ViewVC logotype

Contents of /code/trunk/ChangeLog

Parent Directory Parent Directory | Revision Log Revision Log

Revision 492 - (show annotations)
Tue Mar 2 11:26:55 2010 UTC (11 years, 6 months ago) by ph10
File size: 181751 byte(s)
Added --libs-cpp to pcre-config, but only when C++ support is configured.
1 ChangeLog for PCRE
2 ------------------
4 Version 8.02 01-Mar-2010
5 ------------------------
7 1. The Unicode data tables have been updated to Unicode 5.2.0.
9 2. Added the option --libs-cpp to pcre-config, but only when C++ support is
10 configured.
13 Version 8.01 19-Jan-2010
14 ------------------------
16 1. If a pattern contained a conditional subpattern with only one branch (in
17 particular, this includes all (*DEFINE) patterns), a call to pcre_study()
18 computed the wrong minimum data length (which is of course zero for such
19 subpatterns). This could cause incorrect "no match" results.
21 2. For patterns such as (?i)a(?-i)b|c where an option setting at the start of
22 the pattern is reset in the first branch, pcre_compile() failed with
23 "internal error: code overflow at offset...". This happened only when
24 the reset was to the original external option setting. (An optimization
25 abstracts leading options settings into an external setting, which was the
26 cause of this.)
28 3. A pattern such as ^(?!a(*SKIP)b) where a negative assertion contained one
29 of the verbs SKIP, PRUNE, or COMMIT, did not work correctly. When the
30 assertion pattern did not match (meaning that the assertion was true), it
31 was incorrectly treated as false if the SKIP had been reached during the
32 matching. This also applied to assertions used as conditions.
34 4. If an item that is not supported by pcre_dfa_exec() was encountered in an
35 assertion subpattern, including such a pattern used as a condition,
36 unpredictable results occurred, instead of the error return
39 5. The C++ GlobalReplace function was not working like Perl for the special
40 situation when an empty string is matched. It now does the fancy magic
41 stuff that is necessary.
43 6. In pcre_internal.h, obsolete includes to setjmp.h and stdarg.h have been
44 removed. (These were left over from very, very early versions of PCRE.)
46 7. Some cosmetic changes to the code to make life easier when compiling it
47 as part of something else:
49 (a) Change DEBUG to PCRE_DEBUG.
51 (b) In pcre_compile(), rename the member of the "branch_chain" structure
52 called "current" as "current_branch", to prevent a collision with the
53 Linux macro when compiled as a kernel module.
55 (c) In pcre_study(), rename the function set_bit() as set_table_bit(), to
56 prevent a collision with the Linux macro when compiled as a kernel
57 module.
59 8. In pcre_compile() there are some checks for integer overflows that used to
60 cast potentially large values to (double). This has been changed to that
61 when building, a check for int64_t is made, and if it is found, it is used
62 instead, thus avoiding the use of floating point arithmetic. (There is no
63 other use of FP in PCRE.) If int64_t is not found, the fallback is to
64 double.
66 9. Added two casts to avoid signed/unsigned warnings from VS Studio Express
67 2005 (difference between two addresses compared to an unsigned value).
69 10. Change the standard AC_CHECK_LIB test for libbz2 in configure.ac to a
70 custom one, because of the following reported problem in Windows:
72 - libbz2 uses the Pascal calling convention (WINAPI) for the functions
73 under Win32.
74 - The standard autoconf AC_CHECK_LIB fails to include "bzlib.h",
75 therefore missing the function definition.
76 - The compiler thus generates a "C" signature for the test function.
77 - The linker fails to find the "C" function.
78 - PCRE fails to configure if asked to do so against libbz2.
80 11. When running libtoolize from libtool-2.2.6b as part of autogen.sh, these
81 messages were output:
83 Consider adding `AC_CONFIG_MACRO_DIR([m4])' to configure.ac and
84 rerunning libtoolize, to keep the correct libtool macros in-tree.
85 Consider adding `-I m4' to ACLOCAL_AMFLAGS in Makefile.am.
87 I have done both of these things.
89 12. Although pcre_dfa_exec() does not use nearly as much stack as pcre_exec()
90 most of the time, it *can* run out if it is given a pattern that contains a
91 runaway infinite recursion. I updated the discussion in the pcrestack man
92 page.
94 13. Now that we have gone to the x.xx style of version numbers, the minor
95 version may start with zero. Using 08 or 09 is a bad idea because users
96 might check the value of PCRE_MINOR in their code, and 08 or 09 may be
97 interpreted as invalid octal numbers. I've updated the previous comment in
98 configure.ac, and also added a check that gives an error if 08 or 09 are
99 used.
101 14. Change 8.00/11 was not quite complete: code had been accidentally omitted,
102 causing partial matching to fail when the end of the subject matched \W
103 in a UTF-8 pattern where \W was quantified with a minimum of 3.
105 15. There were some discrepancies between the declarations in pcre_internal.h
106 of _pcre_is_newline(), _pcre_was_newline(), and _pcre_valid_utf8() and
107 their definitions. The declarations used "const uschar *" and the
108 definitions used USPTR. Even though USPTR is normally defined as "const
109 unsigned char *" (and uschar is typedeffed as "unsigned char"), it was
110 reported that: "This difference in casting confuses some C++ compilers, for
111 example, SunCC recognizes above declarations as different functions and
112 generates broken code for hbpcre." I have changed the declarations to use
113 USPTR.
115 16. GNU libtool is named differently on some systems. The autogen.sh script now
116 tries several variants such as glibtoolize (MacOSX) and libtoolize1x
117 (FreeBSD).
119 17. Applied Craig's patch that fixes an HP aCC compile error in pcre 8.00
120 (strtoXX undefined when compiling pcrecpp.cc). The patch contains this
121 comment: "Figure out how to create a longlong from a string: strtoll and
122 equivalent. It's not enough to call AC_CHECK_FUNCS: hpux has a strtoll, for
123 instance, but it only takes 2 args instead of 3!"
125 18. A subtle bug concerned with back references has been fixed by a change of
126 specification, with a corresponding code fix. A pattern such as
127 ^(xa|=?\1a)+$ which contains a back reference inside the group to which it
128 refers, was giving matches when it shouldn't. For example, xa=xaaa would
129 match that pattern. Interestingly, Perl (at least up to 5.11.3) has the
130 same bug. Such groups have to be quantified to be useful, or contained
131 inside another quantified group. (If there's no repetition, the reference
132 can never match.) The problem arises because, having left the group and
133 moved on to the rest of the pattern, a later failure that backtracks into
134 the group uses the captured value from the final iteration of the group
135 rather than the correct earlier one. I have fixed this in PCRE by forcing
136 any group that contains a reference to itself to be an atomic group; that
137 is, there cannot be any backtracking into it once it has completed. This is
138 similar to recursive and subroutine calls.
141 Version 8.00 19-Oct-09
142 ----------------------
144 1. The table for translating pcre_compile() error codes into POSIX error codes
145 was out-of-date, and there was no check on the pcre_compile() error code
146 being within the table. This could lead to an OK return being given in
147 error.
149 2. Changed the call to open a subject file in pcregrep from fopen(pathname,
150 "r") to fopen(pathname, "rb"), which fixed a problem with some of the tests
151 in a Windows environment.
153 3. The pcregrep --count option prints the count for each file even when it is
154 zero, as does GNU grep. However, pcregrep was also printing all files when
155 --files-with-matches was added. Now, when both options are given, it prints
156 counts only for those files that have at least one match. (GNU grep just
157 prints the file name in this circumstance, but including the count seems
158 more useful - otherwise, why use --count?) Also ensured that the
159 combination -clh just lists non-zero counts, with no names.
161 4. The long form of the pcregrep -F option was incorrectly implemented as
162 --fixed_strings instead of --fixed-strings. This is an incompatible change,
163 but it seems right to fix it, and I didn't think it was worth preserving
164 the old behaviour.
166 5. The command line items --regex=pattern and --regexp=pattern were not
167 recognized by pcregrep, which required --regex pattern or --regexp pattern
168 (with a space rather than an '='). The man page documented the '=' forms,
169 which are compatible with GNU grep; these now work.
171 6. No libpcreposix.pc file was created for pkg-config; there was just
172 libpcre.pc and libpcrecpp.pc. The omission has been rectified.
174 7. Added #ifndef SUPPORT_UCP into the pcre_ucd.c module, to reduce its size
175 when UCP support is not needed, by modifying the Python script that
176 generates it from Unicode data files. This should not matter if the module
177 is correctly used as a library, but I received one complaint about 50K of
178 unwanted data. My guess is that the person linked everything into his
179 program rather than using a library. Anyway, it does no harm.
181 8. A pattern such as /\x{123}{2,2}+/8 was incorrectly compiled; the trigger
182 was a minimum greater than 1 for a wide character in a possessive
183 repetition. The same bug could also affect patterns like /(\x{ff}{0,2})*/8
184 which had an unlimited repeat of a nested, fixed maximum repeat of a wide
185 character. Chaos in the form of incorrect output or a compiling loop could
186 result.
188 9. The restrictions on what a pattern can contain when partial matching is
189 requested for pcre_exec() have been removed. All patterns can now be
190 partially matched by this function. In addition, if there are at least two
191 slots in the offset vector, the offset of the earliest inspected character
192 for the match and the offset of the end of the subject are set in them when
193 PCRE_ERROR_PARTIAL is returned.
195 10. Partial matching has been split into two forms: PCRE_PARTIAL_SOFT, which is
196 synonymous with PCRE_PARTIAL, for backwards compatibility, and
197 PCRE_PARTIAL_HARD, which causes a partial match to supersede a full match,
198 and may be more useful for multi-segment matching.
200 11. Partial matching with pcre_exec() is now more intuitive. A partial match
201 used to be given if ever the end of the subject was reached; now it is
202 given only if matching could not proceed because another character was
203 needed. This makes a difference in some odd cases such as Z(*FAIL) with the
204 string "Z", which now yields "no match" instead of "partial match". In the
205 case of pcre_dfa_exec(), "no match" is given if every matching path for the
206 final character ended with (*FAIL).
208 12. Restarting a match using pcre_dfa_exec() after a partial match did not work
209 if the pattern had a "must contain" character that was already found in the
210 earlier partial match, unless partial matching was again requested. For
211 example, with the pattern /dog.(body)?/, the "must contain" character is
212 "g". If the first part-match was for the string "dog", restarting with
213 "sbody" failed. This bug has been fixed.
215 13. The string returned by pcre_dfa_exec() after a partial match has been
216 changed so that it starts at the first inspected character rather than the
217 first character of the match. This makes a difference only if the pattern
218 starts with a lookbehind assertion or \b or \B (\K is not supported by
219 pcre_dfa_exec()). It's an incompatible change, but it makes the two
220 matching functions compatible, and I think it's the right thing to do.
222 14. Added a pcredemo man page, created automatically from the pcredemo.c file,
223 so that the demonstration program is easily available in environments where
224 PCRE has not been installed from source.
226 15. Arranged to add -DPCRE_STATIC to cflags in libpcre.pc, libpcreposix.cp,
227 libpcrecpp.pc and pcre-config when PCRE is not compiled as a shared
228 library.
230 16. Added REG_UNGREEDY to the pcreposix interface, at the request of a user.
231 It maps to PCRE_UNGREEDY. It is not, of course, POSIX-compatible, but it
232 is not the first non-POSIX option to be added. Clearly some people find
233 these options useful.
235 17. If a caller to the POSIX matching function regexec() passes a non-zero
236 value for nmatch with a NULL value for pmatch, the value of
237 nmatch is forced to zero.
239 18. RunGrepTest did not have a test for the availability of the -u option of
240 the diff command, as RunTest does. It now checks in the same way as
241 RunTest, and also checks for the -b option.
243 19. If an odd number of negated classes containing just a single character
244 interposed, within parentheses, between a forward reference to a named
245 subpattern and the definition of the subpattern, compilation crashed with
246 an internal error, complaining that it could not find the referenced
247 subpattern. An example of a crashing pattern is /(?&A)(([^m])(?<A>))/.
248 [The bug was that it was starting one character too far in when skipping
249 over the character class, thus treating the ] as data rather than
250 terminating the class. This meant it could skip too much.]
252 20. Added PCRE_NOTEMPTY_ATSTART in order to be able to correctly implement the
253 /g option in pcretest when the pattern contains \K, which makes it possible
254 to have an empty string match not at the start, even when the pattern is
255 anchored. Updated pcretest and pcredemo to use this option.
257 21. If the maximum number of capturing subpatterns in a recursion was greater
258 than the maximum at the outer level, the higher number was returned, but
259 with unset values at the outer level. The correct (outer level) value is
260 now given.
262 22. If (*ACCEPT) appeared inside capturing parentheses, previous releases of
263 PCRE did not set those parentheses (unlike Perl). I have now found a way to
264 make it do so. The string so far is captured, making this feature
265 compatible with Perl.
267 23. The tests have been re-organized, adding tests 11 and 12, to make it
268 possible to check the Perl 5.10 features against Perl 5.10.
270 24. Perl 5.10 allows subroutine calls in lookbehinds, as long as the subroutine
271 pattern matches a fixed length string. PCRE did not allow this; now it
272 does. Neither allows recursion.
274 25. I finally figured out how to implement a request to provide the minimum
275 length of subject string that was needed in order to match a given pattern.
276 (It was back references and recursion that I had previously got hung up
277 on.) This code has now been added to pcre_study(); it finds a lower bound
278 to the length of subject needed. It is not necessarily the greatest lower
279 bound, but using it to avoid searching strings that are too short does give
280 some useful speed-ups. The value is available to calling programs via
281 pcre_fullinfo().
283 26. While implementing 25, I discovered to my embarrassment that pcretest had
284 not been passing the result of pcre_study() to pcre_dfa_exec(), so the
285 study optimizations had never been tested with that matching function.
286 Oops. What is worse, even when it was passed study data, there was a bug in
287 pcre_dfa_exec() that meant it never actually used it. Double oops. There
288 were also very few tests of studied patterns with pcre_dfa_exec().
290 27. If (?| is used to create subpatterns with duplicate numbers, they are now
291 allowed to have the same name, even if PCRE_DUPNAMES is not set. However,
292 on the other side of the coin, they are no longer allowed to have different
293 names, because these cannot be distinguished in PCRE, and this has caused
294 confusion. (This is a difference from Perl.)
296 28. When duplicate subpattern names are present (necessarily with different
297 numbers, as required by 27 above), and a test is made by name in a
298 conditional pattern, either for a subpattern having been matched, or for
299 recursion in such a pattern, all the associated numbered subpatterns are
300 tested, and the overall condition is true if the condition is true for any
301 one of them. This is the way Perl works, and is also more like the way
302 testing by number works.
305 Version 7.9 11-Apr-09
306 ---------------------
308 1. When building with support for bzlib/zlib (pcregrep) and/or readline
309 (pcretest), all targets were linked against these libraries. This included
310 libpcre, libpcreposix, and libpcrecpp, even though they do not use these
311 libraries. This caused unwanted dependencies to be created. This problem
312 has been fixed, and now only pcregrep is linked with bzlib/zlib and only
313 pcretest is linked with readline.
315 2. The "typedef int BOOL" in pcre_internal.h that was included inside the
316 "#ifndef FALSE" condition by an earlier change (probably 7.8/18) has been
317 moved outside it again, because FALSE and TRUE are already defined in AIX,
318 but BOOL is not.
320 3. The pcre_config() function was treating the PCRE_MATCH_LIMIT and
321 PCRE_MATCH_LIMIT_RECURSION values as ints, when they should be long ints.
323 4. The pcregrep documentation said spaces were inserted as well as colons (or
324 hyphens) following file names and line numbers when outputting matching
325 lines. This is not true; no spaces are inserted. I have also clarified the
326 wording for the --colour (or --color) option.
328 5. In pcregrep, when --colour was used with -o, the list of matching strings
329 was not coloured; this is different to GNU grep, so I have changed it to be
330 the same.
332 6. When --colo(u)r was used in pcregrep, only the first matching substring in
333 each matching line was coloured. Now it goes on to look for further matches
334 of any of the test patterns, which is the same behaviour as GNU grep.
336 7. A pattern that could match an empty string could cause pcregrep to loop; it
337 doesn't make sense to accept an empty string match in pcregrep, so I have
338 locked it out (using PCRE's PCRE_NOTEMPTY option). By experiment, this
339 seems to be how GNU grep behaves.
341 8. The pattern (?(?=.*b)b|^) was incorrectly compiled as "match must be at
342 start or after a newline", because the conditional assertion was not being
343 correctly handled. The rule now is that both the assertion and what follows
344 in the first alternative must satisfy the test.
346 9. If auto-callout was enabled in a pattern with a conditional group whose
347 condition was an assertion, PCRE could crash during matching, both with
348 pcre_exec() and pcre_dfa_exec().
350 10. The PCRE_DOLLAR_ENDONLY option was not working when pcre_dfa_exec() was
351 used for matching.
353 11. Unicode property support in character classes was not working for
354 characters (bytes) greater than 127 when not in UTF-8 mode.
356 12. Added the -M command line option to pcretest.
358 14. Added the non-standard REG_NOTEMPTY option to the POSIX interface.
360 15. Added the PCRE_NO_START_OPTIMIZE match-time option.
362 16. Added comments and documentation about mis-use of no_arg in the C++
363 wrapper.
365 17. Implemented support for UTF-8 encoding in EBCDIC environments, a patch
366 from Martin Jerabek that uses macro names for all relevant character and
367 string constants.
369 18. Added to pcre_internal.h two configuration checks: (a) If both EBCDIC and
370 SUPPORT_UTF8 are set, give an error; (b) If SUPPORT_UCP is set without
371 SUPPORT_UTF8, define SUPPORT_UTF8. The "configure" script handles both of
372 these, but not everybody uses configure.
374 19. A conditional group that had only one branch was not being correctly
375 recognized as an item that could match an empty string. This meant that an
376 enclosing group might also not be so recognized, causing infinite looping
377 (and probably a segfault) for patterns such as ^"((?(?=[a])[^"])|b)*"$
378 with the subject "ab", where knowledge that the repeated group can match
379 nothing is needed in order to break the loop.
381 20. If a pattern that was compiled with callouts was matched using pcre_dfa_
382 exec(), but without supplying a callout function, matching went wrong.
384 21. If PCRE_ERROR_MATCHLIMIT occurred during a recursion, there was a memory
385 leak if the size of the offset vector was greater than 30. When the vector
386 is smaller, the saved offsets during recursion go onto a local stack
387 vector, but for larger vectors malloc() is used. It was failing to free
388 when the recursion yielded PCRE_ERROR_MATCH_LIMIT (or any other "abnormal"
389 error, in fact).
391 22. There was a missing #ifdef SUPPORT_UTF8 round one of the variables in the
392 heapframe that is used only when UTF-8 support is enabled. This caused no
393 problem, but was untidy.
395 23. Steven Van Ingelgem's patch to CMakeLists.txt to change the name
396 CMAKE_BINARY_DIR to PROJECT_BINARY_DIR so that it works when PCRE is
397 included within another project.
399 24. Steven Van Ingelgem's patches to add more options to the CMake support,
400 slightly modified by me:
402 (a) PCRE_BUILD_TESTS can be set OFF not to build the tests, including
403 not building pcregrep.
405 (b) PCRE_BUILD_PCREGREP can be see OFF not to build pcregrep, but only
406 if PCRE_BUILD_TESTS is also set OFF, because the tests use pcregrep.
408 25. Forward references, both numeric and by name, in patterns that made use of
409 duplicate group numbers, could behave incorrectly or give incorrect errors,
410 because when scanning forward to find the reference group, PCRE was not
411 taking into account the duplicate group numbers. A pattern such as
412 ^X(?3)(a)(?|(b)|(q))(Y) is an example.
414 26. Changed a few more instances of "const unsigned char *" to USPTR, making
415 the feature of a custom pointer more persuasive (as requested by a user).
417 27. Wrapped the definitions of fileno and isatty for Windows, which appear in
418 pcretest.c, inside #ifndefs, because it seems they are sometimes already
419 pre-defined.
421 28. Added support for (*UTF8) at the start of a pattern.
423 29. Arrange for flags added by the "release type" setting in CMake to be shown
424 in the configuration summary.
427 Version 7.8 05-Sep-08
428 ---------------------
430 1. Replaced UCP searching code with optimized version as implemented for Ad
431 Muncher (http://www.admuncher.com/) by Peter Kankowski. This uses a two-
432 stage table and inline lookup instead of a function, giving speed ups of 2
433 to 5 times on some simple patterns that I tested. Permission was given to
434 distribute the MultiStage2.py script that generates the tables (it's not in
435 the tarball, but is in the Subversion repository).
437 2. Updated the Unicode datatables to Unicode 5.1.0. This adds yet more
438 scripts.
440 3. Change 12 for 7.7 introduced a bug in pcre_study() when a pattern contained
441 a group with a zero qualifier. The result of the study could be incorrect,
442 or the function might crash, depending on the pattern.
444 4. Caseless matching was not working for non-ASCII characters in back
445 references. For example, /(\x{de})\1/8i was not matching \x{de}\x{fe}.
446 It now works when Unicode Property Support is available.
448 5. In pcretest, an escape such as \x{de} in the data was always generating
449 a UTF-8 string, even in non-UTF-8 mode. Now it generates a single byte in
450 non-UTF-8 mode. If the value is greater than 255, it gives a warning about
451 truncation.
453 6. Minor bugfix in pcrecpp.cc (change "" == ... to NULL == ...).
455 7. Added two (int) casts to pcregrep when printing the difference of two
456 pointers, in case they are 64-bit values.
458 8. Added comments about Mac OS X stack usage to the pcrestack man page and to
459 test 2 if it fails.
461 9. Added PCRE_CALL_CONVENTION just before the names of all exported functions,
462 and a #define of that name to empty if it is not externally set. This is to
463 allow users of MSVC to set it if necessary.
465 10. The PCRE_EXP_DEFN macro which precedes exported functions was missing from
466 the convenience functions in the pcre_get.c source file.
468 11. An option change at the start of a pattern that had top-level alternatives
469 could cause overwriting and/or a crash. This command provoked a crash in
470 some environments:
472 printf "/(?i)[\xc3\xa9\xc3\xbd]|[\xc3\xa9\xc3\xbdA]/8\n" | pcretest
474 This potential security problem was recorded as CVE-2008-2371.
476 12. For a pattern where the match had to start at the beginning or immediately
477 after a newline (e.g /.*anything/ without the DOTALL flag), pcre_exec() and
478 pcre_dfa_exec() could read past the end of the passed subject if there was
479 no match. To help with detecting such bugs (e.g. with valgrind), I modified
480 pcretest so that it places the subject at the end of its malloc-ed buffer.
482 13. The change to pcretest in 12 above threw up a couple more cases when pcre_
483 exec() might read past the end of the data buffer in UTF-8 mode.
485 14. A similar bug to 7.3/2 existed when the PCRE_FIRSTLINE option was set and
486 the data contained the byte 0x85 as part of a UTF-8 character within its
487 first line. This applied both to normal and DFA matching.
489 15. Lazy qualifiers were not working in some cases in UTF-8 mode. For example,
490 /^[^d]*?$/8 failed to match "abc".
492 16. Added a missing copyright notice to pcrecpp_internal.h.
494 17. Make it more clear in the documentation that values returned from
495 pcre_exec() in ovector are byte offsets, not character counts.
497 18. Tidied a few places to stop certain compilers from issuing warnings.
499 19. Updated the Virtual Pascal + BCC files to compile the latest v7.7, as
500 supplied by Stefan Weber. I made a further small update for 7.8 because
501 there is a change of source arrangements: the pcre_searchfuncs.c module is
502 replaced by pcre_ucd.c.
505 Version 7.7 07-May-08
506 ---------------------
508 1. Applied Craig's patch to sort out a long long problem: "If we can't convert
509 a string to a long long, pretend we don't even have a long long." This is
510 done by checking for the strtoq, strtoll, and _strtoi64 functions.
512 2. Applied Craig's patch to pcrecpp.cc to restore ABI compatibility with
513 pre-7.6 versions, which defined a global no_arg variable instead of putting
514 it in the RE class. (See also #8 below.)
516 3. Remove a line of dead code, identified by coverity and reported by Nuno
517 Lopes.
519 4. Fixed two related pcregrep bugs involving -r with --include or --exclude:
521 (1) The include/exclude patterns were being applied to the whole pathnames
522 of files, instead of just to the final components.
524 (2) If there was more than one level of directory, the subdirectories were
525 skipped unless they satisfied the include/exclude conditions. This is
526 inconsistent with GNU grep (and could even be seen as contrary to the
527 pcregrep specification - which I improved to make it absolutely clear).
528 The action now is always to scan all levels of directory, and just
529 apply the include/exclude patterns to regular files.
531 5. Added the --include_dir and --exclude_dir patterns to pcregrep, and used
532 --exclude_dir in the tests to avoid scanning .svn directories.
534 6. Applied Craig's patch to the QuoteMeta function so that it escapes the
535 NUL character as backslash + 0 rather than backslash + NUL, because PCRE
536 doesn't support NULs in patterns.
538 7. Added some missing "const"s to declarations of static tables in
539 pcre_compile.c and pcre_dfa_exec.c.
541 8. Applied Craig's patch to pcrecpp.cc to fix a problem in OS X that was
542 caused by fix #2 above. (Subsequently also a second patch to fix the
543 first patch. And a third patch - this was a messy problem.)
545 9. Applied Craig's patch to remove the use of push_back().
547 10. Applied Alan Lehotsky's patch to add REG_STARTEND support to the POSIX
548 matching function regexec().
550 11. Added support for the Oniguruma syntax \g<name>, \g<n>, \g'name', \g'n',
551 which, however, unlike Perl's \g{...}, are subroutine calls, not back
552 references. PCRE supports relative numbers with this syntax (I don't think
553 Oniguruma does).
555 12. Previously, a group with a zero repeat such as (...){0} was completely
556 omitted from the compiled regex. However, this means that if the group
557 was called as a subroutine from elsewhere in the pattern, things went wrong
558 (an internal error was given). Such groups are now left in the compiled
559 pattern, with a new opcode that causes them to be skipped at execution
560 time.
562 13. Added the PCRE_JAVASCRIPT_COMPAT option. This makes the following changes
563 to the way PCRE behaves:
565 (a) A lone ] character is dis-allowed (Perl treats it as data).
567 (b) A back reference to an unmatched subpattern matches an empty string
568 (Perl fails the current match path).
570 (c) A data ] in a character class must be notated as \] because if the
571 first data character in a class is ], it defines an empty class. (In
572 Perl it is not possible to have an empty class.) The empty class []
573 never matches; it forces failure and is equivalent to (*FAIL) or (?!).
574 The negative empty class [^] matches any one character, independently
575 of the DOTALL setting.
577 14. A pattern such as /(?2)[]a()b](abc)/ which had a forward reference to a
578 non-existent subpattern following a character class starting with ']' and
579 containing () gave an internal compiling error instead of "reference to
580 non-existent subpattern". Fortunately, when the pattern did exist, the
581 compiled code was correct. (When scanning forwards to check for the
582 existencd of the subpattern, it was treating the data ']' as terminating
583 the class, so got the count wrong. When actually compiling, the reference
584 was subsequently set up correctly.)
586 15. The "always fail" assertion (?!) is optimzed to (*FAIL) by pcre_compile;
587 it was being rejected as not supported by pcre_dfa_exec(), even though
588 other assertions are supported. I have made pcre_dfa_exec() support
589 (*FAIL).
591 16. The implementation of 13c above involved the invention of a new opcode,
592 OP_ALLANY, which is like OP_ANY but doesn't check the /s flag. Since /s
593 cannot be changed at match time, I realized I could make a small
594 improvement to matching performance by compiling OP_ALLANY instead of
595 OP_ANY for "." when DOTALL was set, and then removing the runtime tests
596 on the OP_ANY path.
598 17. Compiling pcretest on Windows with readline support failed without the
599 following two fixes: (1) Make the unistd.h include conditional on
600 HAVE_UNISTD_H; (2) #define isatty and fileno as _isatty and _fileno.
602 18. Changed CMakeLists.txt and cmake/FindReadline.cmake to arrange for the
603 ncurses library to be included for pcretest when ReadLine support is
604 requested, but also to allow for it to be overridden. This patch came from
605 Daniel Bergström.
607 19. There was a typo in the file ucpinternal.h where f0_rangeflag was defined
608 as 0x00f00000 instead of 0x00800000. Luckily, this would not have caused
609 any errors with the current Unicode tables. Thanks to Peter Kankowski for
610 spotting this.
613 Version 7.6 28-Jan-08
614 ---------------------
616 1. A character class containing a very large number of characters with
617 codepoints greater than 255 (in UTF-8 mode, of course) caused a buffer
618 overflow.
620 2. Patch to cut out the "long long" test in pcrecpp_unittest when
621 HAVE_LONG_LONG is not defined.
623 3. Applied Christian Ehrlicher's patch to update the CMake build files to
624 bring them up to date and include new features. This patch includes:
626 - Fixed PH's badly added libz and libbz2 support.
627 - Fixed a problem with static linking.
628 - Added pcredemo. [But later removed - see 7 below.]
629 - Fixed dftables problem and added an option.
630 - Added a number of HAVE_XXX tests, including HAVE_WINDOWS_H and
632 - Added readline support for pcretest.
633 - Added an listing of the option settings after cmake has run.
635 4. A user submitted a patch to Makefile that makes it easy to create
636 "pcre.dll" under mingw when using Configure/Make. I added stuff to
637 Makefile.am that cause it to include this special target, without
638 affecting anything else. Note that the same mingw target plus all
639 the other distribution libraries and programs are now supported
640 when configuring with CMake (see 6 below) instead of with
641 Configure/Make.
643 5. Applied Craig's patch that moves no_arg into the RE class in the C++ code.
644 This is an attempt to solve the reported problem "pcrecpp::no_arg is not
645 exported in the Windows port". It has not yet been confirmed that the patch
646 solves the problem, but it does no harm.
648 6. Applied Sheri's patch to CMakeLists.txt to add NON_STANDARD_LIB_PREFIX and
649 NON_STANDARD_LIB_SUFFIX for dll names built with mingw when configured
650 with CMake, and also correct the comment about stack recursion.
652 7. Remove the automatic building of pcredemo from the ./configure system and
653 from CMakeLists.txt. The whole idea of pcredemo.c is that it is an example
654 of a program that users should build themselves after PCRE is installed, so
655 building it automatically is not really right. What is more, it gave
656 trouble in some build environments.
658 8. Further tidies to CMakeLists.txt from Sheri and Christian.
661 Version 7.5 10-Jan-08
662 ---------------------
664 1. Applied a patch from Craig: "This patch makes it possible to 'ignore'
665 values in parens when parsing an RE using the C++ wrapper."
667 2. Negative specials like \S did not work in character classes in UTF-8 mode.
668 Characters greater than 255 were excluded from the class instead of being
669 included.
671 3. The same bug as (2) above applied to negated POSIX classes such as
672 [:^space:].
674 4. PCRECPP_STATIC was referenced in pcrecpp_internal.h, but nowhere was it
675 defined or documented. It seems to have been a typo for PCRE_STATIC, so
676 I have changed it.
678 5. The construct (?&) was not diagnosed as a syntax error (it referenced the
679 first named subpattern) and a construct such as (?&a) would reference the
680 first named subpattern whose name started with "a" (in other words, the
681 length check was missing). Both these problems are fixed. "Subpattern name
682 expected" is now given for (?&) (a zero-length name), and this patch also
683 makes it give the same error for \k'' (previously it complained that that
684 was a reference to a non-existent subpattern).
686 6. The erroneous patterns (?+-a) and (?-+a) give different error messages;
687 this is right because (?- can be followed by option settings as well as by
688 digits. I have, however, made the messages clearer.
690 7. Patterns such as (?(1)a|b) (a pattern that contains fewer subpatterns
691 than the number used in the conditional) now cause a compile-time error.
692 This is actually not compatible with Perl, which accepts such patterns, but
693 treats the conditional as always being FALSE (as PCRE used to), but it
694 seems to me that giving a diagnostic is better.
696 8. Change "alphameric" to the more common word "alphanumeric" in comments
697 and messages.
699 9. Fix two occurrences of "backslash" in comments that should have been
700 "backspace".
702 10. Remove two redundant lines of code that can never be obeyed (their function
703 was moved elsewhere).
705 11. The program that makes PCRE's Unicode character property table had a bug
706 which caused it to generate incorrect table entries for sequences of
707 characters that have the same character type, but are in different scripts.
708 It amalgamated them into a single range, with the script of the first of
709 them. In other words, some characters were in the wrong script. There were
710 thirteen such cases, affecting characters in the following ranges:
712 U+002b0 - U+002c1
713 U+0060c - U+0060d
714 U+0061e - U+00612
715 U+0064b - U+0065e
716 U+0074d - U+0076d
717 U+01800 - U+01805
718 U+01d00 - U+01d77
719 U+01d9b - U+01dbf
720 U+0200b - U+0200f
721 U+030fc - U+030fe
722 U+03260 - U+0327f
723 U+0fb46 - U+0fbb1
724 U+10450 - U+1049d
726 12. The -o option (show only the matching part of a line) for pcregrep was not
727 compatible with GNU grep in that, if there was more than one match in a
728 line, it showed only the first of them. It now behaves in the same way as
729 GNU grep.
731 13. If the -o and -v options were combined for pcregrep, it printed a blank
732 line for every non-matching line. GNU grep prints nothing, and pcregrep now
733 does the same. The return code can be used to tell if there were any
734 non-matching lines.
736 14. Added --file-offsets and --line-offsets to pcregrep.
738 15. The pattern (?=something)(?R) was not being diagnosed as a potentially
739 infinitely looping recursion. The bug was that positive lookaheads were not
740 being skipped when checking for a possible empty match (negative lookaheads
741 and both kinds of lookbehind were skipped).
743 16. Fixed two typos in the Windows-only code in pcregrep.c, and moved the
744 inclusion of <windows.h> to before rather than after the definition of
745 INVALID_FILE_ATTRIBUTES (patch from David Byron).
747 17. Specifying a possessive quantifier with a specific limit for a Unicode
748 character property caused pcre_compile() to compile bad code, which led at
749 runtime to PCRE_ERROR_INTERNAL (-14). Examples of patterns that caused this
750 are: /\p{Zl}{2,3}+/8 and /\p{Cc}{2}+/8. It was the possessive "+" that
751 caused the error; without that there was no problem.
753 18. Added --enable-pcregrep-libz and --enable-pcregrep-libbz2.
755 19. Added --enable-pcretest-libreadline.
757 20. In pcrecpp.cc, the variable 'count' was incremented twice in
758 RE::GlobalReplace(). As a result, the number of replacements returned was
759 double what it should be. I removed one of the increments, but Craig sent a
760 later patch that removed the other one (the right fix) and added unit tests
761 that check the return values (which was not done before).
763 21. Several CMake things:
765 (1) Arranged that, when cmake is used on Unix, the libraries end up with
766 the names libpcre and libpcreposix, not just pcre and pcreposix.
768 (2) The above change means that pcretest and pcregrep are now correctly
769 linked with the newly-built libraries, not previously installed ones.
773 22. In UTF-8 mode, with newline set to "any", a pattern such as .*a.*=.b.*
774 crashed when matching a string such as a\x{2029}b (note that \x{2029} is a
775 UTF-8 newline character). The key issue is that the pattern starts .*;
776 this means that the match must be either at the beginning, or after a
777 newline. The bug was in the code for advancing after a failed match and
778 checking that the new position followed a newline. It was not taking
779 account of UTF-8 characters correctly.
781 23. PCRE was behaving differently from Perl in the way it recognized POSIX
782 character classes. PCRE was not treating the sequence [:...:] as a
783 character class unless the ... were all letters. Perl, however, seems to
784 allow any characters between [: and :], though of course it rejects as
785 unknown any "names" that contain non-letters, because all the known class
786 names consist only of letters. Thus, Perl gives an error for [[:1234:]],
787 for example, whereas PCRE did not - it did not recognize a POSIX character
788 class. This seemed a bit dangerous, so the code has been changed to be
789 closer to Perl. The behaviour is not identical to Perl, because PCRE will
790 diagnose an unknown class for, for example, [[:l\ower:]] where Perl will
791 treat it as [[:lower:]]. However, PCRE does now give "unknown" errors where
792 Perl does, and where it didn't before.
794 24. Rewrite so as to remove the single use of %n from pcregrep because in some
795 Windows environments %n is disabled by default.
798 Version 7.4 21-Sep-07
799 ---------------------
801 1. Change 7.3/28 was implemented for classes by looking at the bitmap. This
802 means that a class such as [\s] counted as "explicit reference to CR or
803 LF". That isn't really right - the whole point of the change was to try to
804 help when there was an actual mention of one of the two characters. So now
805 the change happens only if \r or \n (or a literal CR or LF) character is
806 encountered.
808 2. The 32-bit options word was also used for 6 internal flags, but the numbers
809 of both had grown to the point where there were only 3 bits left.
810 Fortunately, there was spare space in the data structure, and so I have
811 moved the internal flags into a new 16-bit field to free up more option
812 bits.
814 3. The appearance of (?J) at the start of a pattern set the DUPNAMES option,
815 but did not set the internal JCHANGED flag - either of these is enough to
816 control the way the "get" function works - but the PCRE_INFO_JCHANGED
817 facility is supposed to tell if (?J) was ever used, so now (?J) at the
818 start sets both bits.
820 4. Added options (at build time, compile time, exec time) to change \R from
821 matching any Unicode line ending sequence to just matching CR, LF, or CRLF.
823 5. doc/pcresyntax.html was missing from the distribution.
825 6. Put back the definition of PCRE_ERROR_NULLWSLIMIT, for backward
826 compatibility, even though it is no longer used.
828 7. Added macro for snprintf to pcrecpp_unittest.cc and also for strtoll and
829 strtoull to pcrecpp.cc to select the available functions in WIN32 when the
830 windows.h file is present (where different names are used). [This was
831 reversed later after testing - see 16 below.]
833 8. Changed all #include <config.h> to #include "config.h". There were also
834 some further <pcre.h> cases that I changed to "pcre.h".
836 9. When pcregrep was used with the --colour option, it missed the line ending
837 sequence off the lines that it output.
839 10. It was pointed out to me that arrays of string pointers cause lots of
840 relocations when a shared library is dynamically loaded. A technique of
841 using a single long string with a table of offsets can drastically reduce
842 these. I have refactored PCRE in four places to do this. The result is
843 dramatic:
845 Originally: 290
846 After changing UCP table: 187
847 After changing error message table: 43
848 After changing table of "verbs" 36
849 After changing table of Posix names 22
851 Thanks to the folks working on Gregex for glib for this insight.
853 11. --disable-stack-for-recursion caused compiling to fail unless -enable-
854 unicode-properties was also set.
856 12. Updated the tests so that they work when \R is defaulted to ANYCRLF.
858 13. Added checks for ANY and ANYCRLF to pcrecpp.cc where it previously
859 checked only for CRLF.
861 14. Added casts to pcretest.c to avoid compiler warnings.
863 15. Added Craig's patch to various pcrecpp modules to avoid compiler warnings.
865 16. Added Craig's patch to remove the WINDOWS_H tests, that were not working,
866 and instead check for _strtoi64 explicitly, and avoid the use of snprintf()
867 entirely. This removes changes made in 7 above.
869 17. The CMake files have been updated, and there is now more information about
870 building with CMake in the NON-UNIX-USE document.
873 Version 7.3 28-Aug-07
874 ---------------------
876 1. In the rejigging of the build system that eventually resulted in 7.1, the
877 line "#include <pcre.h>" was included in pcre_internal.h. The use of angle
878 brackets there is not right, since it causes compilers to look for an
879 installed pcre.h, not the version that is in the source that is being
880 compiled (which of course may be different). I have changed it back to:
882 #include "pcre.h"
884 I have a vague recollection that the change was concerned with compiling in
885 different directories, but in the new build system, that is taken care of
886 by the VPATH setting the Makefile.
888 2. The pattern .*$ when run in not-DOTALL UTF-8 mode with newline=any failed
889 when the subject happened to end in the byte 0x85 (e.g. if the last
890 character was \x{1ec5}). *Character* 0x85 is one of the "any" newline
891 characters but of course it shouldn't be taken as a newline when it is part
892 of another character. The bug was that, for an unlimited repeat of . in
893 not-DOTALL UTF-8 mode, PCRE was advancing by bytes rather than by
894 characters when looking for a newline.
896 3. A small performance improvement in the DOTALL UTF-8 mode .* case.
898 4. Debugging: adjusted the names of opcodes for different kinds of parentheses
899 in debug output.
901 5. Arrange to use "%I64d" instead of "%lld" and "%I64u" instead of "%llu" for
902 long printing in the pcrecpp unittest when running under MinGW.
904 6. ESC_K was left out of the EBCDIC table.
906 7. Change 7.0/38 introduced a new limit on the number of nested non-capturing
907 parentheses; I made it 1000, which seemed large enough. Unfortunately, the
908 limit also applies to "virtual nesting" when a pattern is recursive, and in
909 this case 1000 isn't so big. I have been able to remove this limit at the
910 expense of backing off one optimization in certain circumstances. Normally,
911 when pcre_exec() would call its internal match() function recursively and
912 immediately return the result unconditionally, it uses a "tail recursion"
913 feature to save stack. However, when a subpattern that can match an empty
914 string has an unlimited repetition quantifier, it no longer makes this
915 optimization. That gives it a stack frame in which to save the data for
916 checking that an empty string has been matched. Previously this was taken
917 from the 1000-entry workspace that had been reserved. So now there is no
918 explicit limit, but more stack is used.
920 8. Applied Daniel's patches to solve problems with the import/export magic
921 syntax that is required for Windows, and which was going wrong for the
922 pcreposix and pcrecpp parts of the library. These were overlooked when this
923 problem was solved for the main library.
925 9. There were some crude static tests to avoid integer overflow when computing
926 the size of patterns that contain repeated groups with explicit upper
927 limits. As the maximum quantifier is 65535, the maximum group length was
928 set at 30,000 so that the product of these two numbers did not overflow a
929 32-bit integer. However, it turns out that people want to use groups that
930 are longer than 30,000 bytes (though not repeat them that many times).
931 Change 7.0/17 (the refactoring of the way the pattern size is computed) has
932 made it possible to implement the integer overflow checks in a much more
933 dynamic way, which I have now done. The artificial limitation on group
934 length has been removed - we now have only the limit on the total length of
935 the compiled pattern, which depends on the LINK_SIZE setting.
937 10. Fixed a bug in the documentation for get/copy named substring when
938 duplicate names are permitted. If none of the named substrings are set, the
939 functions return PCRE_ERROR_NOSUBSTRING (7); the doc said they returned an
940 empty string.
942 11. Because Perl interprets \Q...\E at a high level, and ignores orphan \E
943 instances, patterns such as [\Q\E] or [\E] or even [^\E] cause an error,
944 because the ] is interpreted as the first data character and the
945 terminating ] is not found. PCRE has been made compatible with Perl in this
946 regard. Previously, it interpreted [\Q\E] as an empty class, and [\E] could
947 cause memory overwriting.
949 10. Like Perl, PCRE automatically breaks an unlimited repeat after an empty
950 string has been matched (to stop an infinite loop). It was not recognizing
951 a conditional subpattern that could match an empty string if that
952 subpattern was within another subpattern. For example, it looped when
953 trying to match (((?(1)X|))*) but it was OK with ((?(1)X|)*) where the
954 condition was not nested. This bug has been fixed.
956 12. A pattern like \X?\d or \P{L}?\d in non-UTF-8 mode could cause a backtrack
957 past the start of the subject in the presence of bytes with the top bit
958 set, for example "\x8aBCD".
960 13. Added Perl 5.10 experimental backtracking controls (*FAIL), (*F), (*PRUNE),
961 (*SKIP), (*THEN), (*COMMIT), and (*ACCEPT).
963 14. Optimized (?!) to (*FAIL).
965 15. Updated the test for a valid UTF-8 string to conform to the later RFC 3629.
966 This restricts code points to be within the range 0 to 0x10FFFF, excluding
967 the "low surrogate" sequence 0xD800 to 0xDFFF. Previously, PCRE allowed the
968 full range 0 to 0x7FFFFFFF, as defined by RFC 2279. Internally, it still
969 does: it's just the validity check that is more restrictive.
971 16. Inserted checks for integer overflows during escape sequence (backslash)
972 processing, and also fixed erroneous offset values for syntax errors during
973 backslash processing.
975 17. Fixed another case of looking too far back in non-UTF-8 mode (cf 12 above)
976 for patterns like [\PPP\x8a]{1,}\x80 with the subject "A\x80".
978 18. An unterminated class in a pattern like (?1)\c[ with a "forward reference"
979 caused an overrun.
981 19. A pattern like (?:[\PPa*]*){8,} which had an "extended class" (one with
982 something other than just ASCII characters) inside a group that had an
983 unlimited repeat caused a loop at compile time (while checking to see
984 whether the group could match an empty string).
986 20. Debugging a pattern containing \p or \P could cause a crash. For example,
987 [\P{Any}] did so. (Error in the code for printing property names.)
989 21. An orphan \E inside a character class could cause a crash.
991 22. A repeated capturing bracket such as (A)? could cause a wild memory
992 reference during compilation.
994 23. There are several functions in pcre_compile() that scan along a compiled
995 expression for various reasons (e.g. to see if it's fixed length for look
996 behind). There were bugs in these functions when a repeated \p or \P was
997 present in the pattern. These operators have additional parameters compared
998 with \d, etc, and these were not being taken into account when moving along
999 the compiled data. Specifically:
1001 (a) A item such as \p{Yi}{3} in a lookbehind was not treated as fixed
1002 length.
1004 (b) An item such as \pL+ within a repeated group could cause crashes or
1005 loops.
1007 (c) A pattern such as \p{Yi}+(\P{Yi}+)(?1) could give an incorrect
1008 "reference to non-existent subpattern" error.
1010 (d) A pattern like (\P{Yi}{2}\277)? could loop at compile time.
1012 24. A repeated \S or \W in UTF-8 mode could give wrong answers when multibyte
1013 characters were involved (for example /\S{2}/8g with "A\x{a3}BC").
1015 25. Using pcregrep in multiline, inverted mode (-Mv) caused it to loop.
1017 26. Patterns such as [\P{Yi}A] which include \p or \P and just one other
1018 character were causing crashes (broken optimization).
1020 27. Patterns such as (\P{Yi}*\277)* (group with possible zero repeat containing
1021 \p or \P) caused a compile-time loop.
1023 28. More problems have arisen in unanchored patterns when CRLF is a valid line
1024 break. For example, the unstudied pattern [\r\n]A does not match the string
1025 "\r\nA" because change 7.0/46 below moves the current point on by two
1026 characters after failing to match at the start. However, the pattern \nA
1027 *does* match, because it doesn't start till \n, and if [\r\n]A is studied,
1028 the same is true. There doesn't seem any very clean way out of this, but
1029 what I have chosen to do makes the common cases work: PCRE now takes note
1030 of whether there can be an explicit match for \r or \n anywhere in the
1031 pattern, and if so, 7.0/46 no longer applies. As part of this change,
1032 there's a new PCRE_INFO_HASCRORLF option for finding out whether a compiled
1033 pattern has explicit CR or LF references.
1035 29. Added (*CR) etc for changing newline setting at start of pattern.
1038 Version 7.2 19-Jun-07
1039 ---------------------
1041 1. If the fr_FR locale cannot be found for test 3, try the "french" locale,
1042 which is apparently normally available under Windows.
1044 2. Re-jig the pcregrep tests with different newline settings in an attempt
1045 to make them independent of the local environment's newline setting.
1047 3. Add code to configure.ac to remove -g from the CFLAGS default settings.
1049 4. Some of the "internals" tests were previously cut out when the link size
1050 was not 2, because the output contained actual offsets. The recent new
1051 "Z" feature of pcretest means that these can be cut out, making the tests
1052 usable with all link sizes.
1054 5. Implemented Stan Switzer's goto replacement for longjmp() when not using
1055 stack recursion. This gives a massive performance boost under BSD, but just
1056 a small improvement under Linux. However, it saves one field in the frame
1057 in all cases.
1059 6. Added more features from the forthcoming Perl 5.10:
1061 (a) (?-n) (where n is a string of digits) is a relative subroutine or
1062 recursion call. It refers to the nth most recently opened parentheses.
1064 (b) (?+n) is also a relative subroutine call; it refers to the nth next
1065 to be opened parentheses.
1067 (c) Conditions that refer to capturing parentheses can be specified
1068 relatively, for example, (?(-2)... or (?(+3)...
1070 (d) \K resets the start of the current match so that everything before
1071 is not part of it.
1073 (e) \k{name} is synonymous with \k<name> and \k'name' (.NET compatible).
1075 (f) \g{name} is another synonym - part of Perl 5.10's unification of
1076 reference syntax.
1078 (g) (?| introduces a group in which the numbering of parentheses in each
1079 alternative starts with the same number.
1081 (h) \h, \H, \v, and \V match horizontal and vertical whitespace.
1083 7. Added two new calls to pcre_fullinfo(): PCRE_INFO_OKPARTIAL and
1086 8. A pattern such as (.*(.)?)* caused pcre_exec() to fail by either not
1087 terminating or by crashing. Diagnosed by Viktor Griph; it was in the code
1088 for detecting groups that can match an empty string.
1090 9. A pattern with a very large number of alternatives (more than several
1091 hundred) was running out of internal workspace during the pre-compile
1092 phase, where pcre_compile() figures out how much memory will be needed. A
1093 bit of new cunning has reduced the workspace needed for groups with
1094 alternatives. The 1000-alternative test pattern now uses 12 bytes of
1095 workspace instead of running out of the 4096 that are available.
1097 10. Inserted some missing (unsigned int) casts to get rid of compiler warnings.
1099 11. Applied patch from Google to remove an optimization that didn't quite work.
1100 The report of the bug said:
1102 pcrecpp::RE("a*").FullMatch("aaa") matches, while
1103 pcrecpp::RE("a*?").FullMatch("aaa") does not, and
1104 pcrecpp::RE("a*?\\z").FullMatch("aaa") does again.
1106 12. If \p or \P was used in non-UTF-8 mode on a character greater than 127
1107 it matched the wrong number of bytes.
1110 Version 7.1 24-Apr-07
1111 ---------------------
1113 1. Applied Bob Rossi and Daniel G's patches to convert the build system to one
1114 that is more "standard", making use of automake and other Autotools. There
1115 is some re-arrangement of the files and adjustment of comments consequent
1116 on this.
1118 2. Part of the patch fixed a problem with the pcregrep tests. The test of -r
1119 for recursive directory scanning broke on some systems because the files
1120 are not scanned in any specific order and on different systems the order
1121 was different. A call to "sort" has been inserted into RunGrepTest for the
1122 approprate test as a short-term fix. In the longer term there may be an
1123 alternative.
1125 3. I had an email from Eric Raymond about problems translating some of PCRE's
1126 man pages to HTML (despite the fact that I distribute HTML pages, some
1127 people do their own conversions for various reasons). The problems
1128 concerned the use of low-level troff macros .br and .in. I have therefore
1129 removed all such uses from the man pages (some were redundant, some could
1130 be replaced by .nf/.fi pairs). The 132html script that I use to generate
1131 HTML has been updated to handle .nf/.fi and to complain if it encounters
1132 .br or .in.
1134 4. Updated comments in configure.ac that get placed in config.h.in and also
1135 arranged for config.h to be included in the distribution, with the name
1136 config.h.generic, for the benefit of those who have to compile without
1137 Autotools (compare pcre.h, which is now distributed as pcre.h.generic).
1139 5. Updated the support (such as it is) for Virtual Pascal, thanks to Stefan
1140 Weber: (1) pcre_internal.h was missing some function renames; (2) updated
1141 makevp.bat for the current PCRE, using the additional files
1142 makevp_c.txt, makevp_l.txt, and pcregexp.pas.
1144 6. A Windows user reported a minor discrepancy with test 2, which turned out
1145 to be caused by a trailing space on an input line that had got lost in his
1146 copy. The trailing space was an accident, so I've just removed it.
1148 7. Add -Wl,-R... flags in pcre-config.in for *BSD* systems, as I'm told
1149 that is needed.
1151 8. Mark ucp_table (in ucptable.h) and ucp_gentype (in pcre_ucp_searchfuncs.c)
1152 as "const" (a) because they are and (b) because it helps the PHP
1153 maintainers who have recently made a script to detect big data structures
1154 in the php code that should be moved to the .rodata section. I remembered
1155 to update Builducptable as well, so it won't revert if ucptable.h is ever
1156 re-created.
1158 9. Added some extra #ifdef SUPPORT_UTF8 conditionals into pcretest.c,
1159 pcre_printint.src, pcre_compile.c, pcre_study.c, and pcre_tables.c, in
1160 order to be able to cut out the UTF-8 tables in the latter when UTF-8
1161 support is not required. This saves 1.5-2K of code, which is important in
1162 some applications.
1164 Later: more #ifdefs are needed in pcre_ord2utf8.c and pcre_valid_utf8.c
1165 so as not to refer to the tables, even though these functions will never be
1166 called when UTF-8 support is disabled. Otherwise there are problems with a
1167 shared library.
1169 10. Fixed two bugs in the emulated memmove() function in pcre_internal.h:
1171 (a) It was defining its arguments as char * instead of void *.
1173 (b) It was assuming that all moves were upwards in memory; this was true
1174 a long time ago when I wrote it, but is no longer the case.
1176 The emulated memove() is provided for those environments that have neither
1177 memmove() nor bcopy(). I didn't think anyone used it these days, but that
1178 is clearly not the case, as these two bugs were recently reported.
1180 11. The script PrepareRelease is now distributed: it calls 132html, CleanTxt,
1181 and Detrail to create the HTML documentation, the .txt form of the man
1182 pages, and it removes trailing spaces from listed files. It also creates
1183 pcre.h.generic and config.h.generic from pcre.h and config.h. In the latter
1184 case, it wraps all the #defines with #ifndefs. This script should be run
1185 before "make dist".
1187 12. Fixed two fairly obscure bugs concerned with quantified caseless matching
1188 with Unicode property support.
1190 (a) For a maximizing quantifier, if the two different cases of the
1191 character were of different lengths in their UTF-8 codings (there are
1192 some cases like this - I found 11), and the matching function had to
1193 back up over a mixture of the two cases, it incorrectly assumed they
1194 were both the same length.
1196 (b) When PCRE was configured to use the heap rather than the stack for
1197 recursion during matching, it was not correctly preserving the data for
1198 the other case of a UTF-8 character when checking ahead for a match
1199 while processing a minimizing repeat. If the check also involved
1200 matching a wide character, but failed, corruption could cause an
1201 erroneous result when trying to check for a repeat of the original
1202 character.
1204 13. Some tidying changes to the testing mechanism:
1206 (a) The RunTest script now detects the internal link size and whether there
1207 is UTF-8 and UCP support by running ./pcretest -C instead of relying on
1208 values substituted by "configure". (The RunGrepTest script already did
1209 this for UTF-8.) The configure.ac script no longer substitutes the
1210 relevant variables.
1212 (b) The debugging options /B and /D in pcretest show the compiled bytecode
1213 with length and offset values. This means that the output is different
1214 for different internal link sizes. Test 2 is skipped for link sizes
1215 other than 2 because of this, bypassing the problem. Unfortunately,
1216 there was also a test in test 3 (the locale tests) that used /B and
1217 failed for link sizes other than 2. Rather than cut the whole test out,
1218 I have added a new /Z option to pcretest that replaces the length and
1219 offset values with spaces. This is now used to make test 3 independent
1220 of link size. (Test 2 will be tidied up later.)
1222 14. If erroroffset was passed as NULL to pcre_compile, it provoked a
1223 segmentation fault instead of returning the appropriate error message.
1225 15. In multiline mode when the newline sequence was set to "any", the pattern
1226 ^$ would give a match between the \r and \n of a subject such as "A\r\nB".
1227 This doesn't seem right; it now treats the CRLF combination as the line
1228 ending, and so does not match in that case. It's only a pattern such as ^$
1229 that would hit this one: something like ^ABC$ would have failed after \r
1230 and then tried again after \r\n.
1232 16. Changed the comparison command for RunGrepTest from "diff -u" to "diff -ub"
1233 in an attempt to make files that differ only in their line terminators
1234 compare equal. This works on Linux.
1236 17. Under certain error circumstances pcregrep might try to free random memory
1237 as it exited. This is now fixed, thanks to valgrind.
1239 19. In pcretest, if the pattern /(?m)^$/g<any> was matched against the string
1240 "abc\r\n\r\n", it found an unwanted second match after the second \r. This
1241 was because its rules for how to advance for /g after matching an empty
1242 string at the end of a line did not allow for this case. They now check for
1243 it specially.
1245 20. pcretest is supposed to handle patterns and data of any length, by
1246 extending its buffers when necessary. It was getting this wrong when the
1247 buffer for a data line had to be extended.
1249 21. Added PCRE_NEWLINE_ANYCRLF which is like ANY, but matches only CR, LF, or
1250 CRLF as a newline sequence.
1252 22. Code for handling Unicode properties in pcre_dfa_exec() wasn't being cut
1253 out by #ifdef SUPPORT_UCP. This did no harm, as it could never be used, but
1254 I have nevertheless tidied it up.
1256 23. Added some casts to kill warnings from HP-UX ia64 compiler.
1258 24. Added a man page for pcre-config.
1261 Version 7.0 19-Dec-06
1262 ---------------------
1264 1. Fixed a signed/unsigned compiler warning in pcre_compile.c, shown up by
1265 moving to gcc 4.1.1.
1267 2. The -S option for pcretest uses setrlimit(); I had omitted to #include
1268 sys/time.h, which is documented as needed for this function. It doesn't
1269 seem to matter on Linux, but it showed up on some releases of OS X.
1271 3. It seems that there are systems where bytes whose values are greater than
1272 127 match isprint() in the "C" locale. The "C" locale should be the
1273 default when a C program starts up. In most systems, only ASCII printing
1274 characters match isprint(). This difference caused the output from pcretest
1275 to vary, making some of the tests fail. I have changed pcretest so that:
1277 (a) When it is outputting text in the compiled version of a pattern, bytes
1278 other than 32-126 are always shown as hex escapes.
1280 (b) When it is outputting text that is a matched part of a subject string,
1281 it does the same, unless a different locale has been set for the match
1282 (using the /L modifier). In this case, it uses isprint() to decide.
1284 4. Fixed a major bug that caused incorrect computation of the amount of memory
1285 required for a compiled pattern when options that changed within the
1286 pattern affected the logic of the preliminary scan that determines the
1287 length. The relevant options are -x, and -i in UTF-8 mode. The result was
1288 that the computed length was too small. The symptoms of this bug were
1289 either the PCRE error "internal error: code overflow" from pcre_compile(),
1290 or a glibc crash with a message such as "pcretest: free(): invalid next
1291 size (fast)". Examples of patterns that provoked this bug (shown in
1292 pcretest format) are:
1294 /(?-x: )/x
1295 /(?x)(?-x: \s*#\s*)/
1296 /((?i)[\x{c0}])/8
1297 /(?i:[\x{c0}])/8
1299 HOWEVER: Change 17 below makes this fix obsolete as the memory computation
1300 is now done differently.
1302 5. Applied patches from Google to: (a) add a QuoteMeta function to the C++
1303 wrapper classes; (b) implement a new function in the C++ scanner that is
1304 more efficient than the old way of doing things because it avoids levels of
1305 recursion in the regex matching; (c) add a paragraph to the documentation
1306 for the FullMatch() function.
1308 6. The escape sequence \n was being treated as whatever was defined as
1309 "newline". Not only was this contrary to the documentation, which states
1310 that \n is character 10 (hex 0A), but it also went horribly wrong when
1311 "newline" was defined as CRLF. This has been fixed.
1313 7. In pcre_dfa_exec.c the value of an unsigned integer (the variable called c)
1314 was being set to -1 for the "end of line" case (supposedly a value that no
1315 character can have). Though this value is never used (the check for end of
1316 line is "zero bytes in current character"), it caused compiler complaints.
1317 I've changed it to 0xffffffff.
1319 8. In pcre_version.c, the version string was being built by a sequence of
1320 C macros that, in the event of PCRE_PRERELEASE being defined as an empty
1321 string (as it is for production releases) called a macro with an empty
1322 argument. The C standard says the result of this is undefined. The gcc
1323 compiler treats it as an empty string (which was what was wanted) but it is
1324 reported that Visual C gives an error. The source has been hacked around to
1325 avoid this problem.
1327 9. On the advice of a Windows user, included <io.h> and <fcntl.h> in Windows
1328 builds of pcretest, and changed the call to _setmode() to use _O_BINARY
1329 instead of 0x8000. Made all the #ifdefs test both _WIN32 and WIN32 (not all
1330 of them did).
1332 10. Originally, pcretest opened its input and output without "b"; then I was
1333 told that "b" was needed in some environments, so it was added for release
1334 5.0 to both the input and output. (It makes no difference on Unix-like
1335 systems.) Later I was told that it is wrong for the input on Windows. I've
1336 now abstracted the modes into two macros, to make it easier to fiddle with
1337 them, and removed "b" from the input mode under Windows.
1339 11. Added pkgconfig support for the C++ wrapper library, libpcrecpp.
1341 12. Added -help and --help to pcretest as an official way of being reminded
1342 of the options.
1344 13. Removed some redundant semicolons after macro calls in pcrecpparg.h.in
1345 and pcrecpp.cc because they annoy compilers at high warning levels.
1347 14. A bit of tidying/refactoring in pcre_exec.c in the main bumpalong loop.
1349 15. Fixed an occurrence of == in configure.ac that should have been = (shell
1350 scripts are not C programs :-) and which was not noticed because it works
1351 on Linux.
1353 16. pcretest is supposed to handle any length of pattern and data line (as one
1354 line or as a continued sequence of lines) by extending its input buffer if
1355 necessary. This feature was broken for very long pattern lines, leading to
1356 a string of junk being passed to pcre_compile() if the pattern was longer
1357 than about 50K.
1359 17. I have done a major re-factoring of the way pcre_compile() computes the
1360 amount of memory needed for a compiled pattern. Previously, there was code
1361 that made a preliminary scan of the pattern in order to do this. That was
1362 OK when PCRE was new, but as the facilities have expanded, it has become
1363 harder and harder to keep it in step with the real compile phase, and there
1364 have been a number of bugs (see for example, 4 above). I have now found a
1365 cunning way of running the real compile function in a "fake" mode that
1366 enables it to compute how much memory it would need, while actually only
1367 ever using a few hundred bytes of working memory and without too many
1368 tests of the mode. This should make future maintenance and development
1369 easier. A side effect of this work is that the limit of 200 on the nesting
1370 depth of parentheses has been removed (though this was never a serious
1371 limitation, I suspect). However, there is a downside: pcre_compile() now
1372 runs more slowly than before (30% or more, depending on the pattern). I
1373 hope this isn't a big issue. There is no effect on runtime performance.
1375 18. Fixed a minor bug in pcretest: if a pattern line was not terminated by a
1376 newline (only possible for the last line of a file) and it was a
1377 pattern that set a locale (followed by /Lsomething), pcretest crashed.
1379 19. Added additional timing features to pcretest. (1) The -tm option now times
1380 matching only, not compiling. (2) Both -t and -tm can be followed, as a
1381 separate command line item, by a number that specifies the number of
1382 repeats to use when timing. The default is 50000; this gives better
1383 precision, but takes uncomfortably long for very large patterns.
1385 20. Extended pcre_study() to be more clever in cases where a branch of a
1386 subpattern has no definite first character. For example, (a*|b*)[cd] would
1387 previously give no result from pcre_study(). Now it recognizes that the
1388 first character must be a, b, c, or d.
1390 21. There was an incorrect error "recursive call could loop indefinitely" if
1391 a subpattern (or the entire pattern) that was being tested for matching an
1392 empty string contained only one non-empty item after a nested subpattern.
1393 For example, the pattern (?>\x{100}*)\d(?R) provoked this error
1394 incorrectly, because the \d was being skipped in the check.
1396 22. The pcretest program now has a new pattern option /B and a command line
1397 option -b, which is equivalent to adding /B to every pattern. This causes
1398 it to show the compiled bytecode, without the additional information that
1399 -d shows. The effect of -d is now the same as -b with -i (and similarly, /D
1400 is the same as /B/I).
1402 23. A new optimization is now able automatically to treat some sequences such
1403 as a*b as a*+b. More specifically, if something simple (such as a character
1404 or a simple class like \d) has an unlimited quantifier, and is followed by
1405 something that cannot possibly match the quantified thing, the quantifier
1406 is automatically "possessified".
1408 24. A recursive reference to a subpattern whose number was greater than 39
1409 went wrong under certain circumstances in UTF-8 mode. This bug could also
1410 have affected the operation of pcre_study().
1412 25. Realized that a little bit of performance could be had by replacing
1413 (c & 0xc0) == 0xc0 with c >= 0xc0 when processing UTF-8 characters.
1415 26. Timing data from pcretest is now shown to 4 decimal places instead of 3.
1417 27. Possessive quantifiers such as a++ were previously implemented by turning
1418 them into atomic groups such as ($>a+). Now they have their own opcodes,
1419 which improves performance. This includes the automatically created ones
1420 from 23 above.
1422 28. A pattern such as (?=(\w+))\1: which simulates an atomic group using a
1423 lookahead was broken if it was not anchored. PCRE was mistakenly expecting
1424 the first matched character to be a colon. This applied both to named and
1425 numbered groups.
1427 29. The ucpinternal.h header file was missing its idempotency #ifdef.
1429 30. I was sent a "project" file called libpcre.a.dev which I understand makes
1430 building PCRE on Windows easier, so I have included it in the distribution.
1432 31. There is now a check in pcretest against a ridiculously large number being
1433 returned by pcre_exec() or pcre_dfa_exec(). If this happens in a /g or /G
1434 loop, the loop is abandoned.
1436 32. Forward references to subpatterns in conditions such as (?(2)...) where
1437 subpattern 2 is defined later cause pcre_compile() to search forwards in
1438 the pattern for the relevant set of parentheses. This search went wrong
1439 when there were unescaped parentheses in a character class, parentheses
1440 escaped with \Q...\E, or parentheses in a #-comment in /x mode.
1442 33. "Subroutine" calls and backreferences were previously restricted to
1443 referencing subpatterns earlier in the regex. This restriction has now
1444 been removed.
1446 34. Added a number of extra features that are going to be in Perl 5.10. On the
1447 whole, these are just syntactic alternatives for features that PCRE had
1448 previously implemented using the Python syntax or my own invention. The
1449 other formats are all retained for compatibility.
1451 (a) Named groups can now be defined as (?<name>...) or (?'name'...) as well
1452 as (?P<name>...). The new forms, as well as being in Perl 5.10, are
1453 also .NET compatible.
1455 (b) A recursion or subroutine call to a named group can now be defined as
1456 (?&name) as well as (?P>name).
1458 (c) A backreference to a named group can now be defined as \k<name> or
1459 \k'name' as well as (?P=name). The new forms, as well as being in Perl
1460 5.10, are also .NET compatible.
1462 (d) A conditional reference to a named group can now use the syntax
1463 (?(<name>) or (?('name') as well as (?(name).
1465 (e) A "conditional group" of the form (?(DEFINE)...) can be used to define
1466 groups (named and numbered) that are never evaluated inline, but can be
1467 called as "subroutines" from elsewhere. In effect, the DEFINE condition
1468 is always false. There may be only one alternative in such a group.
1470 (f) A test for recursion can be given as (?(R1).. or (?(R&name)... as well
1471 as the simple (?(R). The condition is true only if the most recent
1472 recursion is that of the given number or name. It does not search out
1473 through the entire recursion stack.
1475 (g) The escape \gN or \g{N} has been added, where N is a positive or
1476 negative number, specifying an absolute or relative reference.
1478 35. Tidied to get rid of some further signed/unsigned compiler warnings and
1479 some "unreachable code" warnings.
1481 36. Updated the Unicode property tables to Unicode version 5.0.0. Amongst other
1482 things, this adds five new scripts.
1484 37. Perl ignores orphaned \E escapes completely. PCRE now does the same.
1485 There were also incompatibilities regarding the handling of \Q..\E inside
1486 character classes, for example with patterns like [\Qa\E-\Qz\E] where the
1487 hyphen was adjacent to \Q or \E. I hope I've cleared all this up now.
1489 38. Like Perl, PCRE detects when an indefinitely repeated parenthesized group
1490 matches an empty string, and forcibly breaks the loop. There were bugs in
1491 this code in non-simple cases. For a pattern such as ^(a()*)* matched
1492 against aaaa the result was just "a" rather than "aaaa", for example. Two
1493 separate and independent bugs (that affected different cases) have been
1494 fixed.
1496 39. Refactored the code to abolish the use of different opcodes for small
1497 capturing bracket numbers. This is a tidy that I avoided doing when I
1498 removed the limit on the number of capturing brackets for 3.5 back in 2001.
1499 The new approach is not only tidier, it makes it possible to reduce the
1500 memory needed to fix the previous bug (38).
1502 40. Implemented PCRE_NEWLINE_ANY to recognize any of the Unicode newline
1503 sequences (http://unicode.org/unicode/reports/tr18/) as "newline" when
1504 processing dot, circumflex, or dollar metacharacters, or #-comments in /x
1505 mode.
1507 41. Add \R to match any Unicode newline sequence, as suggested in the Unicode
1508 report.
1510 42. Applied patch, originally from Ari Pollak, modified by Google, to allow
1511 copy construction and assignment in the C++ wrapper.
1513 43. Updated pcregrep to support "--newline=any". In the process, I fixed a
1514 couple of bugs that could have given wrong results in the "--newline=crlf"
1515 case.
1517 44. Added a number of casts and did some reorganization of signed/unsigned int
1518 variables following suggestions from Dair Grant. Also renamed the variable
1519 "this" as "item" because it is a C++ keyword.
1521 45. Arranged for dftables to add
1523 #include "pcre_internal.h"
1525 to pcre_chartables.c because without it, gcc 4.x may remove the array
1526 definition from the final binary if PCRE is built into a static library and
1527 dead code stripping is activated.
1529 46. For an unanchored pattern, if a match attempt fails at the start of a
1530 newline sequence, and the newline setting is CRLF or ANY, and the next two
1531 characters are CRLF, advance by two characters instead of one.
1534 Version 6.7 04-Jul-06
1535 ---------------------
1537 1. In order to handle tests when input lines are enormously long, pcretest has
1538 been re-factored so that it automatically extends its buffers when
1539 necessary. The code is crude, but this _is_ just a test program. The
1540 default size has been increased from 32K to 50K.
1542 2. The code in pcre_study() was using the value of the re argument before
1543 testing it for NULL. (Of course, in any sensible call of the function, it
1544 won't be NULL.)
1546 3. The memmove() emulation function in pcre_internal.h, which is used on
1547 systems that lack both memmove() and bcopy() - that is, hardly ever -
1548 was missing a "static" storage class specifier.
1550 4. When UTF-8 mode was not set, PCRE looped when compiling certain patterns
1551 containing an extended class (one that cannot be represented by a bitmap
1552 because it contains high-valued characters or Unicode property items, e.g.
1553 [\pZ]). Almost always one would set UTF-8 mode when processing such a
1554 pattern, but PCRE should not loop if you do not (it no longer does).
1555 [Detail: two cases were found: (a) a repeated subpattern containing an
1556 extended class; (b) a recursive reference to a subpattern that followed a
1557 previous extended class. It wasn't skipping over the extended class
1558 correctly when UTF-8 mode was not set.]
1560 5. A negated single-character class was not being recognized as fixed-length
1561 in lookbehind assertions such as (?<=[^f]), leading to an incorrect
1562 compile error "lookbehind assertion is not fixed length".
1564 6. The RunPerlTest auxiliary script was showing an unexpected difference
1565 between PCRE and Perl for UTF-8 tests. It turns out that it is hard to
1566 write a Perl script that can interpret lines of an input file either as
1567 byte characters or as UTF-8, which is what "perltest" was being required to
1568 do for the non-UTF-8 and UTF-8 tests, respectively. Essentially what you
1569 can't do is switch easily at run time between having the "use utf8;" pragma
1570 or not. In the end, I fudged it by using the RunPerlTest script to insert
1571 "use utf8;" explicitly for the UTF-8 tests.
1573 7. In multiline (/m) mode, PCRE was matching ^ after a terminating newline at
1574 the end of the subject string, contrary to the documentation and to what
1575 Perl does. This was true of both matching functions. Now it matches only at
1576 the start of the subject and immediately after *internal* newlines.
1578 8. A call of pcre_fullinfo() from pcretest to get the option bits was passing
1579 a pointer to an int instead of a pointer to an unsigned long int. This
1580 caused problems on 64-bit systems.
1582 9. Applied a patch from the folks at Google to pcrecpp.cc, to fix "another
1583 instance of the 'standard' template library not being so standard".
1585 10. There was no check on the number of named subpatterns nor the maximum
1586 length of a subpattern name. The product of these values is used to compute
1587 the size of the memory block for a compiled pattern. By supplying a very
1588 long subpattern name and a large number of named subpatterns, the size
1589 computation could be caused to overflow. This is now prevented by limiting
1590 the length of names to 32 characters, and the number of named subpatterns
1591 to 10,000.
1593 11. Subpatterns that are repeated with specific counts have to be replicated in
1594 the compiled pattern. The size of memory for this was computed from the
1595 length of the subpattern and the repeat count. The latter is limited to
1596 65535, but there was no limit on the former, meaning that integer overflow
1597 could in principle occur. The compiled length of a repeated subpattern is
1598 now limited to 30,000 bytes in order to prevent this.
1600 12. Added the optional facility to have named substrings with the same name.
1602 13. Added the ability to use a named substring as a condition, using the
1603 Python syntax: (?(name)yes|no). This overloads (?(R)... and names that
1604 are numbers (not recommended). Forward references are permitted.
1606 14. Added forward references in named backreferences (if you see what I mean).
1608 15. In UTF-8 mode, with the PCRE_DOTALL option set, a quantified dot in the
1609 pattern could run off the end of the subject. For example, the pattern
1610 "(?s)(.{1,5})"8 did this with the subject "ab".
1612 16. If PCRE_DOTALL or PCRE_MULTILINE were set, pcre_dfa_exec() behaved as if
1613 PCRE_CASELESS was set when matching characters that were quantified with ?
1614 or *.
1616 17. A character class other than a single negated character that had a minimum
1617 but no maximum quantifier - for example [ab]{6,} - was not handled
1618 correctly by pce_dfa_exec(). It would match only one character.
1620 18. A valid (though odd) pattern that looked like a POSIX character
1621 class but used an invalid character after [ (for example [[,abc,]]) caused
1622 pcre_compile() to give the error "Failed: internal error: code overflow" or
1623 in some cases to crash with a glibc free() error. This could even happen if
1624 the pattern terminated after [[ but there just happened to be a sequence of
1625 letters, a binary zero, and a closing ] in the memory that followed.
1627 19. Perl's treatment of octal escapes in the range \400 to \777 has changed
1628 over the years. Originally (before any Unicode support), just the bottom 8
1629 bits were taken. Thus, for example, \500 really meant \100. Nowadays the
1630 output from "man perlunicode" includes this:
1632 The regular expression compiler produces polymorphic opcodes. That
1633 is, the pattern adapts to the data and automatically switches to
1634 the Unicode character scheme when presented with Unicode data--or
1635 instead uses a traditional byte scheme when presented with byte
1636 data.
1638 Sadly, a wide octal escape does not cause a switch, and in a string with
1639 no other multibyte characters, these octal escapes are treated as before.
1640 Thus, in Perl, the pattern /\500/ actually matches \100 but the pattern
1641 /\500|\x{1ff}/ matches \500 or \777 because the whole thing is treated as a
1642 Unicode string.
1644 I have not perpetrated such confusion in PCRE. Up till now, it took just
1645 the bottom 8 bits, as in old Perl. I have now made octal escapes with
1646 values greater than \377 illegal in non-UTF-8 mode. In UTF-8 mode they
1647 translate to the appropriate multibyte character.
1649 29. Applied some refactoring to reduce the number of warnings from Microsoft
1650 and Borland compilers. This has included removing the fudge introduced
1651 seven years ago for the OS/2 compiler (see 2.02/2 below) because it caused
1652 a warning about an unused variable.
1654 21. PCRE has not included VT (character 0x0b) in the set of whitespace
1655 characters since release 4.0, because Perl (from release 5.004) does not.
1656 [Or at least, is documented not to: some releases seem to be in conflict
1657 with the documentation.] However, when a pattern was studied with
1658 pcre_study() and all its branches started with \s, PCRE still included VT
1659 as a possible starting character. Of course, this did no harm; it just
1660 caused an unnecessary match attempt.
1662 22. Removed a now-redundant internal flag bit that recorded the fact that case
1663 dependency changed within the pattern. This was once needed for "required
1664 byte" processing, but is no longer used. This recovers a now-scarce options
1665 bit. Also moved the least significant internal flag bit to the most-
1666 significant bit of the word, which was not previously used (hangover from
1667 the days when it was an int rather than a uint) to free up another bit for
1668 the future.
1670 23. Added support for CRLF line endings as well as CR and LF. As well as the
1671 default being selectable at build time, it can now be changed at runtime
1672 via the PCRE_NEWLINE_xxx flags. There are now options for pcregrep to
1673 specify that it is scanning data with non-default line endings.
1675 24. Changed the definition of CXXLINK to make it agree with the definition of
1676 LINK in the Makefile, by replacing LDFLAGS to CXXFLAGS.
1678 25. Applied Ian Taylor's patches to avoid using another stack frame for tail
1679 recursions. This makes a big different to stack usage for some patterns.
1681 26. If a subpattern containing a named recursion or subroutine reference such
1682 as (?P>B) was quantified, for example (xxx(?P>B)){3}, the calculation of
1683 the space required for the compiled pattern went wrong and gave too small a
1684 value. Depending on the environment, this could lead to "Failed: internal
1685 error: code overflow at offset 49" or "glibc detected double free or
1686 corruption" errors.
1688 27. Applied patches from Google (a) to support the new newline modes and (b) to
1689 advance over multibyte UTF-8 characters in GlobalReplace.
1691 28. Change free() to pcre_free() in pcredemo.c. Apparently this makes a
1692 difference for some implementation of PCRE in some Windows version.
1694 29. Added some extra testing facilities to pcretest:
1696 \q<number> in a data line sets the "match limit" value
1697 \Q<number> in a data line sets the "match recursion limt" value
1698 -S <number> sets the stack size, where <number> is in megabytes
1700 The -S option isn't available for Windows.
1703 Version 6.6 06-Feb-06
1704 ---------------------
1706 1. Change 16(a) for 6.5 broke things, because PCRE_DATA_SCOPE was not defined
1707 in pcreposix.h. I have copied the definition from pcre.h.
1709 2. Change 25 for 6.5 broke compilation in a build directory out-of-tree
1710 because pcre.h is no longer a built file.
1712 3. Added Jeff Friedl's additional debugging patches to pcregrep. These are
1713 not normally included in the compiled code.
1716 Version 6.5 01-Feb-06
1717 ---------------------
1719 1. When using the partial match feature with pcre_dfa_exec(), it was not
1720 anchoring the second and subsequent partial matches at the new starting
1721 point. This could lead to incorrect results. For example, with the pattern
1722 /1234/, partially matching against "123" and then "a4" gave a match.
1724 2. Changes to pcregrep:
1726 (a) All non-match returns from pcre_exec() were being treated as failures
1727 to match the line. Now, unless the error is PCRE_ERROR_NOMATCH, an
1728 error message is output. Some extra information is given for the
1730 probably the only errors that are likely to be caused by users (by
1731 specifying a regex that has nested indefinite repeats, for instance).
1732 If there are more than 20 of these errors, pcregrep is abandoned.
1734 (b) A binary zero was treated as data while matching, but terminated the
1735 output line if it was written out. This has been fixed: binary zeroes
1736 are now no different to any other data bytes.
1738 (c) Whichever of the LC_ALL or LC_CTYPE environment variables is set is
1739 used to set a locale for matching. The --locale=xxxx long option has
1740 been added (no short equivalent) to specify a locale explicitly on the
1741 pcregrep command, overriding the environment variables.
1743 (d) When -B was used with -n, some line numbers in the output were one less
1744 than they should have been.
1746 (e) Added the -o (--only-matching) option.
1748 (f) If -A or -C was used with -c (count only), some lines of context were
1749 accidentally printed for the final match.
1751 (g) Added the -H (--with-filename) option.
1753 (h) The combination of options -rh failed to suppress file names for files
1754 that were found from directory arguments.
1756 (i) Added the -D (--devices) and -d (--directories) options.
1758 (j) Added the -F (--fixed-strings) option.
1760 (k) Allow "-" to be used as a file name for -f as well as for a data file.
1762 (l) Added the --colo(u)r option.
1764 (m) Added Jeffrey Friedl's -S testing option, but within #ifdefs so that it
1765 is not present by default.
1767 3. A nasty bug was discovered in the handling of recursive patterns, that is,
1768 items such as (?R) or (?1), when the recursion could match a number of
1769 alternatives. If it matched one of the alternatives, but subsequently,
1770 outside the recursion, there was a failure, the code tried to back up into
1771 the recursion. However, because of the way PCRE is implemented, this is not
1772 possible, and the result was an incorrect result from the match.
1774 In order to prevent this happening, the specification of recursion has
1775 been changed so that all such subpatterns are automatically treated as
1776 atomic groups. Thus, for example, (?R) is treated as if it were (?>(?R)).
1778 4. I had overlooked the fact that, in some locales, there are characters for
1779 which isalpha() is true but neither isupper() nor islower() are true. In
1780 the fr_FR locale, for instance, the \xAA and \xBA characters (ordmasculine
1781 and ordfeminine) are like this. This affected the treatment of \w and \W
1782 when they appeared in character classes, but not when they appeared outside
1783 a character class. The bit map for "word" characters is now created
1784 separately from the results of isalnum() instead of just taking it from the
1785 upper, lower, and digit maps. (Plus the underscore character, of course.)
1787 5. The above bug also affected the handling of POSIX character classes such as
1788 [[:alpha:]] and [[:alnum:]]. These do not have their own bit maps in PCRE's
1789 permanent tables. Instead, the bit maps for such a class were previously
1790 created as the appropriate unions of the upper, lower, and digit bitmaps.
1791 Now they are created by subtraction from the [[:word:]] class, which has
1792 its own bitmap.
1794 6. The [[:blank:]] character class matches horizontal, but not vertical space.
1795 It is created by subtracting the vertical space characters (\x09, \x0a,
1796 \x0b, \x0c) from the [[:space:]] bitmap. Previously, however, the
1797 subtraction was done in the overall bitmap for a character class, meaning
1798 that a class such as [\x0c[:blank:]] was incorrect because \x0c would not
1799 be recognized. This bug has been fixed.
1801 7. Patches from the folks at Google:
1803 (a) pcrecpp.cc: "to handle a corner case that may or may not happen in
1804 real life, but is still worth protecting against".
1806 (b) pcrecpp.cc: "corrects a bug when negative radixes are used with
1807 regular expressions".
1809 (c) pcre_scanner.cc: avoid use of std::count() because not all systems
1810 have it.
1812 (d) Split off pcrecpparg.h from pcrecpp.h and had the former built by
1813 "configure" and the latter not, in order to fix a problem somebody had
1814 with compiling the Arg class on HP-UX.
1816 (e) Improve the error-handling of the C++ wrapper a little bit.
1818 (f) New tests for checking recursion limiting.
1820 8. The pcre_memmove() function, which is used only if the environment does not
1821 have a standard memmove() function (and is therefore rarely compiled),
1822 contained two bugs: (a) use of int instead of size_t, and (b) it was not
1823 returning a result (though PCRE never actually uses the result).
1825 9. In the POSIX regexec() interface, if nmatch is specified as a ridiculously
1826 large number - greater than INT_MAX/(3*sizeof(int)) - REG_ESPACE is
1827 returned instead of calling malloc() with an overflowing number that would
1828 most likely cause subsequent chaos.
1830 10. The debugging option of pcretest was not showing the NO_AUTO_CAPTURE flag.
1832 11. The POSIX flag REG_NOSUB is now supported. When a pattern that was compiled
1833 with this option is matched, the nmatch and pmatch options of regexec() are
1834 ignored.
1836 12. Added REG_UTF8 to the POSIX interface. This is not defined by POSIX, but is
1837 provided in case anyone wants to the the POSIX interface with UTF-8
1838 strings.
1840 13. Added CXXLDFLAGS to the Makefile parameters to provide settings only on the
1841 C++ linking (needed for some HP-UX environments).
1843 14. Avoid compiler warnings in get_ucpname() when compiled without UCP support
1844 (unused parameter) and in the pcre_printint() function (omitted "default"
1845 switch label when the default is to do nothing).
1847 15. Added some code to make it possible, when PCRE is compiled as a C++
1848 library, to replace subject pointers for pcre_exec() with a smart pointer
1849 class, thus making it possible to process discontinuous strings.
1851 16. The two macros PCRE_EXPORT and PCRE_DATA_SCOPE are confusing, and perform
1852 much the same function. They were added by different people who were trying
1853 to make PCRE easy to compile on non-Unix systems. It has been suggested
1854 that PCRE_EXPORT be abolished now that there is more automatic apparatus
1855 for compiling on Windows systems. I have therefore replaced it with
1856 PCRE_DATA_SCOPE. This is set automatically for Windows; if not set it
1857 defaults to "extern" for C or "extern C" for C++, which works fine on
1858 Unix-like systems. It is now possible to override the value of PCRE_DATA_
1859 SCOPE with something explicit in config.h. In addition:
1861 (a) pcreposix.h still had just "extern" instead of either of these macros;
1862 I have replaced it with PCRE_DATA_SCOPE.
1864 (b) Functions such as _pcre_xclass(), which are internal to the library,
1865 but external in the C sense, all had PCRE_EXPORT in their definitions.
1866 This is apparently wrong for the Windows case, so I have removed it.
1867 (It makes no difference on Unix-like systems.)
1869 17. Added a new limit, MATCH_LIMIT_RECURSION, which limits the depth of nesting
1870 of recursive calls to match(). This is different to MATCH_LIMIT because
1871 that limits the total number of calls to match(), not all of which increase
1872 the depth of recursion. Limiting the recursion depth limits the amount of
1873 stack (or heap if NO_RECURSE is set) that is used. The default can be set
1874 when PCRE is compiled, and changed at run time. A patch from Google adds
1875 this functionality to the C++ interface.
1877 18. Changes to the handling of Unicode character properties:
1879 (a) Updated the table to Unicode 4.1.0.
1881 (b) Recognize characters that are not in the table as "Cn" (undefined).
1883 (c) I revised the way the table is implemented to a much improved format
1884 which includes recognition of ranges. It now supports the ranges that
1885 are defined in UnicodeData.txt, and it also amalgamates other
1886 characters into ranges. This has reduced the number of entries in the
1887 table from around 16,000 to around 3,000, thus reducing its size
1888 considerably. I realized I did not need to use a tree structure after
1889 all - a binary chop search is just as efficient. Having reduced the
1890 number of entries, I extended their size from 6 bytes to 8 bytes to
1891 allow for more data.
1893 (d) Added support for Unicode script names via properties such as \p{Han}.
1895 19. In UTF-8 mode, a backslash followed by a non-Ascii character was not
1896 matching that character.
1898 20. When matching a repeated Unicode property with a minimum greater than zero,
1899 (for example \pL{2,}), PCRE could look past the end of the subject if it
1900 reached it while seeking the minimum number of characters. This could
1901 happen only if some of the characters were more than one byte long, because
1902 there is a check for at least the minimum number of bytes.
1904 21. Refactored the implementation of \p and \P so as to be more general, to
1905 allow for more different types of property in future. This has changed the
1906 compiled form incompatibly. Anybody with saved compiled patterns that use
1907 \p or \P will have to recompile them.
1909 22. Added "Any" and "L&" to the supported property types.
1911 23. Recognize \x{...} as a code point specifier, even when not in UTF-8 mode,
1912 but give a compile time error if the value is greater than 0xff.
1914 24. The man pages for pcrepartial, pcreprecompile, and pcre_compile2 were
1915 accidentally not being installed or uninstalled.
1917 25. The pcre.h file was built from pcre.h.in, but the only changes that were
1918 made were to insert the current release number. This seemed silly, because
1919 it made things harder for people building PCRE on systems that don't run
1920 "configure". I have turned pcre.h into a distributed file, no longer built
1921 by "configure", with the version identification directly included. There is
1922 no longer a pcre.h.in file.
1924 However, this change necessitated a change to the pcre-config script as
1925 well. It is built from pcre-config.in, and one of the substitutions was the
1926 release number. I have updated configure.ac so that ./configure now finds
1927 the release number by grepping pcre.h.
1929 26. Added the ability to run the tests under valgrind.
1932 Version 6.4 05-Sep-05
1933 ---------------------
1935 1. Change 6.0/10/(l) to pcregrep introduced a bug that caused separator lines
1936 "--" to be printed when multiple files were scanned, even when none of the
1937 -A, -B, or -C options were used. This is not compatible with Gnu grep, so I
1938 consider it to be a bug, and have restored the previous behaviour.
1940 2. A couple of code tidies to get rid of compiler warnings.
1942 3. The pcretest program used to cheat by referring to symbols in the library
1943 whose names begin with _pcre_. These are internal symbols that are not
1944 really supposed to be visible externally, and in some environments it is
1945 possible to suppress them. The cheating is now confined to including
1946 certain files from the library's source, which is a bit cleaner.
1948 4. Renamed pcre.in as pcre.h.in to go with pcrecpp.h.in; it also makes the
1949 file's purpose clearer.
1951 5. Reorganized pcre_ucp_findchar().
1954 Version 6.3 15-Aug-05
1955 ---------------------
1957 1. The file libpcre.pc.in did not have general read permission in the tarball.
1959 2. There were some problems when building without C++ support:
1961 (a) If C++ support was not built, "make install" and "make test" still
1962 tried to test it.
1964 (b) There were problems when the value of CXX was explicitly set. Some
1965 changes have been made to try to fix these, and ...
1967 (c) --disable-cpp can now be used to explicitly disable C++ support.
1969 (d) The use of @CPP_OBJ@ directly caused a blank line preceded by a
1970 backslash in a target when C++ was disabled. This confuses some
1971 versions of "make", apparently. Using an intermediate variable solves
1972 this. (Same for CPP_LOBJ.)
1974 3. $(LINK_FOR_BUILD) now includes $(CFLAGS_FOR_BUILD) and $(LINK)
1975 (non-Windows) now includes $(CFLAGS) because these flags are sometimes
1976 necessary on certain architectures.
1978 4. Added a setting of -export-symbols-regex to the link command to remove
1979 those symbols that are exported in the C sense, but actually are local
1980 within the library, and not documented. Their names all begin with
1981 "_pcre_". This is not a perfect job, because (a) we have to except some
1982 symbols that pcretest ("illegally") uses, and (b) the facility isn't always
1983 available (and never for static libraries). I have made a note to try to
1984 find a way round (a) in the future.
1987 Version 6.2 01-Aug-05
1988 ---------------------
1990 1. There was no test for integer overflow of quantifier values. A construction
1991 such as {1111111111111111} would give undefined results. What is worse, if
1992 a minimum quantifier for a parenthesized subpattern overflowed and became
1993 negative, the calculation of the memory size went wrong. This could have
1994 led to memory overwriting.
1996 2. Building PCRE using VPATH was broken. Hopefully it is now fixed.
1998 3. Added "b" to the 2nd argument of fopen() in dftables.c, for non-Unix-like
1999 operating environments where this matters.
2001 4. Applied Giuseppe Maxia's patch to add additional features for controlling
2002 PCRE options from within the C++ wrapper.
2004 5. Named capturing subpatterns were not being correctly counted when a pattern
2005 was compiled. This caused two problems: (a) If there were more than 100
2006 such subpatterns, the calculation of the memory needed for the whole
2007 compiled pattern went wrong, leading to an overflow error. (b) Numerical
2008 back references of the form \12, where the number was greater than 9, were
2009 not recognized as back references, even though there were sufficient
2010 previous subpatterns.
2012 6. Two minor patches to pcrecpp.cc in order to allow it to compile on older
2013 versions of gcc, e.g. 2.95.4.
2016 Version 6.1 21-Jun-05
2017 ---------------------
2019 1. There was one reference to the variable "posix" in pcretest.c that was not
2020 surrounded by "#if !defined NOPOSIX".
2022 2. Make it possible to compile pcretest without DFA support, UTF8 support, or
2023 the cross-check on the old pcre_info() function, for the benefit of the
2024 cut-down version of PCRE that is currently imported into Exim.
2026 3. A (silly) pattern starting with (?i)(?-i) caused an internal space
2027 allocation error. I've done the easy fix, which wastes 2 bytes for sensible
2028 patterns that start (?i) but I don't think that matters. The use of (?i) is
2029 just an example; this all applies to the other options as well.
2031 4. Since libtool seems to echo the compile commands it is issuing, the output
2032 from "make" can be reduced a bit by putting "@" in front of each libtool
2033 compile command.
2035 5. Patch from the folks at Google for configure.in to be a bit more thorough
2036 in checking for a suitable C++ installation before trying to compile the
2037 C++ stuff. This should fix a reported problem when a compiler was present,
2038 but no suitable headers.
2040 6. The man pages all had just "PCRE" as their title. I have changed them to
2041 be the relevant file name. I have also arranged that these names are
2042 retained in the file doc/pcre.txt, which is a concatenation in text format
2043 of all the man pages except the little individual ones for each function.
2045 7. The NON-UNIX-USE file had not been updated for the different set of source
2046 files that come with release 6. I also added a few comments about the C++
2047 wrapper.
2050 Version 6.0 07-Jun-05
2051 ---------------------
2053 1. Some minor internal re-organization to help with my DFA experiments.
2055 2. Some missing #ifdef SUPPORT_UCP conditionals in pcretest and printint that
2056 didn't matter for the library itself when fully configured, but did matter
2057 when compiling without UCP support, or within Exim, where the ucp files are
2058 not imported.
2060 3. Refactoring of the library code to split up the various functions into
2061 different source modules. The addition of the new DFA matching code (see
2062 below) to a single monolithic source would have made it really too
2063 unwieldy, quite apart from causing all the code to be include in a
2064 statically linked application, when only some functions are used. This is
2065 relevant even without the DFA addition now that patterns can be compiled in
2066 one application and matched in another.
2068 The downside of splitting up is that there have to be some external
2069 functions and data tables that are used internally in different modules of
2070 the library but which are not part of the API. These have all had their
2071 names changed to start with "_pcre_" so that they are unlikely to clash
2072 with other external names.
2074 4. Added an alternate matching function, pcre_dfa_exec(), which matches using
2075 a different (DFA) algorithm. Although it is slower than the original
2076 function, it does have some advantages for certain types of matching
2077 problem.
2079 5. Upgrades to pcretest in order to test the features of pcre_dfa_exec(),
2080 including restarting after a partial match.
2082 6. A patch for pcregrep that defines INVALID_FILE_ATTRIBUTES if it is not
2083 defined when compiling for Windows was sent to me. I have put it into the
2084 code, though I have no means of testing or verifying it.
2086 7. Added the pcre_refcount() auxiliary function.
2088 8. Added the PCRE_FIRSTLINE option. This constrains an unanchored pattern to
2089 match before or at the first newline in the subject string. In pcretest,
2090 the /f option on a pattern can be used to set this.
2092 9. A repeated \w when used in UTF-8 mode with characters greater than 256
2093 would behave wrongly. This has been present in PCRE since release 4.0.
2095 10. A number of changes to the pcregrep command:
2097 (a) Refactored how -x works; insert ^(...)$ instead of setting
2098 PCRE_ANCHORED and checking the length, in preparation for adding
2099 something similar for -w.
2101 (b) Added the -w (match as a word) option.
2103 (c) Refactored the way lines are read and buffered so as to have more
2104 than one at a time available.
2106 (d) Implemented a pcregrep test script.
2108 (e) Added the -M (multiline match) option. This allows patterns to match
2109 over several lines of the subject. The buffering ensures that at least
2110 8K, or the rest of the document (whichever is the shorter) is available
2111 for matching (and similarly the previous 8K for lookbehind assertions).
2113 (f) Changed the --help output so that it now says
2115 -w, --word-regex(p)
2117 instead of two lines, one with "regex" and the other with "regexp"
2118 because that confused at least one person since the short forms are the
2119 same. (This required a bit of code, as the output is generated
2120 automatically from a table. It wasn't just a text change.)
2122 (g) -- can be used to terminate pcregrep options if the next thing isn't an
2123 option but starts with a hyphen. Could be a pattern or a path name
2124 starting with a hyphen, for instance.
2126 (h) "-" can be given as a file name to represent stdin.
2128 (i) When file names are being printed, "(standard input)" is used for
2129 the standard input, for compatibility with GNU grep. Previously
2130 "<stdin>" was used.
2132 (j) The option --label=xxx can be used to supply a name to be used for
2133 stdin when file names are being printed. There is no short form.
2135 (k) Re-factored the options decoding logic because we are going to add
2136 two more options that take data. Such options can now be given in four
2137 different ways, e.g. "-fname", "-f name", "--file=name", "--file name".
2139 (l) Added the -A, -B, and -C options for requesting that lines of context
2140 around matches be printed.
2142 (m) Added the -L option to print the names of files that do not contain
2143 any matching lines, that is, the complement of -l.
2145 (n) The return code is 2 if any file cannot be opened, but pcregrep does
2146 continue to scan other files.
2148 (o) The -s option was incorrectly implemented. For compatibility with other
2149 greps, it now suppresses the error message for a non-existent or non-
2150 accessible file (but not the return code). There is a new option called
2151 -q that suppresses the output of matching lines, which was what -s was
2152 previously doing.
2154 (p) Added --include and --exclude options to specify files for inclusion
2155 and exclusion when recursing.
2157 11. The Makefile was not using the Autoconf-supported LDFLAGS macro properly.
2158 Hopefully, it now does.
2160 12. Missing cast in pcre_study().
2162 13. Added an "uninstall" target to the makefile.
2164 14. Replaced "extern" in the function prototypes in Makefile.in with
2165 "PCRE_DATA_SCOPE", which defaults to 'extern' or 'extern "C"' in the Unix
2166 world, but is set differently for Windows.
2168 15. Added a second compiling function called pcre_compile2(). The only
2169 difference is that it has an extra argument, which is a pointer to an
2170 integer error code. When there is a compile-time failure, this is set
2171 non-zero, in addition to the error test pointer being set to point to an
2172 error message. The new argument may be NULL if no error number is required
2173 (but then you may as well call pcre_compile(), which is now just a
2174 wrapper). This facility is provided because some applications need a
2175 numeric error indication, but it has also enabled me to tidy up the way
2176 compile-time errors are handled in the POSIX wrapper.
2178 16. Added VPATH=.libs to the makefile; this should help when building with one
2179 prefix path and installing with another. (Or so I'm told by someone who
2180 knows more about this stuff than I do.)
2182 17. Added a new option, REG_DOTALL, to the POSIX function regcomp(). This
2183 passes PCRE_DOTALL to the pcre_compile() function, making the "." character
2184 match everything, including newlines. This is not POSIX-compatible, but
2185 somebody wanted the feature. From pcretest it can be activated by using
2186 both the P and the s flags.
2188 18. AC_PROG_LIBTOOL appeared twice in Makefile.in. Removed one.
2190 19. libpcre.pc was being incorrectly installed as executable.
2192 20. A couple of places in pcretest check for end-of-line by looking for '\n';
2193 it now also looks for '\r' so that it will work unmodified on Windows.
2195 21. Added Google's contributed C++ wrapper to the distribution.
2197 22. Added some untidy missing memory free() calls in pcretest, to keep
2198 Electric Fence happy when testing.
2202 Version 5.0 13-Sep-04
2203 ---------------------
2205 1. Internal change: literal characters are no longer packed up into items
2206 containing multiple characters in a single byte-string. Each character
2207 is now matched using a separate opcode. However, there may be more than one
2208 byte in the character in UTF-8 mode.
2210 2. The pcre_callout_block structure has two new fields: pattern_position and
2211 next_item_length. These contain the offset in the pattern to the next match
2212 item, and its length, respectively.
2214 3. The PCRE_AUTO_CALLOUT option for pcre_compile() requests the automatic
2215 insertion of callouts before each pattern item. Added the /C option to
2216 pcretest to make use of this.
2218 4. On the advice of a Windows user, the lines
2220 #if defined(_WIN32) || defined(WIN32)
2221 _setmode( _fileno( stdout ), 0x8000 );
2222 #endif /* defined(_WIN32) || defined(WIN32) */
2224 have been added to the source of pcretest. This apparently does useful
2225 magic in relation to line terminators.
2227 5. Changed "r" and "w" in the calls to fopen() in pcretest to "rb" and "wb"
2228 for the benefit of those environments where the "b" makes a difference.
2230 6. The icc compiler has the same options as gcc, but "configure" doesn't seem
2231 to know about it. I have put a hack into configure.in that adds in code
2232 to set GCC=yes if CC=icc. This seems to end up at a point in the
2233 generated configure script that is early enough to affect the setting of
2234 compiler options, which is what is needed, but I have no means of testing
2235 whether it really works. (The user who reported this had patched the
2236 generated configure script, which of course I cannot do.)
2238 LATER: After change 22 below (new libtool files), the configure script
2239 seems to know about icc (and also ecc). Therefore, I have commented out
2240 this hack in configure.in.
2242 7. Added support for pkg-config (2 patches were sent in).
2244 8. Negated POSIX character classes that used a combination of internal tables
2245 were completely broken. These were [[:^alpha:]], [[:^alnum:]], and
2246 [[:^ascii]]. Typically, they would match almost any characters. The other
2247 POSIX classes were not broken in this way.
2249 9. Matching the pattern "\b.*?" against "ab cd", starting at offset 1, failed
2250 to find the match, as PCRE was deluded into thinking that the match had to
2251 start at the start point or following a newline. The same bug applied to
2252 patterns with negative forward assertions or any backward assertions
2253 preceding ".*" at the start, unless the pattern required a fixed first
2254 character. This was a failing pattern: "(?!.bcd).*". The bug is now fixed.
2256 10. In UTF-8 mode, when moving forwards in the subject after a failed match
2257 starting at the last subject character, bytes beyond the end of the subject
2258 string were read.
2260 11. Renamed the variable "class" as "classbits" to make life easier for C++
2261 users. (Previously there was a macro definition, but it apparently wasn't
2262 enough.)
2264 12. Added the new field "tables" to the extra data so that tables can be passed
2265 in at exec time, or the internal tables can be re-selected. This allows
2266 a compiled regex to be saved and re-used at a later time by a different
2267 program that might have everything at different addresses.
2269 13. Modified the pcre-config script so that, when run on Solaris, it shows a
2270 -R library as well as a -L library.
2272 14. The debugging options of pcretest (-d on the command line or D on a
2273 pattern) showed incorrect output for anything following an extended class
2274 that contained multibyte characters and which was followed by a quantifier.
2276 15. Added optional support for general category Unicode character properties
2277 via the \p, \P, and \X escapes. Unicode property support implies UTF-8
2278 support. It adds about 90K to the size of the library. The meanings of the
2279 inbuilt class escapes such as \d and \s have NOT been changed.
2281 16. Updated pcredemo.c to include calls to free() to release the memory for the
2282 compiled pattern.
2284 17. The generated file chartables.c was being created in the source directory
2285 instead of in the building directory. This caused the build to fail if the
2286 source directory was different from the building directory, and was
2287 read-only.
2289 18. Added some sample Win commands from Mark Tetrode into the NON-UNIX-USE
2290 file. No doubt somebody will tell me if they don't make sense... Also added
2291 Dan Mooney's comments about building on OpenVMS.
2293 19. Added support for partial matching via the PCRE_PARTIAL option for
2294 pcre_exec() and the \P data escape in pcretest.
2296 20. Extended pcretest with 3 new pattern features:
2298 (i) A pattern option of the form ">rest-of-line" causes pcretest to
2299 write the compiled pattern to the file whose name is "rest-of-line".
2300 This is a straight binary dump of the data, with the saved pointer to
2301 the character tables forced to be NULL. The study data, if any, is
2302 written too. After writing, pcretest reads a new pattern.
2304 (ii) If, instead of a pattern, "<rest-of-line" is given, pcretest reads a
2305 compiled pattern from the given file. There must not be any
2306 occurrences of "<" in the file name (pretty unlikely); if there are,
2307 pcretest will instead treat the initial "<" as a pattern delimiter.
2308 After reading in the pattern, pcretest goes on to read data lines as
2309 usual.
2311 (iii) The F pattern option causes pcretest to flip the bytes in the 32-bit
2312 and 16-bit fields in a compiled pattern, to simulate a pattern that
2313 was compiled on a host of opposite endianness.
2315 21. The pcre-exec() function can now cope with patterns that were compiled on
2316 hosts of opposite endianness, with this restriction:
2318 As for any compiled expression that is saved and used later, the tables
2319 pointer field cannot be preserved; the extra_data field in the arguments
2320 to pcre_exec() should be used to pass in a tables address if a value
2321 other than the default internal tables were used at compile time.
2323 22. Calling pcre_exec() with a negative value of the "ovecsize" parameter is
2324 now diagnosed as an error. Previously, most of the time, a negative number
2325 would have been treated as zero, but if in addition "ovector" was passed as
2326 NULL, a crash could occur.
2328 23. Updated the files ltmain.sh, config.sub, config.guess, and aclocal.m4 with
2329 new versions from the libtool 1.5 distribution (the last one is a copy of
2330 a file called libtool.m4). This seems to have fixed the need to patch
2331 "configure" to support Darwin 1.3 (which I used to do). However, I still
2332 had to patch ltmain.sh to ensure that ${SED} is set (it isn't on my
2333 workstation).
2335 24. Changed the PCRE licence to be the more standard "BSD" licence.
2338 Version 4.5 01-Dec-03
2339 ---------------------
2341 1. There has been some re-arrangement of the code for the match() function so
2342 that it can be compiled in a version that does not call itself recursively.
2343 Instead, it keeps those local variables that need separate instances for
2344 each "recursion" in a frame on the heap, and gets/frees frames whenever it
2345 needs to "recurse". Keeping track of where control must go is done by means
2346 of setjmp/longjmp. The whole thing is implemented by a set of macros that
2347 hide most of the details from the main code, and operates only if
2348 NO_RECURSE is defined while compiling pcre.c. If PCRE is built using the
2349 "configure" mechanism, "--disable-stack-for-recursion" turns on this way of
2350 operating.
2352 To make it easier for callers to provide specially tailored get/free
2353 functions for this usage, two new functions, pcre_stack_malloc, and
2354 pcre_stack_free, are used. They are always called in strict stacking order,
2355 and the size of block requested is always the same.
2357 The PCRE_CONFIG_STACKRECURSE info parameter can be used to find out whether
2358 PCRE has been compiled to use the stack or the heap for recursion. The
2359 -C option of pcretest uses this to show which version is compiled.
2361 A new data escape \S, is added to pcretest; it causes the amounts of store
2362 obtained and freed by both kinds of malloc/free at match time to be added
2363 to the output.
2365 2. Changed the locale test to use "fr_FR" instead of "fr" because that's
2366 what's available on my current Linux desktop machine.
2368 3. When matching a UTF-8 string, the test for a valid string at the start has
2369 been extended. If start_offset is not zero, PCRE now checks that it points
2370 to a byte that is the start of a UTF-8 character. If not, it returns
2371 PCRE_ERROR_BADUTF8_OFFSET (-11). Note: the whole string is still checked;
2372 this is necessary because there may be backward assertions in the pattern.
2373 When matching the same subject several times, it may save resources to use
2374 PCRE_NO_UTF8_CHECK on all but the first call if the string is long.
2376 4. The code for checking the validity of UTF-8 strings has been tightened so
2377 that it rejects (a) strings containing 0xfe or 0xff bytes and (b) strings
2378 containing "overlong sequences".
2380 5. Fixed a bug (appearing twice) that I could not find any way of exploiting!
2381 I had written "if ((digitab[*p++] && chtab_digit) == 0)" where the "&&"
2382 should have been "&", but it just so happened that all the cases this let
2383 through by mistake were picked up later in the function.
2385 6. I had used a variable called "isblank" - this is a C99 function, causing
2386 some compilers to warn. To avoid this, I renamed it (as "blankclass").
2388 7. Cosmetic: (a) only output another newline at the end of pcretest if it is
2389 prompting; (b) run "./pcretest /dev/null" at the start of the test script
2390 so the version is shown; (c) stop "make test" echoing "./RunTest".
2392 8. Added patches from David Burgess to enable PCRE to run on EBCDIC systems.
2394 9. The prototype for memmove() for systems that don't have it was using
2395 size_t, but the inclusion of the header that defines size_t was later. I've
2396 moved the #includes for the C headers earlier to avoid this.
2398 10. Added some adjustments to the code to make it easier to compiler on certain
2399 special systems:
2401 (a) Some "const" qualifiers were missing.
2402 (b) Added the macro EXPORT before all exported functions; by default this
2403 is defined to be empty.
2404 (c) Changed the dftables auxiliary program (that builds chartables.c) so
2405 that it reads its output file name as an argument instead of writing
2406 to the standard output and assuming this can be redirected.
2408 11. In UTF-8 mode, if a recursive reference (e.g. (?1)) followed a character
2409 class containing characters with values greater than 255, PCRE compilation
2410 went into a loop.
2412 12. A recursive reference to a subpattern that was within another subpattern
2413 that had a minimum quantifier of zero caused PCRE to crash. For example,
2414 (x(y(?2))z)? provoked this bug with a subject that got as far as the
2415 recursion. If the recursively-called subpattern itself had a zero repeat,
2416 that was OK.
2418 13. In pcretest, the buffer for reading a data line was set at 30K, but the
2419 buffer into which it was copied (for escape processing) was still set at
2420 1024, so long lines caused crashes.
2422 14. A pattern such as /[ab]{1,3}+/ failed to compile, giving the error
2423 "internal error: code overflow...". This applied to any character class
2424 that was followed by a possessive quantifier.
2426 15. Modified the Makefile to add libpcre.la as a prerequisite for
2427 libpcreposix.la because I was told this is needed for a parallel build to
2428 work.
2430 16. If a pattern that contained .* following optional items at the start was
2431 studied, the wrong optimizing data was generated, leading to matching
2432 errors. For example, studying /[ab]*.*c/ concluded, erroneously, that any
2433 matching string must start with a or b or c. The correct conclusion for
2434 this pattern is that a match can start with any character.
2437 Version 4.4 13-Aug-03
2438 ---------------------
2440 1. In UTF-8 mode, a character class containing characters with values between
2441 127 and 255 was not handled correctly if the compiled pattern was studied.
2442 In fixing this, I have also improved the studying algorithm for such
2443 classes (slightly).
2445 2. Three internal functions had redundant arguments passed to them. Removal
2446 might give a very teeny performance improvement.
2448 3. Documentation bug: the value of the capture_top field in a callout is *one
2449 more than* the number of the hightest numbered captured substring.
2451 4. The Makefile linked pcretest and pcregrep with -lpcre, which could result
2452 in incorrectly linking with a previously installed version. They now link
2453 explicitly with libpcre.la.
2455 5. configure.in no longer needs to recognize Cygwin specially.
2457 6. A problem in pcre.in for Windows platforms is fixed.
2459 7. If a pattern was successfully studied, and the -d (or /D) flag was given to
2460 pcretest, it used to include the size of the study block as part of its
2461 output. Unfortunately, the structure contains a field that has a different
2462 size on different hardware architectures. This meant that the tests that
2463 showed this size failed. As the block is currently always of a fixed size,
2464 this information isn't actually particularly useful in pcretest output, so
2465 I have just removed it.
2467 8. Three pre-processor statements accidentally did not start in column 1.
2468 Sadly, there are *still* compilers around that complain, even though
2469 standard C has not required this for well over a decade. Sigh.
2471 9. In pcretest, the code for checking callouts passed small integers in the
2472 callout_data field, which is a void * field. However, some picky compilers
2473 complained about the casts involved for this on 64-bit systems. Now
2474 pcretest passes the address of the small integer instead, which should get
2475 rid of the warnings.
2477 10. By default, when in UTF-8 mode, PCRE now checks for valid UTF-8 strings at
2478 both compile and run time, and gives an error if an invalid UTF-8 sequence
2479 is found. There is a option for disabling this check in cases where the
2480 string is known to be correct and/or the maximum performance is wanted.
2482 11. In response to a bug report, I changed one line in Makefile.in from
2484 -Wl,--out-implib,.libs/lib@WIN_PREFIX@pcreposix.dll.a \
2485 to
2486 -Wl,--out-implib,.libs/@WIN_PREFIX@libpcreposix.dll.a \
2488 to look similar to other lines, but I have no way of telling whether this
2489 is the right thing to do, as I do not use Windows. No doubt I'll get told
2490 if it's wrong...
2493 Version 4.3 21-May-03
2494 ---------------------
2496 1. Two instances of @WIN_PREFIX@ omitted from the Windows targets in the
2497 Makefile.
2499 2. Some refactoring to improve the quality of the code:
2501 (i) The utf8_table... variables are now declared "const".
2503 (ii) The code for \cx, which used the "case flipping" table to upper case
2504 lower case letters, now just substracts 32. This is ASCII-specific,
2505 but the whole concept of \cx is ASCII-specific, so it seems
2506 reasonable.
2508 (iii) PCRE was using its character types table to recognize decimal and
2509 hexadecimal digits in the pattern. This is silly, because it handles
2510 only 0-9, a-f, and A-F, but the character types table is locale-
2511 specific, which means strange things might happen. A private
2512 table is now used for this - though it costs 256 bytes, a table is
2513 much faster than multiple explicit tests. Of course, the standard
2514 character types table is still used for matching digits in subject
2515 strings against \d.
2517 (iv) Strictly, the identifier ESC_t is reserved by POSIX (all identifiers
2518 ending in _t are). So I've renamed it as ESC_tee.
2520 3. The first argument for regexec() in the POSIX wrapper should have been
2521 defined as "const".
2523 4. Changed pcretest to use malloc() for its buffers so that they can be
2524 Electric Fenced for debugging.
2526 5. There were several places in the code where, in UTF-8 mode, PCRE would try
2527 to read one or more bytes before the start of the subject string. Often this
2528 had no effect on PCRE's behaviour, but in some circumstances it could
2529 provoke a segmentation fault.
2531 6. A lookbehind at the start of a pattern in UTF-8 mode could also cause PCRE
2532 to try to read one or more bytes before the start of the subject string.
2534 7. A lookbehind in a pattern matched in non-UTF-8 mode on a PCRE compiled with
2535 UTF-8 support could misbehave in various ways if the subject string
2536 contained bytes with the 0x80 bit set and the 0x40 bit unset in a lookbehind
2537 area. (PCRE was not checking for the UTF-8 mode flag, and trying to move
2538 back over UTF-8 characters.)
2541 Version 4.2 14-Apr-03
2542 ---------------------
2544 1. Typo "#if SUPPORT_UTF8" instead of "#ifdef SUPPORT_UTF8" fixed.
2546 2. Changes to the building process, supplied by Ronald Landheer-Cieslak
2547 [ON_WINDOWS]: new variable, "#" on non-Windows platforms
2548 [NOT_ON_WINDOWS]: new variable, "#" on Windows platforms
2549 [WIN_PREFIX]: new variable, "cyg" for Cygwin
2550 * Makefile.in: use autoconf substitution for OBJEXT, EXEEXT, BUILD_OBJEXT
2552 Note: automatic setting of the BUILD variables is not yet working
2553 set CPPFLAGS and BUILD_CPPFLAGS (but don't use yet) - should be used at
2554 compile-time but not at link-time
2555 [LINK]: use for linking executables only
2556 make different versions for Windows and non-Windows
2557 [LINKLIB]: new variable, copy of UNIX-style LINK, used for linking
2558 libraries
2559 [LINK_FOR_BUILD]: new variable
2560 [OBJEXT]: use throughout
2561 [EXEEXT]: use throughout
2562 <winshared>: new target
2563 <wininstall>: new target
2564 <dftables.o>: use native compiler
2565 <dftables>: use native linker
2566 <install>: handle Windows platform correctly
2567 <clean>: ditto
2568 <check>: ditto
2569 copy DLL to top builddir before testing
2571 As part of these changes, -no-undefined was removed again. This was reported
2572 to give trouble on HP-UX 11.0, so getting rid of it seems like a good idea
2573 in any case.
2575 3. Some tidies to get rid of compiler warnings:
2577 . In the match_data structure, match_limit was an unsigned long int, whereas
2578 match_call_count was an int. I've made them both unsigned long ints.
2580 . In pcretest the fact that a const uschar * doesn't automatically cast to
2581 a void * provoked a warning.
2583 . Turning on some more compiler warnings threw up some "shadow" variables
2584 and a few more missing casts.
2586 4. If PCRE was complied with UTF-8 support, but called without the PCRE_UTF8
2587 option, a class that contained a single character with a value between 128
2588 and 255 (e.g. /[\xFF]/) caused PCRE to crash.
2590 5. If PCRE was compiled with UTF-8 support, but called without the PCRE_UTF8
2591 option, a class that contained several characters, but with at least one
2592 whose value was between 128 and 255 caused PCRE to crash.
2595 Version 4.1 12-Mar-03
2596 ---------------------
2598 1. Compiling with gcc -pedantic found a couple of places where casts were
2599 needed, and a string in dftables.c that was longer than standard compilers are
2600 required to support.
2602 2. Compiling with Sun's compiler found a few more places where the code could
2603 be tidied up in order to avoid warnings.
2605 3. The variables for cross-compiling were called HOST_CC and HOST_CFLAGS; the
2606 first of these names is deprecated in the latest Autoconf in favour of the name
2607 CC_FOR_BUILD, because "host" is typically used to mean the system on which the
2608 compiled code will be run. I can't find a reference for HOST_CFLAGS, but by
2609 analogy I have changed it to CFLAGS_FOR_BUILD.
2611 4. Added -no-undefined to the linking command in the Makefile, because this is
2612 apparently helpful for Windows. To make it work, also added "-L. -lpcre" to the
2613 linking step for the pcreposix library.
2615 5. PCRE was failing to diagnose the case of two named groups with the same
2616 name.
2618 6. A problem with one of PCRE's optimizations was discovered. PCRE remembers a
2619 literal character that is needed in the subject for a match, and scans along to
2620 ensure that it is present before embarking on the full matching process. This
2621 saves time in cases of nested unlimited repeats that are never going to match.
2622 Problem: the scan can take a lot of time if the subject is very long (e.g.
2623 megabytes), thus penalizing straightforward matches. It is now done only if the
2624 amount of subject to be scanned is less than 1000 bytes.
2626 7. A lesser problem with the same optimization is that it was recording the
2627 first character of an anchored pattern as "needed", thus provoking a search
2628 right along the subject, even when the first match of the pattern was going to
2629 fail. The "needed" character is now not set for anchored patterns, unless it
2630 follows something in the pattern that is of non-fixed length. Thus, it still
2631 fulfils its original purpose of finding quick non-matches in cases of nested
2632 unlimited repeats, but isn't used for simple anchored patterns such as /^abc/.
2635 Version 4.0 17-Feb-03
2636 ---------------------
2638 1. If a comment in an extended regex that started immediately after a meta-item
2639 extended to the end of string, PCRE compiled incorrect data. This could lead to
2640 all kinds of weird effects. Example: /#/ was bad; /()#/ was bad; /a#/ was not.
2642 2. Moved to autoconf 2.53 and libtool 1.4.2.
2644 3. Perl 5.8 no longer needs "use utf8" for doing UTF-8 things. Consequently,
2645 the special perltest8 script is no longer needed - all the tests can be run
2646 from a single perltest script.
2648 4. From 5.004, Perl has not included the VT character (0x0b) in the set defined
2649 by \s. It has now been removed in PCRE. This means it isn't recognized as
2650 whitespace in /x regexes too, which is the same as Perl. Note that the POSIX
2651 class [:space:] *does* include VT, thereby creating a mess.
2653 5. Added the class [:blank:] (a GNU extension from Perl 5.8) to match only
2654 space and tab.
2656 6. Perl 5.005 was a long time ago. It's time to amalgamate the tests that use
2657 its new features into the main test script, reducing the number of scripts.
2659 7. Perl 5.8 has changed the meaning of patterns like /a(?i)b/. Earlier versions
2660 were backward compatible, and made the (?i) apply to the whole pattern, as if
2661 /i were given. Now it behaves more logically, and applies the option setting
2662 only to what follows. PCRE has been changed to follow suit. However, if it
2663 finds options settings right at the start of the pattern, it extracts them into
2664 the global options, as before. Thus, they show up in the info data.
2666 8. Added support for the \Q...\E escape sequence. Characters in between are
2667 treated as literals. This is slightly different from Perl in that $ and @ are
2668 also handled as literals inside the quotes. In Perl, they will cause variable
2669 interpolation. Note the following examples:
2671 Pattern PCRE matches Perl matches
2673 \Qabc$xyz\E abc$xyz abc followed by the contents of $xyz
2674 \Qabc\$xyz\E abc\$xyz abc\$xyz
2675 \Qabc\E\$\Qxyz\E abc$xyz abc$xyz
2677 For compatibility with Perl, \Q...\E sequences are recognized inside character
2678 classes as well as outside them.
2680 9. Re-organized 3 code statements in pcretest to avoid "overflow in
2681 floating-point constant arithmetic" warnings from a Microsoft compiler. Added a
2682 (size_t) cast to one statement in pcretest and one in pcreposix to avoid
2683 signed/unsigned warnings.
2685 10. SunOS4 doesn't have strtoul(). This was used only for unpicking the -o
2686 option for pcretest, so I've replaced it by a simple function that does just
2687 that job.
2689 11. pcregrep was ending with code 0 instead of 2 for the commands "pcregrep" or
2690 "pcregrep -".
2692 12. Added "possessive quantifiers" ?+, *+, ++, and {,}+ which come from Sun's
2693 Java package. This provides some syntactic sugar for simple cases of what my
2694 documentation calls "once-only subpatterns". A pattern such as x*+ is the same
2695 as (?>x*). In other words, if what is inside (?>...) is just a single repeated
2696 item, you can use this simplified notation. Note that only makes sense with
2697 greedy quantifiers. Consequently, the use of the possessive quantifier forces
2698 greediness, whatever the setting of the PCRE_UNGREEDY option.
2700 13. A change of greediness default within a pattern was not taking effect at
2701 the current level for patterns like /(b+(?U)a+)/. It did apply to parenthesized
2702 subpatterns that followed. Patterns like /b+(?U)a+/ worked because the option
2703 was abstracted outside.
2705 14. PCRE now supports the \G assertion. It is true when the current matching
2706 position is at the start point of the match. This differs from \A when the
2707 starting offset is non-zero. Used with the /g option of pcretest (or similar
2708 code), it works in the same way as it does for Perl's /g option. If all
2709 alternatives of a regex begin with \G, the expression is anchored to the start
2710 match position, and the "anchored" flag is set in the compiled expression.
2712 15. Some bugs concerning the handling of certain option changes within patterns
2713 have been fixed. These applied to options other than (?ims). For example,
2714 "a(?x: b c )d" did not match "XabcdY" but did match "Xa b c dY". It should have
2715 been the other way round. Some of this was related to change 7 above.
2717 16. PCRE now gives errors for /[.x.]/ and /[=x=]/ as unsupported POSIX
2718 features, as Perl does. Previously, PCRE gave the warnings only for /[[.x.]]/
2719 and /[[=x=]]/. PCRE now also gives an error for /[:name:]/ because it supports
2720 POSIX classes only within a class (e.g. /[[:alpha:]]/).
2722 17. Added support for Perl's \C escape. This matches one byte, even in UTF8
2723 mode. Unlike ".", it always matches newline, whatever the setting of
2724 PCRE_DOTALL. However, PCRE does not permit \C to appear in lookbehind
2725 assertions. Perl allows it, but it doesn't (in general) work because it can't
2726 calculate the length of the lookbehind. At least, that's the case for Perl
2727 5.8.0 - I've been told they are going to document that it doesn't work in
2728 future.
2730 18. Added an error diagnosis for escapes that PCRE does not support: these are
2731 \L, \l, \N, \P, \p, \U, \u, and \X.
2733 19. Although correctly diagnosing a missing ']' in a character class, PCRE was
2734 reading past the end of the pattern in cases such as /[abcd/.
2736 20. PCRE was getting more memory than necessary for patterns with classes that
2737 contained both POSIX named classes and other characters, e.g. /[[:space:]abc/.
2739 21. Added some code, conditional on #ifdef VPCOMPAT, to make life easier for
2740 compiling PCRE for use with Virtual Pascal.
2742 22. Small fix to the Makefile to make it work properly if the build is done
2743 outside the source tree.
2745 23. Added a new extension: a condition to go with recursion. If a conditional
2746 subpattern starts with (?(R) the "true" branch is used if recursion has
2747 happened, whereas the "false" branch is used only at the top level.
2749 24. When there was a very long string of literal characters (over 255 bytes
2750 without UTF support, over 250 bytes with UTF support), the computation of how
2751 much memory was required could be incorrect, leading to segfaults or other
2752 strange effects.
2754 25. PCRE was incorrectly assuming anchoring (either to start of subject or to
2755 start of line for a non-DOTALL pattern) when a pattern started with (.*) and
2756 there was a subsequent back reference to those brackets. This meant that, for
2757 example, /(.*)\d+\1/ failed to match "abc123bc". Unfortunately, it isn't
2758 possible to check for precisely this case. All we can do is abandon the
2759 optimization if .* occurs inside capturing brackets when there are any back
2760 references whatsoever. (See below for a better fix that came later.)
2762 26. The handling of the optimization for finding the first character of a
2763 non-anchored pattern, and for finding a character that is required later in the
2764 match were failing in some cases. This didn't break the matching; it just
2765 failed to optimize when it could. The way this is done has been re-implemented.
2767 27. Fixed typo in error message for invalid (?R item (it said "(?p").
2769 28. Added a new feature that provides some of the functionality that Perl
2770 provides with (?{...}). The facility is termed a "callout". The way it is done
2771 in PCRE is for the caller to provide an optional function, by setting
2772 pcre_callout to its entry point. Like pcre_malloc and pcre_free, this is a
2773 global variable. By default it is unset, which disables all calling out. To get
2774 the function called, the regex must include (?C) at appropriate points. This
2775 is, in fact, equivalent to (?C0), and any number <= 255 may be given with (?C).
2776 This provides a means of identifying different callout points. When PCRE
2777 reaches such a point in the regex, if pcre_callout has been set, the external
2778 function is called. It is provided with data in a structure called
2779 pcre_callout_block, which is defined in pcre.h. If the function returns 0,
2780 matching continues; if it returns a non-zero value, the match at the current
2781 point fails. However, backtracking will occur if possible. [This was changed
2782 later and other features added - see item 49 below.]
2784 29. pcretest is upgraded to test the callout functionality. It provides a
2785 callout function that displays information. By default, it shows the start of
2786 the match and the current position in the text. There are some new data escapes
2787 to vary what happens:
2789 \C+ in addition, show current contents of captured substrings
2790 \C- do not supply a callout function
2791 \C!n return 1 when callout number n is reached
2792 \C!n!m return 1 when callout number n is reached for the mth time
2794 30. If pcregrep was called with the -l option and just a single file name, it
2795 output "<stdin>" if a match was found, instead of the file name.
2797 31. Improve the efficiency of the POSIX API to PCRE. If the number of capturing
2798 slots is less than POSIX_MALLOC_THRESHOLD, use a block on the stack to pass to
2799 pcre_exec(). This saves a malloc/free per call. The default value of
2800 POSIX_MALLOC_THRESHOLD is 10; it can be changed by --with-posix-malloc-threshold
2801 when configuring.
2803 32. The default maximum size of a compiled pattern is 64K. There have been a
2804 few cases of people hitting this limit. The code now uses macros to handle the
2805 storing of links as offsets within the compiled pattern. It defaults to 2-byte
2806 links, but this can be changed to 3 or 4 bytes by --with-link-size when
2807 configuring. Tests 2 and 5 work only with 2-byte links because they output
2808 debugging information about compiled patterns.
2810 33. Internal code re-arrangements:
2812 (a) Moved the debugging function for printing out a compiled regex into
2813 its own source file (printint.c) and used #include to pull it into
2814 pcretest.c and, when DEBUG is defined, into pcre.c, instead of having two
2815 separate copies.
2817 (b) Defined the list of op-code names for debugging as a macro in
2818 internal.h so that it is next to the definition of the opcodes.
2820 (c) Defined a table of op-code lengths for simpler skipping along compiled
2821 code. This is again a macro in internal.h so that it is next to the
2822 definition of the opcodes.
2824 34. Added support for recursive calls to individual subpatterns, along the
2825 lines of Robin Houston's patch (but implemented somewhat differently).
2827 35. Further mods to the Makefile to help Win32. Also, added code to pcregrep to
2828 allow it to read and process whole directories in Win32. This code was
2829 contributed by Lionel Fourquaux; it has not been tested by me.
2831 36. Added support for named subpatterns. The Python syntax (?P<name>...) is
2832 used to name a group. Names consist of alphanumerics and underscores, and must
2833 be unique. Back references use the syntax (?P=name) and recursive calls use
2834 (?P>name) which is a PCRE extension to the Python extension. Groups still have
2835 numbers. The function pcre_fullinfo() can be used after compilation to extract
2836 a name/number map. There are three relevant calls:
2838 PCRE_INFO_NAMEENTRYSIZE yields the size of each entry in the map
2839 PCRE_INFO_NAMECOUNT yields the number of entries
2840 PCRE_INFO_NAMETABLE yields a pointer to the map.
2842 The map is a vector of fixed-size entries. The size of each entry depends on
2843 the length of the longest name used. The first two bytes of each entry are the
2844 group number, most significant byte first. There follows the corresponding
2845 name, zero terminated. The names are in alphabetical order.
2847 37. Make the maximum literal string in the compiled code 250 for the non-UTF-8
2848 case instead of 255. Making it the same both with and without UTF-8 support
2849 means that the same test output works with both.
2851 38. There was a case of malloc(0) in the POSIX testing code in pcretest. Avoid
2852 calling malloc() with a zero argument.
2854 39. Change 25 above had to resort to a heavy-handed test for the .* anchoring
2855 optimization. I've improved things by keeping a bitmap of backreferences with
2856 numbers 1-31 so that if .* occurs inside capturing brackets that are not in
2857 fact referenced, the optimization can be applied. It is unlikely that a
2858 relevant occurrence of .* (i.e. one which might indicate anchoring or forcing
2859 the match to follow \n) will appear inside brackets with a number greater than
2860 31, but if it does, any back reference > 31 suppresses the optimization.
2862 40. Added a new compile-time option PCRE_NO_AUTO_CAPTURE. This has the effect
2863 of disabling numbered capturing parentheses. Any opening parenthesis that is
2864 not followed by ? behaves as if it were followed by ?: but named parentheses
2865 can still be used for capturing (and they will acquire numbers in the usual
2866 way).
2868 41. Redesigned the return codes from the match() function into yes/no/error so
2869 that errors can be passed back from deep inside the nested calls. A malloc
2870 failure while inside a recursive subpattern call now causes the
2871 PCRE_ERROR_NOMEMORY return instead of quietly going wrong.
2873 42. It is now possible to set a limit on the number of times the match()
2874 function is called in a call to pcre_exec(). This facility makes it possible to
2875 limit the amount of recursion and backtracking, though not in a directly
2876 obvious way, because the match() function is used in a number of different
2877 circumstances. The count starts from zero for each position in the subject
2878 string (for non-anchored patterns). The default limit is, for compatibility, a
2879 large number, namely 10 000 000. You can change this in two ways:
2881 (a) When configuring PCRE before making, you can use --with-match-limit=n
2882 to set a default value for the compiled library.
2884 (b) For each call to pcre_exec(), you can pass a pcre_extra block in which
2885 a different value is set. See 45 below.
2887 If the limit is exceeded, pcre_exec() returns PCRE_ERROR_MATCHLIMIT.
2889 43. Added a new function pcre_config(int, void *) to enable run-time extraction
2890 of things that can be changed at compile time. The first argument specifies
2891 what is wanted and the second points to where the information is to be placed.
2892 The current list of available information is:
2896 The output is an integer that is set to one if UTF-8 support is available;
2897 otherwise it is set to zero.
2901 The output is an integer that it set to the value of the code that is used for
2902 newline. It is either LF (10) or CR (13).
2906 The output is an integer that contains the number of bytes used for internal
2907 linkage in compiled expressions. The value is 2, 3, or 4. See item 32 above.
2911 The output is an integer that contains the threshold above which the POSIX
2912 interface uses malloc() for output vectors. See item 31 above.
2916 The output is an unsigned integer that contains the default limit of the number
2917 of match() calls in a pcre_exec() execution. See 42 above.
2919 44. pcretest has been upgraded by the addition of the -C option. This causes it
2920 to extract all the available output from the new pcre_config() function, and to
2921 output it. The program then exits immediately.
2923 45. A need has arisen to pass over additional data with calls to pcre_exec() in
2924 order to support additional features. One way would have been to define
2925 pcre_exec2() (for example) with extra arguments, but this would not have been
2926 extensible, and would also have required all calls to the original function to
2927 be mapped to the new one. Instead, I have chosen to extend the mechanism that
2928 is used for passing in "extra" data from pcre_study().
2930 The pcre_extra structure is now exposed and defined in pcre.h. It currently
2931 contains the following fields:
2933 flags a bitmap indicating which of the following fields are set
2934 study_data opaque data from pcre_study()
2935 match_limit a way of specifying a limit on match() calls for a specific
2936 call to pcre_exec()
2937 callout_data data for callouts (see 49 below)
2939 The flag bits are also defined in pcre.h, and are
2945 The pcre_study() function now returns one of these new pcre_extra blocks, with
2946 the actual study data pointed to by the study_data field, and the
2947 PCRE_EXTRA_STUDY_DATA flag set. This can be passed directly to pcre_exec() as
2948 before. That is, this change is entirely upwards-compatible and requires no
2949 change to existing code.
2951 If you want to pass in additional data to pcre_exec(), you can either place it
2952 in a pcre_extra block provided by pcre_study(), or create your own pcre_extra
2953 block.
2955 46. pcretest has been extended to test the PCRE_EXTRA_MATCH_LIMIT feature. If a
2956 data string contains the escape sequence \M, pcretest calls pcre_exec() several
2957 times with different match limits, until it finds the minimum value needed for
2958 pcre_exec() to complete. The value is then output. This can be instructive; for
2959 most simple matches the number is quite small, but for pathological cases it
2960 gets very large very quickly.
2962 47. There's a new option for pcre_fullinfo() called PCRE_INFO_STUDYSIZE. It
2963 returns the size of the data block pointed to by the study_data field in a
2964 pcre_extra block, that is, the value that was passed as the argument to
2965 pcre_malloc() when PCRE was getting memory in which to place the information
2966 created by pcre_study(). The fourth argument should point to a size_t variable.
2967 pcretest has been extended so that this information is shown after a successful
2968 pcre_study() call when information about the compiled regex is being displayed.
2970 48. Cosmetic change to Makefile: there's no need to have / after $(DESTDIR)
2971 because what follows is always an absolute path. (Later: it turns out that this
2972 is more than cosmetic for MinGW, because it doesn't like empty path
2973 components.)
2975 49. Some changes have been made to the callout feature (see 28 above):
2977 (i) A callout function now has three choices for what it returns:
2979 0 => success, carry on matching
2980 > 0 => failure at this point, but backtrack if possible
2981 < 0 => serious error, return this value from pcre_exec()
2983 Negative values should normally be chosen from the set of PCRE_ERROR_xxx
2984 values. In particular, returning PCRE_ERROR_NOMATCH forces a standard
2985 "match failed" error. The error number PCRE_ERROR_CALLOUT is reserved for
2986 use by callout functions. It will never be used by PCRE itself.
2988 (ii) The pcre_extra structure (see 45 above) has a void * field called
2989 callout_data, with corresponding flag bit PCRE_EXTRA_CALLOUT_DATA. The
2990 pcre_callout_block structure has a field of the same name. The contents of
2991 the field passed in the pcre_extra structure are passed to the callout
2992 function in the corresponding field in the callout block. This makes it
2993 easier to use the same callout-containing regex from multiple threads. For
2994 testing, the pcretest program has a new data escape
2996 \C*n pass the number n (may be negative) as callout_data
2998 If the callout function in pcretest receives a non-zero value as
2999 callout_data, it returns that value.
3001 50. Makefile wasn't handling CFLAGS properly when compiling dftables. Also,
3002 there were some redundant $(CFLAGS) in commands that are now specified as
3003 $(LINK), which already includes $(CFLAGS).
3005 51. Extensions to UTF-8 support are listed below. These all apply when (a) PCRE
3006 has been compiled with UTF-8 support *and* pcre_compile() has been compiled
3007 with the PCRE_UTF8 flag. Patterns that are compiled without that flag assume
3008 one-byte characters throughout. Note that case-insensitive matching applies
3009 only to characters whose values are less than 256. PCRE doesn't support the
3010 notion of cases for higher-valued characters.
3012 (i) A character class whose characters are all within 0-255 is handled as
3013 a bit map, and the map is inverted for negative classes. Previously, a
3014 character > 255 always failed to match such a class; however it should
3015 match if the class was a negative one (e.g. [^ab]). This has been fixed.
3017 (ii) A negated character class with a single character < 255 is coded as
3018 "not this character" (OP_NOT). This wasn't working properly when the test
3019 character was multibyte, either singly or repeated.
3021 (iii) Repeats of multibyte characters are now handled correctly in UTF-8
3022 mode, for example: \x{100}{2,3}.
3024 (iv) The character escapes \b, \B, \d, \D, \s, \S, \w, and \W (either
3025 singly or repeated) now correctly test multibyte characters. However,
3026 PCRE doesn't recognize any characters with values greater than 255 as
3027 digits, spaces, or word characters. Such characters always match \D, \S,
3028 and \W, and never match \d, \s, or \w.
3030 (v) Classes may now contain characters and character ranges with values
3031 greater than 255. For example: [ab\x{100}-\x{400}].
3033 (vi) pcregrep now has a --utf-8 option (synonym -u) which makes it call
3034 PCRE in UTF-8 mode.
3036 52. The info request value PCRE_INFO_FIRSTCHAR has been renamed
3037 PCRE_INFO_FIRSTBYTE because it is a byte value. However, the old name is
3038 retained for backwards compatibility. (Note that LASTLITERAL is also a byte
3039 value.)
3041 53. The single man page has become too large. I have therefore split it up into
3042 a number of separate man pages. These also give rise to individual HTML pages;
3043 these are now put in a separate directory, and there is an index.html page that
3044 lists them all. Some hyperlinking between the pages has been installed.
3046 54. Added convenience functions for handling named capturing parentheses.
3048 55. Unknown escapes inside character classes (e.g. [\M]) and escapes that
3049 aren't interpreted therein (e.g. [\C]) are literals in Perl. This is now also
3050 true in PCRE, except when the PCRE_EXTENDED option is set, in which case they
3051 are faulted.
3053 56. Introduced HOST_CC and HOST_CFLAGS which can be set in the environment when
3054 calling configure. These values are used when compiling the dftables.c program
3055 which is run to generate the source of the default character tables. They
3056 default to the values of CC and CFLAGS. If you are cross-compiling PCRE,
3057 you will need to set these values.
3059 57. Updated the building process for Windows DLL, as provided by Fred Cox.
3062 Version 3.9 02-Jan-02
3063 ---------------------
3065 1. A bit of extraneous text had somehow crept into the pcregrep documentation.
3067 2. If --disable-static was given, the building process failed when trying to
3068 build pcretest and pcregrep. (For some reason it was using libtool to compile
3069 them, which is not right, as they aren't part of the library.)
3072 Version 3.8 18-Dec-01
3073 ---------------------
3075 1. The experimental UTF-8 code was completely screwed up. It was packing the
3076 bytes in the wrong order. How dumb can you get?
3079 Version 3.7 29-Oct-01
3080 ---------------------
3082 1. In updating pcretest to check change 1 of version 3.6, I screwed up.
3083 This caused pcretest, when used on the test data, to segfault. Unfortunately,
3084 this didn't happen under Solaris 8, where I normally test things.
3086 2. The Makefile had to be changed to make it work on BSD systems, where 'make'
3087 doesn't seem to recognize that ./xxx and xxx are the same file. (This entry
3088 isn't in ChangeLog distributed with 3.7 because I forgot when I hastily made
3089 this fix an hour or so after the initial 3.7 release.)
3092 Version 3.6 23-Oct-01
3093 ---------------------
3095 1. Crashed with /(sens|respons)e and \1ibility/ and "sense and sensibility" if
3096 offsets passed as NULL with zero offset count.
3098 2. The config.guess and config.sub files had not been updated when I moved to
3099 the latest autoconf.
3102 Version 3.5 15-Aug-01
3103 ---------------------
3105 1. Added some missing #if !defined NOPOSIX conditionals in pcretest.c that
3106 had been forgotten.
3108 2. By using declared but undefined structures, we can avoid using "void"
3109 definitions in pcre.h while keeping the internal definitions of the structures
3110 private.
3112 3. The distribution is now built using autoconf 2.50 and libtool 1.4. From a
3113 user point of view, this means that both static and shared libraries are built
3114 by default, but this can be individually controlled. More of the work of
3115 handling this static/shared cases is now inside libtool instead of PCRE's make
3116 file.
3118 4. The pcretest utility is now installed along with pcregrep because it is
3119 useful for users (to test regexs) and by doing this, it automatically gets
3120 relinked by libtool. The documentation has been turned into a man page, so
3121 there are now .1, .txt, and .html versions in /doc.
3123 5. Upgrades to pcregrep:
3124 (i) Added long-form option names like gnu grep.
3125 (ii) Added --help to list all options with an explanatory phrase.
3126 (iii) Added -r, --recursive to recurse into sub-directories.
3127 (iv) Added -f, --file to read patterns from a file.
3129 6. pcre_exec() was referring to its "code" argument before testing that
3130 argument for NULL (and giving an error if it was NULL).
3132 7. Upgraded Makefile.in to allow for compiling in a different directory from
3133 the source directory.
3135 8. Tiny buglet in pcretest: when pcre_fullinfo() was called to retrieve the
3136 options bits, the pointer it was passed was to an int instead of to an unsigned
3137 long int. This mattered only on 64-bit systems.
3139 9. Fixed typo (3.4/1) in pcre.h again. Sigh. I had changed pcre.h (which is
3140 generated) instead of pcre.in, which it its source. Also made the same change
3141 in several of the .c files.
3143 10. A new release of gcc defines printf() as a macro, which broke pcretest
3144 because it had an ifdef in the middle of a string argument for printf(). Fixed
3145 by using separate calls to printf().
3147 11. Added --enable-newline-is-cr and --enable-newline-is-lf to the configure
3148 script, to force use of CR or LF instead of \n in the source. On non-Unix
3149 systems, the value can be set in config.h.
3151 12. The limit of 200 on non-capturing parentheses is a _nesting_ limit, not an
3152 absolute limit. Changed the text of the error message to make this clear, and
3153 likewise updated the man page.
3155 13. The limit of 99 on the number of capturing subpatterns has been removed.
3156 The new limit is 65535, which I hope will not be a "real" limit.
3159 Version 3.4 22-Aug-00
3160 ---------------------
3162 1. Fixed typo in pcre.h: unsigned const char * changed to const unsigned char *.
3164 2. Diagnose condition (?(0) as an error instead of crashing on matching.
3167 Version 3.3 01-Aug-00
3168 ---------------------
3170 1. If an octal character was given, but the value was greater than \377, it
3171 was not getting masked to the least significant bits, as documented. This could
3172 lead to crashes in some systems.
3174 2. Perl 5.6 (if not earlier versions) accepts classes like [a-\d] and treats
3175 the hyphen as a literal. PCRE used to give an error; it now behaves like Perl.
3177 3. Added the functions pcre_free_substring() and pcre_free_substring_list().
3178 These just pass their arguments on to (pcre_free)(), but they are provided
3179 because some uses of PCRE bind it to non-C systems that can call its functions,
3180 but cannot call free() or pcre_free() directly.
3182 4. Add "make test" as a synonym for "make check". Corrected some comments in
3183 the Makefile.
3185 5. Add $(DESTDIR)/ in front of all the paths in the "install" target in the
3186 Makefile.
3188 6. Changed the name of pgrep to pcregrep, because Solaris has introduced a
3189 command called pgrep for grepping around the active processes.
3191 7. Added the beginnings of support for UTF-8 character strings.
3193 8. Arranged for the Makefile to pass over the settings of CC, CFLAGS, and
3194 RANLIB to ./ltconfig so that they are used by libtool. I think these are all
3195 the relevant ones. (AR is not passed because ./ltconfig does its own figuring
3196 out for the ar command.)
3199 Version 3.2 12-May-00
3200 ---------------------
3202 This is purely a bug fixing release.
3204 1. If the pattern /((Z)+|A)*/ was matched agained ZABCDEFG it matched Z instead
3205 of ZA. This was just one example of several cases that could provoke this bug,
3206 which was introduced by change 9 of version 2.00. The code for breaking
3207 infinite loops after an iteration that matches an empty string was't working
3208 correctly.
3210 2. The pcretest program was not imitating Perl correctly for the pattern /a*/g
3211 when matched against abbab (for example). After matching an empty string, it
3212 wasn't forcing anchoring when setting PCRE_NOTEMPTY for the next attempt; this
3213 caused it to match further down the string than it should.
3215 3. The code contained an inclusion of sys/types.h. It isn't clear why this
3216 was there because it doesn't seem to be needed, and it causes trouble on some
3217 systems, as it is not a Standard C header. It has been removed.
3219 4. Made 4 silly changes to the source to avoid stupid compiler warnings that
3220 were reported on the Macintosh. The changes were from
3222 while ((c = *(++ptr)) != 0 && c != '\n');
3223 to
3224 while ((c = *(++ptr)) != 0 && c != '\n') ;
3226 Totally extraordinary, but if that's what it takes...
3228 5. PCRE is being used in one environment where neither memmove() nor bcopy() is
3229 available. Added HAVE_BCOPY and an autoconf test for it; if neither
3230 HAVE_MEMMOVE nor HAVE_BCOPY is set, use a built-in emulation function which
3231 assumes the way PCRE uses memmove() (always moving upwards).
3233 6. PCRE is being used in one environment where strchr() is not available. There
3234 was only one use in pcre.c, and writing it out to avoid strchr() probably gives
3235 faster code anyway.
3238 Version 3.1 09-Feb-00
3239 ---------------------
3241 The only change in this release is the fixing of some bugs in Makefile.in for
3242 the "install" target:
3244 (1) It was failing to install pcreposix.h.
3246 (2) It was overwriting the pcre.3 man page with the pcreposix.3 man page.
3249 Version 3.0 01-Feb-00
3250 ---------------------
3252 1. Add support for the /+ modifier to perltest (to output $` like it does in
3253 pcretest).
3255 2. Add support for the /g modifier to perltest.
3257 3. Fix pcretest so that it behaves even more like Perl for /g when the pattern
3258 matches null strings.
3260 4. Fix perltest so that it doesn't do unwanted things when fed an empty
3261 pattern. Perl treats empty patterns specially - it reuses the most recent
3262 pattern, which is not what we want. Replace // by /(?#)/ in order to avoid this
3263 effect.
3265 5. The POSIX interface was broken in that it was just handing over the POSIX
3266 captured string vector to pcre_exec(), but (since release 2.00) PCRE has
3267 required a bigger vector, with some working space on the end. This means that
3268 the POSIX wrapper now has to get and free some memory, and copy the results.
3270 6. Added some simple autoconf support, placing the test data and the
3271 documentation in separate directories, re-organizing some of the
3272 information files, and making it build pcre-config (a GNU standard). Also added
3273 libtool support for building PCRE as a shared library, which is now the
3274 default.
3276 7. Got rid of the leading zero in the definition of PCRE_MINOR because 08 and
3277 09 are not valid octal constants. Single digits will be used for minor values
3278 less than 10.
3280 8. Defined REG_EXTENDED and REG_NOSUB as zero in the POSIX header, so that
3281 existing programs that set these in the POSIX interface can use PCRE without
3282 modification.
3284 9. Added a new function, pcre_fullinfo() with an extensible interface. It can
3285 return all that pcre_info() returns, plus additional data. The pcre_info()
3286 function is retained for compatibility, but is considered to be obsolete.
3288 10. Added experimental recursion feature (?R) to handle one common case that
3289 Perl 5.6 will be able to do with (?p{...}).
3291 11. Added support for POSIX character classes like [:alpha:], which Perl is
3292 adopting.
3295 Version 2.08 31-Aug-99
3296 ----------------------
3298 1. When startoffset was not zero and the pattern began with ".*", PCRE was not
3299 trying to match at the startoffset position, but instead was moving forward to
3300 the next newline as if a previous match had failed.
3302 2. pcretest was not making use of PCRE_NOTEMPTY when repeating for /g and /G,
3303 and could get into a loop if a null string was matched other than at the start
3304 of the subject.
3306 3. Added definitions of PCRE_MAJOR and PCRE_MINOR to pcre.h so the version can
3307 be distinguished at compile time, and for completeness also added PCRE_DATE.
3309 5. Added Paul Sokolovsky's minor changes to make it easy to compile a Win32 DLL
3310 in GnuWin32 environments.
3313 Version 2.07 29-Jul-99
3314 ----------------------
3316 1. The documentation is now supplied in plain text form and HTML as well as in
3317 the form of man page sources.
3319 2. C++ compilers don't like assigning (void *) values to other pointer types.
3320 In particular this affects malloc(). Although there is no problem in Standard
3321 C, I've put in casts to keep C++ compilers happy.
3323 3. Typo on pcretest.c; a cast of (unsigned char *) in the POSIX regexec() call
3324 should be (const char *).
3326 4. If NOPOSIX is defined, pcretest.c compiles without POSIX support. This may
3327 be useful for non-Unix systems who don't want to bother with the POSIX stuff.
3328 However, I haven't made this a standard facility. The documentation doesn't
3329 mention it, and the Makefile doesn't support it.
3331 5. The Makefile now contains an "install" target, with editable destinations at
3332 the top of the file. The pcretest program is not installed.
3334 6. pgrep -V now gives the PCRE version number and date.
3336 7. Fixed bug: a zero repetition after a literal string (e.g. /abcde{0}/) was
3337 causing the entire string to be ignored, instead of just the last character.
3339 8. If a pattern like /"([^\\"]+|\\.)*"/ is applied in the normal way to a
3340 non-matching string, it can take a very, very long time, even for strings of
3341 quite modest length, because of the nested recursion. PCRE now does better in
3342 some of these cases. It does this by remembering the last required literal
3343 character in the pattern, and pre-searching the subject to ensure it is present
3344 before running the real match. In other words, it applies a heuristic to detect
3345 some types of certain failure quickly, and in the above example, if presented
3346 with a string that has no trailing " it gives "no match" very quickly.
3348 9. A new runtime option PCRE_NOTEMPTY causes null string matches to be ignored;
3349 other alternatives are tried instead.
3352 Version 2.06 09-Jun-99
3353 ----------------------
3355 1. Change pcretest's output for amount of store used to show just the code
3356 space, because the remainder (the data block) varies in size between 32-bit and
3357 64-bit systems.
3359 2. Added an extra argument to pcre_exec() to supply an offset in the subject to
3360 start matching at. This allows lookbehinds to work when searching for multiple
3361 occurrences in a string.
3363 3. Added additional options to pcretest for testing multiple occurrences:
3365 /+ outputs the rest of the string that follows a match
3366 /g loops for multiple occurrences, using the new startoffset argument
3367 /G loops for multiple occurrences by passing an incremented pointer
3369 4. PCRE wasn't doing the "first character" optimization for patterns starting
3370 with \b or \B, though it was doing it for other lookbehind assertions. That is,
3371 it wasn't noticing that a match for a pattern such as /\bxyz/ has to start with
3372 the letter 'x'. On long subject strings, this gives a significant speed-up.
3375 Version 2.05 21-Apr-99
3376 ----------------------
3378 1. Changed the type of magic_number from int to long int so that it works
3379 properly on 16-bit systems.
3381 2. Fixed a bug which caused patterns starting with .* not to work correctly
3382 when the subject string contained newline characters. PCRE was assuming
3383 anchoring for such patterns in all cases, which is not correct because .* will
3384 not pass a newline unless PCRE_DOTALL is set. It now assumes anchoring only if
3385 DOTALL is set at top level; otherwise it knows that patterns starting with .*
3386 must be retried after every newline in the subject.
3389 Version 2.04 18-Feb-99
3390 ----------------------
3392 1. For parenthesized subpatterns with repeats whose minimum was zero, the
3393 computation of the store needed to hold the pattern was incorrect (too large).
3394 If such patterns were nested a few deep, this could multiply and become a real
3395 problem.
3397 2. Added /M option to pcretest to show the memory requirement of a specific
3398 pattern. Made -m a synonym of -s (which does this globally) for compatibility.
3400 3. Subpatterns of the form (regex){n,m} (i.e. limited maximum) were being
3401 compiled in such a way that the backtracking after subsequent failure was
3402 pessimal. Something like (a){0,3} was compiled as (a)?(a)?(a)? instead of
3403 ((a)((a)(a)?)?)? with disastrous performance if the maximum was of any size.
3406 Version 2.03 02-Feb-99
3407 ----------------------
3409 1. Fixed typo and small mistake in man page.
3411 2. Added 4th condition (GPL supersedes if conflict) and created separate
3412 LICENCE file containing the conditions.
3414 3. Updated pcretest so that patterns such as /abc\/def/ work like they do in
3415 Perl, that is the internal \ allows the delimiter to be included in the
3416 pattern. Locked out the use of \ as a delimiter. If \ immediately follows
3417 the final delimiter, add \ to the end of the pattern (to test the error).
3419 4. Added the convenience functions for extracting substrings after a successful
3420 match. Updated pcretest to make it able to test these functions.
3423 Version 2.02 14-Jan-99
3424 ----------------------
3426 1. Initialized the working variables associated with each extraction so that
3427 their saving and restoring doesn't refer to uninitialized store.
3429 2. Put dummy code into study.c in order to trick the optimizer of the IBM C
3430 compiler for OS/2 into generating correct code. Apparently IBM isn't going to
3431 fix the problem.
3433 3. Pcretest: the timing code wasn't using LOOPREPEAT for timing execution
3434 calls, and wasn't printing the correct value for compiling calls. Increased the
3435 default value of LOOPREPEAT, and the number of significant figures in the
3436 times.
3438 4. Changed "/bin/rm" in the Makefile to "-rm" so it works on Windows NT.
3440 5. Renamed "deftables" as "dftables" to get it down to 8 characters, to avoid
3441 a building problem on Windows NT with a FAT file system.
3444 Version 2.01 21-Oct-98
3445 ----------------------
3447 1. Changed the API for pcre_compile() to allow for the provision of a pointer
3448 to character tables built by pcre_maketables() in the current locale. If NULL
3449 is passed, the default tables are used.
3452 Version 2.00 24-Sep-98
3453 ----------------------
3455 1. Since the (>?) facility is in Perl 5.005, don't require PCRE_EXTRA to enable
3456 it any more.
3458 2. Allow quantification of (?>) groups, and make it work correctly.
3460 3. The first character computation wasn't working for (?>) groups.
3462 4. Correct the implementation of \Z (it is permitted to match on the \n at the
3463 end of the subject) and add 5.005's \z, which really does match only at the
3464 very end of the subject.
3466 5. Remove the \X "cut" facility; Perl doesn't have it, and (?> is neater.
3468 6. Remove the ability to specify CASELESS, MULTILINE, DOTALL, and
3469 DOLLAR_END_ONLY at runtime, to make it possible to implement the Perl 5.005
3470 localized options. All options to pcre_study() were also removed.
3472 7. Add other new features from 5.005:
3474 $(?<= positive lookbehind
3475 $(?<! negative lookbehind
3476 (?imsx-imsx) added the unsetting capability
3477 such a setting is global if at outer level; local otherwise
3478 (?imsx-imsx:) non-capturing groups with option setting
3479 (?(cond)re|re) conditional pattern matching
3481 A backreference to itself in a repeated group matches the previous
3482 captured string.
3484 8. General tidying up of studying (both automatic and via "study")
3485 consequential on the addition of new assertions.
3487 9. As in 5.005, unlimited repeated groups that could match an empty substring
3488 are no longer faulted at compile time. Instead, the loop is forcibly broken at
3489 runtime if any iteration does actually match an empty substring.
3491 10. Include the RunTest script in the distribution.
3493 11. Added tests from the Perl 5.005_02 distribution. This showed up a few
3494 discrepancies, some of which were old and were also with respect to 5.004. They
3495 have now been fixed.
3498 Version 1.09 28-Apr-98
3499 ----------------------
3501 1. A negated single character class followed by a quantifier with a minimum
3502 value of one (e.g. [^x]{1,6} ) was not compiled correctly. This could lead to
3503 program crashes, or just wrong answers. This did not apply to negated classes
3504 containing more than one character, or to minima other than one.
3507 Version 1.08 27-Mar-98
3508 ----------------------
3510 1. Add PCRE_UNGREEDY to invert the greediness of quantifiers.
3512 2. Add (?U) and (?X) to set PCRE_UNGREEDY and PCRE_EXTRA respectively. The
3513 latter must appear before anything that relies on it in the pattern.
3516 Version 1.07 16-Feb-98
3517 ----------------------
3519 1. A pattern such as /((a)*)*/ was not being diagnosed as in error (unlimited
3520 repeat of a potentially empty string).
3523 Version 1.06 23-Jan-98
3524 ----------------------
3526 1. Added Markus Oberhumer's little patches for C++.
3528 2. Literal strings longer than 255 characters were broken.
3531 Version 1.05 23-Dec-97
3532 ----------------------
3534 1. Negated character classes containing more than one character were failing if
3535 PCRE_CASELESS was set at run time.
3538 Version 1.04 19-Dec-97
3539 ----------------------
3541 1. Corrected the man page, where some "const" qualifiers had been omitted.
3543 2. Made debugging output print "{0,xxx}" instead of just "{,xxx}" to agree with
3544 input syntax.
3546 3. Fixed memory leak which occurred when a regex with back references was
3547 matched with an offsets vector that wasn't big enough. The temporary memory
3548 that is used in this case wasn't being freed if the match failed.
3550 4. Tidied pcretest to ensure it frees memory that it gets.
3552 5. Temporary memory was being obtained in the case where the passed offsets
3553 vector was exactly big enough.
3555 6. Corrected definition of offsetof() from change 5 below.
3557 7. I had screwed up change 6 below and broken the rules for the use of
3558 setjmp(). Now fixed.
3561 Version 1.03 18-Dec-97
3562 ----------------------
3564 1. A erroneous regex with a missing opening parenthesis was correctly
3565 diagnosed, but PCRE attempted to access brastack[-1], which could cause crashes
3566 on some systems.
3568 2. Replaced offsetof(real_pcre, code) by offsetof(real_pcre, code[0]) because
3569 it was reported that one broken compiler failed on the former because "code" is
3570 also an independent variable.
3572 3. The erroneous regex a[]b caused an array overrun reference.
3574 4. A regex ending with a one-character negative class (e.g. /[^k]$/) did not
3575 fail on data ending with that character. (It was going on too far, and checking
3576 the next character, typically a binary zero.) This was specific to the
3577 optimized code for single-character negative classes.
3579 5. Added a contributed patch from the TIN world which does the following:
3581 + Add an undef for memmove, in case the the system defines a macro for it.
3583 + Add a definition of offsetof(), in case there isn't one. (I don't know
3584 the reason behind this - offsetof() is part of the ANSI standard - but
3585 it does no harm).
3587 + Reduce the ifdef's in pcre.c using macro DPRINTF, thereby eliminating
3588 most of the places where whitespace preceded '#'. I have given up and
3589 allowed the remaining 2 cases to be at the margin.
3591 + Rename some variables in pcre to eliminate shadowing. This seems very
3592 pedantic, but does no harm, of course.
3594 6. Moved the call to setjmp() into its own function, to get rid of warnings
3595 from gcc -Wall, and avoided calling it at all unless PCRE_EXTRA is used.
3597 7. Constructs such as \d{8,} were compiling into the equivalent of
3598 \d{8}\d{0,65527} instead of \d{8}\d* which didn't make much difference to the
3599 outcome, but in this particular case used more store than had been allocated,
3600 which caused the bug to be discovered because it threw up an internal error.
3602 8. The debugging code in both pcre and pcretest for outputting the compiled
3603 form of a regex was going wrong in the case of back references followed by
3604 curly-bracketed repeats.
3607 Version 1.02 12-Dec-97
3608 ----------------------
3610 1. Typos in pcre.3 and comments in the source fixed.
3612 2. Applied a contributed patch to get rid of places where it used to remove
3613 'const' from variables, and fixed some signed/unsigned and uninitialized
3614 variable warnings.
3616 3. Added the "runtest" target to Makefile.
3618 4. Set default compiler flag to -O2 rather than just -O.
3621 Version 1.01 19-Nov-97
3622 ----------------------
3624 1. PCRE was failing to diagnose unlimited repeat of empty string for patterns
3625 like /([ab]*)*/, that is, for classes with more than one character in them.
3627 2. Likewise, it wasn't diagnosing patterns with "once-only" subpatterns, such
3628 as /((?>a*))*/ (a PCRE_EXTRA facility).
3631 Version 1.00 18-Nov-97
3632 ----------------------
3634 1. Added compile-time macros to support systems such as SunOS4 which don't have
3635 memmove() or strerror() but have other things that can be used instead.
3637 2. Arranged that "make clean" removes the executables.
3640 Version 0.99 27-Oct-97
3641 ----------------------
3643 1. Fixed bug in code for optimizing classes with only one character. It was
3644 initializing a 32-byte map regardless, which could cause it to run off the end
3645 of the memory it had got.
3647 2. Added, conditional on PCRE_EXTRA, the proposed (?>REGEX) construction.
3650 Version 0.98 22-Oct-97
3651 ----------------------
3653 1. Fixed bug in code for handling temporary memory usage when there are more
3654 back references than supplied space in the ovector. This could cause segfaults.
3657 Version 0.97 21-Oct-97
3658 ----------------------
3660 1. Added the \X "cut" facility, conditional on PCRE_EXTRA.
3662 2. Optimized negated single characters not to use a bit map.
3664 3. Brought error texts together as macro definitions; clarified some of them;
3665 fixed one that was wrong - it said "range out of order" when it meant "invalid
3666 escape sequence".
3668 4. Changed some char * arguments to const char *.
3670 5. Added PCRE_NOTBOL and PCRE_NOTEOL (from POSIX).
3672 6. Added the POSIX-style API wrapper in pcreposix.a and testing facilities in
3673 pcretest.
3676 Version 0.96 16-Oct-97
3677 ----------------------
3679 1. Added a simple "pgrep" utility to the distribution.
3681 2. Fixed an incompatibility with Perl: "{" is now treated as a normal character
3682 unless it appears in one of the precise forms "{ddd}", "{ddd,}", or "{ddd,ddd}"
3683 where "ddd" means "one or more decimal digits".
3685 3. Fixed serious bug. If a pattern had a back reference, but the call to
3686 pcre_exec() didn't supply a large enough ovector to record the related
3687 identifying subpattern, the match always failed. PCRE now remembers the number
3688 of the largest back reference, and gets some temporary memory in which to save
3689 the offsets during matching if necessary, in order to ensure that
3690 backreferences always work.
3692 4. Increased the compatibility with Perl in a number of ways:
3694 (a) . no longer matches \n by default; an option PCRE_DOTALL is provided
3695 to request this handling. The option can be set at compile or exec time.
3697 (b) $ matches before a terminating newline by default; an option
3698 PCRE_DOLLAR_ENDONLY is provided to override this (but not in multiline
3699 mode). The option can be set at compile or exec time.
3701 (c) The handling of \ followed by a digit other than 0 is now supposed to be
3702 the same as Perl's. If the decimal number it represents is less than 10
3703 or there aren't that many previous left capturing parentheses, an octal
3704 escape is read. Inside a character class, it's always an octal escape,
3705 even if it is a single digit.
3707 (d) An escaped but undefined alphabetic character is taken as a literal,
3708 unless PCRE_EXTRA is set. Currently this just reserves the remaining
3709 escapes.
3711 (e) {0} is now permitted. (The previous item is removed from the compiled
3712 pattern).
3714 5. Changed all the names of code files so that the basic parts are no longer
3715 than 10 characters, and abolished the teeny "globals.c" file.
3717 6. Changed the handling of character classes; they are now done with a 32-byte
3718 bit map always.
3720 7. Added the -d and /D options to pcretest to make it possible to look at the
3721 internals of compilation without having to recompile pcre.
3724 Version 0.95 23-Sep-97
3725 ----------------------
3727 1. Fixed bug in pre-pass concerning escaped "normal" characters such as \x5c or
3728 \x20 at the start of a run of normal characters. These were being treated as
3729 real characters, instead of the source characters being re-checked.
3732 Version 0.94 18-Sep-97
3733 ----------------------
3735 1. The functions are now thread-safe, with the caveat that the global variables
3736 containing pointers to malloc() and free() or alternative functions are the
3737 same for all threads.
3739 2. Get pcre_study() to generate a bitmap of initial characters for non-
3740 anchored patterns when this is possible, and use it if passed to pcre_exec().
3743 Version 0.93 15-Sep-97
3744 ----------------------
3746 1. /(b)|(:+)/ was computing an incorrect first character.
3748 2. Add pcre_study() to the API and the passing of pcre_extra to pcre_exec(),
3749 but not actually doing anything yet.
3751 3. Treat "-" characters in classes that cannot be part of ranges as literals,
3752 as Perl does (e.g. [-az] or [az-]).
3754 4. Set the anchored flag if a branch starts with .* or .*? because that tests
3755 all possible positions.
3757 5. Split up into different modules to avoid including unneeded functions in a
3758 compiled binary. However, compile and exec are still in one module. The "study"
3759 function is split off.
3761 6. The character tables are now in a separate module whose source is generated
3762 by an auxiliary program - but can then be edited by hand if required. There are
3763 now no calls to isalnum(), isspace(), isdigit(), isxdigit(), tolower() or
3764 toupper() in the code.
3766 7. Turn the malloc/free funtions variables into pcre_malloc and pcre_free and
3767 make them global. Abolish the function for setting them, as the caller can now
3768 set them directly.
3771 Version 0.92 11-Sep-97
3772 ----------------------
3774 1. A repeat with a fixed maximum and a minimum of 1 for an ordinary character
3775 (e.g. /a{1,3}/) was broken (I mis-optimized it).
3777 2. Caseless matching was not working in character classes if the characters in
3778 the pattern were in upper case.
3780 3. Make ranges like [W-c] work in the same way as Perl for caseless matching.
3782 4. Make PCRE_ANCHORED public and accept as a compile option.
3784 5. Add an options word to pcre_exec() and accept PCRE_ANCHORED and
3785 PCRE_CASELESS at run time. Add escapes \A and \I to pcretest to cause it to
3786 pass them.
3788 6. Give an error if bad option bits passed at compile or run time.
3790 7. Add PCRE_MULTILINE at compile and exec time, and (?m) as well. Add \M to
3791 pcretest to cause it to pass that flag.
3793 8. Add pcre_info(), to get the number of identifying subpatterns, the stored
3794 options, and the first character, if set.
3796 9. Recognize C+ or C{n,m} where n >= 1 as providing a fixed starting character.
3799 Version 0.91 10-Sep-97
3800 ----------------------
3802 1. PCRE was failing to diagnose unlimited repeats of subpatterns that could
3803 match the empty string as in /(a*)*/. It was looping and ultimately crashing.
3805 2. PCRE was looping on encountering an indefinitely repeated back reference to
3806 a subpattern that had matched an empty string, e.g. /(a|)\1*/. It now does what
3807 Perl does - treats the match as successful.
3809 ****


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

  ViewVC Help
Powered by ViewVC 1.1.5