ViewVC logotype

Contents of /code/trunk/ChangeLog

Parent Directory Parent Directory | Revision Log Revision Log

Revision 499 - (show annotations)
Wed Mar 3 20:09:39 2010 UTC (5 years, 9 months ago) by ph10
File size: 183343 byte(s)
Add some checks for the eint vector size and the list of compile-time error 
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.
12 3. Updated the licensing terms in the pcregexp.pas file, as agreed with the
13 original author of that file, following a query about its status.
15 4. On systems that do not have stdint.h (e.g. Solaris), check for and include
16 inttypes.h instead. This fixes a bug that was introduced by change 8.01/8.
18 5. A pattern such as (?&t)*+(?(DEFINE)(?<t>.)) which has a possessive
19 quantifier applied to a forward-referencing subroutine call, could compile
20 incorrect code or give the error "internal error: previously-checked
21 referenced subpattern not found".
23 6. Both MS Visual Studio and Symbian OS have problems with initializing
24 variables to point to external functions. For these systems, therefore,
25 pcre_malloc etc. are now initialized to local functions that call the
26 relevant global functions.
28 7. There were two entries missing in the vectors called coptable and poptable
29 in pcre_dfa_exec.c. This could lead to memory accesses outsize the vectors.
30 I've fixed the data, and added a kludgy way of testing at compile time that
31 the lengths are correct (equal to the number of opcodes).
33 8. Following on from 7, I added a similar kludge to check the length of the
34 eint vector in pcreposix.c.
36 9. Error texts for pcre_compile() are held as one long string to avoid too
37 much relocation at load time. To find a text, the string is searched,
38 counting zeros. There was no check for running off the end of the string,
39 which could happen if a new error number was added without updating the
40 string.
43 Version 8.01 19-Jan-2010
44 ------------------------
46 1. If a pattern contained a conditional subpattern with only one branch (in
47 particular, this includes all (*DEFINE) patterns), a call to pcre_study()
48 computed the wrong minimum data length (which is of course zero for such
49 subpatterns). This could cause incorrect "no match" results.
51 2. For patterns such as (?i)a(?-i)b|c where an option setting at the start of
52 the pattern is reset in the first branch, pcre_compile() failed with
53 "internal error: code overflow at offset...". This happened only when
54 the reset was to the original external option setting. (An optimization
55 abstracts leading options settings into an external setting, which was the
56 cause of this.)
58 3. A pattern such as ^(?!a(*SKIP)b) where a negative assertion contained one
59 of the verbs SKIP, PRUNE, or COMMIT, did not work correctly. When the
60 assertion pattern did not match (meaning that the assertion was true), it
61 was incorrectly treated as false if the SKIP had been reached during the
62 matching. This also applied to assertions used as conditions.
64 4. If an item that is not supported by pcre_dfa_exec() was encountered in an
65 assertion subpattern, including such a pattern used as a condition,
66 unpredictable results occurred, instead of the error return
69 5. The C++ GlobalReplace function was not working like Perl for the special
70 situation when an empty string is matched. It now does the fancy magic
71 stuff that is necessary.
73 6. In pcre_internal.h, obsolete includes to setjmp.h and stdarg.h have been
74 removed. (These were left over from very, very early versions of PCRE.)
76 7. Some cosmetic changes to the code to make life easier when compiling it
77 as part of something else:
79 (a) Change DEBUG to PCRE_DEBUG.
81 (b) In pcre_compile(), rename the member of the "branch_chain" structure
82 called "current" as "current_branch", to prevent a collision with the
83 Linux macro when compiled as a kernel module.
85 (c) In pcre_study(), rename the function set_bit() as set_table_bit(), to
86 prevent a collision with the Linux macro when compiled as a kernel
87 module.
89 8. In pcre_compile() there are some checks for integer overflows that used to
90 cast potentially large values to (double). This has been changed to that
91 when building, a check for int64_t is made, and if it is found, it is used
92 instead, thus avoiding the use of floating point arithmetic. (There is no
93 other use of FP in PCRE.) If int64_t is not found, the fallback is to
94 double.
96 9. Added two casts to avoid signed/unsigned warnings from VS Studio Express
97 2005 (difference between two addresses compared to an unsigned value).
99 10. Change the standard AC_CHECK_LIB test for libbz2 in configure.ac to a
100 custom one, because of the following reported problem in Windows:
102 - libbz2 uses the Pascal calling convention (WINAPI) for the functions
103 under Win32.
104 - The standard autoconf AC_CHECK_LIB fails to include "bzlib.h",
105 therefore missing the function definition.
106 - The compiler thus generates a "C" signature for the test function.
107 - The linker fails to find the "C" function.
108 - PCRE fails to configure if asked to do so against libbz2.
110 11. When running libtoolize from libtool-2.2.6b as part of autogen.sh, these
111 messages were output:
113 Consider adding `AC_CONFIG_MACRO_DIR([m4])' to configure.ac and
114 rerunning libtoolize, to keep the correct libtool macros in-tree.
115 Consider adding `-I m4' to ACLOCAL_AMFLAGS in Makefile.am.
117 I have done both of these things.
119 12. Although pcre_dfa_exec() does not use nearly as much stack as pcre_exec()
120 most of the time, it *can* run out if it is given a pattern that contains a
121 runaway infinite recursion. I updated the discussion in the pcrestack man
122 page.
124 13. Now that we have gone to the x.xx style of version numbers, the minor
125 version may start with zero. Using 08 or 09 is a bad idea because users
126 might check the value of PCRE_MINOR in their code, and 08 or 09 may be
127 interpreted as invalid octal numbers. I've updated the previous comment in
128 configure.ac, and also added a check that gives an error if 08 or 09 are
129 used.
131 14. Change 8.00/11 was not quite complete: code had been accidentally omitted,
132 causing partial matching to fail when the end of the subject matched \W
133 in a UTF-8 pattern where \W was quantified with a minimum of 3.
135 15. There were some discrepancies between the declarations in pcre_internal.h
136 of _pcre_is_newline(), _pcre_was_newline(), and _pcre_valid_utf8() and
137 their definitions. The declarations used "const uschar *" and the
138 definitions used USPTR. Even though USPTR is normally defined as "const
139 unsigned char *" (and uschar is typedeffed as "unsigned char"), it was
140 reported that: "This difference in casting confuses some C++ compilers, for
141 example, SunCC recognizes above declarations as different functions and
142 generates broken code for hbpcre." I have changed the declarations to use
143 USPTR.
145 16. GNU libtool is named differently on some systems. The autogen.sh script now
146 tries several variants such as glibtoolize (MacOSX) and libtoolize1x
147 (FreeBSD).
149 17. Applied Craig's patch that fixes an HP aCC compile error in pcre 8.00
150 (strtoXX undefined when compiling pcrecpp.cc). The patch contains this
151 comment: "Figure out how to create a longlong from a string: strtoll and
152 equivalent. It's not enough to call AC_CHECK_FUNCS: hpux has a strtoll, for
153 instance, but it only takes 2 args instead of 3!"
155 18. A subtle bug concerned with back references has been fixed by a change of
156 specification, with a corresponding code fix. A pattern such as
157 ^(xa|=?\1a)+$ which contains a back reference inside the group to which it
158 refers, was giving matches when it shouldn't. For example, xa=xaaa would
159 match that pattern. Interestingly, Perl (at least up to 5.11.3) has the
160 same bug. Such groups have to be quantified to be useful, or contained
161 inside another quantified group. (If there's no repetition, the reference
162 can never match.) The problem arises because, having left the group and
163 moved on to the rest of the pattern, a later failure that backtracks into
164 the group uses the captured value from the final iteration of the group
165 rather than the correct earlier one. I have fixed this in PCRE by forcing
166 any group that contains a reference to itself to be an atomic group; that
167 is, there cannot be any backtracking into it once it has completed. This is
168 similar to recursive and subroutine calls.
171 Version 8.00 19-Oct-09
172 ----------------------
174 1. The table for translating pcre_compile() error codes into POSIX error codes
175 was out-of-date, and there was no check on the pcre_compile() error code
176 being within the table. This could lead to an OK return being given in
177 error.
179 2. Changed the call to open a subject file in pcregrep from fopen(pathname,
180 "r") to fopen(pathname, "rb"), which fixed a problem with some of the tests
181 in a Windows environment.
183 3. The pcregrep --count option prints the count for each file even when it is
184 zero, as does GNU grep. However, pcregrep was also printing all files when
185 --files-with-matches was added. Now, when both options are given, it prints
186 counts only for those files that have at least one match. (GNU grep just
187 prints the file name in this circumstance, but including the count seems
188 more useful - otherwise, why use --count?) Also ensured that the
189 combination -clh just lists non-zero counts, with no names.
191 4. The long form of the pcregrep -F option was incorrectly implemented as
192 --fixed_strings instead of --fixed-strings. This is an incompatible change,
193 but it seems right to fix it, and I didn't think it was worth preserving
194 the old behaviour.
196 5. The command line items --regex=pattern and --regexp=pattern were not
197 recognized by pcregrep, which required --regex pattern or --regexp pattern
198 (with a space rather than an '='). The man page documented the '=' forms,
199 which are compatible with GNU grep; these now work.
201 6. No libpcreposix.pc file was created for pkg-config; there was just
202 libpcre.pc and libpcrecpp.pc. The omission has been rectified.
204 7. Added #ifndef SUPPORT_UCP into the pcre_ucd.c module, to reduce its size
205 when UCP support is not needed, by modifying the Python script that
206 generates it from Unicode data files. This should not matter if the module
207 is correctly used as a library, but I received one complaint about 50K of
208 unwanted data. My guess is that the person linked everything into his
209 program rather than using a library. Anyway, it does no harm.
211 8. A pattern such as /\x{123}{2,2}+/8 was incorrectly compiled; the trigger
212 was a minimum greater than 1 for a wide character in a possessive
213 repetition. The same bug could also affect patterns like /(\x{ff}{0,2})*/8
214 which had an unlimited repeat of a nested, fixed maximum repeat of a wide
215 character. Chaos in the form of incorrect output or a compiling loop could
216 result.
218 9. The restrictions on what a pattern can contain when partial matching is
219 requested for pcre_exec() have been removed. All patterns can now be
220 partially matched by this function. In addition, if there are at least two
221 slots in the offset vector, the offset of the earliest inspected character
222 for the match and the offset of the end of the subject are set in them when
223 PCRE_ERROR_PARTIAL is returned.
225 10. Partial matching has been split into two forms: PCRE_PARTIAL_SOFT, which is
226 synonymous with PCRE_PARTIAL, for backwards compatibility, and
227 PCRE_PARTIAL_HARD, which causes a partial match to supersede a full match,
228 and may be more useful for multi-segment matching.
230 11. Partial matching with pcre_exec() is now more intuitive. A partial match
231 used to be given if ever the end of the subject was reached; now it is
232 given only if matching could not proceed because another character was
233 needed. This makes a difference in some odd cases such as Z(*FAIL) with the
234 string "Z", which now yields "no match" instead of "partial match". In the
235 case of pcre_dfa_exec(), "no match" is given if every matching path for the
236 final character ended with (*FAIL).
238 12. Restarting a match using pcre_dfa_exec() after a partial match did not work
239 if the pattern had a "must contain" character that was already found in the
240 earlier partial match, unless partial matching was again requested. For
241 example, with the pattern /dog.(body)?/, the "must contain" character is
242 "g". If the first part-match was for the string "dog", restarting with
243 "sbody" failed. This bug has been fixed.
245 13. The string returned by pcre_dfa_exec() after a partial match has been
246 changed so that it starts at the first inspected character rather than the
247 first character of the match. This makes a difference only if the pattern
248 starts with a lookbehind assertion or \b or \B (\K is not supported by
249 pcre_dfa_exec()). It's an incompatible change, but it makes the two
250 matching functions compatible, and I think it's the right thing to do.
252 14. Added a pcredemo man page, created automatically from the pcredemo.c file,
253 so that the demonstration program is easily available in environments where
254 PCRE has not been installed from source.
256 15. Arranged to add -DPCRE_STATIC to cflags in libpcre.pc, libpcreposix.cp,
257 libpcrecpp.pc and pcre-config when PCRE is not compiled as a shared
258 library.
260 16. Added REG_UNGREEDY to the pcreposix interface, at the request of a user.
261 It maps to PCRE_UNGREEDY. It is not, of course, POSIX-compatible, but it
262 is not the first non-POSIX option to be added. Clearly some people find
263 these options useful.
265 17. If a caller to the POSIX matching function regexec() passes a non-zero
266 value for nmatch with a NULL value for pmatch, the value of
267 nmatch is forced to zero.
269 18. RunGrepTest did not have a test for the availability of the -u option of
270 the diff command, as RunTest does. It now checks in the same way as
271 RunTest, and also checks for the -b option.
273 19. If an odd number of negated classes containing just a single character
274 interposed, within parentheses, between a forward reference to a named
275 subpattern and the definition of the subpattern, compilation crashed with
276 an internal error, complaining that it could not find the referenced
277 subpattern. An example of a crashing pattern is /(?&A)(([^m])(?<A>))/.
278 [The bug was that it was starting one character too far in when skipping
279 over the character class, thus treating the ] as data rather than
280 terminating the class. This meant it could skip too much.]
282 20. Added PCRE_NOTEMPTY_ATSTART in order to be able to correctly implement the
283 /g option in pcretest when the pattern contains \K, which makes it possible
284 to have an empty string match not at the start, even when the pattern is
285 anchored. Updated pcretest and pcredemo to use this option.
287 21. If the maximum number of capturing subpatterns in a recursion was greater
288 than the maximum at the outer level, the higher number was returned, but
289 with unset values at the outer level. The correct (outer level) value is
290 now given.
292 22. If (*ACCEPT) appeared inside capturing parentheses, previous releases of
293 PCRE did not set those parentheses (unlike Perl). I have now found a way to
294 make it do so. The string so far is captured, making this feature
295 compatible with Perl.
297 23. The tests have been re-organized, adding tests 11 and 12, to make it
298 possible to check the Perl 5.10 features against Perl 5.10.
300 24. Perl 5.10 allows subroutine calls in lookbehinds, as long as the subroutine
301 pattern matches a fixed length string. PCRE did not allow this; now it
302 does. Neither allows recursion.
304 25. I finally figured out how to implement a request to provide the minimum
305 length of subject string that was needed in order to match a given pattern.
306 (It was back references and recursion that I had previously got hung up
307 on.) This code has now been added to pcre_study(); it finds a lower bound
308 to the length of subject needed. It is not necessarily the greatest lower
309 bound, but using it to avoid searching strings that are too short does give
310 some useful speed-ups. The value is available to calling programs via
311 pcre_fullinfo().
313 26. While implementing 25, I discovered to my embarrassment that pcretest had
314 not been passing the result of pcre_study() to pcre_dfa_exec(), so the
315 study optimizations had never been tested with that matching function.
316 Oops. What is worse, even when it was passed study data, there was a bug in
317 pcre_dfa_exec() that meant it never actually used it. Double oops. There
318 were also very few tests of studied patterns with pcre_dfa_exec().
320 27. If (?| is used to create subpatterns with duplicate numbers, they are now
321 allowed to have the same name, even if PCRE_DUPNAMES is not set. However,
322 on the other side of the coin, they are no longer allowed to have different
323 names, because these cannot be distinguished in PCRE, and this has caused
324 confusion. (This is a difference from Perl.)
326 28. When duplicate subpattern names are present (necessarily with different
327 numbers, as required by 27 above), and a test is made by name in a
328 conditional pattern, either for a subpattern having been matched, or for
329 recursion in such a pattern, all the associated numbered subpatterns are
330 tested, and the overall condition is true if the condition is true for any
331 one of them. This is the way Perl works, and is also more like the way
332 testing by number works.
335 Version 7.9 11-Apr-09
336 ---------------------
338 1. When building with support for bzlib/zlib (pcregrep) and/or readline
339 (pcretest), all targets were linked against these libraries. This included
340 libpcre, libpcreposix, and libpcrecpp, even though they do not use these
341 libraries. This caused unwanted dependencies to be created. This problem
342 has been fixed, and now only pcregrep is linked with bzlib/zlib and only
343 pcretest is linked with readline.
345 2. The "typedef int BOOL" in pcre_internal.h that was included inside the
346 "#ifndef FALSE" condition by an earlier change (probably 7.8/18) has been
347 moved outside it again, because FALSE and TRUE are already defined in AIX,
348 but BOOL is not.
350 3. The pcre_config() function was treating the PCRE_MATCH_LIMIT and
351 PCRE_MATCH_LIMIT_RECURSION values as ints, when they should be long ints.
353 4. The pcregrep documentation said spaces were inserted as well as colons (or
354 hyphens) following file names and line numbers when outputting matching
355 lines. This is not true; no spaces are inserted. I have also clarified the
356 wording for the --colour (or --color) option.
358 5. In pcregrep, when --colour was used with -o, the list of matching strings
359 was not coloured; this is different to GNU grep, so I have changed it to be
360 the same.
362 6. When --colo(u)r was used in pcregrep, only the first matching substring in
363 each matching line was coloured. Now it goes on to look for further matches
364 of any of the test patterns, which is the same behaviour as GNU grep.
366 7. A pattern that could match an empty string could cause pcregrep to loop; it
367 doesn't make sense to accept an empty string match in pcregrep, so I have
368 locked it out (using PCRE's PCRE_NOTEMPTY option). By experiment, this
369 seems to be how GNU grep behaves.
371 8. The pattern (?(?=.*b)b|^) was incorrectly compiled as "match must be at
372 start or after a newline", because the conditional assertion was not being
373 correctly handled. The rule now is that both the assertion and what follows
374 in the first alternative must satisfy the test.
376 9. If auto-callout was enabled in a pattern with a conditional group whose
377 condition was an assertion, PCRE could crash during matching, both with
378 pcre_exec() and pcre_dfa_exec().
380 10. The PCRE_DOLLAR_ENDONLY option was not working when pcre_dfa_exec() was
381 used for matching.
383 11. Unicode property support in character classes was not working for
384 characters (bytes) greater than 127 when not in UTF-8 mode.
386 12. Added the -M command line option to pcretest.
388 14. Added the non-standard REG_NOTEMPTY option to the POSIX interface.
390 15. Added the PCRE_NO_START_OPTIMIZE match-time option.
392 16. Added comments and documentation about mis-use of no_arg in the C++
393 wrapper.
395 17. Implemented support for UTF-8 encoding in EBCDIC environments, a patch
396 from Martin Jerabek that uses macro names for all relevant character and
397 string constants.
399 18. Added to pcre_internal.h two configuration checks: (a) If both EBCDIC and
400 SUPPORT_UTF8 are set, give an error; (b) If SUPPORT_UCP is set without
401 SUPPORT_UTF8, define SUPPORT_UTF8. The "configure" script handles both of
402 these, but not everybody uses configure.
404 19. A conditional group that had only one branch was not being correctly
405 recognized as an item that could match an empty string. This meant that an
406 enclosing group might also not be so recognized, causing infinite looping
407 (and probably a segfault) for patterns such as ^"((?(?=[a])[^"])|b)*"$
408 with the subject "ab", where knowledge that the repeated group can match
409 nothing is needed in order to break the loop.
411 20. If a pattern that was compiled with callouts was matched using pcre_dfa_
412 exec(), but without supplying a callout function, matching went wrong.
414 21. If PCRE_ERROR_MATCHLIMIT occurred during a recursion, there was a memory
415 leak if the size of the offset vector was greater than 30. When the vector
416 is smaller, the saved offsets during recursion go onto a local stack
417 vector, but for larger vectors malloc() is used. It was failing to free
418 when the recursion yielded PCRE_ERROR_MATCH_LIMIT (or any other "abnormal"
419 error, in fact).
421 22. There was a missing #ifdef SUPPORT_UTF8 round one of the variables in the
422 heapframe that is used only when UTF-8 support is enabled. This caused no
423 problem, but was untidy.
425 23. Steven Van Ingelgem's patch to CMakeLists.txt to change the name
426 CMAKE_BINARY_DIR to PROJECT_BINARY_DIR so that it works when PCRE is
427 included within another project.
429 24. Steven Van Ingelgem's patches to add more options to the CMake support,
430 slightly modified by me:
432 (a) PCRE_BUILD_TESTS can be set OFF not to build the tests, including
433 not building pcregrep.
435 (b) PCRE_BUILD_PCREGREP can be see OFF not to build pcregrep, but only
436 if PCRE_BUILD_TESTS is also set OFF, because the tests use pcregrep.
438 25. Forward references, both numeric and by name, in patterns that made use of
439 duplicate group numbers, could behave incorrectly or give incorrect errors,
440 because when scanning forward to find the reference group, PCRE was not
441 taking into account the duplicate group numbers. A pattern such as
442 ^X(?3)(a)(?|(b)|(q))(Y) is an example.
444 26. Changed a few more instances of "const unsigned char *" to USPTR, making
445 the feature of a custom pointer more persuasive (as requested by a user).
447 27. Wrapped the definitions of fileno and isatty for Windows, which appear in
448 pcretest.c, inside #ifndefs, because it seems they are sometimes already
449 pre-defined.
451 28. Added support for (*UTF8) at the start of a pattern.
453 29. Arrange for flags added by the "release type" setting in CMake to be shown
454 in the configuration summary.
457 Version 7.8 05-Sep-08
458 ---------------------
460 1. Replaced UCP searching code with optimized version as implemented for Ad
461 Muncher (http://www.admuncher.com/) by Peter Kankowski. This uses a two-
462 stage table and inline lookup instead of a function, giving speed ups of 2
463 to 5 times on some simple patterns that I tested. Permission was given to
464 distribute the MultiStage2.py script that generates the tables (it's not in
465 the tarball, but is in the Subversion repository).
467 2. Updated the Unicode datatables to Unicode 5.1.0. This adds yet more
468 scripts.
470 3. Change 12 for 7.7 introduced a bug in pcre_study() when a pattern contained
471 a group with a zero qualifier. The result of the study could be incorrect,
472 or the function might crash, depending on the pattern.
474 4. Caseless matching was not working for non-ASCII characters in back
475 references. For example, /(\x{de})\1/8i was not matching \x{de}\x{fe}.
476 It now works when Unicode Property Support is available.
478 5. In pcretest, an escape such as \x{de} in the data was always generating
479 a UTF-8 string, even in non-UTF-8 mode. Now it generates a single byte in
480 non-UTF-8 mode. If the value is greater than 255, it gives a warning about
481 truncation.
483 6. Minor bugfix in pcrecpp.cc (change "" == ... to NULL == ...).
485 7. Added two (int) casts to pcregrep when printing the difference of two
486 pointers, in case they are 64-bit values.
488 8. Added comments about Mac OS X stack usage to the pcrestack man page and to
489 test 2 if it fails.
491 9. Added PCRE_CALL_CONVENTION just before the names of all exported functions,
492 and a #define of that name to empty if it is not externally set. This is to
493 allow users of MSVC to set it if necessary.
495 10. The PCRE_EXP_DEFN macro which precedes exported functions was missing from
496 the convenience functions in the pcre_get.c source file.
498 11. An option change at the start of a pattern that had top-level alternatives
499 could cause overwriting and/or a crash. This command provoked a crash in
500 some environments:
502 printf "/(?i)[\xc3\xa9\xc3\xbd]|[\xc3\xa9\xc3\xbdA]/8\n" | pcretest
504 This potential security problem was recorded as CVE-2008-2371.
506 12. For a pattern where the match had to start at the beginning or immediately
507 after a newline (e.g /.*anything/ without the DOTALL flag), pcre_exec() and
508 pcre_dfa_exec() could read past the end of the passed subject if there was
509 no match. To help with detecting such bugs (e.g. with valgrind), I modified
510 pcretest so that it places the subject at the end of its malloc-ed buffer.
512 13. The change to pcretest in 12 above threw up a couple more cases when pcre_
513 exec() might read past the end of the data buffer in UTF-8 mode.
515 14. A similar bug to 7.3/2 existed when the PCRE_FIRSTLINE option was set and
516 the data contained the byte 0x85 as part of a UTF-8 character within its
517 first line. This applied both to normal and DFA matching.
519 15. Lazy qualifiers were not working in some cases in UTF-8 mode. For example,
520 /^[^d]*?$/8 failed to match "abc".
522 16. Added a missing copyright notice to pcrecpp_internal.h.
524 17. Make it more clear in the documentation that values returned from
525 pcre_exec() in ovector are byte offsets, not character counts.
527 18. Tidied a few places to stop certain compilers from issuing warnings.
529 19. Updated the Virtual Pascal + BCC files to compile the latest v7.7, as
530 supplied by Stefan Weber. I made a further small update for 7.8 because
531 there is a change of source arrangements: the pcre_searchfuncs.c module is
532 replaced by pcre_ucd.c.
535 Version 7.7 07-May-08
536 ---------------------
538 1. Applied Craig's patch to sort out a long long problem: "If we can't convert
539 a string to a long long, pretend we don't even have a long long." This is
540 done by checking for the strtoq, strtoll, and _strtoi64 functions.
542 2. Applied Craig's patch to pcrecpp.cc to restore ABI compatibility with
543 pre-7.6 versions, which defined a global no_arg variable instead of putting
544 it in the RE class. (See also #8 below.)
546 3. Remove a line of dead code, identified by coverity and reported by Nuno
547 Lopes.
549 4. Fixed two related pcregrep bugs involving -r with --include or --exclude:
551 (1) The include/exclude patterns were being applied to the whole pathnames
552 of files, instead of just to the final components.
554 (2) If there was more than one level of directory, the subdirectories were
555 skipped unless they satisfied the include/exclude conditions. This is
556 inconsistent with GNU grep (and could even be seen as contrary to the
557 pcregrep specification - which I improved to make it absolutely clear).
558 The action now is always to scan all levels of directory, and just
559 apply the include/exclude patterns to regular files.
561 5. Added the --include_dir and --exclude_dir patterns to pcregrep, and used
562 --exclude_dir in the tests to avoid scanning .svn directories.
564 6. Applied Craig's patch to the QuoteMeta function so that it escapes the
565 NUL character as backslash + 0 rather than backslash + NUL, because PCRE
566 doesn't support NULs in patterns.
568 7. Added some missing "const"s to declarations of static tables in
569 pcre_compile.c and pcre_dfa_exec.c.
571 8. Applied Craig's patch to pcrecpp.cc to fix a problem in OS X that was
572 caused by fix #2 above. (Subsequently also a second patch to fix the
573 first patch. And a third patch - this was a messy problem.)
575 9. Applied Craig's patch to remove the use of push_back().
577 10. Applied Alan Lehotsky's patch to add REG_STARTEND support to the POSIX
578 matching function regexec().
580 11. Added support for the Oniguruma syntax \g<name>, \g<n>, \g'name', \g'n',
581 which, however, unlike Perl's \g{...}, are subroutine calls, not back
582 references. PCRE supports relative numbers with this syntax (I don't think
583 Oniguruma does).
585 12. Previously, a group with a zero repeat such as (...){0} was completely
586 omitted from the compiled regex. However, this means that if the group
587 was called as a subroutine from elsewhere in the pattern, things went wrong
588 (an internal error was given). Such groups are now left in the compiled
589 pattern, with a new opcode that causes them to be skipped at execution
590 time.
592 13. Added the PCRE_JAVASCRIPT_COMPAT option. This makes the following changes
593 to the way PCRE behaves:
595 (a) A lone ] character is dis-allowed (Perl treats it as data).
597 (b) A back reference to an unmatched subpattern matches an empty string
598 (Perl fails the current match path).
600 (c) A data ] in a character class must be notated as \] because if the
601 first data character in a class is ], it defines an empty class. (In
602 Perl it is not possible to have an empty class.) The empty class []
603 never matches; it forces failure and is equivalent to (*FAIL) or (?!).
604 The negative empty class [^] matches any one character, independently
605 of the DOTALL setting.
607 14. A pattern such as /(?2)[]a()b](abc)/ which had a forward reference to a
608 non-existent subpattern following a character class starting with ']' and
609 containing () gave an internal compiling error instead of "reference to
610 non-existent subpattern". Fortunately, when the pattern did exist, the
611 compiled code was correct. (When scanning forwards to check for the
612 existencd of the subpattern, it was treating the data ']' as terminating
613 the class, so got the count wrong. When actually compiling, the reference
614 was subsequently set up correctly.)
616 15. The "always fail" assertion (?!) is optimzed to (*FAIL) by pcre_compile;
617 it was being rejected as not supported by pcre_dfa_exec(), even though
618 other assertions are supported. I have made pcre_dfa_exec() support
619 (*FAIL).
621 16. The implementation of 13c above involved the invention of a new opcode,
622 OP_ALLANY, which is like OP_ANY but doesn't check the /s flag. Since /s
623 cannot be changed at match time, I realized I could make a small
624 improvement to matching performance by compiling OP_ALLANY instead of
625 OP_ANY for "." when DOTALL was set, and then removing the runtime tests
626 on the OP_ANY path.
628 17. Compiling pcretest on Windows with readline support failed without the
629 following two fixes: (1) Make the unistd.h include conditional on
630 HAVE_UNISTD_H; (2) #define isatty and fileno as _isatty and _fileno.
632 18. Changed CMakeLists.txt and cmake/FindReadline.cmake to arrange for the
633 ncurses library to be included for pcretest when ReadLine support is
634 requested, but also to allow for it to be overridden. This patch came from
635 Daniel Bergström.
637 19. There was a typo in the file ucpinternal.h where f0_rangeflag was defined
638 as 0x00f00000 instead of 0x00800000. Luckily, this would not have caused
639 any errors with the current Unicode tables. Thanks to Peter Kankowski for
640 spotting this.
643 Version 7.6 28-Jan-08
644 ---------------------
646 1. A character class containing a very large number of characters with
647 codepoints greater than 255 (in UTF-8 mode, of course) caused a buffer
648 overflow.
650 2. Patch to cut out the "long long" test in pcrecpp_unittest when
651 HAVE_LONG_LONG is not defined.
653 3. Applied Christian Ehrlicher's patch to update the CMake build files to
654 bring them up to date and include new features. This patch includes:
656 - Fixed PH's badly added libz and libbz2 support.
657 - Fixed a problem with static linking.
658 - Added pcredemo. [But later removed - see 7 below.]
659 - Fixed dftables problem and added an option.
660 - Added a number of HAVE_XXX tests, including HAVE_WINDOWS_H and
662 - Added readline support for pcretest.
663 - Added an listing of the option settings after cmake has run.
665 4. A user submitted a patch to Makefile that makes it easy to create
666 "pcre.dll" under mingw when using Configure/Make. I added stuff to
667 Makefile.am that cause it to include this special target, without
668 affecting anything else. Note that the same mingw target plus all
669 the other distribution libraries and programs are now supported
670 when configuring with CMake (see 6 below) instead of with
671 Configure/Make.
673 5. Applied Craig's patch that moves no_arg into the RE class in the C++ code.
674 This is an attempt to solve the reported problem "pcrecpp::no_arg is not
675 exported in the Windows port". It has not yet been confirmed that the patch
676 solves the problem, but it does no harm.
678 6. Applied Sheri's patch to CMakeLists.txt to add NON_STANDARD_LIB_PREFIX and
679 NON_STANDARD_LIB_SUFFIX for dll names built with mingw when configured
680 with CMake, and also correct the comment about stack recursion.
682 7. Remove the automatic building of pcredemo from the ./configure system and
683 from CMakeLists.txt. The whole idea of pcredemo.c is that it is an example
684 of a program that users should build themselves after PCRE is installed, so
685 building it automatically is not really right. What is more, it gave
686 trouble in some build environments.
688 8. Further tidies to CMakeLists.txt from Sheri and Christian.
691 Version 7.5 10-Jan-08
692 ---------------------
694 1. Applied a patch from Craig: "This patch makes it possible to 'ignore'
695 values in parens when parsing an RE using the C++ wrapper."
697 2. Negative specials like \S did not work in character classes in UTF-8 mode.
698 Characters greater than 255 were excluded from the class instead of being
699 included.
701 3. The same bug as (2) above applied to negated POSIX classes such as
702 [:^space:].
704 4. PCRECPP_STATIC was referenced in pcrecpp_internal.h, but nowhere was it
705 defined or documented. It seems to have been a typo for PCRE_STATIC, so
706 I have changed it.
708 5. The construct (?&) was not diagnosed as a syntax error (it referenced the
709 first named subpattern) and a construct such as (?&a) would reference the
710 first named subpattern whose name started with "a" (in other words, the
711 length check was missing). Both these problems are fixed. "Subpattern name
712 expected" is now given for (?&) (a zero-length name), and this patch also
713 makes it give the same error for \k'' (previously it complained that that
714 was a reference to a non-existent subpattern).
716 6. The erroneous patterns (?+-a) and (?-+a) give different error messages;
717 this is right because (?- can be followed by option settings as well as by
718 digits. I have, however, made the messages clearer.
720 7. Patterns such as (?(1)a|b) (a pattern that contains fewer subpatterns
721 than the number used in the conditional) now cause a compile-time error.
722 This is actually not compatible with Perl, which accepts such patterns, but
723 treats the conditional as always being FALSE (as PCRE used to), but it
724 seems to me that giving a diagnostic is better.
726 8. Change "alphameric" to the more common word "alphanumeric" in comments
727 and messages.
729 9. Fix two occurrences of "backslash" in comments that should have been
730 "backspace".
732 10. Remove two redundant lines of code that can never be obeyed (their function
733 was moved elsewhere).
735 11. The program that makes PCRE's Unicode character property table had a bug
736 which caused it to generate incorrect table entries for sequences of
737 characters that have the same character type, but are in different scripts.
738 It amalgamated them into a single range, with the script of the first of
739 them. In other words, some characters were in the wrong script. There were
740 thirteen such cases, affecting characters in the following ranges:
742 U+002b0 - U+002c1
743 U+0060c - U+0060d
744 U+0061e - U+00612
745 U+0064b - U+0065e
746 U+0074d - U+0076d
747 U+01800 - U+01805
748 U+01d00 - U+01d77
749 U+01d9b - U+01dbf
750 U+0200b - U+0200f
751 U+030fc - U+030fe
752 U+03260 - U+0327f
753 U+0fb46 - U+0fbb1
754 U+10450 - U+1049d
756 12. The -o option (show only the matching part of a line) for pcregrep was not
757 compatible with GNU grep in that, if there was more than one match in a
758 line, it showed only the first of them. It now behaves in the same way as
759 GNU grep.
761 13. If the -o and -v options were combined for pcregrep, it printed a blank
762 line for every non-matching line. GNU grep prints nothing, and pcregrep now
763 does the same. The return code can be used to tell if there were any
764 non-matching lines.
766 14. Added --file-offsets and --line-offsets to pcregrep.
768 15. The pattern (?=something)(?R) was not being diagnosed as a potentially
769 infinitely looping recursion. The bug was that positive lookaheads were not
770 being skipped when checking for a possible empty match (negative lookaheads
771 and both kinds of lookbehind were skipped).
773 16. Fixed two typos in the Windows-only code in pcregrep.c, and moved the
774 inclusion of <windows.h> to before rather than after the definition of
775 INVALID_FILE_ATTRIBUTES (patch from David Byron).
777 17. Specifying a possessive quantifier with a specific limit for a Unicode
778 character property caused pcre_compile() to compile bad code, which led at
779 runtime to PCRE_ERROR_INTERNAL (-14). Examples of patterns that caused this
780 are: /\p{Zl}{2,3}+/8 and /\p{Cc}{2}+/8. It was the possessive "+" that
781 caused the error; without that there was no problem.
783 18. Added --enable-pcregrep-libz and --enable-pcregrep-libbz2.
785 19. Added --enable-pcretest-libreadline.
787 20. In pcrecpp.cc, the variable 'count' was incremented twice in
788 RE::GlobalReplace(). As a result, the number of replacements returned was
789 double what it should be. I removed one of the increments, but Craig sent a
790 later patch that removed the other one (the right fix) and added unit tests
791 that check the return values (which was not done before).
793 21. Several CMake things:
795 (1) Arranged that, when cmake is used on Unix, the libraries end up with
796 the names libpcre and libpcreposix, not just pcre and pcreposix.
798 (2) The above change means that pcretest and pcregrep are now correctly
799 linked with the newly-built libraries, not previously installed ones.
803 22. In UTF-8 mode, with newline set to "any", a pattern such as .*a.*=.b.*
804 crashed when matching a string such as a\x{2029}b (note that \x{2029} is a
805 UTF-8 newline character). The key issue is that the pattern starts .*;
806 this means that the match must be either at the beginning, or after a
807 newline. The bug was in the code for advancing after a failed match and
808 checking that the new position followed a newline. It was not taking
809 account of UTF-8 characters correctly.
811 23. PCRE was behaving differently from Perl in the way it recognized POSIX
812 character classes. PCRE was not treating the sequence [:...:] as a
813 character class unless the ... were all letters. Perl, however, seems to
814 allow any characters between [: and :], though of course it rejects as
815 unknown any "names" that contain non-letters, because all the known class
816 names consist only of letters. Thus, Perl gives an error for [[:1234:]],
817 for example, whereas PCRE did not - it did not recognize a POSIX character
818 class. This seemed a bit dangerous, so the code has been changed to be
819 closer to Perl. The behaviour is not identical to Perl, because PCRE will
820 diagnose an unknown class for, for example, [[:l\ower:]] where Perl will
821 treat it as [[:lower:]]. However, PCRE does now give "unknown" errors where
822 Perl does, and where it didn't before.
824 24. Rewrite so as to remove the single use of %n from pcregrep because in some
825 Windows environments %n is disabled by default.
828 Version 7.4 21-Sep-07
829 ---------------------
831 1. Change 7.3/28 was implemented for classes by looking at the bitmap. This
832 means that a class such as [\s] counted as "explicit reference to CR or
833 LF". That isn't really right - the whole point of the change was to try to
834 help when there was an actual mention of one of the two characters. So now
835 the change happens only if \r or \n (or a literal CR or LF) character is
836 encountered.
838 2. The 32-bit options word was also used for 6 internal flags, but the numbers
839 of both had grown to the point where there were only 3 bits left.
840 Fortunately, there was spare space in the data structure, and so I have
841 moved the internal flags into a new 16-bit field to free up more option
842 bits.
844 3. The appearance of (?J) at the start of a pattern set the DUPNAMES option,
845 but did not set the internal JCHANGED flag - either of these is enough to
846 control the way the "get" function works - but the PCRE_INFO_JCHANGED
847 facility is supposed to tell if (?J) was ever used, so now (?J) at the
848 start sets both bits.
850 4. Added options (at build time, compile time, exec time) to change \R from
851 matching any Unicode line ending sequence to just matching CR, LF, or CRLF.
853 5. doc/pcresyntax.html was missing from the distribution.
855 6. Put back the definition of PCRE_ERROR_NULLWSLIMIT, for backward
856 compatibility, even though it is no longer used.
858 7. Added macro for snprintf to pcrecpp_unittest.cc and also for strtoll and
859 strtoull to pcrecpp.cc to select the available functions in WIN32 when the
860 windows.h file is present (where different names are used). [This was
861 reversed later after testing - see 16 below.]
863 8. Changed all #include <config.h> to #include "config.h". There were also
864 some further <pcre.h> cases that I changed to "pcre.h".
866 9. When pcregrep was used with the --colour option, it missed the line ending
867 sequence off the lines that it output.
869 10. It was pointed out to me that arrays of string pointers cause lots of
870 relocations when a shared library is dynamically loaded. A technique of
871 using a single long string with a table of offsets can drastically reduce
872 these. I have refactored PCRE in four places to do this. The result is
873 dramatic:
875 Originally: 290
876 After changing UCP table: 187
877 After changing error message table: 43
878 After changing table of "verbs" 36
879 After changing table of Posix names 22
881 Thanks to the folks working on Gregex for glib for this insight.
883 11. --disable-stack-for-recursion caused compiling to fail unless -enable-
884 unicode-properties was also set.
886 12. Updated the tests so that they work when \R is defaulted to ANYCRLF.
888 13. Added checks for ANY and ANYCRLF to pcrecpp.cc where it previously
889 checked only for CRLF.
891 14. Added casts to pcretest.c to avoid compiler warnings.
893 15. Added Craig's patch to various pcrecpp modules to avoid compiler warnings.
895 16. Added Craig's patch to remove the WINDOWS_H tests, that were not working,
896 and instead check for _strtoi64 explicitly, and avoid the use of snprintf()
897 entirely. This removes changes made in 7 above.
899 17. The CMake files have been updated, and there is now more information about
900 building with CMake in the NON-UNIX-USE document.
903 Version 7.3 28-Aug-07
904 ---------------------
906 1. In the rejigging of the build system that eventually resulted in 7.1, the
907 line "#include <pcre.h>" was included in pcre_internal.h. The use of angle
908 brackets there is not right, since it causes compilers to look for an
909 installed pcre.h, not the version that is in the source that is being
910 compiled (which of course may be different). I have changed it back to:
912 #include "pcre.h"
914 I have a vague recollection that the change was concerned with compiling in
915 different directories, but in the new build system, that is taken care of
916 by the VPATH setting the Makefile.
918 2. The pattern .*$ when run in not-DOTALL UTF-8 mode with newline=any failed
919 when the subject happened to end in the byte 0x85 (e.g. if the last
920 character was \x{1ec5}). *Character* 0x85 is one of the "any" newline
921 characters but of course it shouldn't be taken as a newline when it is part
922 of another character. The bug was that, for an unlimited repeat of . in
923 not-DOTALL UTF-8 mode, PCRE was advancing by bytes rather than by
924 characters when looking for a newline.
926 3. A small performance improvement in the DOTALL UTF-8 mode .* case.
928 4. Debugging: adjusted the names of opcodes for different kinds of parentheses
929 in debug output.
931 5. Arrange to use "%I64d" instead of "%lld" and "%I64u" instead of "%llu" for
932 long printing in the pcrecpp unittest when running under MinGW.
934 6. ESC_K was left out of the EBCDIC table.
936 7. Change 7.0/38 introduced a new limit on the number of nested non-capturing
937 parentheses; I made it 1000, which seemed large enough. Unfortunately, the
938 limit also applies to "virtual nesting" when a pattern is recursive, and in
939 this case 1000 isn't so big. I have been able to remove this limit at the
940 expense of backing off one optimization in certain circumstances. Normally,
941 when pcre_exec() would call its internal match() function recursively and
942 immediately return the result unconditionally, it uses a "tail recursion"
943 feature to save stack. However, when a subpattern that can match an empty
944 string has an unlimited repetition quantifier, it no longer makes this
945 optimization. That gives it a stack frame in which to save the data for
946 checking that an empty string has been matched. Previously this was taken
947 from the 1000-entry workspace that had been reserved. So now there is no
948 explicit limit, but more stack is used.
950 8. Applied Daniel's patches to solve problems with the import/export magic
951 syntax that is required for Windows, and which was going wrong for the
952 pcreposix and pcrecpp parts of the library. These were overlooked when this
953 problem was solved for the main library.
955 9. There were some crude static tests to avoid integer overflow when computing
956 the size of patterns that contain repeated groups with explicit upper
957 limits. As the maximum quantifier is 65535, the maximum group length was
958 set at 30,000 so that the product of these two numbers did not overflow a
959 32-bit integer. However, it turns out that people want to use groups that
960 are longer than 30,000 bytes (though not repeat them that many times).
961 Change 7.0/17 (the refactoring of the way the pattern size is computed) has
962 made it possible to implement the integer overflow checks in a much more
963 dynamic way, which I have now done. The artificial limitation on group
964 length has been removed - we now have only the limit on the total length of
965 the compiled pattern, which depends on the LINK_SIZE setting.
967 10. Fixed a bug in the documentation for get/copy named substring when
968 duplicate names are permitted. If none of the named substrings are set, the
969 functions return PCRE_ERROR_NOSUBSTRING (7); the doc said they returned an
970 empty string.
972 11. Because Perl interprets \Q...\E at a high level, and ignores orphan \E
973 instances, patterns such as [\Q\E] or [\E] or even [^\E] cause an error,
974 because the ] is interpreted as the first data character and the
975 terminating ] is not found. PCRE has been made compatible with Perl in this
976 regard. Previously, it interpreted [\Q\E] as an empty class, and [\E] could
977 cause memory overwriting.
979 10. Like Perl, PCRE automatically breaks an unlimited repeat after an empty
980 string has been matched (to stop an infinite loop). It was not recognizing
981 a conditional subpattern that could match an empty string if that
982 subpattern was within another subpattern. For example, it looped when
983 trying to match (((?(1)X|))*) but it was OK with ((?(1)X|)*) where the
984 condition was not nested. This bug has been fixed.
986 12. A pattern like \X?\d or \P{L}?\d in non-UTF-8 mode could cause a backtrack
987 past the start of the subject in the presence of bytes with the top bit
988 set, for example "\x8aBCD".
990 13. Added Perl 5.10 experimental backtracking controls (*FAIL), (*F), (*PRUNE),
991 (*SKIP), (*THEN), (*COMMIT), and (*ACCEPT).
993 14. Optimized (?!) to (*FAIL).
995 15. Updated the test for a valid UTF-8 string to conform to the later RFC 3629.
996 This restricts code points to be within the range 0 to 0x10FFFF, excluding
997 the "low surrogate" sequence 0xD800 to 0xDFFF. Previously, PCRE allowed the
998 full range 0 to 0x7FFFFFFF, as defined by RFC 2279. Internally, it still
999 does: it's just the validity check that is more restrictive.
1001 16. Inserted checks for integer overflows during escape sequence (backslash)
1002 processing, and also fixed erroneous offset values for syntax errors during
1003 backslash processing.
1005 17. Fixed another case of looking too far back in non-UTF-8 mode (cf 12 above)
1006 for patterns like [\PPP\x8a]{1,}\x80 with the subject "A\x80".
1008 18. An unterminated class in a pattern like (?1)\c[ with a "forward reference"
1009 caused an overrun.
1011 19. A pattern like (?:[\PPa*]*){8,} which had an "extended class" (one with
1012 something other than just ASCII characters) inside a group that had an
1013 unlimited repeat caused a loop at compile time (while checking to see
1014 whether the group could match an empty string).
1016 20. Debugging a pattern containing \p or \P could cause a crash. For example,
1017 [\P{Any}] did so. (Error in the code for printing property names.)
1019 21. An orphan \E inside a character class could cause a crash.
1021 22. A repeated capturing bracket such as (A)? could cause a wild memory
1022 reference during compilation.
1024 23. There are several functions in pcre_compile() that scan along a compiled
1025 expression for various reasons (e.g. to see if it's fixed length for look
1026 behind). There were bugs in these functions when a repeated \p or \P was
1027 present in the pattern. These operators have additional parameters compared
1028 with \d, etc, and these were not being taken into account when moving along
1029 the compiled data. Specifically:
1031 (a) A item such as \p{Yi}{3} in a lookbehind was not treated as fixed
1032 length.
1034 (b) An item such as \pL+ within a repeated group could cause crashes or
1035 loops.
1037 (c) A pattern such as \p{Yi}+(\P{Yi}+)(?1) could give an incorrect
1038 "reference to non-existent subpattern" error.
1040 (d) A pattern like (\P{Yi}{2}\277)? could loop at compile time.
1042 24. A repeated \S or \W in UTF-8 mode could give wrong answers when multibyte
1043 characters were involved (for example /\S{2}/8g with "A\x{a3}BC").
1045 25. Using pcregrep in multiline, inverted mode (-Mv) caused it to loop.
1047 26. Patterns such as [\P{Yi}A] which include \p or \P and just one other
1048 character were causing crashes (broken optimization).
1050 27. Patterns such as (\P{Yi}*\277)* (group with possible zero repeat containing
1051 \p or \P) caused a compile-time loop.
1053 28. More problems have arisen in unanchored patterns when CRLF is a valid line
1054 break. For example, the unstudied pattern [\r\n]A does not match the string
1055 "\r\nA" because change 7.0/46 below moves the current point on by two
1056 characters after failing to match at the start. However, the pattern \nA
1057 *does* match, because it doesn't start till \n, and if [\r\n]A is studied,
1058 the same is true. There doesn't seem any very clean way out of this, but
1059 what I have chosen to do makes the common cases work: PCRE now takes note
1060 of whether there can be an explicit match for \r or \n anywhere in the
1061 pattern, and if so, 7.0/46 no longer applies. As part of this change,
1062 there's a new PCRE_INFO_HASCRORLF option for finding out whether a compiled
1063 pattern has explicit CR or LF references.
1065 29. Added (*CR) etc for changing newline setting at start of pattern.
1068 Version 7.2 19-Jun-07
1069 ---------------------
1071 1. If the fr_FR locale cannot be found for test 3, try the "french" locale,
1072 which is apparently normally available under Windows.
1074 2. Re-jig the pcregrep tests with different newline settings in an attempt
1075 to make them independent of the local environment's newline setting.
1077 3. Add code to configure.ac to remove -g from the CFLAGS default settings.
1079 4. Some of the "internals" tests were previously cut out when the link size
1080 was not 2, because the output contained actual offsets. The recent new
1081 "Z" feature of pcretest means that these can be cut out, making the tests
1082 usable with all link sizes.
1084 5. Implemented Stan Switzer's goto replacement for longjmp() when not using
1085 stack recursion. This gives a massive performance boost under BSD, but just
1086 a small improvement under Linux. However, it saves one field in the frame
1087 in all cases.
1089 6. Added more features from the forthcoming Perl 5.10:
1091 (a) (?-n) (where n is a string of digits) is a relative subroutine or
1092 recursion call. It refers to the nth most recently opened parentheses.
1094 (b) (?+n) is also a relative subroutine call; it refers to the nth next
1095 to be opened parentheses.
1097 (c) Conditions that refer to capturing parentheses can be specified
1098 relatively, for example, (?(-2)... or (?(+3)...
1100 (d) \K resets the start of the current match so that everything before
1101 is not part of it.
1103 (e) \k{name} is synonymous with \k<name> and \k'name' (.NET compatible).
1105 (f) \g{name} is another synonym - part of Perl 5.10's unification of
1106 reference syntax.
1108 (g) (?| introduces a group in which the numbering of parentheses in each
1109 alternative starts with the same number.
1111 (h) \h, \H, \v, and \V match horizontal and vertical whitespace.
1113 7. Added two new calls to pcre_fullinfo(): PCRE_INFO_OKPARTIAL and
1116 8. A pattern such as (.*(.)?)* caused pcre_exec() to fail by either not
1117 terminating or by crashing. Diagnosed by Viktor Griph; it was in the code
1118 for detecting groups that can match an empty string.
1120 9. A pattern with a very large number of alternatives (more than several
1121 hundred) was running out of internal workspace during the pre-compile
1122 phase, where pcre_compile() figures out how much memory will be needed. A
1123 bit of new cunning has reduced the workspace needed for groups with
1124 alternatives. The 1000-alternative test pattern now uses 12 bytes of
1125 workspace instead of running out of the 4096 that are available.
1127 10. Inserted some missing (unsigned int) casts to get rid of compiler warnings.
1129 11. Applied patch from Google to remove an optimization that didn't quite work.
1130 The report of the bug said:
1132 pcrecpp::RE("a*").FullMatch("aaa") matches, while
1133 pcrecpp::RE("a*?").FullMatch("aaa") does not, and
1134 pcrecpp::RE("a*?\\z").FullMatch("aaa") does again.
1136 12. If \p or \P was used in non-UTF-8 mode on a character greater than 127
1137 it matched the wrong number of bytes.
1140 Version 7.1 24-Apr-07
1141 ---------------------
1143 1. Applied Bob Rossi and Daniel G's patches to convert the build system to one
1144 that is more "standard", making use of automake and other Autotools. There
1145 is some re-arrangement of the files and adjustment of comments consequent
1146 on this.
1148 2. Part of the patch fixed a problem with the pcregrep tests. The test of -r
1149 for recursive directory scanning broke on some systems because the files
1150 are not scanned in any specific order and on different systems the order
1151 was different. A call to "sort" has been inserted into RunGrepTest for the
1152 approprate test as a short-term fix. In the longer term there may be an
1153 alternative.
1155 3. I had an email from Eric Raymond about problems translating some of PCRE's
1156 man pages to HTML (despite the fact that I distribute HTML pages, some
1157 people do their own conversions for various reasons). The problems
1158 concerned the use of low-level troff macros .br and .in. I have therefore
1159 removed all such uses from the man pages (some were redundant, some could
1160 be replaced by .nf/.fi pairs). The 132html script that I use to generate
1161 HTML has been updated to handle .nf/.fi and to complain if it encounters
1162 .br or .in.
1164 4. Updated comments in configure.ac that get placed in config.h.in and also
1165 arranged for config.h to be included in the distribution, with the name
1166 config.h.generic, for the benefit of those who have to compile without
1167 Autotools (compare pcre.h, which is now distributed as pcre.h.generic).
1169 5. Updated the support (such as it is) for Virtual Pascal, thanks to Stefan
1170 Weber: (1) pcre_internal.h was missing some function renames; (2) updated
1171 makevp.bat for the current PCRE, using the additional files
1172 makevp_c.txt, makevp_l.txt, and pcregexp.pas.
1174 6. A Windows user reported a minor discrepancy with test 2, which turned out
1175 to be caused by a trailing space on an input line that had got lost in his
1176 copy. The trailing space was an accident, so I've just removed it.
1178 7. Add -Wl,-R... flags in pcre-config.in for *BSD* systems, as I'm told
1179 that is needed.
1181 8. Mark ucp_table (in ucptable.h) and ucp_gentype (in pcre_ucp_searchfuncs.c)
1182 as "const" (a) because they are and (b) because it helps the PHP
1183 maintainers who have recently made a script to detect big data structures
1184 in the php code that should be moved to the .rodata section. I remembered
1185 to update Builducptable as well, so it won't revert if ucptable.h is ever
1186 re-created.
1188 9. Added some extra #ifdef SUPPORT_UTF8 conditionals into pcretest.c,
1189 pcre_printint.src, pcre_compile.c, pcre_study.c, and pcre_tables.c, in
1190 order to be able to cut out the UTF-8 tables in the latter when UTF-8
1191 support is not required. This saves 1.5-2K of code, which is important in
1192 some applications.
1194 Later: more #ifdefs are needed in pcre_ord2utf8.c and pcre_valid_utf8.c
1195 so as not to refer to the tables, even though these functions will never be
1196 called when UTF-8 support is disabled. Otherwise there are problems with a
1197 shared library.
1199 10. Fixed two bugs in the emulated memmove() function in pcre_internal.h:
1201 (a) It was defining its arguments as char * instead of void *.
1203 (b) It was assuming that all moves were upwards in memory; this was true
1204 a long time ago when I wrote it, but is no longer the case.
1206 The emulated memove() is provided for those environments that have neither
1207 memmove() nor bcopy(). I didn't think anyone used it these days, but that
1208 is clearly not the case, as these two bugs were recently reported.
1210 11. The script PrepareRelease is now distributed: it calls 132html, CleanTxt,
1211 and Detrail to create the HTML documentation, the .txt form of the man
1212 pages, and it removes trailing spaces from listed files. It also creates
1213 pcre.h.generic and config.h.generic from pcre.h and config.h. In the latter
1214 case, it wraps all the #defines with #ifndefs. This script should be run
1215 before "make dist".
1217 12. Fixed two fairly obscure bugs concerned with quantified caseless matching
1218 with Unicode property support.
1220 (a) For a maximizing quantifier, if the two different cases of the
1221 character were of different lengths in their UTF-8 codings (there are
1222 some cases like this - I found 11), and the matching function had to
1223 back up over a mixture of the two cases, it incorrectly assumed they
1224 were both the same length.
1226 (b) When PCRE was configured to use the heap rather than the stack for
1227 recursion during matching, it was not correctly preserving the data for
1228 the other case of a UTF-8 character when checking ahead for a match
1229 while processing a minimizing repeat. If the check also involved
1230 matching a wide character, but failed, corruption could cause an
1231 erroneous result when trying to check for a repeat of the original
1232 character.
1234 13. Some tidying changes to the testing mechanism:
1236 (a) The RunTest script now detects the internal link size and whether there
1237 is UTF-8 and UCP support by running ./pcretest -C instead of relying on
1238 values substituted by "configure". (The RunGrepTest script already did
1239 this for UTF-8.) The configure.ac script no longer substitutes the
1240 relevant variables.
1242 (b) The debugging options /B and /D in pcretest show the compiled bytecode
1243 with length and offset values. This means that the output is different
1244 for different internal link sizes. Test 2 is skipped for link sizes
1245 other than 2 because of this, bypassing the problem. Unfortunately,
1246 there was also a test in test 3 (the locale tests) that used /B and
1247 failed for link sizes other than 2. Rather than cut the whole test out,
1248 I have added a new /Z option to pcretest that replaces the length and
1249 offset values with spaces. This is now used to make test 3 independent
1250 of link size. (Test 2 will be tidied up later.)
1252 14. If erroroffset was passed as NULL to pcre_compile, it provoked a
1253 segmentation fault instead of returning the appropriate error message.
1255 15. In multiline mode when the newline sequence was set to "any", the pattern
1256 ^$ would give a match between the \r and \n of a subject such as "A\r\nB".
1257 This doesn't seem right; it now treats the CRLF combination as the line
1258 ending, and so does not match in that case. It's only a pattern such as ^$
1259 that would hit this one: something like ^ABC$ would have failed after \r
1260 and then tried again after \r\n.
1262 16. Changed the comparison command for RunGrepTest from "diff -u" to "diff -ub"
1263 in an attempt to make files that differ only in their line terminators
1264 compare equal. This works on Linux.
1266 17. Under certain error circumstances pcregrep might try to free random memory
1267 as it exited. This is now fixed, thanks to valgrind.
1269 19. In pcretest, if the pattern /(?m)^$/g<any> was matched against the string
1270 "abc\r\n\r\n", it found an unwanted second match after the second \r. This
1271 was because its rules for how to advance for /g after matching an empty
1272 string at the end of a line did not allow for this case. They now check for
1273 it specially.
1275 20. pcretest is supposed to handle patterns and data of any length, by
1276 extending its buffers when necessary. It was getting this wrong when the
1277 buffer for a data line had to be extended.
1279 21. Added PCRE_NEWLINE_ANYCRLF which is like ANY, but matches only CR, LF, or
1280 CRLF as a newline sequence.
1282 22. Code for handling Unicode properties in pcre_dfa_exec() wasn't being cut
1283 out by #ifdef SUPPORT_UCP. This did no harm, as it could never be used, but
1284 I have nevertheless tidied it up.
1286 23. Added some casts to kill warnings from HP-UX ia64 compiler.
1288 24. Added a man page for pcre-config.
1291 Version 7.0 19-Dec-06
1292 ---------------------
1294 1. Fixed a signed/unsigned compiler warning in pcre_compile.c, shown up by
1295 moving to gcc 4.1.1.
1297 2. The -S option for pcretest uses setrlimit(); I had omitted to #include
1298 sys/time.h, which is documented as needed for this function. It doesn't
1299 seem to matter on Linux, but it showed up on some releases of OS X.
1301 3. It seems that there are systems where bytes whose values are greater than
1302 127 match isprint() in the "C" locale. The "C" locale should be the
1303 default when a C program starts up. In most systems, only ASCII printing
1304 characters match isprint(). This difference caused the output from pcretest
1305 to vary, making some of the tests fail. I have changed pcretest so that:
1307 (a) When it is outputting text in the compiled version of a pattern, bytes
1308 other than 32-126 are always shown as hex escapes.
1310 (b) When it is outputting text that is a matched part of a subject string,
1311 it does the same, unless a different locale has been set for the match
1312 (using the /L modifier). In this case, it uses isprint() to decide.
1314 4. Fixed a major bug that caused incorrect computation of the amount of memory
1315 required for a compiled pattern when options that changed within the
1316 pattern affected the logic of the preliminary scan that determines the
1317 length. The relevant options are -x, and -i in UTF-8 mode. The result was
1318 that the computed length was too small. The symptoms of this bug were
1319 either the PCRE error "internal error: code overflow" from pcre_compile(),
1320 or a glibc crash with a message such as "pcretest: free(): invalid next
1321 size (fast)". Examples of patterns that provoked this bug (shown in
1322 pcretest format) are:
1324 /(?-x: )/x
1325 /(?x)(?-x: \s*#\s*)/
1326 /((?i)[\x{c0}])/8
1327 /(?i:[\x{c0}])/8
1329 HOWEVER: Change 17 below makes this fix obsolete as the memory computation
1330 is now done differently.
1332 5. Applied patches from Google to: (a) add a QuoteMeta function to the C++
1333 wrapper classes; (b) implement a new function in the C++ scanner that is
1334 more efficient than the old way of doing things because it avoids levels of
1335 recursion in the regex matching; (c) add a paragraph to the documentation
1336 for the FullMatch() function.
1338 6. The escape sequence \n was being treated as whatever was defined as
1339 "newline". Not only was this contrary to the documentation, which states
1340 that \n is character 10 (hex 0A), but it also went horribly wrong when
1341 "newline" was defined as CRLF. This has been fixed.
1343 7. In pcre_dfa_exec.c the value of an unsigned integer (the variable called c)
1344 was being set to -1 for the "end of line" case (supposedly a value that no
1345 character can have). Though this value is never used (the check for end of
1346 line is "zero bytes in current character"), it caused compiler complaints.
1347 I've changed it to 0xffffffff.
1349 8. In pcre_version.c, the version string was being built by a sequence of
1350 C macros that, in the event of PCRE_PRERELEASE being defined as an empty
1351 string (as it is for production releases) called a macro with an empty
1352 argument. The C standard says the result of this is undefined. The gcc
1353 compiler treats it as an empty string (which was what was wanted) but it is
1354 reported that Visual C gives an error. The source has been hacked around to
1355 avoid this problem.
1357 9. On the advice of a Windows user, included <io.h> and <fcntl.h> in Windows
1358 builds of pcretest, and changed the call to _setmode() to use _O_BINARY
1359 instead of 0x8000. Made all the #ifdefs test both _WIN32 and WIN32 (not all
1360 of them did).
1362 10. Originally, pcretest opened its input and output without "b"; then I was
1363 told that "b" was needed in some environments, so it was added for release
1364 5.0 to both the input and output. (It makes no difference on Unix-like
1365 systems.) Later I was told that it is wrong for the input on Windows. I've
1366 now abstracted the modes into two macros, to make it easier to fiddle with
1367 them, and removed "b" from the input mode under Windows.
1369 11. Added pkgconfig support for the C++ wrapper library, libpcrecpp.
1371 12. Added -help and --help to pcretest as an official way of being reminded
1372 of the options.
1374 13. Removed some redundant semicolons after macro calls in pcrecpparg.h.in
1375 and pcrecpp.cc because they annoy compilers at high warning levels.
1377 14. A bit of tidying/refactoring in pcre_exec.c in the main bumpalong loop.
1379 15. Fixed an occurrence of == in configure.ac that should have been = (shell
1380 scripts are not C programs :-) and which was not noticed because it works
1381 on Linux.
1383 16. pcretest is supposed to handle any length of pattern and data line (as one
1384 line or as a continued sequence of lines) by extending its input buffer if
1385 necessary. This feature was broken for very long pattern lines, leading to
1386 a string of junk being passed to pcre_compile() if the pattern was longer
1387 than about 50K.
1389 17. I have done a major re-factoring of the way pcre_compile() computes the
1390 amount of memory needed for a compiled pattern. Previously, there was code
1391 that made a preliminary scan of the pattern in order to do this. That was
1392 OK when PCRE was new, but as the facilities have expanded, it has become
1393 harder and harder to keep it in step with the real compile phase, and there
1394 have been a number of bugs (see for example, 4 above). I have now found a
1395 cunning way of running the real compile function in a "fake" mode that
1396 enables it to compute how much memory it would need, while actually only
1397 ever using a few hundred bytes of working memory and without too many
1398 tests of the mode. This should make future maintenance and development
1399 easier. A side effect of this work is that the limit of 200 on the nesting
1400 depth of parentheses has been removed (though this was never a serious
1401 limitation, I suspect). However, there is a downside: pcre_compile() now
1402 runs more slowly than before (30% or more, depending on the pattern). I
1403 hope this isn't a big issue. There is no effect on runtime performance.
1405 18. Fixed a minor bug in pcretest: if a pattern line was not terminated by a
1406 newline (only possible for the last line of a file) and it was a
1407 pattern that set a locale (followed by /Lsomething), pcretest crashed.
1409 19. Added additional timing features to pcretest. (1) The -tm option now times
1410 matching only, not compiling. (2) Both -t and -tm can be followed, as a
1411 separate command line item, by a number that specifies the number of
1412 repeats to use when timing. The default is 50000; this gives better
1413 precision, but takes uncomfortably long for very large patterns.
1415 20. Extended pcre_study() to be more clever in cases where a branch of a
1416 subpattern has no definite first character. For example, (a*|b*)[cd] would
1417 previously give no result from pcre_study(). Now it recognizes that the
1418 first character must be a, b, c, or d.
1420 21. There was an incorrect error "recursive call could loop indefinitely" if
1421 a subpattern (or the entire pattern) that was being tested for matching an
1422 empty string contained only one non-empty item after a nested subpattern.
1423 For example, the pattern (?>\x{100}*)\d(?R) provoked this error
1424 incorrectly, because the \d was being skipped in the check.
1426 22. The pcretest program now has a new pattern option /B and a command line
1427 option -b, which is equivalent to adding /B to every pattern. This causes
1428 it to show the compiled bytecode, without the additional information that
1429 -d shows. The effect of -d is now the same as -b with -i (and similarly, /D
1430 is the same as /B/I).
1432 23. A new optimization is now able automatically to treat some sequences such
1433 as a*b as a*+b. More specifically, if something simple (such as a character
1434 or a simple class like \d) has an unlimited quantifier, and is followed by
1435 something that cannot possibly match the quantified thing, the quantifier
1436 is automatically "possessified".
1438 24. A recursive reference to a subpattern whose number was greater than 39
1439 went wrong under certain circumstances in UTF-8 mode. This bug could also
1440 have affected the operation of pcre_study().
1442 25. Realized that a little bit of performance could be had by replacing
1443 (c & 0xc0) == 0xc0 with c >= 0xc0 when processing UTF-8 characters.
1445 26. Timing data from pcretest is now shown to 4 decimal places instead of 3.
1447 27. Possessive quantifiers such as a++ were previously implemented by turning
1448 them into atomic groups such as ($>a+). Now they have their own opcodes,
1449 which improves performance. This includes the automatically created ones
1450 from 23 above.
1452 28. A pattern such as (?=(\w+))\1: which simulates an atomic group using a
1453 lookahead was broken if it was not anchored. PCRE was mistakenly expecting
1454 the first matched character to be a colon. This applied both to named and
1455 numbered groups.
1457 29. The ucpinternal.h header file was missing its idempotency #ifdef.
1459 30. I was sent a "project" file called libpcre.a.dev which I understand makes
1460 building PCRE on Windows easier, so I have included it in the distribution.
1462 31. There is now a check in pcretest against a ridiculously large number being
1463 returned by pcre_exec() or pcre_dfa_exec(). If this happens in a /g or /G
1464 loop, the loop is abandoned.
1466 32. Forward references to subpatterns in conditions such as (?(2)...) where
1467 subpattern 2 is defined later cause pcre_compile() to search forwards in
1468 the pattern for the relevant set of parentheses. This search went wrong
1469 when there were unescaped parentheses in a character class, parentheses
1470 escaped with \Q...\E, or parentheses in a #-comment in /x mode.
1472 33. "Subroutine" calls and backreferences were previously restricted to
1473 referencing subpatterns earlier in the regex. This restriction has now
1474 been removed.
1476 34. Added a number of extra features that are going to be in Perl 5.10. On the
1477 whole, these are just syntactic alternatives for features that PCRE had
1478 previously implemented using the Python syntax or my own invention. The
1479 other formats are all retained for compatibility.
1481 (a) Named groups can now be defined as (?<name>...) or (?'name'...) as well
1482 as (?P<name>...). The new forms, as well as being in Perl 5.10, are
1483 also .NET compatible.
1485 (b) A recursion or subroutine call to a named group can now be defined as
1486 (?&name) as well as (?P>name).
1488 (c) A backreference to a named group can now be defined as \k<name> or
1489 \k'name' as well as (?P=name). The new forms, as well as being in Perl
1490 5.10, are also .NET compatible.
1492 (d) A conditional reference to a named group can now use the syntax
1493 (?(<name>) or (?('name') as well as (?(name).
1495 (e) A "conditional group" of the form (?(DEFINE)...) can be used to define
1496 groups (named and numbered) that are never evaluated inline, but can be
1497 called as "subroutines" from elsewhere. In effect, the DEFINE condition
1498 is always false. There may be only one alternative in such a group.
1500 (f) A test for recursion can be given as (?(R1).. or (?(R&name)... as well
1501 as the simple (?(R). The condition is true only if the most recent
1502 recursion is that of the given number or name. It does not search out
1503 through the entire recursion stack.
1505 (g) The escape \gN or \g{N} has been added, where N is a positive or
1506 negative number, specifying an absolute or relative reference.
1508 35. Tidied to get rid of some further signed/unsigned compiler warnings and
1509 some "unreachable code" warnings.
1511 36. Updated the Unicode property tables to Unicode version 5.0.0. Amongst other
1512 things, this adds five new scripts.
1514 37. Perl ignores orphaned \E escapes completely. PCRE now does the same.
1515 There were also incompatibilities regarding the handling of \Q..\E inside
1516 character classes, for example with patterns like [\Qa\E-\Qz\E] where the
1517 hyphen was adjacent to \Q or \E. I hope I've cleared all this up now.
1519 38. Like Perl, PCRE detects when an indefinitely repeated parenthesized group
1520 matches an empty string, and forcibly breaks the loop. There were bugs in
1521 this code in non-simple cases. For a pattern such as ^(a()*)* matched
1522 against aaaa the result was just "a" rather than "aaaa", for example. Two
1523 separate and independent bugs (that affected different cases) have been
1524 fixed.
1526 39. Refactored the code to abolish the use of different opcodes for small
1527 capturing bracket numbers. This is a tidy that I avoided doing when I
1528 removed the limit on the number of capturing brackets for 3.5 back in 2001.
1529 The new approach is not only tidier, it makes it possible to reduce the
1530 memory needed to fix the previous bug (38).
1532 40. Implemented PCRE_NEWLINE_ANY to recognize any of the Unicode newline
1533 sequences (http://unicode.org/unicode/reports/tr18/) as "newline" when
1534 processing dot, circumflex, or dollar metacharacters, or #-comments in /x
1535 mode.
1537 41. Add \R to match any Unicode newline sequence, as suggested in the Unicode
1538 report.
1540 42. Applied patch, originally from Ari Pollak, modified by Google, to allow
1541 copy construction and assignment in the C++ wrapper.
1543 43. Updated pcregrep to support "--newline=any". In the process, I fixed a
1544 couple of bugs that could have given wrong results in the "--newline=crlf"
1545 case.
1547 44. Added a number of casts and did some reorganization of signed/unsigned int
1548 variables following suggestions from Dair Grant. Also renamed the variable
1549 "this" as "item" because it is a C++ keyword.
1551 45. Arranged for dftables to add
1553 #include "pcre_internal.h"
1555 to pcre_chartables.c because without it, gcc 4.x may remove the array
1556 definition from the final binary if PCRE is built into a static library and
1557 dead code stripping is activated.
1559 46. For an unanchored pattern, if a match attempt fails at the start of a
1560 newline sequence, and the newline setting is CRLF or ANY, and the next two
1561 characters are CRLF, advance by two characters instead of one.
1564 Version 6.7 04-Jul-06
1565 ---------------------
1567 1. In order to handle tests when input lines are enormously long, pcretest has
1568 been re-factored so that it automatically extends its buffers when
1569 necessary. The code is crude, but this _is_ just a test program. The
1570 default size has been increased from 32K to 50K.
1572 2. The code in pcre_study() was using the value of the re argument before
1573 testing it for NULL. (Of course, in any sensible call of the function, it
1574 won't be NULL.)
1576 3. The memmove() emulation function in pcre_internal.h, which is used on
1577 systems that lack both memmove() and bcopy() - that is, hardly ever -
1578 was missing a "static" storage class specifier.
1580 4. When UTF-8 mode was not set, PCRE looped when compiling certain patterns
1581 containing an extended class (one that cannot be represented by a bitmap
1582 because it contains high-valued characters or Unicode property items, e.g.
1583 [\pZ]). Almost always one would set UTF-8 mode when processing such a
1584 pattern, but PCRE should not loop if you do not (it no longer does).
1585 [Detail: two cases were found: (a) a repeated subpattern containing an
1586 extended class; (b) a recursive reference to a subpattern that followed a
1587 previous extended class. It wasn't skipping over the extended class
1588 correctly when UTF-8 mode was not set.]
1590 5. A negated single-character class was not being recognized as fixed-length
1591 in lookbehind assertions such as (?<=[^f]), leading to an incorrect
1592 compile error "lookbehind assertion is not fixed length".
1594 6. The RunPerlTest auxiliary script was showing an unexpected difference
1595 between PCRE and Perl for UTF-8 tests. It turns out that it is hard to
1596 write a Perl script that can interpret lines of an input file either as
1597 byte characters or as UTF-8, which is what "perltest" was being required to
1598 do for the non-UTF-8 and UTF-8 tests, respectively. Essentially what you
1599 can't do is switch easily at run time between having the "use utf8;" pragma
1600 or not. In the end, I fudged it by using the RunPerlTest script to insert
1601 "use utf8;" explicitly for the UTF-8 tests.
1603 7. In multiline (/m) mode, PCRE was matching ^ after a terminating newline at
1604 the end of the subject string, contrary to the documentation and to what
1605 Perl does. This was true of both matching functions. Now it matches only at
1606 the start of the subject and immediately after *internal* newlines.
1608 8. A call of pcre_fullinfo() from pcretest to get the option bits was passing
1609 a pointer to an int instead of a pointer to an unsigned long int. This
1610 caused problems on 64-bit systems.
1612 9. Applied a patch from the folks at Google to pcrecpp.cc, to fix "another
1613 instance of the 'standard' template library not being so standard".
1615 10. There was no check on the number of named subpatterns nor the maximum
1616 length of a subpattern name. The product of these values is used to compute
1617 the size of the memory block for a compiled pattern. By supplying a very
1618 long subpattern name and a large number of named subpatterns, the size
1619 computation could be caused to overflow. This is now prevented by limiting
1620 the length of names to 32 characters, and the number of named subpatterns
1621 to 10,000.
1623 11. Subpatterns that are repeated with specific counts have to be replicated in
1624 the compiled pattern. The size of memory for this was computed from the
1625 length of the subpattern and the repeat count. The latter is limited to
1626 65535, but there was no limit on the former, meaning that integer overflow
1627 could in principle occur. The compiled length of a repeated subpattern is
1628 now limited to 30,000 bytes in order to prevent this.
1630 12. Added the optional facility to have named substrings with the same name.
1632 13. Added the ability to use a named substring as a condition, using the
1633 Python syntax: (?(name)yes|no). This overloads (?(R)... and names that
1634 are numbers (not recommended). Forward references are permitted.
1636 14. Added forward references in named backreferences (if you see what I mean).
1638 15. In UTF-8 mode, with the PCRE_DOTALL option set, a quantified dot in the
1639 pattern could run off the end of the subject. For example, the pattern
1640 "(?s)(.{1,5})"8 did this with the subject "ab".
1642 16. If PCRE_DOTALL or PCRE_MULTILINE were set, pcre_dfa_exec() behaved as if
1643 PCRE_CASELESS was set when matching characters that were quantified with ?
1644 or *.
1646 17. A character class other than a single negated character that had a minimum
1647 but no maximum quantifier - for example [ab]{6,} - was not handled
1648 correctly by pce_dfa_exec(). It would match only one character.
1650 18. A valid (though odd) pattern that looked like a POSIX character
1651 class but used an invalid character after [ (for example [[,abc,]]) caused
1652 pcre_compile() to give the error "Failed: internal error: code overflow" or
1653 in some cases to crash with a glibc free() error. This could even happen if
1654 the pattern terminated after [[ but there just happened to be a sequence of
1655 letters, a binary zero, and a closing ] in the memory that followed.
1657 19. Perl's treatment of octal escapes in the range \400 to \777 has changed
1658 over the years. Originally (before any Unicode support), just the bottom 8
1659 bits were taken. Thus, for example, \500 really meant \100. Nowadays the
1660 output from "man perlunicode" includes this:
1662 The regular expression compiler produces polymorphic opcodes. That
1663 is, the pattern adapts to the data and automatically switches to
1664 the Unicode character scheme when presented with Unicode data--or
1665 instead uses a traditional byte scheme when presented with byte
1666 data.
1668 Sadly, a wide octal escape does not cause a switch, and in a string with
1669 no other multibyte characters, these octal escapes are treated as before.
1670 Thus, in Perl, the pattern /\500/ actually matches \100 but the pattern
1671 /\500|\x{1ff}/ matches \500 or \777 because the whole thing is treated as a
1672 Unicode string.
1674 I have not perpetrated such confusion in PCRE. Up till now, it took just
1675 the bottom 8 bits, as in old Perl. I have now made octal escapes with
1676 values greater than \377 illegal in non-UTF-8 mode. In UTF-8 mode they
1677 translate to the appropriate multibyte character.
1679 29. Applied some refactoring to reduce the number of warnings from Microsoft
1680 and Borland compilers. This has included removing the fudge introduced
1681 seven years ago for the OS/2 compiler (see 2.02/2 below) because it caused
1682 a warning about an unused variable.
1684 21. PCRE has not included VT (character 0x0b) in the set of whitespace
1685 characters since release 4.0, because Perl (from release 5.004) does not.
1686 [Or at least, is documented not to: some releases seem to be in conflict
1687 with the documentation.] However, when a pattern was studied with
1688 pcre_study() and all its branches started with \s, PCRE still included VT
1689 as a possible starting character. Of course, this did no harm; it just
1690 caused an unnecessary match attempt.
1692 22. Removed a now-redundant internal flag bit that recorded the fact that case
1693 dependency changed within the pattern. This was once needed for "required
1694 byte" processing, but is no longer used. This recovers a now-scarce options
1695 bit. Also moved the least significant internal flag bit to the most-
1696 significant bit of the word, which was not previously used (hangover from
1697 the days when it was an int rather than a uint) to free up another bit for
1698 the future.
1700 23. Added support for CRLF line endings as well as CR and LF. As well as the
1701 default being selectable at build time, it can now be changed at runtime
1702 via the PCRE_NEWLINE_xxx flags. There are now options for pcregrep to
1703 specify that it is scanning data with non-default line endings.
1705 24. Changed the definition of CXXLINK to make it agree with the definition of
1706 LINK in the Makefile, by replacing LDFLAGS to CXXFLAGS.
1708 25. Applied Ian Taylor's patches to avoid using another stack frame for tail
1709 recursions. This makes a big different to stack usage for some patterns.
1711 26. If a subpattern containing a named recursion or subroutine reference such
1712 as (?P>B) was quantified, for example (xxx(?P>B)){3}, the calculation of
1713 the space required for the compiled pattern went wrong and gave too small a
1714 value. Depending on the environment, this could lead to "Failed: internal
1715 error: code overflow at offset 49" or "glibc detected double free or
1716 corruption" errors.
1718 27. Applied patches from Google (a) to support the new newline modes and (b) to
1719 advance over multibyte UTF-8 characters in GlobalReplace.
1721 28. Change free() to pcre_free() in pcredemo.c. Apparently this makes a
1722 difference for some implementation of PCRE in some Windows version.
1724 29. Added some extra testing facilities to pcretest:
1726 \q<number> in a data line sets the "match limit" value
1727 \Q<number> in a data line sets the "match recursion limt" value
1728 -S <number> sets the stack size, where <number> is in megabytes
1730 The -S option isn't available for Windows.
1733 Version 6.6 06-Feb-06
1734 ---------------------
1736 1. Change 16(a) for 6.5 broke things, because PCRE_DATA_SCOPE was not defined
1737 in pcreposix.h. I have copied the definition from pcre.h.
1739 2. Change 25 for 6.5 broke compilation in a build directory out-of-tree
1740 because pcre.h is no longer a built file.
1742 3. Added Jeff Friedl's additional debugging patches to pcregrep. These are
1743 not normally included in the compiled code.
1746 Version 6.5 01-Feb-06
1747 ---------------------
1749 1. When using the partial match feature with pcre_dfa_exec(), it was not
1750 anchoring the second and subsequent partial matches at the new starting
1751 point. This could lead to incorrect results. For example, with the pattern
1752 /1234/, partially matching against "123" and then "a4" gave a match.
1754 2. Changes to pcregrep:
1756 (a) All non-match returns from pcre_exec() were being treated as failures
1757 to match the line. Now, unless the error is PCRE_ERROR_NOMATCH, an
1758 error message is output. Some extra information is given for the
1760 probably the only errors that are likely to be caused by users (by
1761 specifying a regex that has nested indefinite repeats, for instance).
1762 If there are more than 20 of these errors, pcregrep is abandoned.
1764 (b) A binary zero was treated as data while matching, but terminated the
1765 output line if it was written out. This has been fixed: binary zeroes
1766 are now no different to any other data bytes.
1768 (c) Whichever of the LC_ALL or LC_CTYPE environment variables is set is
1769 used to set a locale for matching. The --locale=xxxx long option has
1770 been added (no short equivalent) to specify a locale explicitly on the
1771 pcregrep command, overriding the environment variables.
1773 (d) When -B was used with -n, some line numbers in the output were one less
1774 than they should have been.
1776 (e) Added the -o (--only-matching) option.
1778 (f) If -A or -C was used with -c (count only), some lines of context were
1779 accidentally printed for the final match.
1781 (g) Added the -H (--with-filename) option.
1783 (h) The combination of options -rh failed to suppress file names for files
1784 that were found from directory arguments.
1786 (i) Added the -D (--devices) and -d (--directories) options.
1788 (j) Added the -F (--fixed-strings) option.
1790 (k) Allow "-" to be used as a file name for -f as well as for a data file.
1792 (l) Added the --colo(u)r option.
1794 (m) Added Jeffrey Friedl's -S testing option, but within #ifdefs so that it
1795 is not present by default.
1797 3. A nasty bug was discovered in the handling of recursive patterns, that is,
1798 items such as (?R) or (?1), when the recursion could match a number of
1799 alternatives. If it matched one of the alternatives, but subsequently,
1800 outside the recursion, there was a failure, the code tried to back up into
1801 the recursion. However, because of the way PCRE is implemented, this is not
1802 possible, and the result was an incorrect result from the match.
1804 In order to prevent this happening, the specification of recursion has
1805 been changed so that all such subpatterns are automatically treated as
1806 atomic groups. Thus, for example, (?R) is treated as if it were (?>(?R)).
1808 4. I had overlooked the fact that, in some locales, there are characters for
1809 which isalpha() is true but neither isupper() nor islower() are true. In
1810 the fr_FR locale, for instance, the \xAA and \xBA characters (ordmasculine
1811 and ordfeminine) are like this. This affected the treatment of \w and \W
1812 when they appeared in character classes, but not when they appeared outside
1813 a character class. The bit map for "word" characters is now created
1814 separately from the results of isalnum() instead of just taking it from the
1815 upper, lower, and digit maps. (Plus the underscore character, of course.)
1817 5. The above bug also affected the handling of POSIX character classes such as
1818 [[:alpha:]] and [[:alnum:]]. These do not have their own bit maps in PCRE's
1819 permanent tables. Instead, the bit maps for such a class were previously
1820 created as the appropriate unions of the upper, lower, and digit bitmaps.
1821 Now they are created by subtraction from the [[:word:]] class, which has
1822 its own bitmap.
1824 6. The [[:blank:]] character class matches horizontal, but not vertical space.
1825 It is created by subtracting the vertical space characters (\x09, \x0a,
1826 \x0b, \x0c) from the [[:space:]] bitmap. Previously, however, the
1827 subtraction was done in the overall bitmap for a character class, meaning
1828 that a class such as [\x0c[:blank:]] was incorrect because \x0c would not
1829 be recognized. This bug has been fixed.
1831 7. Patches from the folks at Google:
1833 (a) pcrecpp.cc: "to handle a corner case that may or may not happen in
1834 real life, but is still worth protecting against".
1836 (b) pcrecpp.cc: "corrects a bug when negative radixes are used with
1837 regular expressions".
1839 (c) pcre_scanner.cc: avoid use of std::count() because not all systems
1840 have it.
1842 (d) Split off pcrecpparg.h from pcrecpp.h and had the former built by
1843 "configure" and the latter not, in order to fix a problem somebody had
1844 with compiling the Arg class on HP-UX.
1846 (e) Improve the error-handling of the C++ wrapper a little bit.
1848 (f) New tests for checking recursion limiting.
1850 8. The pcre_memmove() function, which is used only if the environment does not
1851 have a standard memmove() function (and is therefore rarely compiled),
1852 contained two bugs: (a) use of int instead of size_t, and (b) it was not
1853 returning a result (though PCRE never actually uses the result).
1855 9. In the POSIX regexec() interface, if nmatch is specified as a ridiculously
1856 large number - greater than INT_MAX/(3*sizeof(int)) - REG_ESPACE is
1857 returned instead of calling malloc() with an overflowing number that would
1858 most likely cause subsequent chaos.
1860 10. The debugging option of pcretest was not showing the NO_AUTO_CAPTURE flag.
1862 11. The POSIX flag REG_NOSUB is now supported. When a pattern that was compiled
1863 with this option is matched, the nmatch and pmatch options of regexec() are
1864 ignored.
1866 12. Added REG_UTF8 to the POSIX interface. This is not defined by POSIX, but is
1867 provided in case anyone wants to the the POSIX interface with UTF-8
1868 strings.
1870 13. Added CXXLDFLAGS to the Makefile parameters to provide settings only on the
1871 C++ linking (needed for some HP-UX environments).
1873 14. Avoid compiler warnings in get_ucpname() when compiled without UCP support
1874 (unused parameter) and in the pcre_printint() function (omitted "default"
1875 switch label when the default is to do nothing).
1877 15. Added some code to make it possible, when PCRE is compiled as a C++
1878 library, to replace subject pointers for pcre_exec() with a smart pointer
1879 class, thus making it possible to process discontinuous strings.
1881 16. The two macros PCRE_EXPORT and PCRE_DATA_SCOPE are confusing, and perform
1882 much the same function. They were added by different people who were trying
1883 to make PCRE easy to compile on non-Unix systems. It has been suggested
1884 that PCRE_EXPORT be abolished now that there is more automatic apparatus
1885 for compiling on Windows systems. I have therefore replaced it with
1886 PCRE_DATA_SCOPE. This is set automatically for Windows; if not set it
1887 defaults to "extern" for C or "extern C" for C++, which works fine on
1888 Unix-like systems. It is now possible to override the value of PCRE_DATA_
1889 SCOPE with something explicit in config.h. In addition:
1891 (a) pcreposix.h still had just "extern" instead of either of these macros;
1892 I have replaced it with PCRE_DATA_SCOPE.
1894 (b) Functions such as _pcre_xclass(), which are internal to the library,
1895 but external in the C sense, all had PCRE_EXPORT in their definitions.
1896 This is apparently wrong for the Windows case, so I have removed it.
1897 (It makes no difference on Unix-like systems.)
1899 17. Added a new limit, MATCH_LIMIT_RECURSION, which limits the depth of nesting
1900 of recursive calls to match(). This is different to MATCH_LIMIT because
1901 that limits the total number of calls to match(), not all of which increase
1902 the depth of recursion. Limiting the recursion depth limits the amount of
1903 stack (or heap if NO_RECURSE is set) that is used. The default can be set
1904 when PCRE is compiled, and changed at run time. A patch from Google adds
1905 this functionality to the C++ interface.
1907 18. Changes to the handling of Unicode character properties:
1909 (a) Updated the table to Unicode 4.1.0.
1911 (b) Recognize characters that are not in the table as "Cn" (undefined).
1913 (c) I revised the way the table is implemented to a much improved format
1914 which includes recognition of ranges. It now supports the ranges that
1915 are defined in UnicodeData.txt, and it also amalgamates other
1916 characters into ranges. This has reduced the number of entries in the
1917 table from around 16,000 to around 3,000, thus reducing its size
1918 considerably. I realized I did not need to use a tree structure after
1919 all - a binary chop search is just as efficient. Having reduced the
1920 number of entries, I extended their size from 6 bytes to 8 bytes to
1921 allow for more data.
1923 (d) Added support for Unicode script names via properties such as \p{Han}.
1925 19. In UTF-8 mode, a backslash followed by a non-Ascii character was not
1926 matching that character.
1928 20. When matching a repeated Unicode property with a minimum greater than zero,
1929 (for example \pL{2,}), PCRE could look past the end of the subject if it
1930 reached it while seeking the minimum number of characters. This could
1931 happen only if some of the characters were more than one byte long, because
1932 there is a check for at least the minimum number of bytes.
1934 21. Refactored the implementation of \p and \P so as to be more general, to
1935 allow for more different types of property in future. This has changed the
1936 compiled form incompatibly. Anybody with saved compiled patterns that use
1937 \p or \P will have to recompile them.
1939 22. Added "Any" and "L&" to the supported property types.
1941 23. Recognize \x{...} as a code point specifier, even when not in UTF-8 mode,
1942 but give a compile time error if the value is greater than 0xff.
1944 24. The man pages for pcrepartial, pcreprecompile, and pcre_compile2 were
1945 accidentally not being installed or uninstalled.
1947 25. The pcre.h file was built from pcre.h.in, but the only changes that were
1948 made were to insert the current release number. This seemed silly, because
1949 it made things harder for people building PCRE on systems that don't run
1950 "configure". I have turned pcre.h into a distributed file, no longer built
1951 by "configure", with the version identification directly included. There is
1952 no longer a pcre.h.in file.
1954 However, this change necessitated a change to the pcre-config script as
1955 well. It is built from pcre-config.in, and one of the substitutions was the
1956 release number. I have updated configure.ac so that ./configure now finds
1957 the release number by grepping pcre.h.
1959 26. Added the ability to run the tests under valgrind.
1962 Version 6.4 05-Sep-05
1963 ---------------------
1965 1. Change 6.0/10/(l) to pcregrep introduced a bug that caused separator lines
1966 "--" to be printed when multiple files were scanned, even when none of the
1967 -A, -B, or -C options were used. This is not compatible with Gnu grep, so I
1968 consider it to be a bug, and have restored the previous behaviour.
1970 2. A couple of code tidies to get rid of compiler warnings.
1972 3. The pcretest program used to cheat by referring to symbols in the library
1973 whose names begin with _pcre_. These are internal symbols that are not
1974 really supposed to be visible externally, and in some environments it is
1975 possible to suppress them. The cheating is now confined to including
1976 certain files from the library's source, which is a bit cleaner.
1978 4. Renamed pcre.in as pcre.h.in to go with pcrecpp.h.in; it also makes the
1979 file's purpose clearer.
1981 5. Reorganized pcre_ucp_findchar().
1984 Version 6.3 15-Aug-05
1985 ---------------------
1987 1. The file libpcre.pc.in did not have general read permission in the tarball.
1989 2. There were some problems when building without C++ support:
1991 (a) If C++ support was not built, "make install" and "make test" still
1992 tried to test it.
1994 (b) There were problems when the value of CXX was explicitly set. Some
1995 changes have been made to try to fix these, and ...
1997 (c) --disable-cpp can now be used to explicitly disable C++ support.
1999 (d) The use of @CPP_OBJ@ directly caused a blank line preceded by a
2000 backslash in a target when C++ was disabled. This confuses some
2001 versions of "make", apparently. Using an intermediate variable solves
2002 this. (Same for CPP_LOBJ.)
2004 3. $(LINK_FOR_BUILD) now includes $(CFLAGS_FOR_BUILD) and $(LINK)
2005 (non-Windows) now includes $(CFLAGS) because these flags are sometimes
2006 necessary on certain architectures.
2008 4. Added a setting of -export-symbols-regex to the link command to remove
2009 those symbols that are exported in the C sense, but actually are local
2010 within the library, and not documented. Their names all begin with
2011 "_pcre_". This is not a perfect job, because (a) we have to except some
2012 symbols that pcretest ("illegally") uses, and (b) the facility isn't always
2013 available (and never for static libraries). I have made a note to try to
2014 find a way round (a) in the future.
2017 Version 6.2 01-Aug-05
2018 ---------------------
2020 1. There was no test for integer overflow of quantifier values. A construction
2021 such as {1111111111111111} would give undefined results. What is worse, if
2022 a minimum quantifier for a parenthesized subpattern overflowed and became
2023 negative, the calculation of the memory size went wrong. This could have
2024 led to memory overwriting.
2026 2. Building PCRE using VPATH was broken. Hopefully it is now fixed.
2028 3. Added "b" to the 2nd argument of fopen() in dftables.c, for non-Unix-like
2029 operating environments where this matters.
2031 4. Applied Giuseppe Maxia's patch to add additional features for controlling
2032 PCRE options from within the C++ wrapper.
2034 5. Named capturing subpatterns were not being correctly counted when a pattern
2035 was compiled. This caused two problems: (a) If there were more than 100
2036 such subpatterns, the calculation of the memory needed for the whole
2037 compiled pattern went wrong, leading to an overflow error. (b) Numerical
2038 back references of the form \12, where the number was greater than 9, were
2039 not recognized as back references, even though there were sufficient
2040 previous subpatterns.
2042 6. Two minor patches to pcrecpp.cc in order to allow it to compile on older
2043 versions of gcc, e.g. 2.95.4.
2046 Version 6.1 21-Jun-05
2047 ---------------------
2049 1. There was one reference to the variable "posix" in pcretest.c that was not
2050 surrounded by "#if !defined NOPOSIX".
2052 2. Make it possible to compile pcretest without DFA support, UTF8 support, or
2053 the cross-check on the old pcre_info() function, for the benefit of the
2054 cut-down version of PCRE that is currently imported into Exim.
2056 3. A (silly) pattern starting with (?i)(?-i) caused an internal space
2057 allocation error. I've done the easy fix, which wastes 2 bytes for sensible
2058 patterns that start (?i) but I don't think that matters. The use of (?i) is
2059 just an example; this all applies to the other options as well.
2061 4. Since libtool seems to echo the compile commands it is issuing, the output
2062 from "make" can be reduced a bit by putting "@" in front of each libtool
2063 compile command.
2065 5. Patch from the folks at Google for configure.in to be a bit more thorough
2066 in checking for a suitable C++ installation before trying to compile the
2067 C++ stuff. This should fix a reported problem when a compiler was present,
2068 but no suitable headers.
2070 6. The man pages all had just "PCRE" as their title. I have changed them to
2071 be the relevant file name. I have also arranged that these names are
2072 retained in the file doc/pcre.txt, which is a concatenation in text format
2073 of all the man pages except the little individual ones for each function.
2075 7. The NON-UNIX-USE file had not been updated for the different set of source
2076 files that come with release 6. I also added a few comments about the C++
2077 wrapper.
2080 Version 6.0 07-Jun-05
2081 ---------------------
2083 1. Some minor internal re-organization to help with my DFA experiments.
2085 2. Some missing #ifdef SUPPORT_UCP conditionals in pcretest and printint that
2086 didn't matter for the library itself when fully configured, but did matter
2087 when compiling without UCP support, or within Exim, where the ucp files are
2088 not imported.
2090 3. Refactoring of the library code to split up the various functions into
2091 different source modules. The addition of the new DFA matching code (see
2092 below) to a single monolithic source would have made it really too
2093 unwieldy, quite apart from causing all the code to be include in a
2094 statically linked application, when only some functions are used. This is
2095 relevant even without the DFA addition now that patterns can be compiled in
2096 one application and matched in another.
2098 The downside of splitting up is that there have to be some external
2099 functions and data tables that are used internally in different modules of
2100 the library but which are not part of the API. These have all had their
2101 names changed to start with "_pcre_" so that they are unlikely to clash
2102 with other external names.
2104 4. Added an alternate matching function, pcre_dfa_exec(), which matches using
2105 a different (DFA) algorithm. Although it is slower than the original
2106 function, it does have some advantages for certain types of matching
2107 problem.
2109 5. Upgrades to pcretest in order to test the features of pcre_dfa_exec(),
2110 including restarting after a partial match.
2112 6. A patch for pcregrep that defines INVALID_FILE_ATTRIBUTES if it is not
2113 defined when compiling for Windows was sent to me. I have put it into the
2114 code, though I have no means of testing or verifying it.
2116 7. Added the pcre_refcount() auxiliary function.
2118 8. Added the PCRE_FIRSTLINE option. This constrains an unanchored pattern to
2119 match before or at the first newline in the subject string. In pcretest,
2120 the /f option on a pattern can be used to set this.
2122 9. A repeated \w when used in UTF-8 mode with characters greater than 256
2123 would behave wrongly. This has been present in PCRE since release 4.0.
2125 10. A number of changes to the pcregrep command:
2127 (a) Refactored how -x works; insert ^(...)$ instead of setting
2128 PCRE_ANCHORED and checking the length, in preparation for adding
2129 something similar for -w.
2131 (b) Added the -w (match as a word) option.
2133 (c) Refactored the way lines are read and buffered so as to have more
2134 than one at a time available.
2136 (d) Implemented a pcregrep test script.
2138 (e) Added the -M (multiline match) option. This allows patterns to match
2139 over several lines of the subject. The buffering ensures that at least
2140 8K, or the rest of the document (whichever is the shorter) is available
2141 for matching (and similarly the previous 8K for lookbehind assertions).
2143 (f) Changed the --help output so that it now says
2145 -w, --word-regex(p)
2147 instead of two lines, one with "regex" and the other with "regexp"
2148 because that confused at least one person since the short forms are the
2149 same. (This required a bit of code, as the output is generated
2150 automatically from a table. It wasn't just a text change.)
2152 (g) -- can be used to terminate pcregrep options if the next thing isn't an
2153 option but starts with a hyphen. Could be a pattern or a path name
2154 starting with a hyphen, for instance.
2156 (h) "-" can be given as a file name to represent stdin.
2158 (i) When file names are being printed, "(standard input)" is used for
2159 the standard input, for compatibility with GNU grep. Previously
2160 "<stdin>" was used.
2162 (j) The option --label=xxx can be used to supply a name to be used for
2163 stdin when file names are being printed. There is no short form.
2165 (k) Re-factored the options decoding logic because we are going to add
2166 two more options that take data. Such options can now be given in four
2167 different ways, e.g. "-fname", "-f name", "--file=name", "--file name".
2169 (l) Added the -A, -B, and -C options for requesting that lines of context
2170 around matches be printed.
2172 (m) Added the -L option to print the names of files that do not contain
2173 any matching lines, that is, the complement of -l.
2175 (n) The return code is 2 if any file cannot be opened, but pcregrep does
2176 continue to scan other files.
2178 (o) The -s option was incorrectly implemented. For compatibility with other
2179 greps, it now suppresses the error message for a non-existent or non-
2180 accessible file (but not the return code). There is a new option called
2181 -q that suppresses the output of matching lines, which was what -s was
2182 previously doing.
2184 (p) Added --include and --exclude options to specify files for inclusion
2185 and exclusion when recursing.
2187 11. The Makefile was not using the Autoconf-supported LDFLAGS macro properly.
2188 Hopefully, it now does.
2190 12. Missing cast in pcre_study().
2192 13. Added an "uninstall" target to the makefile.
2194 14. Replaced "extern" in the function prototypes in Makefile.in with
2195 "PCRE_DATA_SCOPE", which defaults to 'extern' or 'extern "C"' in the Unix
2196 world, but is set differently for Windows.
2198 15. Added a second compiling function called pcre_compile2(). The only
2199 difference is that it has an extra argument, which is a pointer to an
2200 integer error code. When there is a compile-time failure, this is set
2201 non-zero, in addition to the error test pointer being set to point to an
2202 error message. The new argument may be NULL if no error number is required
2203 (but then you may as well call pcre_compile(), which is now just a
2204 wrapper). This facility is provided because some applications need a
2205 numeric error indication, but it has also enabled me to tidy up the way
2206 compile-time errors are handled in the POSIX wrapper.
2208 16. Added VPATH=.libs to the makefile; this should help when building with one
2209 prefix path and installing with another. (Or so I'm told by someone who
2210 knows more about this stuff than I do.)
2212 17. Added a new option, REG_DOTALL, to the POSIX function regcomp(). This
2213 passes PCRE_DOTALL to the pcre_compile() function, making the "." character
2214 match everything, including newlines. This is not POSIX-compatible, but
2215 somebody wanted the feature. From pcretest it can be activated by using
2216 both the P and the s flags.
2218 18. AC_PROG_LIBTOOL appeared twice in Makefile.in. Removed one.
2220 19. libpcre.pc was being incorrectly installed as executable.
2222 20. A couple of places in pcretest check for end-of-line by looking for '\n';
2223 it now also looks for '\r' so that it will work unmodified on Windows.
2225 21. Added Google's contributed C++ wrapper to the distribution.
2227 22. Added some untidy missing memory free() calls in pcretest, to keep
2228 Electric Fence happy when testing.
2232 Version 5.0 13-Sep-04
2233 ---------------------
2235 1. Internal change: literal characters are no longer packed up into items
2236 containing multiple characters in a single byte-string. Each character
2237 is now matched using a separate opcode. However, there may be more than one
2238 byte in the character in UTF-8 mode.
2240 2. The pcre_callout_block structure has two new fields: pattern_position and
2241 next_item_length. These contain the offset in the pattern to the next match
2242 item, and its length, respectively.
2244 3. The PCRE_AUTO_CALLOUT option for pcre_compile() requests the automatic
2245 insertion of callouts before each pattern item. Added the /C option to
2246 pcretest to make use of this.
2248 4. On the advice of a Windows user, the lines
2250 #if defined(_WIN32) || defined(WIN32)
2251 _setmode( _fileno( stdout ), 0x8000 );
2252 #endif /* defined(_WIN32) || defined(WIN32) */
2254 have been added to the source of pcretest. This apparently does useful
2255 magic in relation to line terminators.
2257 5. Changed "r" and "w" in the calls to fopen() in pcretest to "rb" and "wb"
2258 for the benefit of those environments where the "b" makes a difference.
2260 6. The icc compiler has the same options as gcc, but "configure" doesn't seem
2261 to know about it. I have put a hack into configure.in that adds in code
2262 to set GCC=yes if CC=icc. This seems to end up at a point in the
2263 generated configure script that is early enough to affect the setting of
2264 compiler options, which is what is needed, but I have no means of testing
2265 whether it really works. (The user who reported this had patched the
2266 generated configure script, which of course I cannot do.)
2268 LATER: After change 22 below (new libtool files), the configure script
2269 seems to know about icc (and also ecc). Therefore, I have commented out
2270 this hack in configure.in.
2272 7. Added support for pkg-config (2 patches were sent in).
2274 8. Negated POSIX character classes that used a combination of internal tables
2275 were completely broken. These were [[:^alpha:]], [[:^alnum:]], and
2276 [[:^ascii]]. Typically, they would match almost any characters. The other
2277 POSIX classes were not broken in this way.
2279 9. Matching the pattern "\b.*?" against "ab cd", starting at offset 1, failed
2280 to find the match, as PCRE was deluded into thinking that the match had to
2281 start at the start point or following a newline. The same bug applied to
2282 patterns with negative forward assertions or any backward assertions
2283 preceding ".*" at the start, unless the pattern required a fixed first
2284 character. This was a failing pattern: "(?!.bcd).*". The bug is now fixed.
2286 10. In UTF-8 mode, when moving forwards in the subject after a failed match
2287 starting at the last subject character, bytes beyond the end of the subject
2288 string were read.
2290 11. Renamed the variable "class" as "classbits" to make life easier for C++
2291 users. (Previously there was a macro definition, but it apparently wasn't
2292 enough.)
2294 12. Added the new field "tables" to the extra data so that tables can be passed
2295 in at exec time, or the internal tables can be re-selected. This allows
2296 a compiled regex to be saved and re-used at a later time by a different
2297 program that might have everything at different addresses.
2299 13. Modified the pcre-config script so that, when run on Solaris, it shows a
2300 -R library as well as a -L library.
2302 14. The debugging options of pcretest (-d on the command line or D on a
2303 pattern) showed incorrect output for anything following an extended class
2304 that contained multibyte characters and which was followed by a quantifier.
2306 15. Added optional support for general category Unicode character properties
2307 via the \p, \P, and \X escapes. Unicode property support implies UTF-8
2308 support. It adds about 90K to the size of the library. The meanings of the
2309 inbuilt class escapes such as \d and \s have NOT been changed.
2311 16. Updated pcredemo.c to include calls to free() to release the memory for the
2312 compiled pattern.
2314 17. The generated file chartables.c was being created in the source directory
2315 instead of in the building directory. This caused the build to fail if the
2316 source directory was different from the building directory, and was
2317 read-only.
2319 18. Added some sample Win commands from Mark Tetrode into the NON-UNIX-USE
2320 file. No doubt somebody will tell me if they don't make sense... Also added
2321 Dan Mooney's comments about building on OpenVMS.
2323 19. Added support for partial matching via the PCRE_PARTIAL option for
2324 pcre_exec() and the \P data escape in pcretest.
2326 20. Extended pcretest with 3 new pattern features:
2328 (i) A pattern option of the form ">rest-of-line" causes pcretest to
2329 write the compiled pattern to the file whose name is "rest-of-line".
2330 This is a straight binary dump of the data, with the saved pointer to
2331 the character tables forced to be NULL. The study data, if any, is
2332 written too. After writing, pcretest reads a new pattern.
2334 (ii) If, instead of a pattern, "<rest-of-line" is given, pcretest reads a
2335 compiled pattern from the given file. There must not be any
2336 occurrences of "<" in the file name (pretty unlikely); if there are,
2337 pcretest will instead treat the initial "<" as a pattern delimiter.
2338 After reading in the pattern, pcretest goes on to read data lines as
2339 usual.
2341 (iii) The F pattern option causes pcretest to flip the bytes in the 32-bit
2342 and 16-bit fields in a compiled pattern, to simulate a pattern that
2343 was compiled on a host of opposite endianness.
2345 21. The pcre-exec() function can now cope with patterns that were compiled on
2346 hosts of opposite endianness, with this restriction:
2348 As for any compiled expression that is saved and used later, the tables
2349 pointer field cannot be preserved; the extra_data field in the arguments
2350 to pcre_exec() should be used to pass in a tables address if a value
2351 other than the default internal tables were used at compile time.
2353 22. Calling pcre_exec() with a negative value of the "ovecsize" parameter is
2354 now diagnosed as an error. Previously, most of the time, a negative number
2355 would have been treated as zero, but if in addition "ovector" was passed as
2356 NULL, a crash could occur.
2358 23. Updated the files ltmain.sh, config.sub, config.guess, and aclocal.m4 with
2359 new versions from the libtool 1.5 distribution (the last one is a copy of
2360 a file called libtool.m4). This seems to have fixed the need to patch
2361 "configure" to support Darwin 1.3 (which I used to do). However, I still
2362 had to patch ltmain.sh to ensure that ${SED} is set (it isn't on my
2363 workstation).
2365 24. Changed the PCRE licence to be the more standard "BSD" licence.
2368 Version 4.5 01-Dec-03
2369 ---------------------
2371 1. There has been some re-arrangement of the code for the match() function so
2372 that it can be compiled in a version that does not call itself recursively.
2373 Instead, it keeps those local variables that need separate instances for
2374 each "recursion" in a frame on the heap, and gets/frees frames whenever it
2375 needs to "recurse". Keeping track of where control must go is done by means
2376 of setjmp/longjmp. The whole thing is implemented by a set of macros that
2377 hide most of the details from the main code, and operates only if
2378 NO_RECURSE is defined while compiling pcre.c. If PCRE is built using the
2379 "configure" mechanism, "--disable-stack-for-recursion" turns on this way of
2380 operating.
2382 To make it easier for callers to provide specially tailored get/free
2383 functions for this usage, two new functions, pcre_stack_malloc, and
2384 pcre_stack_free, are used. They are always called in strict stacking order,
2385 and the size of block requested is always the same.
2387 The PCRE_CONFIG_STACKRECURSE info parameter can be used to find out whether
2388 PCRE has been compiled to use the stack or the heap for recursion. The
2389 -C option of pcretest uses this to show which version is compiled.
2391 A new data escape \S, is added to pcretest; it causes the amounts of store
2392 obtained and freed by both kinds of malloc/free at match time to be added
2393 to the output.
2395 2. Changed the locale test to use "fr_FR" instead of "fr" because that's
2396 what's available on my current Linux desktop machine.
2398 3. When matching a UTF-8 string, the test for a valid string at the start has
2399 been extended. If start_offset is not zero, PCRE now checks that it points
2400 to a byte that is the start of a UTF-8 character. If not, it returns
2401 PCRE_ERROR_BADUTF8_OFFSET (-11). Note: the whole string is still checked;
2402 this is necessary because there may be backward assertions in the pattern.
2403 When matching the same subject several times, it may save resources to use
2404 PCRE_NO_UTF8_CHECK on all but the first call if the string is long.
2406 4. The code for checking the validity of UTF-8 strings has been tightened so
2407 that it rejects (a) strings containing 0xfe or 0xff bytes and (b) strings
2408 containing "overlong sequences".
2410 5. Fixed a bug (appearing twice) that I could not find any way of exploiting!
2411 I had written "if ((digitab[*p++] && chtab_digit) == 0)" where the "&&"
2412 should have been "&", but it just so happened that all the cases this let
2413 through by mistake were picked up later in the function.
2415 6. I had used a variable called "isblank" - this is a C99 function, causing
2416 some compilers to warn. To avoid this, I renamed it (as "blankclass").
2418 7. Cosmetic: (a) only output another newline at the end of pcretest if it is
2419 prompting; (b) run "./pcretest /dev/null" at the start of the test script
2420 so the version is shown; (c) stop "make test" echoing "./RunTest".
2422 8. Added patches from David Burgess to enable PCRE to run on EBCDIC systems.
2424 9. The prototype for memmove() for systems that don't have it was using
2425 size_t, but the inclusion of the header that defines size_t was later. I've
2426 moved the #includes for the C headers earlier to avoid this.
2428 10. Added some adjustments to the code to make it easier to compiler on certain
2429 special systems:
2431 (a) Some "const" qualifiers were missing.
2432 (b) Added the macro EXPORT before all exported functions; by default this
2433 is defined to be empty.
2434 (c) Changed the dftables auxiliary program (that builds chartables.c) so
2435 that it reads its output file name as an argument instead of writing
2436 to the standard output and assuming this can be redirected.
2438 11. In UTF-8 mode, if a recursive reference (e.g. (?1)) followed a character
2439 class containing characters with values greater than 255, PCRE compilation
2440 went into a loop.
2442 12. A recursive reference to a subpattern that was within another subpattern
2443 that had a minimum quantifier of zero caused PCRE to crash. For example,
2444 (x(y(?2))z)? provoked this bug with a subject that got as far as the
2445 recursion. If the recursively-called subpattern itself had a zero repeat,
2446 that was OK.
2448 13. In pcretest, the buffer for reading a data line was set at 30K, but the
2449 buffer into which it was copied (for escape processing) was still set at
2450 1024, so long lines caused crashes.
2452 14. A pattern such as /[ab]{1,3}+/ failed to compile, giving the error
2453 "internal error: code overflow...". This applied to any character class
2454 that was followed by a possessive quantifier.
2456 15. Modified the Makefile to add libpcre.la as a prerequisite for
2457 libpcreposix.la because I was told this is needed for a parallel build to
2458 work.
2460 16. If a pattern that contained .* following optional items at the start was
2461 studied, the wrong optimizing data was generated, leading to matching
2462 errors. For example, studying /[ab]*.*c/ concluded, erroneously, that any
2463 matching string must start with a or b or c. The correct conclusion for
2464 this pattern is that a match can start with any character.
2467 Version 4.4 13-Aug-03
2468 ---------------------
2470 1. In UTF-8 mode, a character class containing characters with values between
2471 127 and 255 was not handled correctly if the compiled pattern was studied.
2472 In fixing this, I have also improved the studying algorithm for such
2473 classes (slightly).
2475 2. Three internal functions had redundant arguments passed to them. Removal
2476 might give a very teeny performance improvement.
2478 3. Documentation bug: the value of the capture_top field in a callout is *one
2479 more than* the number of the hightest numbered captured substring.
2481 4. The Makefile linked pcretest and pcregrep with -lpcre, which could result
2482 in incorrectly linking with a previously installed version. They now link
2483 explicitly with libpcre.la.
2485 5. configure.in no longer needs to recognize Cygwin specially.
2487 6. A problem in pcre.in for Windows platforms is fixed.
2489 7. If a pattern was successfully studied, and the -d (or /D) flag was given to
2490 pcretest, it used to include the size of the study block as part of its
2491 output. Unfortunately, the structure contains a field that has a different
2492 size on different hardware architectures. This meant that the tests that
2493 showed this size failed. As the block is currently always of a fixed size,
2494 this information isn't actually particularly useful in pcretest output, so
2495 I have just removed it.
2497 8. Three pre-processor statements accidentally did not start in column 1.
2498 Sadly, there are *still* compilers around that complain, even though
2499 standard C has not required this for well over a decade. Sigh.
2501 9. In pcretest, the code for checking callouts passed small integers in the
2502 callout_data field, which is a void * field. However, some picky compilers
2503 complained about the casts involved for this on 64-bit systems. Now
2504 pcretest passes the address of the small integer instead, which should get
2505 rid of the warnings.
2507 10. By default, when in UTF-8 mode, PCRE now checks for valid UTF-8 strings at
2508 both compile and run time, and gives an error if an invalid UTF-8 sequence
2509 is found. There is a option for disabling this check in cases where the
2510 string is known to be correct and/or the maximum performance is wanted.
2512 11. In response to a bug report, I changed one line in Makefile.in from
2514 -Wl,--out-implib,.libs/lib@WIN_PREFIX@pcreposix.dll.a \
2515 to
2516 -Wl,--out-implib,.libs/@WIN_PREFIX@libpcreposix.dll.a \
2518 to look similar to other lines, but I have no way of telling whether this
2519 is the right thing to do, as I do not use Windows. No doubt I'll get told
2520 if it's wrong...
2523 Version 4.3 21-May-03
2524 ---------------------
2526 1. Two instances of @WIN_PREFIX@ omitted from the Windows targets in the
2527 Makefile.
2529 2. Some refactoring to improve the quality of the code:
2531 (i) The utf8_table... variables are now declared "const".
2533 (ii) The code for \cx, which used the "case flipping" table to upper case
2534 lower case letters, now just substracts 32. This is ASCII-specific,
2535 but the whole concept of \cx is ASCII-specific, so it seems
2536 reasonable.
2538 (iii) PCRE was using its character types table to recognize decimal and
2539 hexadecimal digits in the pattern. This is silly, because it handles
2540 only 0-9, a-f, and A-F, but the character types table is locale-
2541 specific, which means strange things might happen. A private
2542 table is now used for this - though it costs 256 bytes, a table is
2543 much faster than multiple explicit tests. Of course, the standard
2544 character types table is still used for matching digits in subject
2545 strings against \d.
2547 (iv) Strictly, the identifier ESC_t is reserved by POSIX (all identifiers
2548 ending in _t are). So I've renamed it as ESC_tee.
2550 3. The first argument for regexec() in the POSIX wrapper should have been
2551 defined as "const".
2553 4. Changed pcretest to use malloc() for its buffers so that they can be
2554 Electric Fenced for debugging.
2556 5. There were several places in the code where, in UTF-8 mode, PCRE would try
2557 to read one or more bytes before the start of the subject string. Often this
2558 had no effect on PCRE's behaviour, but in some circumstances it could
2559 provoke a segmentation fault.
2561 6. A lookbehind at the start of a pattern in UTF-8 mode could also cause PCRE
2562 to try to read one or more bytes before the start of the subject string.
2564 7. A lookbehind in a pattern matched in non-UTF-8 mode on a PCRE compiled with
2565 UTF-8 support could misbehave in various ways if the subject string
2566 contained bytes with the 0x80 bit set and the 0x40 bit unset in a lookbehind
2567 area. (PCRE was not checking for the UTF-8 mode flag, and trying to move
2568 back over UTF-8 characters.)
2571 Version 4.2 14-Apr-03
2572 ---------------------
2574 1. Typo "#if SUPPORT_UTF8" instead of "#ifdef SUPPORT_UTF8" fixed.
2576 2. Changes to the building process, supplied by Ronald Landheer-Cieslak
2577 [ON_WINDOWS]: new variable, "#" on non-Windows platforms
2578 [NOT_ON_WINDOWS]: new variable, "#" on Windows platforms
2579 [WIN_PREFIX]: new variable, "cyg" for Cygwin
2580 * Makefile.in: use autoconf substitution for OBJEXT, EXEEXT, BUILD_OBJEXT
2582 Note: automatic setting of the BUILD variables is not yet working
2583 set CPPFLAGS and BUILD_CPPFLAGS (but don't use yet) - should be used at
2584 compile-time but not at link-time
2585 [LINK]: use for linking executables only
2586 make different versions for Windows and non-Windows
2587 [LINKLIB]: new variable, copy of UNIX-style LINK, used for linking
2588 libraries
2589 [LINK_FOR_BUILD]: new variable
2590 [OBJEXT]: use throughout
2591 [EXEEXT]: use throughout
2592 <winshared>: new target
2593 <wininstall>: new target
2594 <dftables.o>: use native compiler
2595 <dftables>: use native linker
2596 <install>: handle Windows platform correctly
2597 <clean>: ditto
2598 <check>: ditto
2599 copy DLL to top builddir before testing
2601 As part of these changes, -no-undefined was removed again. This was reported
2602 to give trouble on HP-UX 11.0, so getting rid of it seems like a good idea
2603 in any case.
2605 3. Some tidies to get rid of compiler warnings:
2607 . In the match_data structure, match_limit was an unsigned long int, whereas
2608 match_call_count was an int. I've made them both unsigned long ints.
2610 . In pcretest the fact that a const uschar * doesn't automatically cast to
2611 a void * provoked a warning.
2613 . Turning on some more compiler warnings threw up some "shadow" variables
2614 and a few more missing casts.
2616 4. If PCRE was complied with UTF-8 support, but called without the PCRE_UTF8
2617 option, a class that contained a single character with a value between 128
2618 and 255 (e.g. /[\xFF]/) caused PCRE to crash.
2620 5. If PCRE was compiled with UTF-8 support, but called without the PCRE_UTF8
2621 option, a class that contained several characters, but with at least one
2622 whose value was between 128 and 255 caused PCRE to crash.
2625 Version 4.1 12-Mar-03
2626 ---------------------
2628 1. Compiling with gcc -pedantic found a couple of places where casts were
2629 needed, and a string in dftables.c that was longer than standard compilers are
2630 required to support.
2632 2. Compiling with Sun's compiler found a few more places where the code could
2633 be tidied up in order to avoid warnings.
2635 3. The variables for cross-compiling were called HOST_CC and HOST_CFLAGS; the
2636 first of these names is deprecated in the latest Autoconf in favour of the name
2637 CC_FOR_BUILD, because "host" is typically used to mean the system on which the
2638 compiled code will be run. I can't find a reference for HOST_CFLAGS, but by
2639 analogy I have changed it to CFLAGS_FOR_BUILD.
2641 4. Added -no-undefined to the linking command in the Makefile, because this is
2642 apparently helpful for Windows. To make it work, also added "-L. -lpcre" to the
2643 linking step for the pcreposix library.
2645 5. PCRE was failing to diagnose the case of two named groups with the same
2646 name.
2648 6. A problem with one of PCRE's optimizations was discovered. PCRE remembers a
2649 literal character that is needed in the subject for a match, and scans along to
2650 ensure that it is present before embarking on the full matching process. This
2651 saves time in cases of nested unlimited repeats that are never going to match.
2652 Problem: the scan can take a lot of time if the subject is very long (e.g.
2653 megabytes), thus penalizing straightforward matches. It is now done only if the
2654 amount of subject to be scanned is less than 1000 bytes.
2656 7. A lesser problem with the same optimization is that it was recording the
2657 first character of an anchored pattern as "needed", thus provoking a search
2658 right along the subject, even when the first match of the pattern was going to
2659 fail. The "needed" character is now not set for anchored patterns, unless it
2660 follows something in the pattern that is of non-fixed length. Thus, it still
2661 fulfils its original purpose of finding quick non-matches in cases of nested
2662 unlimited repeats, but isn't used for simple anchored patterns such as /^abc/.
2665 Version 4.0 17-Feb-03
2666 ---------------------
2668 1. If a comment in an extended regex that started immediately after a meta-item
2669 extended to the end of string, PCRE compiled incorrect data. This could lead to
2670 all kinds of weird effects. Example: /#/ was bad; /()#/ was bad; /a#/ was not.
2672 2. Moved to autoconf 2.53 and libtool 1.4.2.
2674 3. Perl 5.8 no longer needs "use utf8" for doing UTF-8 things. Consequently,
2675 the special perltest8 script is no longer needed - all the tests can be run
2676 from a single perltest script.
2678 4. From 5.004, Perl has not included the VT character (0x0b) in the set defined
2679 by \s. It has now been removed in PCRE. This means it isn't recognized as
2680 whitespace in /x regexes too, which is the same as Perl. Note that the POSIX
2681 class [:space:] *does* include VT, thereby creating a mess.
2683 5. Added the class [:blank:] (a GNU extension from Perl 5.8) to match only
2684 space and tab.
2686 6. Perl 5.005 was a long time ago. It's time to amalgamate the tests that use
2687 its new features into the main test script, reducing the number of scripts.
2689 7. Perl 5.8 has changed the meaning of patterns like /a(?i)b/. Earlier versions
2690 were backward compatible, and made the (?i) apply to the whole pattern, as if
2691 /i were given. Now it behaves more logically, and applies the option setting
2692 only to what follows. PCRE has been changed to follow suit. However, if it
2693 finds options settings right at the start of the pattern, it extracts them into
2694 the global options, as before. Thus, they show up in the info data.
2696 8. Added support for the \Q...\E escape sequence. Characters in between are
2697 treated as literals. This is slightly different from Perl in that $ and @ are
2698 also handled as literals inside the quotes. In Perl, they will cause variable
2699 interpolation. Note the following examples:
2701 Pattern PCRE matches Perl matches
2703 \Qabc$xyz\E abc$xyz abc followed by the contents of $xyz
2704 \Qabc\$xyz\E abc\$xyz abc\$xyz
2705 \Qabc\E\$\Qxyz\E abc$xyz abc$xyz
2707 For compatibility with Perl, \Q...\E sequences are recognized inside character
2708 classes as well as outside them.
2710 9. Re-organized 3 code statements in pcretest to avoid "overflow in
2711 floating-point constant arithmetic" warnings from a Microsoft compiler. Added a
2712 (size_t) cast to one statement in pcretest and one in pcreposix to avoid
2713 signed/unsigned warnings.
2715 10. SunOS4 doesn't have strtoul(). This was used only for unpicking the -o
2716 option for pcretest, so I've replaced it by a simple function that does just
2717 that job.
2719 11. pcregrep was ending with code 0 instead of 2 for the commands "pcregrep" or
2720 "pcregrep -".
2722 12. Added "possessive quantifiers" ?+, *+, ++, and {,}+ which come from Sun's
2723 Java package. This provides some syntactic sugar for simple cases of what my
2724 documentation calls "once-only subpatterns". A pattern such as x*+ is the same
2725 as (?>x*). In other words, if what is inside (?>...) is just a single repeated
2726 item, you can use this simplified notation. Note that only makes sense with
2727 greedy quantifiers. Consequently, the use of the possessive quantifier forces
2728 greediness, whatever the setting of the PCRE_UNGREEDY option.
2730 13. A change of greediness default within a pattern was not taking effect at
2731 the current level for patterns like /(b+(?U)a+)/. It did apply to parenthesized
2732 subpatterns that followed. Patterns like /b+(?U)a+/ worked because the option
2733 was abstracted outside.
2735 14. PCRE now supports the \G assertion. It is true when the current matching
2736 position is at the start point of the match. This differs from \A when the
2737 starting offset is non-zero. Used with the /g option of pcretest (or similar
2738 code), it works in the same way as it does for Perl's /g option. If all
2739 alternatives of a regex begin with \G, the expression is anchored to the start
2740 match position, and the "anchored" flag is set in the compiled expression.
2742 15. Some bugs concerning the handling of certain option changes within patterns
2743 have been fixed. These applied to options other than (?ims). For example,
2744 "a(?x: b c )d" did not match "XabcdY" but did match "Xa b c dY". It should have
2745 been the other way round. Some of this was related to change 7 above.
2747 16. PCRE now gives errors for /[.x.]/ and /[=x=]/ as unsupported POSIX
2748 features, as Perl does. Previously, PCRE gave the warnings only for /[[.x.]]/
2749 and /[[=x=]]/. PCRE now also gives an error for /[:name:]/ because it supports
2750 POSIX classes only within a class (e.g. /[[:alpha:]]/).
2752 17. Added support for Perl's \C escape. This matches one byte, even in UTF8
2753 mode. Unlike ".", it always matches newline, whatever the setting of
2754 PCRE_DOTALL. However, PCRE does not permit \C to appear in lookbehind
2755 assertions. Perl allows it, but it doesn't (in general) work because it can't
2756 calculate the length of the lookbehind. At least, that's the case for Perl
2757 5.8.0 - I've been told they are going to document that it doesn't work in
2758 future.
2760 18. Added an error diagnosis for escapes that PCRE does not support: these are
2761 \L, \l, \N, \P, \p, \U, \u, and \X.
2763 19. Although correctly diagnosing a missing ']' in a character class, PCRE was
2764 reading past the end of the pattern in cases such as /[abcd/.
2766 20. PCRE was getting more memory than necessary for patterns with classes that
2767 contained both POSIX named classes and other characters, e.g. /[[:space:]abc/.
2769 21. Added some code, conditional on #ifdef VPCOMPAT, to make life easier for
2770 compiling PCRE for use with Virtual Pascal.
2772 22. Small fix to the Makefile to make it work properly if the build is done
2773 outside the source tree.
2775 23. Added a new extension: a condition to go with recursion. If a conditional
2776 subpattern starts with (?(R) the "true" branch is used if recursion has
2777 happened, whereas the "false" branch is used only at the top level.
2779 24. When there was a very long string of literal characters (over 255 bytes
2780 without UTF support, over 250 bytes with UTF support), the computation of how
2781 much memory was required could be incorrect, leading to segfaults or other
2782 strange effects.
2784 25. PCRE was incorrectly assuming anchoring (either to start of subject or to
2785 start of line for a non-DOTALL pattern) when a pattern started with (.*) and
2786 there was a subsequent back reference to those brackets. This meant that, for
2787 example, /(.*)\d+\1/ failed to match "abc123bc". Unfortunately, it isn't
2788 possible to check for precisely this case. All we can do is abandon the
2789 optimization if .* occurs inside capturing brackets when there are any back
2790 references whatsoever. (See below for a better fix that came later.)
2792 26. The handling of the optimization for finding the first character of a
2793 non-anchored pattern, and for finding a character that is required later in the
2794 match were failing in some cases. This didn't break the matching; it just
2795 failed to optimize when it could. The way this is done has been re-implemented.
2797 27. Fixed typo in error message for invalid (?R item (it said "(?p").
2799 28. Added a new feature that provides some of the functionality that Perl
2800 provides with (?{...}). The facility is termed a "callout". The way it is done
2801 in PCRE is for the caller to provide an optional function, by setting
2802 pcre_callout to its entry point. Like pcre_malloc and pcre_free, this is a
2803 global variable. By default it is unset, which disables all calling out. To get
2804 the function called, the regex must include (?C) at appropriate points. This
2805 is, in fact, equivalent to (?C0), and any number <= 255 may be given with (?C).
2806 This provides a means of identifying different callout points. When PCRE
2807 reaches such a point in the regex, if pcre_callout has been set, the external
2808 function is called. It is provided with data in a structure called
2809 pcre_callout_block, which is defined in pcre.h. If the function returns 0,
2810 matching continues; if it returns a non-zero value, the match at the current
2811 point fails. However, backtracking will occur if possible. [This was changed
2812 later and other features added - see item 49 below.]
2814 29. pcretest is upgraded to test the callout functionality. It provides a
2815 callout function that displays information. By default, it shows the start of
2816 the match and the current position in the text. There are some new data escapes
2817 to vary what happens:
2819 \C+ in addition, show current contents of captured substrings
2820 \C- do not supply a callout function
2821 \C!n return 1 when callout number n is reached
2822 \C!n!m return 1 when callout number n is reached for the mth time
2824 30. If pcregrep was called with the -l option and just a single file name, it
2825 output "<stdin>" if a match was found, instead of the file name.
2827 31. Improve the efficiency of the POSIX API to PCRE. If the number of capturing
2828 slots is less than POSIX_MALLOC_THRESHOLD, use a block on the stack to pass to
2829 pcre_exec(). This saves a malloc/free per call. The default value of
2830 POSIX_MALLOC_THRESHOLD is 10; it can be changed by --with-posix-malloc-threshold
2831 when configuring.
2833 32. The default maximum size of a compiled pattern is 64K. There have been a
2834 few cases of people hitting this limit. The code now uses macros to handle the
2835 storing of links as offsets within the compiled pattern. It defaults to 2-byte
2836 links, but this can be changed to 3 or 4 bytes by --with-link-size when
2837 configuring. Tests 2 and 5 work only with 2-byte links because they output
2838 debugging information about compiled patterns.
2840 33. Internal code re-arrangements:
2842 (a) Moved the debugging function for printing out a compiled regex into
2843 its own source file (printint.c) and used #include to pull it into
2844 pcretest.c and, when DEBUG is defined, into pcre.c, instead of having two
2845 separate copies.
2847 (b) Defined the list of op-code names for debugging as a macro in
2848 internal.h so that it is next to the definition of the opcodes.
2850 (c) Defined a table of op-code lengths for simpler skipping along compiled
2851 code. This is again a macro in internal.h so that it is next to the
2852 definition of the opcodes.
2854 34. Added support for recursive calls to individual subpatterns, along the
2855 lines of Robin Houston's patch (but implemented somewhat differently).
2857 35. Further mods to the Makefile to help Win32. Also, added code to pcregrep to
2858 allow it to read and process whole directories in Win32. This code was
2859 contributed by Lionel Fourquaux; it has not been tested by me.
2861 36. Added support for named subpatterns. The Python syntax (?P<name>...) is
2862 used to name a group. Names consist of alphanumerics and underscores, and must
2863 be unique. Back references use the syntax (?P=name) and recursive calls use
2864 (?P>name) which is a PCRE extension to the Python extension. Groups still have
2865 numbers. The function pcre_fullinfo() can be used after compilation to extract
2866 a name/number map. There are three relevant calls:
2868 PCRE_INFO_NAMEENTRYSIZE yields the size of each entry in the map
2869 PCRE_INFO_NAMECOUNT yields the number of entries
2870 PCRE_INFO_NAMETABLE yields a pointer to the map.
2872 The map is a vector of fixed-size entries. The size of each entry depends on
2873 the length of the longest name used. The first two bytes of each entry are the
2874 group number, most significant byte first. There follows the corresponding
2875 name, zero terminated. The names are in alphabetical order.
2877 37. Make the maximum literal string in the compiled code 250 for the non-UTF-8
2878 case instead of 255. Making it the same both with and without UTF-8 support
2879 means that the same test output works with both.
2881 38. There was a case of malloc(0) in the POSIX testing code in pcretest. Avoid
2882 calling malloc() with a zero argument.
2884 39. Change 25 above had to resort to a heavy-handed test for the .* anchoring
2885 optimization. I've improved things by keeping a bitmap of backreferences with
2886 numbers 1-31 so that if .* occurs inside capturing brackets that are not in
2887 fact referenced, the optimization can be applied. It is unlikely that a
2888 relevant occurrence of .* (i.e. one which might indicate anchoring or forcing
2889 the match to follow \n) will appear inside brackets with a number greater than
2890 31, but if it does, any back reference > 31 suppresses the optimization.
2892 40. Added a new compile-time option PCRE_NO_AUTO_CAPTURE. This has the effect
2893 of disabling numbered capturing parentheses. Any opening parenthesis that is
2894 not followed by ? behaves as if it were followed by ?: but named parentheses
2895 can still be used for capturing (and they will acquire numbers in the usual
2896 way).
2898 41. Redesigned the return codes from the match() function into yes/no/error so
2899 that errors can be passed back from deep inside the nested calls. A malloc
2900 failure while inside a recursive subpattern call now causes the
2901 PCRE_ERROR_NOMEMORY return instead of quietly going wrong.
2903 42. It is now possible to set a limit on the number of times the match()
2904 function is called in a call to pcre_exec(). This facility makes it possible to
2905 limit the amount of recursion and backtracking, though not in a directly
2906 obvious way, because the match() function is used in a number of different
2907 circumstances. The count starts from zero for each position in the subject
2908 string (for non-anchored patterns). The default limit is, for compatibility, a
2909 large number, namely 10 000 000. You can change this in two ways:
2911 (a) When configuring PCRE before making, you can use --with-match-limit=n
2912 to set a default value for the compiled library.
2914 (b) For each call to pcre_exec(), you can pass a pcre_extra block in which
2915 a different value is set. See 45 below.
2917 If the limit is exceeded, pcre_exec() returns PCRE_ERROR_MATCHLIMIT.
2919 43. Added a new function pcre_config(int, void *) to enable run-time extraction
2920 of things that can be changed at compile time. The first argument specifies
2921 what is wanted and the second points to where the information is to be placed.
2922 The current list of available information is:
2926 The output is an integer that is set to one if UTF-8 support is available;
2927 otherwise it is set to zero.
2931 The output is an integer that it set to the value of the code that is used for
2932 newline. It is either LF (10) or CR (13).
2936 The output is an integer that contains the number of bytes used for internal
2937 linkage in compiled expressions. The value is 2, 3, or 4. See item 32 above.
2941 The output is an integer that contains the threshold above which the POSIX
2942 interface uses malloc() for output vectors. See item 31 above.
2946 The output is an unsigned integer that contains the default limit of the number
2947 of match() calls in a pcre_exec() execution. See 42 above.
2949 44. pcretest has been upgraded by the addition of the -C option. This causes it
2950 to extract all the available output from the new pcre_config() function, and to
2951 output it. The program then exits immediately.
2953 45. A need has arisen to pass over additional data with calls to pcre_exec() in
2954 order to support additional features. One way would have been to define
2955 pcre_exec2() (for example) with extra arguments, but this would not have been
2956 extensible, and would also have required all calls to the original function to
2957 be mapped to the new one. Instead, I have chosen to extend the mechanism that
2958 is used for passing in "extra" data from pcre_study().
2960 The pcre_extra structure is now exposed and defined in pcre.h. It currently
2961 contains the following fields:
2963 flags a bitmap indicating which of the following fields are set
2964 study_data opaque data from pcre_study()
2965 match_limit a way of specifying a limit on match() calls for a specific
2966 call to pcre_exec()
2967 callout_data data for callouts (see 49 below)
2969 The flag bits are also defined in pcre.h, and are
2975 The pcre_study() function now returns one of these new pcre_extra blocks, with
2976 the actual study data pointed to by the study_data field, and the
2977 PCRE_EXTRA_STUDY_DATA flag set. This can be passed directly to pcre_exec() as
2978 before. That is, this change is entirely upwards-compatible and requires no
2979 change to existing code.
2981 If you want to pass in additional data to pcre_exec(), you can either place it
2982 in a pcre_extra block provided by pcre_study(), or create your own pcre_extra
2983 block.
2985 46. pcretest has been extended to test the PCRE_EXTRA_MATCH_LIMIT feature. If a
2986 data string contains the escape sequence \M, pcretest calls pcre_exec() several
2987 times with different match limits, until it finds the minimum value needed for
2988 pcre_exec() to complete. The value is then output. This can be instructive; for
2989 most simple matches the number is quite small, but for pathological cases it
2990 gets very large very quickly.
2992 47. There's a new option for pcre_fullinfo() called PCRE_INFO_STUDYSIZE. It
2993 returns the size of the data block pointed to by the study_data field in a
2994 pcre_extra block, that is, the value that was passed as the argument to
2995 pcre_malloc() when PCRE was getting memory in which to place the information
2996 created by pcre_study(). The fourth argument should point to a size_t variable.
2997 pcretest has been extended so that this information is shown after a successful
2998 pcre_study() call when information about the compiled regex is being displayed.
3000 48. Cosmetic change to Makefile: there's no need to have / after $(DESTDIR)
3001 because what follows is always an absolute path. (Later: it turns out that this
3002 is more than cosmetic for MinGW, because it doesn't like empty path
3003 components.)
3005 49. Some changes have been made to the callout feature (see 28 above):
3007 (i) A callout function now has three choices for what it returns:
3009 0 => success, carry on matching
3010 > 0 => failure at this point, but backtrack if possible
3011 < 0 => serious error, return this value from pcre_exec()
3013 Negative values should normally be chosen from the set of PCRE_ERROR_xxx
3014 values. In particular, returning PCRE_ERROR_NOMATCH forces a standard
3015 "match failed" error. The error number PCRE_ERROR_CALLOUT is reserved for
3016 use by callout functions. It will never be used by PCRE itself.
3018 (ii) The pcre_extra structure (see 45 above) has a void * field called
3019 callout_data, with corresponding flag bit PCRE_EXTRA_CALLOUT_DATA. The
3020 pcre_callout_block structure has a field of the same name. The contents of
3021 the field passed in the pcre_extra structure are passed to the callout
3022 function in the corresponding field in the callout block. This makes it
3023 easier to use the same callout-containing regex from multiple threads. For
3024 testing, the pcretest program has a new data escape
3026 \C*n pass the number n (may be negative) as callout_data
3028 If the callout function in pcretest receives a non-zero value as
3029 callout_data, it returns that value.
3031 50. Makefile wasn't handling CFLAGS properly when compiling dftables. Also,
3032 there were some redundant $(CFLAGS) in commands that are now specified as
3033 $(LINK), which already includes $(CFLAGS).
3035 51. Extensions to UTF-8 support are listed below. These all apply when (a) PCRE
3036 has been compiled with UTF-8 support *and* pcre_compile() has been compiled
3037 with the PCRE_UTF8 flag. Patterns that are compiled without that flag assume
3038 one-byte characters throughout. Note that case-insensitive matching applies
3039 only to characters whose values are less than 256. PCRE doesn't support the
3040 notion of cases for higher-valued characters.
3042 (i) A character class whose characters are all within 0-255 is handled as
3043 a bit map, and the map is inverted for negative classes. Previously, a
3044 character > 255 always failed to match such a class; however it should
3045 match if the class was a negative one (e.g. [^ab]). This has been fixed.
3047 (ii) A negated character class with a single character < 255 is coded as
3048 "not this character" (OP_NOT). This wasn't working properly when the test
3049 character was multibyte, either singly or repeated.
3051 (iii) Repeats of multibyte characters are now handled correctly in UTF-8
3052 mode, for example: \x{100}{2,3}.
3054 (iv) The character escapes \b, \B, \d, \D, \s, \S, \w, and \W (either
3055 singly or repeated) now correctly test multibyte characters. However,
3056 PCRE doesn't recognize any characters with values greater than 255 as
3057 digits, spaces, or word characters. Such characters always match \D, \S,
3058 and \W, and never match \d, \s, or \w.
3060 (v) Classes may now contain characters and character ranges with values
3061 greater than 255. For example: [ab\x{100}-\x{400}].
3063 (vi) pcregrep now has a --utf-8 option (synonym -u) which makes it call
3064 PCRE in UTF-8 mode.
3066 52. The info request value PCRE_INFO_FIRSTCHAR has been renamed
3067 PCRE_INFO_FIRSTBYTE because it is a byte value. However, the old name is
3068 retained for backwards compatibility. (Note that LASTLITERAL is also a byte
3069 value.)
3071 53. The single man page has become too large. I have therefore split it up into
3072 a number of separate man pages. These also give rise to individual HTML pages;
3073 these are now put in a separate directory, and there is an index.html page that
3074 lists them all. Some hyperlinking between the pages has been installed.
3076 54. Added convenience functions for handling named capturing parentheses.
3078 55. Unknown escapes inside character classes (e.g. [\M]) and escapes that
3079 aren't interpreted therein (e.g. [\C]) are literals in Perl. This is now also
3080 true in PCRE, except when the PCRE_EXTENDED option is set, in which case they
3081 are faulted.
3083 56. Introduced HOST_CC and HOST_CFLAGS which can be set in the environment when
3084 calling configure. These values are used when compiling the dftables.c program
3085 which is run to generate the source of the default character tables. They
3086 default to the values of CC and CFLAGS. If you are cross-compiling PCRE,
3087 you will need to set these values.
3089 57. Updated the building process for Windows DLL, as provided by Fred Cox.
3092 Version 3.9 02-Jan-02
3093 ---------------------
3095 1. A bit of extraneous text had somehow crept into the pcregrep documentation.
3097 2. If --disable-static was given, the building process failed when trying to
3098 build pcretest and pcregrep. (For some reason it was using libtool to compile
3099 them, which is not right, as they aren't part of the library.)
3102 Version 3.8 18-Dec-01
3103 ---------------------
3105 1. The experimental UTF-8 code was completely screwed up. It was packing the
3106 bytes in the wrong order. How dumb can you get?
3109 Version 3.7 29-Oct-01
3110 ---------------------
3112 1. In updating pcretest to check change 1 of version 3.6, I screwed up.
3113 This caused pcretest, when used on the test data, to segfault. Unfortunately,
3114 this didn't happen under Solaris 8, where I normally test things.
3116 2. The Makefile had to be changed to make it work on BSD systems, where 'make'
3117 doesn't seem to recognize that ./xxx and xxx are the same file. (This entry
3118 isn't in ChangeLog distributed with 3.7 because I forgot when I hastily made
3119 this fix an hour or so after the initial 3.7 release.)
3122 Version 3.6 23-Oct-01
3123 ---------------------
3125 1. Crashed with /(sens|respons)e and \1ibility/ and "sense and sensibility" if
3126 offsets passed as NULL with zero offset count.
3128 2. The config.guess and config.sub files had not been updated when I moved to
3129 the latest autoconf.
3132 Version 3.5 15-Aug-01
3133 ---------------------
3135 1. Added some missing #if !defined NOPOSIX conditionals in pcretest.c that
3136 had been forgotten.
3138 2. By using declared but undefined structures, we can avoid using "void"
3139 definitions in pcre.h while keeping the internal definitions of the structures
3140 private.
3142 3. The distribution is now built using autoconf 2.50 and libtool 1.4. From a
3143 user point of view, this means that both static and shared libraries are built
3144 by default, but this can be individually controlled. More of the work of
3145 handling this static/shared cases is now inside libtool instead of PCRE's make
3146 file.
3148 4. The pcretest utility is now installed along with pcregrep because it is
3149 useful for users (to test regexs) and by doing this, it automatically gets
3150 relinked by libtool. The documentation has been turned into a man page, so
3151 there are now .1, .txt, and .html versions in /doc.
3153 5. Upgrades to pcregrep:
3154 (i) Added long-form option names like gnu grep.
3155 (ii) Added --help to list all options with an explanatory phrase.
3156 (iii) Added -r, --recursive to recurse into sub-directories.
3157 (iv) Added -f, --file to read patterns from a file.
3159 6. pcre_exec() was referring to its "code" argument before testing that
3160 argument for NULL (and giving an error if it was NULL).
3162 7. Upgraded Makefile.in to allow for compiling in a different directory from
3163 the source directory.
3165 8. Tiny buglet in pcretest: when pcre_fullinfo() was called to retrieve the
3166 options bits, the pointer it was passed was to an int instead of to an unsigned
3167 long int. This mattered only on 64-bit systems.
3169 9. Fixed typo (3.4/1) in pcre.h again. Sigh. I had changed pcre.h (which is
3170 generated) instead of pcre.in, which it its source. Also made the same change
3171 in several of the .c files.
3173 10. A new release of gcc defines printf() as a macro, which broke pcretest
3174 because it had an ifdef in the middle of a string argument for printf(). Fixed
3175 by using separate calls to printf().
3177 11. Added --enable-newline-is-cr and --enable-newline-is-lf to the configure
3178 script, to force use of CR or LF instead of \n in the source. On non-Unix
3179 systems, the value can be set in config.h.
3181 12. The limit of 200 on non-capturing parentheses is a _nesting_ limit, not an
3182 absolute limit. Changed the text of the error message to make this clear, and
3183 likewise updated the man page.
3185 13. The limit of 99 on the number of capturing subpatterns has been removed.
3186 The new limit is 65535, which I hope will not be a "real" limit.
3189 Version 3.4 22-Aug-00
3190 ---------------------
3192 1. Fixed typo in pcre.h: unsigned const char * changed to const unsigned char *.
3194 2. Diagnose condition (?(0) as an error instead of crashing on matching.
3197 Version 3.3 01-Aug-00
3198 ---------------------
3200 1. If an octal character was given, but the value was greater than \377, it
3201 was not getting masked to the least significant bits, as documented. This could
3202 lead to crashes in some systems.
3204 2. Perl 5.6 (if not earlier versions) accepts classes like [a-\d] and treats
3205 the hyphen as a literal. PCRE used to give an error; it now behaves like Perl.
3207 3. Added the functions pcre_free_substring() and pcre_free_substring_list().
3208 These just pass their arguments on to (pcre_free)(), but they are provided
3209 because some uses of PCRE bind it to non-C systems that can call its functions,
3210 but cannot call free() or pcre_free() directly.
3212 4. Add "make test" as a synonym for "make check". Corrected some comments in
3213 the Makefile.
3215 5. Add $(DESTDIR)/ in front of all the paths in the "install" target in the
3216 Makefile.
3218 6. Changed the name of pgrep to pcregrep, because Solaris has introduced a
3219 command called pgrep for grepping around the active processes.
3221 7. Added the beginnings of support for UTF-8 character strings.
3223 8. Arranged for the Makefile to pass over the settings of CC, CFLAGS, and
3224 RANLIB to ./ltconfig so that they are used by libtool. I think these are all
3225 the relevant ones. (AR is not passed because ./ltconfig does its own figuring
3226 out for the ar command.)
3229 Version 3.2 12-May-00
3230 ---------------------
3232 This is purely a bug fixing release.
3234 1. If the pattern /((Z)+|A)*/ was matched agained ZABCDEFG it matched Z instead
3235 of ZA. This was just one example of several cases that could provoke this bug,
3236 which was introduced by change 9 of version 2.00. The code for breaking
3237 infinite loops after an iteration that matches an empty string was't working
3238 correctly.
3240 2. The pcretest program was not imitating Perl correctly for the pattern /a*/g
3241 when matched against abbab (for example). After matching an empty string, it
3242 wasn't forcing anchoring when setting PCRE_NOTEMPTY for the next attempt; this
3243 caused it to match further down the string than it should.
3245 3. The code contained an inclusion of sys/types.h. It isn't clear why this
3246 was there because it doesn't seem to be needed, and it causes trouble on some
3247 systems, as it is not a Standard C header. It has been removed.
3249 4. Made 4 silly changes to the source to avoid stupid compiler warnings that
3250 were reported on the Macintosh. The changes were from
3252 while ((c = *(++ptr)) != 0 && c != '\n');
3253 to
3254 while ((c = *(++ptr)) != 0 && c != '\n') ;
3256 Totally extraordinary, but if that's what it takes...
3258 5. PCRE is being used in one environment where neither memmove() nor bcopy() is
3259 available. Added HAVE_BCOPY and an autoconf test for it; if neither
3260 HAVE_MEMMOVE nor HAVE_BCOPY is set, use a built-in emulation function which
3261 assumes the way PCRE uses memmove() (always moving upwards).
3263 6. PCRE is being used in one environment where strchr() is not available. There
3264 was only one use in pcre.c, and writing it out to avoid strchr() probably gives
3265 faster code anyway.
3268 Version 3.1 09-Feb-00
3269 ---------------------
3271 The only change in this release is the fixing of some bugs in Makefile.in for
3272 the "install" target:
3274 (1) It was failing to install pcreposix.h.
3276 (2) It was overwriting the pcre.3 man page with the pcreposix.3 man page.
3279 Version 3.0 01-Feb-00
3280 ---------------------
3282 1. Add support for the /+ modifier to perltest (to output $` like it does in
3283 pcretest).
3285 2. Add support for the /g modifier to perltest.
3287 3. Fix pcretest so that it behaves even more like Perl for /g when the pattern
3288 matches null strings.
3290 4. Fix perltest so that it doesn't do unwanted things when fed an empty
3291 pattern. Perl treats empty patterns specially - it reuses the most recent
3292 pattern, which is not what we want. Replace // by /(?#)/ in order to avoid this
3293 effect.
3295 5. The POSIX interface was broken in that it was just handing over the POSIX
3296 captured string vector to pcre_exec(), but (since release 2.00) PCRE has
3297 required a bigger vector, with some working space on the end. This means that
3298 the POSIX wrapper now has to get and free some memory, and copy the results.
3300 6. Added some simple autoconf support, placing the test data and the
3301 documentation in separate directories, re-organizing some of the
3302 information files, and making it build pcre-config (a GNU standard). Also added
3303 libtool support for building PCRE as a shared library, which is now the
3304 default.
3306 7. Got rid of the leading zero in the definition of PCRE_MINOR because 08 and
3307 09 are not valid octal constants. Single digits will be used for minor values
3308 less than 10.
3310 8. Defined REG_EXTENDED and REG_NOSUB as zero in the POSIX header, so that
3311 existing programs that set these in the POSIX interface can use PCRE without
3312 modification.
3314 9. Added a new function, pcre_fullinfo() with an extensible interface. It can
3315 return all that pcre_info() returns, plus additional data. The pcre_info()
3316 function is retained for compatibility, but is considered to be obsolete.
3318 10. Added experimental recursion feature (?R) to handle one common case that
3319 Perl 5.6 will be able to do with (?p{...}).
3321 11. Added support for POSIX character classes like [:alpha:], which Perl is
3322 adopting.
3325 Version 2.08 31-Aug-99
3326 ----------------------
3328 1. When startoffset was not zero and the pattern began with ".*", PCRE was not
3329 trying to match at the startoffset position, but instead was moving forward to
3330 the next newline as if a previous match had failed.
3332 2. pcretest was not making use of PCRE_NOTEMPTY when repeating for /g and /G,
3333 and could get into a loop if a null string was matched other than at the start
3334 of the subject.
3336 3. Added definitions of PCRE_MAJOR and PCRE_MINOR to pcre.h so the version can
3337 be distinguished at compile time, and for completeness also added PCRE_DATE.
3339 5. Added Paul Sokolovsky's minor changes to make it easy to compile a Win32 DLL
3340 in GnuWin32 environments.
3343 Version 2.07 29-Jul-99
3344 ----------------------
3346 1. The documentation is now supplied in plain text form and HTML as well as in
3347 the form of man page sources.
3349 2. C++ compilers don't like assigning (void *) values to other pointer types.
3350 In particular this affects malloc(). Although there is no problem in Standard
3351 C, I've put in casts to keep C++ compilers happy.
3353 3. Typo on pcretest.c; a cast of (unsigned char *) in the POSIX regexec() call
3354 should be (const char *).
3356 4. If NOPOSIX is defined, pcretest.c compiles without POSIX support. This may
3357 be useful for non-Unix systems who don't want to bother with the POSIX stuff.
3358 However, I haven't made this a standard facility. The documentation doesn't
3359 mention it, and the Makefile doesn't support it.
3361 5. The Makefile now contains an "install" target, with editable destinations at
3362 the top of the file. The pcretest program is not installed.
3364 6. pgrep -V now gives the PCRE version number and date.
3366 7. Fixed bug: a zero repetition after a literal string (e.g. /abcde{0}/) was
3367 causing the entire string to be ignored, instead of just the last character.
3369 8. If a pattern like /"([^\\"]+|\\.)*"/ is applied in the normal way to a
3370 non-matching string, it can take a very, very long time, even for strings of
3371 quite modest length, because of the nested recursion. PCRE now does better in
3372 some of these cases. It does this by remembering the last required literal
3373 character in the pattern, and pre-searching the subject to ensure it is present
3374 before running the real match. In other words, it applies a heuristic to detect
3375 some types of certain failure quickly, and in the above example, if presented
3376 with a string that has no trailing " it gives "no match" very quickly.
3378 9. A new runtime option PCRE_NOTEMPTY causes null string matches to be ignored;
3379 other alternatives are tried instead.
3382 Version 2.06 09-Jun-99
3383 ----------------------
3385 1. Change pcretest's output for amount of store used to show just the code
3386 space, because the remainder (the data block) varies in size between 32-bit and
3387 64-bit systems.
3389 2. Added an extra argument to pcre_exec() to supply an offset in the subject to
3390 start matching at. This allows lookbehinds to work when searching for multiple
3391 occurrences in a string.
3393 3. Added additional options to pcretest for testing multiple occurrences:
3395 /+ outputs the rest of the string that follows a match
3396 /g loops for multiple occurrences, using the new startoffset argument
3397 /G loops for multiple occurrences by passing an incremented pointer
3399 4. PCRE wasn't doing the "first character" optimization for patterns starting
3400 with \b or \B, though it was doing it for other lookbehind assertions. That is,
3401 it wasn't noticing that a match for a pattern such as /\bxyz/ has to start with
3402 the letter 'x'. On long subject strings, this gives a significant speed-up.
3405 Version 2.05 21-Apr-99
3406 ----------------------
3408 1. Changed the type of magic_number from int to long int so that it works
3409 properly on 16-bit systems.
3411 2. Fixed a bug which caused patterns starting with .* not to work correctly
3412 when the subject string contained newline characters. PCRE was assuming
3413 anchoring for such patterns in all cases, which is not correct because .* will
3414 not pass a newline unless PCRE_DOTALL is set. It now assumes anchoring only if
3415 DOTALL is set at top level; otherwise it knows that patterns starting with .*
3416 must be retried after every newline in the subject.
3419 Version 2.04 18-Feb-99
3420 ----------------------
3422 1. For parenthesized subpatterns with repeats whose minimum was zero, the
3423 computation of the store needed to hold the pattern was incorrect (too large).
3424 If such patterns were nested a few deep, this could multiply and become a real
3425 problem.
3427 2. Added /M option to pcretest to show the memory requirement of a specific
3428 pattern. Made -m a synonym of -s (which does this globally) for compatibility.
3430 3. Subpatterns of the form (regex){n,m} (i.e. limited maximum) were being
3431 compiled in such a way that the backtracking after subsequent failure was
3432 pessimal. Something like (a){0,3} was compiled as (a)?(a)?(a)? instead of
3433 ((a)((a)(a)?)?)? with disastrous performance if the maximum was of any size.
3436 Version 2.03 02-Feb-99
3437 ----------------------
3439 1. Fixed typo and small mistake in man page.
3441 2. Added 4th condition (GPL supersedes if conflict) and created separate
3442 LICENCE file containing the conditions.
3444 3. Updated pcretest so that patterns such as /abc\/def/ work like they do in
3445 Perl, that is the internal \ allows the delimiter to be included in the
3446 pattern. Locked out the use of \ as a delimiter. If \ immediately follows
3447 the final delimiter, add \ to the end of the pattern (to test the error).
3449 4. Added the convenience functions for extracting substrings after a successful
3450 match. Updated pcretest to make it able to test these functions.
3453 Version 2.02 14-Jan-99
3454 ----------------------
3456 1. Initialized the working variables associated with each extraction so that
3457 their saving and restoring doesn't refer to uninitialized store.
3459 2. Put dummy code into study.c in order to trick the optimizer of the IBM C
3460 compiler for OS/2 into generating correct code. Apparently IBM isn't going to
3461 fix the problem.
3463 3. Pcretest: the timing code wasn't using LOOPREPEAT for timing execution
3464 calls, and wasn't printing the correct value for compiling calls. Increased the
3465 default value of LOOPREPEAT, and the number of significant figures in the
3466 times.
3468 4. Changed "/bin/rm" in the Makefile to "-rm" so it works on Windows NT.
3470 5. Renamed "deftables" as "dftables" to get it down to 8 characters, to avoid
3471 a building problem on Windows NT with a FAT file system.
3474 Version 2.01 21-Oct-98
3475 ----------------------
3477 1. Changed the API for pcre_compile() to allow for the provision of a pointer
3478 to character tables built by pcre_maketables() in the current locale. If NULL
3479 is passed, the default tables are used.
3482 Version 2.00 24-Sep-98
3483 ----------------------
3485 1. Since the (>?) facility is in Perl 5.005, don't require PCRE_EXTRA to enable
3486 it any more.
3488 2. Allow quantification of (?>) groups, and make it work correctly.
3490 3. The first character computation wasn't working for (?>) groups.
3492 4. Correct the implementation of \Z (it is permitted to match on the \n at the
3493 end of the subject) and add 5.005's \z, which really does match only at the
3494 very end of the subject.
3496 5. Remove the \X "cut" facility; Perl doesn't have it, and (?> is neater.
3498 6. Remove the ability to specify CASELESS, MULTILINE, DOTALL, and
3499 DOLLAR_END_ONLY at runtime, to make it possible to implement the Perl 5.005
3500 localized options. All options to pcre_study() were also removed.
3502 7. Add other new features from 5.005:
3504 $(?<= positive lookbehind
3505 $(?<! negative lookbehind
3506 (?imsx-imsx) added the unsetting capability
3507 such a setting is global if at outer level; local otherwise
3508 (?imsx-imsx:) non-capturing groups with option setting
3509 (?(cond)re|re) conditional pattern matching
3511 A backreference to itself in a repeated group matches the previous
3512 captured string.
3514 8. General tidying up of studying (both automatic and via "study")
3515 consequential on the addition of new assertions.
3517 9. As in 5.005, unlimited repeated groups that could match an empty substring
3518 are no longer faulted at compile time. Instead, the loop is forcibly broken at
3519 runtime if any iteration does actually match an empty substring.
3521 10. Include the RunTest script in the distribution.
3523 11. Added tests from the Perl 5.005_02 distribution. This showed up a few
3524 discrepancies, some of which were old and were also with respect to 5.004. They
3525 have now been fixed.
3528 Version 1.09 28-Apr-98
3529 ----------------------
3531 1. A negated single character class followed by a quantifier with a minimum
3532 value of one (e.g. [^x]{1,6} ) was not compiled correctly. This could lead to
3533 program crashes, or just wrong answers. This did not apply to negated classes
3534 containing more than one character, or to minima other than one.
3537 Version 1.08 27-Mar-98
3538 ----------------------
3540 1. Add PCRE_UNGREEDY to invert the greediness of quantifiers.
3542 2. Add (?U) and (?X) to set PCRE_UNGREEDY and PCRE_EXTRA respectively. The
3543 latter must appear before anything that relies on it in the pattern.
3546 Version 1.07 16-Feb-98
3547 ----------------------
3549 1. A pattern such as /((a)*)*/ was not being diagnosed as in error (unlimited
3550 repeat of a potentially empty string).
3553 Version 1.06 23-Jan-98
3554 ----------------------
3556 1. Added Markus Oberhumer's little patches for C++.
3558 2. Literal strings longer than 255 characters were broken.
3561 Version 1.05 23-Dec-97
3562 ----------------------
3564 1. Negated character classes containing more than one character were failing if
3565 PCRE_CASELESS was set at run time.
3568 Version 1.04 19-Dec-97
3569 ----------------------
3571 1. Corrected the man page, where some "const" qualifiers had been omitted.
3573 2. Made debugging output print "{0,xxx}" instead of just "{,xxx}" to agree with
3574 input syntax.
3576 3. Fixed memory leak which occurred when a regex with back references was
3577 matched with an offsets vector that wasn't big enough. The temporary memory
3578 that is used in this case wasn't being freed if the match failed.
3580 4. Tidied pcretest to ensure it frees memory that it gets.
3582 5. Temporary memory was being obtained in the case where the passed offsets
3583 vector was exactly big enough.
3585 6. Corrected definition of offsetof() from change 5 below.
3587 7. I had screwed up change 6 below and broken the rules for the use of
3588 setjmp(). Now fixed.
3591 Version 1.03 18-Dec-97
3592 ----------------------
3594 1. A erroneous regex with a missing opening parenthesis was correctly
3595 diagnosed, but PCRE attempted to access brastack[-1], which could cause crashes
3596 on some systems.
3598 2. Replaced offsetof(real_pcre, code) by offsetof(real_pcre, code[0]) because
3599 it was reported that one broken compiler failed on the former because "code" is
3600 also an independent variable.
3602 3. The erroneous regex a[]b caused an array overrun reference.
3604 4. A regex ending with a one-character negative class (e.g. /[^k]$/) did not
3605 fail on data ending with that character. (It was going on too far, and checking
3606 the next character, typically a binary zero.) This was specific to the
3607 optimized code for single-character negative classes.
3609 5. Added a contributed patch from the TIN world which does the following:
3611 + Add an undef for memmove, in case the the system defines a macro for it.
3613 + Add a definition of offsetof(), in case there isn't one. (I don't know
3614 the reason behind this - offsetof() is part of the ANSI standard - but
3615 it does no harm).
3617 + Reduce the ifdef's in pcre.c using macro DPRINTF, thereby eliminating
3618 most of the places where whitespace preceded '#'. I have given up and
3619 allowed the remaining 2 cases to be at the margin.
3621 + Rename some variables in pcre to eliminate shadowing. This seems very
3622 pedantic, but does no harm, of course.
3624 6. Moved the call to setjmp() into its own function, to get rid of warnings
3625 from gcc -Wall, and avoided calling it at all unless PCRE_EXTRA is used.
3627 7. Constructs such as \d{8,} were compiling into the equivalent of
3628 \d{8}\d{0,65527} instead of \d{8}\d* which didn't make much difference to the
3629 outcome, but in this particular case used more store than had been allocated,
3630 which caused the bug to be discovered because it threw up an internal error.
3632 8. The debugging code in both pcre and pcretest for outputting the compiled
3633 form of a regex was going wrong in the case of back references followed by
3634 curly-bracketed repeats.
3637 Version 1.02 12-Dec-97
3638 ----------------------
3640 1. Typos in pcre.3 and comments in the source fixed.
3642 2. Applied a contributed patch to get rid of places where it used to remove
3643 'const' from variables, and fixed some signed/unsigned and uninitialized
3644 variable warnings.
3646 3. Added the "runtest" target to Makefile.
3648 4. Set default compiler flag to -O2 rather than just -O.
3651 Version 1.01 19-Nov-97
3652 ----------------------
3654 1. PCRE was failing to diagnose unlimited repeat of empty string for patterns
3655 like /([ab]*)*/, that is, for classes with more than one character in them.
3657 2. Likewise, it wasn't diagnosing patterns with "once-only" subpatterns, such
3658 as /((?>a*))*/ (a PCRE_EXTRA facility).
3661 Version 1.00 18-Nov-97
3662 ----------------------
3664 1. Added compile-time macros to support systems such as SunOS4 which don't have
3665 memmove() or strerror() but have other things that can be used instead.
3667 2. Arranged that "make clean" removes the executables.
3670 Version 0.99 27-Oct-97
3671 ----------------------
3673 1. Fixed bug in code for optimizing classes with only one character. It was
3674 initializing a 32-byte map regardless, which could cause it to run off the end
3675 of the memory it had got.
3677 2. Added, conditional on PCRE_EXTRA, the proposed (?>REGEX) construction.
3680 Version 0.98 22-Oct-97
3681 ----------------------
3683 1. Fixed bug in code for handling temporary memory usage when there are more
3684 back references than supplied space in the ovector. This could cause segfaults.
3687 Version 0.97 21-Oct-97
3688 ----------------------
3690 1. Added the \X "cut" facility, conditional on PCRE_EXTRA.
3692 2. Optimized negated single characters not to use a bit map.
3694 3. Brought error texts together as macro definitions; clarified some of them;
3695 fixed one that was wrong - it said "range out of order" when it meant "invalid
3696 escape sequence".
3698 4. Changed some char * arguments to const char *.
3700 5. Added PCRE_NOTBOL and PCRE_NOTEOL (from POSIX).
3702 6. Added the POSIX-style API wrapper in pcreposix.a and testing facilities in
3703 pcretest.
3706 Version 0.96 16-Oct-97
3707 ----------------------
3709 1. Added a simple "pgrep" utility to the distribution.
3711 2. Fixed an incompatibility with Perl: "{" is now treated as a normal character
3712 unless it appears in one of the precise forms "{ddd}", "{ddd,}", or "{ddd,ddd}"
3713 where "ddd" means "one or more decimal digits".
3715 3. Fixed serious bug. If a pattern had a back reference, but the call to
3716 pcre_exec() didn't supply a large enough ovector to record the related
3717 identifying subpattern, the match always failed. PCRE now remembers the number
3718 of the largest back reference, and gets some temporary memory in which to save
3719 the offsets during matching if necessary, in order to ensure that
3720 backreferences always work.
3722 4. Increased the compatibility with Perl in a number of ways:
3724 (a) . no longer matches \n by default; an option PCRE_DOTALL is provided
3725 to request this handling. The option can be set at compile or exec time.
3727 (b) $ matches before a terminating newline by default; an option
3728 PCRE_DOLLAR_ENDONLY is provided to override this (but not in multiline
3729 mode). The option can be set at compile or exec time.
3731 (c) The handling of \ followed by a digit other than 0 is now supposed to be
3732 the same as Perl's. If the decimal number it represents is less than 10
3733 or there aren't that many previous left capturing parentheses, an octal
3734 escape is read. Inside a character class, it's always an octal escape,
3735 even if it is a single digit.
3737 (d) An escaped but undefined alphabetic character is taken as a literal,
3738 unless PCRE_EXTRA is set. Currently this just reserves the remaining
3739 escapes.
3741 (e) {0} is now permitted. (The previous item is removed from the compiled
3742 pattern).
3744 5. Changed all the names of code files so that the basic parts are no longer
3745 than 10 characters, and abolished the teeny "globals.c" file.
3747 6. Changed the handling of character classes; they are now done with a 32-byte
3748 bit map always.
3750 7. Added the -d and /D options to pcretest to make it possible to look at the
3751 internals of compilation without having to recompile pcre.
3754 Version 0.95 23-Sep-97
3755 ----------------------
3757 1. Fixed bug in pre-pass concerning escaped "normal" characters such as \x5c or
3758 \x20 at the start of a run of normal characters. These were being treated as
3759 real characters, instead of the source characters being re-checked.
3762 Version 0.94 18-Sep-97
3763 ----------------------
3765 1. The functions are now thread-safe, with the caveat that the global variables
3766 containing pointers to malloc() and free() or alternative functions are the
3767 same for all threads.
3769 2. Get pcre_study() to generate a bitmap of initial characters for non-
3770 anchored patterns when this is possible, and use it if passed to pcre_exec().
3773 Version 0.93 15-Sep-97
3774 ----------------------
3776 1. /(b)|(:+)/ was computing an incorrect first character.
3778 2. Add pcre_study() to the API and the passing of pcre_extra to pcre_exec(),
3779 but not actually doing anything yet.
3781 3. Treat "-" characters in classes that cannot be part of ranges as literals,
3782 as Perl does (e.g. [-az] or [az-]).
3784 4. Set the anchored flag if a branch starts with .* or .*? because that tests
3785 all possible positions.
3787 5. Split up into different modules to avoid including unneeded functions in a
3788 compiled binary. However, compile and exec are still in one module. The "study"
3789 function is split off.
3791 6. The character tables are now in a separate module whose source is generated
3792 by an auxiliary program - but can then be edited by hand if required. There are
3793 now no calls to isalnum(), isspace(), isdigit(), isxdigit(), tolower() or
3794 toupper() in the code.
3796 7. Turn the malloc/free funtions variables into pcre_malloc and pcre_free and
3797 make them global. Abolish the function for setting them, as the caller can now
3798 set them directly.
3801 Version 0.92 11-Sep-97
3802 ----------------------
3804 1. A repeat with a fixed maximum and a minimum of 1 for an ordinary character
3805 (e.g. /a{1,3}/) was broken (I mis-optimized it).
3807 2. Caseless matching was not working in character classes if the characters in
3808 the pattern were in upper case.
3810 3. Make ranges like [W-c] work in the same way as Perl for caseless matching.
3812 4. Make PCRE_ANCHORED public and accept as a compile option.
3814 5. Add an options word to pcre_exec() and accept PCRE_ANCHORED and
3815 PCRE_CASELESS at run time. Add escapes \A and \I to pcretest to cause it to
3816 pass them.
3818 6. Give an error if bad option bits passed at compile or run time.
3820 7. Add PCRE_MULTILINE at compile and exec time, and (?m) as well. Add \M to
3821 pcretest to cause it to pass that flag.
3823 8. Add pcre_info(), to get the number of identifying subpatterns, the stored
3824 options, and the first character, if set.
3826 9. Recognize C+ or C{n,m} where n >= 1 as providing a fixed starting character.
3829 Version 0.91 10-Sep-97
3830 ----------------------
3832 1. PCRE was failing to diagnose unlimited repeats of subpatterns that could
3833 match the empty string as in /(a*)*/. It was looping and ultimately crashing.
3835 2. PCRE was looping on encountering an indefinitely repeated back reference to
3836 a subpattern that had matched an empty string, e.g. /(a|)\1*/. It now does what
3837 Perl does - treats the match as successful.
3839 ****


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

  ViewVC Help
Powered by ViewVC 1.1.5