3495 2. Added 4th condition (GPL supersedes if conflict) and created separate
3496 LICENCE file containing the conditions.
3498 3. Updated pcretest so that patterns such as /abc\/def/ work like they do in
3499 Perl, that is the internal \ allows the delimiter to be included in the
3500 pattern. Locked out the use of \ as a delimiter. If \ immediately follows
3501 the final delimiter, add \ to the end of the pattern (to test the error).
3503 4. Added the convenience functions for extracting substrings after a successful
3504 match. Updated pcretest to make it able to test these functions.
3507 Version 2.02 14-Jan-99
3508 ----------------------
3510 1. Initialized the working variables associated with each extraction so that
3511 their saving and restoring doesn't refer to uninitialized store.
3513 2. Put dummy code into study.c in order to trick the optimizer of the IBM C
3514 compiler for OS/2 into generating correct code. Apparently IBM isn't going to
3515 fix the problem.
3517 3. Pcretest: the timing code wasn't using LOOPREPEAT for timing execution
3518 calls, and wasn't printing the correct value for compiling calls. Increased the
3519 default value of LOOPREPEAT, and the number of significant figures in the
3520 times.
3522 4. Changed "/bin/rm" in the Makefile to "-rm" so it works on Windows NT.
3524 5. Renamed "deftables" as "dftables" to get it down to 8 characters, to avoid
3525 a building problem on Windows NT with a FAT file system.
3528 Version 2.01 21-Oct-98
3529 ----------------------
3531 1. Changed the API for pcre_compile() to allow for the provision of a pointer
3532 to character tables built by pcre_maketables() in the current locale. If NULL
3533 is passed, the default tables are used.
3536 Version 2.00 24-Sep-98
3537 ----------------------
3539 1. Since the (>?) facility is in Perl 5.005, don't require PCRE_EXTRA to enable
3540 it any more.
3542 2. Allow quantification of (?>) groups, and make it work correctly.
3544 3. The first character computation wasn't working for (?>) groups.
3546 4. Correct the implementation of \Z (it is permitted to match on the \n at the
3547 end of the subject) and add 5.005's \z, which really does match only at the
3548 very end of the subject.
3550 5. Remove the \X "cut" facility; Perl doesn't have it, and (?> is neater.
3552 6. Remove the ability to specify CASELESS, MULTILINE, DOTALL, and
3553 DOLLAR_END_ONLY at runtime, to make it possible to implement the Perl 5.005
3554 localized options. All options to pcre_study() were also removed.
3556 7. Add other new features from 5.005:
3558 $(?<= positive lookbehind
3559 $(?<! negative lookbehind
3560 (?imsx-imsx) added the unsetting capability
3561 such a setting is global if at outer level; local otherwise
3562 (?imsx-imsx:) non-capturing groups with option setting
3563 (?(cond)re|re) conditional pattern matching
3565 A backreference to itself in a repeated group matches the previous
3566 captured string.
3568 8. General tidying up of studying (both automatic and via "study")
3569 consequential on the addition of new assertions.
3571 9. As in 5.005, unlimited repeated groups that could match an empty substring
3572 are no longer faulted at compile time. Instead, the loop is forcibly broken at
3573 runtime if any iteration does actually match an empty substring.
3575 10. Include the RunTest script in the distribution.
3577 11. Added tests from the Perl 5.005_02 distribution. This showed up a few
3578 discrepancies, some of which were old and were also with respect to 5.004. They
3579 have now been fixed.
3582 Version 1.09 28-Apr-98
3583 ----------------------
3585 1. A negated single character class followed by a quantifier with a minimum
3586 value of one (e.g. [^x]{1,6} ) was not compiled correctly. This could lead to
3587 program crashes, or just wrong answers. This did not apply to negated classes
3588 containing more than one character, or to minima other than one.
3591 Version 1.08 27-Mar-98
3592 ----------------------
3594 1. Add PCRE_UNGREEDY to invert the greediness of quantifiers.
3596 2. Add (?U) and (?X) to set PCRE_UNGREEDY and PCRE_EXTRA respectively. The
3597 latter must appear before anything that relies on it in the pattern.
3600 Version 1.07 16-Feb-98
3601 ----------------------
3603 1. A pattern such as /((a)*)*/ was not being diagnosed as in error (unlimited
3604 repeat of a potentially empty string).
3607 Version 1.06 23-Jan-98
3608 ----------------------
3610 1. Added Markus Oberhumer's little patches for C++.
3612 2. Literal strings longer than 255 characters were broken.
3615 Version 1.05 23-Dec-97
3616 ----------------------
3618 1. Negated character classes containing more than one character were failing if
3619 PCRE_CASELESS was set at run time.
3622 Version 1.04 19-Dec-97
3623 ----------------------
3625 1. Corrected the man page, where some "const" qualifiers had been omitted.
3627 2. Made debugging output print "{0,xxx}" instead of just "{,xxx}" to agree with
3628 input syntax.
3630 3. Fixed memory leak which occurred when a regex with back references was
3631 matched with an offsets vector that wasn't big enough. The temporary memory
3632 that is used in this case wasn't being freed if the match failed.
3634 4. Tidied pcretest to ensure it frees memory that it gets.
3636 5. Temporary memory was being obtained in the case where the passed offsets
3637 vector was exactly big enough.
3639 6. Corrected definition of offsetof() from change 5 below.
3641 7. I had screwed up change 6 below and broken the rules for the use of
3642 setjmp(). Now fixed.
3645 Version 1.03 18-Dec-97
3646 ----------------------
3648 1. A erroneous regex with a missing opening parenthesis was correctly
3649 diagnosed, but PCRE attempted to access brastack[-1], which could cause crashes
3650 on some systems.
3652 2. Replaced offsetof(real_pcre, code) by offsetof(real_pcre, code[0]) because
3653 it was reported that one broken compiler failed on the former because "code" is
3654 also an independent variable.
3656 3. The erroneous regex a[]b caused an array overrun reference.
3658 4. A regex ending with a one-character negative class (e.g. /[^k]$/) did not
3659 fail on data ending with that character. (It was going on too far, and checking
3660 the next character, typically a binary zero.) This was specific to the
3661 optimized code for single-character negative classes.
3663 5. Added a contributed patch from the TIN world which does the following:
3665 + Add an undef for memmove, in case the the system defines a macro for it.
3667 + Add a definition of offsetof(), in case there isn't one. (I don't know
3668 the reason behind this - offsetof() is part of the ANSI standard - but
3669 it does no harm).
3671 + Reduce the ifdef's in pcre.c using macro DPRINTF, thereby eliminating
3672 most of the places where whitespace preceded '#'. I have given up and
3673 allowed the remaining 2 cases to be at the margin.
3675 + Rename some variables in pcre to eliminate shadowing. This seems very
3676 pedantic, but does no harm, of course.
3678 6. Moved the call to setjmp() into its own function, to get rid of warnings
3679 from gcc -Wall, and avoided calling it at all unless PCRE_EXTRA is used.
3681 7. Constructs such as \d{8,} were compiling into the equivalent of
3682 \d{8}\d{0,65527} instead of \d{8}\d* which didn't make much difference to the
3683 outcome, but in this particular case used more store than had been allocated,
3684 which caused the bug to be discovered because it threw up an internal error.
3686 8. The debugging code in both pcre and pcretest for outputting the compiled
3687 form of a regex was going wrong in the case of back references followed by
3688 curly-bracketed repeats.
3691 Version 1.02 12-Dec-97
3692 ----------------------
3694 1. Typos in pcre.3 and comments in the source fixed.
3696 2. Applied a contributed patch to get rid of places where it used to remove
3697 'const' from variables, and fixed some signed/unsigned and uninitialized
3698 variable warnings.
3700 3. Added the "runtest" target to Makefile.
3702 4. Set default compiler flag to -O2 rather than just -O.
3705 Version 1.01 19-Nov-97
3706 ----------------------
3708 1. PCRE was failing to diagnose unlimited repeat of empty string for patterns
3709 like /([ab]*)*/, that is, for classes with more than one character in them.
3711 2. Likewise, it wasn't diagnosing patterns with "once-only" subpatterns, such
3712 as /((?>a*))*/ (a PCRE_EXTRA facility).
3715 Version 1.00 18-Nov-97
3716 ----------------------
3718 1. Added compile-time macros to support systems such as SunOS4 which don't have
3719 memmove() or strerror() but have other things that can be used instead.
3721 2. Arranged that "make clean" removes the executables.
3724 Version 0.99 27-Oct-97
3725 ----------------------
3727 1. Fixed bug in code for optimizing classes with only one character. It was
3728 initializing a 32-byte map regardless, which could cause it to run off the end
3729 of the memory it had got.
3731 2. Added, conditional on PCRE_EXTRA, the proposed (?>REGEX) construction.
3734 Version 0.98 22-Oct-97
3735 ----------------------
3737 1. Fixed bug in code for handling temporary memory usage when there are more
3738 back references than supplied space in the ovector. This could cause segfaults.
3741 Version 0.97 21-Oct-97
3742 ----------------------
3744 1. Added the \X "cut" facility, conditional on PCRE_EXTRA.
3746 2. Optimized negated single characters not to use a bit map.
3748 3. Brought error texts together as macro definitions; clarified some of them;
3749 fixed one that was wrong - it said "range out of order" when it meant "invalid
3750 escape sequence".
3752 4. Changed some char * arguments to const char *.
3754 5. Added PCRE_NOTBOL and PCRE_NOTEOL (from POSIX).
3756 6. Added the POSIX-style API wrapper in pcreposix.a and testing facilities in
3757 pcretest.
3760 Version 0.96 16-Oct-97
3761 ----------------------
3763 1. Added a simple "pgrep" utility to the distribution.
3765 2. Fixed an incompatibility with Perl: "{" is now treated as a normal character
3766 unless it appears in one of the precise forms "{ddd}", "{ddd,}", or "{ddd,ddd}"
3767 where "ddd" means "one or more decimal digits".
3769 3. Fixed serious bug. If a pattern had a back reference, but the call to
3770 pcre_exec() didn't supply a large enough ovector to record the related
3771 identifying subpattern, the match always failed. PCRE now remembers the number
3772 of the largest back reference, and gets some temporary memory in which to save
3773 the offsets during matching if necessary, in order to ensure that
3774 backreferences always work.
3776 4. Increased the compatibility with Perl in a number of ways:
3778 (a) . no longer matches \n by default; an option PCRE_DOTALL is provided
3779 to request this handling. The option can be set at compile or exec time.
3781 (b) $ matches before a terminating newline by default; an option
3782 PCRE_DOLLAR_ENDONLY is provided to override this (but not in multiline
3783 mode). The option can be set at compile or exec time.
3785 (c) The handling of \ followed by a digit other than 0 is now supposed to be
3786 the same as Perl's. If the decimal number it represents is less than 10
3787 or there aren't that many previous left capturing parentheses, an octal
3788 escape is read. Inside a character class, it's always an octal escape,
3789 even if it is a single digit.
3791 (d) An escaped but undefined alphabetic character is taken as a literal,
3792 unless PCRE_EXTRA is set. Currently this just reserves the remaining
3793 escapes.
3795 (e) {0} is now permitted. (The previous item is removed from the compiled
3796 pattern).
3798 5. Changed all the names of code files so that the basic parts are no longer
3799 than 10 characters, and abolished the teeny "globals.c" file.
3801 6. Changed the handling of character classes; they are now done with a 32-byte
3802 bit map always.
3804 7. Added the -d and /D options to pcretest to make it possible to look at the
3805 internals of compilation without having to recompile pcre.
3808 Version 0.95 23-Sep-97
3809 ----------------------
3811 1. Fixed bug in pre-pass concerning escaped "normal" characters such as \x5c or
3812 \x20 at the start of a run of normal characters. These were being treated as
3813 real characters, instead of the source characters being re-checked.
3816 Version 0.94 18-Sep-97
3817 ----------------------
3819 1. The functions are now thread-safe, with the caveat that the global variables
3820 containing pointers to malloc() and free() or alternative functions are the
3821 same for all threads.
3823 2. Get pcre_study() to generate a bitmap of initial characters for non-
3824 anchored patterns when this is possible, and use it if passed to pcre_exec().
3827 Version 0.93 15-Sep-97
3828 ----------------------
3830 1. /(b)|(:+)/ was computing an incorrect first character.
3832 2. Add pcre_study() to the API and the passing of pcre_extra to pcre_exec(),
3833 but not actually doing anything yet.
3835 3. Treat "-" characters in classes that cannot be part of ranges as literals,
3836 as Perl does (e.g. [-az] or [az-]).
3838 4. Set the anchored flag if a branch starts with .* or .*? because that tests
3839 all possible positions.
3841 5. Split up into different modules to avoid including unneeded functions in a
3842 compiled binary. However, compile and exec are still in one module. The "study"
3843 function is split off.
3845 6. The character tables are now in a separate module whose source is generated
3846 by an auxiliary program - but can then be edited by hand if required. There are
3847 now no calls to isalnum(), isspace(), isdigit(), isxdigit(), tolower() or
3848 toupper() in the code.
3850 7. Turn the malloc/free funtions variables into pcre_malloc and pcre_free and
3851 make them global. Abolish the function for setting them, as the caller can now
3852 set them directly.
3855 Version 0.92 11-Sep-97
3856 ----------------------
3858 1. A repeat with a fixed maximum and a minimum of 1 for an ordinary character
3859 (e.g. /a{1,3}/) was broken (I mis-optimized it).
3861 2. Caseless matching was not working in character classes if the characters in
3862 the pattern were in upper case.
3864 3. Make ranges like [W-c] work in the same way as Perl for caseless matching.
3866 4. Make PCRE_ANCHORED public and accept as a compile option.
3868 5. Add an options word to pcre_exec() and accept PCRE_ANCHORED and
3869 PCRE_CASELESS at run time. Add escapes \A and \I to pcretest to cause it to
3870 pass them.
3872 6. Give an error if bad option bits passed at compile or run time.
3874 7. Add PCRE_MULTILINE at compile and exec time, and (?m) as well. Add \M to
3875 pcretest to cause it to pass that flag.
3877 8. Add pcre_info(), to get the number of identifying subpatterns, the stored
3878 options, and the first character, if set.
3880 9. Recognize C+ or C{n,m} where n >= 1 as providing a fixed starting character.
3883 Version 0.91 10-Sep-97
3884 ----------------------
3886 1. PCRE was failing to diagnose unlimited repeats of subpatterns that could
3887 match the empty string as in /(a*)*/. It was looping and ultimately crashing.
3889 2. PCRE was looping on encountering an indefinitely repeated back reference to
3890 a subpattern that had matched an empty string, e.g. /(a|)\1*/. It now does what
3891 Perl does - treats the match as successful.
3893 ****


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

  ViewVC Help
Powered by ViewVC 1.1.5