ViewVC logotype

Contents of /code/trunk/ChangeLog

Parent Directory Parent Directory | Revision Log Revision Log

Revision 506 - (show annotations)
Tue Mar 9 17:01:40 2010 UTC (5 years, 7 months ago) by ph10
File size: 184801 byte(s)
Error occurred while calculating annotation data.
Fix "srtoq" typo in configure.ac.
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.
42 10. \K gave a compile-time error if it appeared in a lookbehind assersion.
44 11. \K was not working if it appeared in an atomic group or in a group that
45 was called as a "subroutine", or in an assertion. Perl 5.11 documents that
46 \K is "not well defined" if used in an assertion. PCRE now accepts it if
47 the assertion is positive, but not if it is negative.
49 12. Change 11 fortuitously reduced the size of the stack frame used in the
50 "match()" function of pcre_exec.c by one pointer. Forthcoming
51 implementation of support for (*MARK) will need an extra pointer on the
52 stack; I have reserved it now, so that the stack frame size does not
53 decrease.
55 13. A pattern such as (?P<L1>(?P<L2>0)|(?P>L2)(?P>L1)) in which the only other
56 item in branch that calls a recursion is a subroutine call - as in the
57 second branch in the above example - was incorrectly given the compile-
58 time error "recursive call could loop indefinitely" because pcre_compile()
59 was not correctly checking the subroutine for matching a non-empty string.
61 14. The checks for overrunning compiling workspace could trigger after an
62 overrun had occurred. This is a "should never occur" error, but it can be
63 triggered by pathological patterns such as hundreds of nested parentheses.
64 The checks now trigger 100 bytes before the end of the workspace.
66 15. Fix typo in configure.ac: "srtoq" should be "strtoq".
69 Version 8.01 19-Jan-2010
70 ------------------------
72 1. If a pattern contained a conditional subpattern with only one branch (in
73 particular, this includes all (*DEFINE) patterns), a call to pcre_study()
74 computed the wrong minimum data length (which is of course zero for such
75 subpatterns). This could cause incorrect "no match" results.
77 2. For patterns such as (?i)a(?-i)b|c where an option setting at the start of
78 the pattern is reset in the first branch, pcre_compile() failed with
79 "internal error: code overflow at offset...". This happened only when
80 the reset was to the original external option setting. (An optimization
81 abstracts leading options settings into an external setting, which was the
82 cause of this.)
84 3. A pattern such as ^(?!a(*SKIP)b) where a negative assertion contained one
85 of the verbs SKIP, PRUNE, or COMMIT, did not work correctly. When the
86 assertion pattern did not match (meaning that the assertion was true), it
87 was incorrectly treated as false if the SKIP had been reached during the
88 matching. This also applied to assertions used as conditions.
90 4. If an item that is not supported by pcre_dfa_exec() was encountered in an
91 assertion subpattern, including such a pattern used as a condition,
92 unpredictable results occurred, instead of the error return
95 5. The C++ GlobalReplace function was not working like Perl for the special
96 situation when an empty string is matched. It now does the fancy magic
97 stuff that is necessary.
99 6. In pcre_internal.h, obsolete includes to setjmp.h and stdarg.h have been
100 removed. (These were left over from very, very early versions of PCRE.)
102 7. Some cosmetic changes to the code to make life easier when compiling it
103 as part of something else:
105 (a) Change DEBUG to PCRE_DEBUG.
107 (b) In pcre_compile(), rename the member of the "branch_chain" structure
108 called "current" as "current_branch", to prevent a collision with the
109 Linux macro when compiled as a kernel module.
111 (c) In pcre_study(), rename the function set_bit() as set_table_bit(), to
112 prevent a collision with the Linux macro when compiled as a kernel
113 module.
115 8. In pcre_compile() there are some checks for integer overflows that used to
116 cast potentially large values to (double). This has been changed to that
117 when building, a check for int64_t is made, and if it is found, it is used
118 instead, thus avoiding the use of floating point arithmetic. (There is no
119 other use of FP in PCRE.) If int64_t is not found, the fallback is to
120 double.
122 9. Added two casts to avoid signed/unsigned warnings from VS Studio Express
123 2005 (difference between two addresses compared to an unsigned value).
125 10. Change the standard AC_CHECK_LIB test for libbz2 in configure.ac to a
126 custom one, because of the following reported problem in Windows:
128 - libbz2 uses the Pascal calling convention (WINAPI) for the functions
129 under Win32.
130 - The standard autoconf AC_CHECK_LIB fails to include "bzlib.h",
131 therefore missing the function definition.
132 - The compiler thus generates a "C" signature for the test function.
133 - The linker fails to find the "C" function.
134 - PCRE fails to configure if asked to do so against libbz2.
136 11. When running libtoolize from libtool-2.2.6b as part of autogen.sh, these
137 messages were output:
139 Consider adding `AC_CONFIG_MACRO_DIR([m4])' to configure.ac and
140 rerunning libtoolize, to keep the correct libtool macros in-tree.
141 Consider adding `-I m4' to ACLOCAL_AMFLAGS in Makefile.am.
143 I have done both of these things.
145 12. Although pcre_dfa_exec() does not use nearly as much stack as pcre_exec()
146 most of the time, it *can* run out if it is given a pattern that contains a
147 runaway infinite recursion. I updated the discussion in the pcrestack man
148 page.
150 13. Now that we have gone to the x.xx style of version numbers, the minor
151 version may start with zero. Using 08 or 09 is a bad idea because users
152 might check the value of PCRE_MINOR in their code, and 08 or 09 may be
153 interpreted as invalid octal numbers. I've updated the previous comment in
154 configure.ac, and also added a check that gives an error if 08 or 09 are
155 used.
157 14. Change 8.00/11 was not quite complete: code had been accidentally omitted,
158 causing partial matching to fail when the end of the subject matched \W
159 in a UTF-8 pattern where \W was quantified with a minimum of 3.
161 15. There were some discrepancies between the declarations in pcre_internal.h
162 of _pcre_is_newline(), _pcre_was_newline(), and _pcre_valid_utf8() and
163 their definitions. The declarations used "const uschar *" and the
164 definitions used USPTR. Even though USPTR is normally defined as "const
165 unsigned char *" (and uschar is typedeffed as "unsigned char"), it was
166 reported that: "This difference in casting confuses some C++ compilers, for
167 example, SunCC recognizes above declarations as different functions and
168 generates broken code for hbpcre." I have changed the declarations to use
169 USPTR.
171 16. GNU libtool is named differently on some systems. The autogen.sh script now
172 tries several variants such as glibtoolize (MacOSX) and libtoolize1x
173 (FreeBSD).
175 17. Applied Craig's patch that fixes an HP aCC compile error in pcre 8.00
176 (strtoXX undefined when compiling pcrecpp.cc). The patch contains this
177 comment: "Figure out how to create a longlong from a string: strtoll and
178 equivalent. It's not enough to call AC_CHECK_FUNCS: hpux has a strtoll, for
179 instance, but it only takes 2 args instead of 3!"
181 18. A subtle bug concerned with back references has been fixed by a change of
182 specification, with a corresponding code fix. A pattern such as
183 ^(xa|=?\1a)+$ which contains a back reference inside the group to which it
184 refers, was giving matches when it shouldn't. For example, xa=xaaa would
185 match that pattern. Interestingly, Perl (at least up to 5.11.3) has the
186 same bug. Such groups have to be quantified to be useful, or contained
187 inside another quantified group. (If there's no repetition, the reference
188 can never match.) The problem arises because, having left the group and
189 moved on to the rest of the pattern, a later failure that backtracks into
190 the group uses the captured value from the final iteration of the group
191 rather than the correct earlier one. I have fixed this in PCRE by forcing
192 any group that contains a reference to itself to be an atomic group; that
193 is, there cannot be any backtracking into it once it has completed. This is
194 similar to recursive and subroutine calls.
197 Version 8.00 19-Oct-09
198 ----------------------
200 1. The table for translating pcre_compile() error codes into POSIX error codes
201 was out-of-date, and there was no check on the pcre_compile() error code
202 being within the table. This could lead to an OK return being given in
203 error.
205 2. Changed the call to open a subject file in pcregrep from fopen(pathname,
206 "r") to fopen(pathname, "rb"), which fixed a problem with some of the tests
207 in a Windows environment.
209 3. The pcregrep --count option prints the count for each file even when it is
210 zero, as does GNU grep. However, pcregrep was also printing all files when
211 --files-with-matches was added. Now, when both options are given, it prints
212 counts only for those files that have at least one match. (GNU grep just
213 prints the file name in this circumstance, but including the count seems
214 more useful - otherwise, why use --count?) Also ensured that the
215 combination -clh just lists non-zero counts, with no names.
217 4. The long form of the pcregrep -F option was incorrectly implemented as
218 --fixed_strings instead of --fixed-strings. This is an incompatible change,
219 but it seems right to fix it, and I didn't think it was worth preserving
220 the old behaviour.
222 5. The command line items --regex=pattern and --regexp=pattern were not
223 recognized by pcregrep, which required --regex pattern or --regexp pattern
224 (with a space rather than an '='). The man page documented the '=' forms,
225 which are compatible with GNU grep; these now work.
227 6. No libpcreposix.pc file was created for pkg-config; there was just
228 libpcre.pc and libpcrecpp.pc. The omission has been rectified.
230 7. Added #ifndef SUPPORT_UCP into the pcre_ucd.c module, to reduce its size
231 when UCP support is not needed, by modifying the Python script that
232 generates it from Unicode data files. This should not matter if the module
233 is correctly used as a library, but I received one complaint about 50K of
234 unwanted data. My guess is that the person linked everything into his
235 program rather than using a library. Anyway, it does no harm.
237 8. A pattern such as /\x{123}{2,2}+/8 was incorrectly compiled; the trigger
238 was a minimum greater than 1 for a wide character in a possessive
239 repetition. The same bug could also affect patterns like /(\x{ff}{0,2})*/8
240 which had an unlimited repeat of a nested, fixed maximum repeat of a wide
241 character. Chaos in the form of incorrect output or a compiling loop could
242 result.
244 9. The restrictions on what a pattern can contain when partial matching is
245 requested for pcre_exec() have been removed. All patterns can now be
246 partially matched by this function. In addition, if there are at least two
247 slots in the offset vector, the offset of the earliest inspected character
248 for the match and the offset of the end of the subject are set in them when
249 PCRE_ERROR_PARTIAL is returned.
251 10. Partial matching has been split into two forms: PCRE_PARTIAL_SOFT, which is
252 synonymous with PCRE_PARTIAL, for backwards compatibility, and
253 PCRE_PARTIAL_HARD, which causes a partial match to supersede a full match,
254 and may be more useful for multi-segment matching.
256 11. Partial matching with pcre_exec() is now more intuitive. A partial match
257 used to be given if ever the end of the subject was reached; now it is
258 given only if matching could not proceed because another character was
259 needed. This makes a difference in some odd cases such as Z(*FAIL) with the
260 string "Z", which now yields "no match" instead of "partial match". In the
261 case of pcre_dfa_exec(), "no match" is given if every matching path for the
262 final character ended with (*FAIL).
264 12. Restarting a match using pcre_dfa_exec() after a partial match did not work
265 if the pattern had a "must contain" character that was already found in the
266 earlier partial match, unless partial matching was again requested. For
267 example, with the pattern /dog.(body)?/, the "must contain" character is
268 "g". If the first part-match was for the string "dog", restarting with
269 "sbody" failed. This bug has been fixed.
271 13. The string returned by pcre_dfa_exec() after a partial match has been
272 changed so that it starts at the first inspected character rather than the
273 first character of the match. This makes a difference only if the pattern
274 starts with a lookbehind assertion or \b or \B (\K is not supported by
275 pcre_dfa_exec()). It's an incompatible change, but it makes the two
276 matching functions compatible, and I think it's the right thing to do.
278 14. Added a pcredemo man page, created automatically from the pcredemo.c file,
279 so that the demonstration program is easily available in environments where
280 PCRE has not been installed from source.
282 15. Arranged to add -DPCRE_STATIC to cflags in libpcre.pc, libpcreposix.cp,
283 libpcrecpp.pc and pcre-config when PCRE is not compiled as a shared
284 library.
286 16. Added REG_UNGREEDY to the pcreposix interface, at the request of a user.
287 It maps to PCRE_UNGREEDY. It is not, of course, POSIX-compatible, but it
288 is not the first non-POSIX option to be added. Clearly some people find
289 these options useful.
291 17. If a caller to the POSIX matching function regexec() passes a non-zero
292 value for nmatch with a NULL value for pmatch, the value of
293 nmatch is forced to zero.
295 18. RunGrepTest did not have a test for the availability of the -u option of
296 the diff command, as RunTest does. It now checks in the same way as
297 RunTest, and also checks for the -b option.
299 19. If an odd number of negated classes containing just a single character
300 interposed, within parentheses, between a forward reference to a named
301 subpattern and the definition of the subpattern, compilation crashed with
302 an internal error, complaining that it could not find the referenced
303 subpattern. An example of a crashing pattern is /(?&A)(([^m])(?<A>))/.
304 [The bug was that it was starting one character too far in when skipping
305 over the character class, thus treating the ] as data rather than
306 terminating the class. This meant it could skip too much.]
308 20. Added PCRE_NOTEMPTY_ATSTART in order to be able to correctly implement the
309 /g option in pcretest when the pattern contains \K, which makes it possible
310 to have an empty string match not at the start, even when the pattern is
311 anchored. Updated pcretest and pcredemo to use this option.
313 21. If the maximum number of capturing subpatterns in a recursion was greater
314 than the maximum at the outer level, the higher number was returned, but
315 with unset values at the outer level. The correct (outer level) value is
316 now given.
318 22. If (*ACCEPT) appeared inside capturing parentheses, previous releases of
319 PCRE did not set those parentheses (unlike Perl). I have now found a way to
320 make it do so. The string so far is captured, making this feature
321 compatible with Perl.
323 23. The tests have been re-organized, adding tests 11 and 12, to make it
324 possible to check the Perl 5.10 features against Perl 5.10.
326 24. Perl 5.10 allows subroutine calls in lookbehinds, as long as the subroutine
327 pattern matches a fixed length string. PCRE did not allow this; now it
328 does. Neither allows recursion.
330 25. I finally figured out how to implement a request to provide the minimum
331 length of subject string that was needed in order to match a given pattern.
332 (It was back references and recursion that I had previously got hung up
333 on.) This code has now been added to pcre_study(); it finds a lower bound
334 to the length of subject needed. It is not necessarily the greatest lower
335 bound, but using it to avoid searching strings that are too short does give
336 some useful speed-ups. The value is available to calling programs via
337 pcre_fullinfo().
339 26. While implementing 25, I discovered to my embarrassment that pcretest had
340 not been passing the result of pcre_study() to pcre_dfa_exec(), so the
341 study optimizations had never been tested with that matching function.
342 Oops. What is worse, even when it was passed study data, there was a bug in
343 pcre_dfa_exec() that meant it never actually used it. Double oops. There
344 were also very few tests of studied patterns with pcre_dfa_exec().
346 27. If (?| is used to create subpatterns with duplicate numbers, they are now
347 allowed to have the same name, even if PCRE_DUPNAMES is not set. However,
348 on the other side of the coin, they are no longer allowed to have different
349 names, because these cannot be distinguished in PCRE, and this has caused
350 confusion. (This is a difference from Perl.)
352 28. When duplicate subpattern names are present (necessarily with different
353 numbers, as required by 27 above), and a test is made by name in a
354 conditional pattern, either for a subpattern having been matched, or for
355 recursion in such a pattern, all the associated numbered subpatterns are
356 tested, and the overall condition is true if the condition is true for any
357 one of them. This is the way Perl works, and is also more like the way
358 testing by number works.
361 Version 7.9 11-Apr-09
362 ---------------------
364 1. When building with support for bzlib/zlib (pcregrep) and/or readline
365 (pcretest), all targets were linked against these libraries. This included
366 libpcre, libpcreposix, and libpcrecpp, even though they do not use these
367 libraries. This caused unwanted dependencies to be created. This problem
368 has been fixed, and now only pcregrep is linked with bzlib/zlib and only
369 pcretest is linked with readline.
371 2. The "typedef int BOOL" in pcre_internal.h that was included inside the
372 "#ifndef FALSE" condition by an earlier change (probably 7.8/18) has been
373 moved outside it again, because FALSE and TRUE are already defined in AIX,
374 but BOOL is not.
376 3. The pcre_config() function was treating the PCRE_MATCH_LIMIT and
377 PCRE_MATCH_LIMIT_RECURSION values as ints, when they should be long ints.
379 4. The pcregrep documentation said spaces were inserted as well as colons (or
380 hyphens) following file names and line numbers when outputting matching
381 lines. This is not true; no spaces are inserted. I have also clarified the
382 wording for the --colour (or --color) option.
384 5. In pcregrep, when --colour was used with -o, the list of matching strings
385 was not coloured; this is different to GNU grep, so I have changed it to be
386 the same.
388 6. When --colo(u)r was used in pcregrep, only the first matching substring in
389 each matching line was coloured. Now it goes on to look for further matches
390 of any of the test patterns, which is the same behaviour as GNU grep.
392 7. A pattern that could match an empty string could cause pcregrep to loop; it
393 doesn't make sense to accept an empty string match in pcregrep, so I have
394 locked it out (using PCRE's PCRE_NOTEMPTY option). By experiment, this
395 seems to be how GNU grep behaves.
397 8. The pattern (?(?=.*b)b|^) was incorrectly compiled as "match must be at
398 start or after a newline", because the conditional assertion was not being
399 correctly handled. The rule now is that both the assertion and what follows
400 in the first alternative must satisfy the test.
402 9. If auto-callout was enabled in a pattern with a conditional group whose
403 condition was an assertion, PCRE could crash during matching, both with
404 pcre_exec() and pcre_dfa_exec().
406 10. The PCRE_DOLLAR_ENDONLY option was not working when pcre_dfa_exec() was
407 used for matching.
409 11. Unicode property support in character classes was not working for
410 characters (bytes) greater than 127 when not in UTF-8 mode.
412 12. Added the -M command line option to pcretest.
414 14. Added the non-standard REG_NOTEMPTY option to the POSIX interface.
416 15. Added the PCRE_NO_START_OPTIMIZE match-time option.
418 16. Added comments and documentation about mis-use of no_arg in the C++
419 wrapper.
421 17. Implemented support for UTF-8 encoding in EBCDIC environments, a patch
422 from Martin Jerabek that uses macro names for all relevant character and
423 string constants.
425 18. Added to pcre_internal.h two configuration checks: (a) If both EBCDIC and
426 SUPPORT_UTF8 are set, give an error; (b) If SUPPORT_UCP is set without
427 SUPPORT_UTF8, define SUPPORT_UTF8. The "configure" script handles both of
428 these, but not everybody uses configure.
430 19. A conditional group that had only one branch was not being correctly
431 recognized as an item that could match an empty string. This meant that an
432 enclosing group might also not be so recognized, causing infinite looping
433 (and probably a segfault) for patterns such as ^"((?(?=[a])[^"])|b)*"$
434 with the subject "ab", where knowledge that the repeated group can match
435 nothing is needed in order to break the loop.
437 20. If a pattern that was compiled with callouts was matched using pcre_dfa_
438 exec(), but without supplying a callout function, matching went wrong.
440 21. If PCRE_ERROR_MATCHLIMIT occurred during a recursion, there was a memory
441 leak if the size of the offset vector was greater than 30. When the vector
442 is smaller, the saved offsets during recursion go onto a local stack
443 vector, but for larger vectors malloc() is used. It was failing to free
444 when the recursion yielded PCRE_ERROR_MATCH_LIMIT (or any other "abnormal"
445 error, in fact).
447 22. There was a missing #ifdef SUPPORT_UTF8 round one of the variables in the
448 heapframe that is used only when UTF-8 support is enabled. This caused no
449 problem, but was untidy.
451 23. Steven Van Ingelgem's patch to CMakeLists.txt to change the name
452 CMAKE_BINARY_DIR to PROJECT_BINARY_DIR so that it works when PCRE is
453 included within another project.
455 24. Steven Van Ingelgem's patches to add more options to the CMake support,
456 slightly modified by me:
458 (a) PCRE_BUILD_TESTS can be set OFF not to build the tests, including
459 not building pcregrep.
461 (b) PCRE_BUILD_PCREGREP can be see OFF not to build pcregrep, but only
462 if PCRE_BUILD_TESTS is also set OFF, because the tests use pcregrep.
464 25. Forward references, both numeric and by name, in patterns that made use of
465 duplicate group numbers, could behave incorrectly or give incorrect errors,
466 because when scanning forward to find the reference group, PCRE was not
467 taking into account the duplicate group numbers. A pattern such as
468 ^X(?3)(a)(?|(b)|(q))(Y) is an example.
470 26. Changed a few more instances of "const unsigned char *" to USPTR, making
471 the feature of a custom pointer more persuasive (as requested by a user).
473 27. Wrapped the definitions of fileno and isatty for Windows, which appear in
474 pcretest.c, inside #ifndefs, because it seems they are sometimes already
475 pre-defined.
477 28. Added support for (*UTF8) at the start of a pattern.
479 29. Arrange for flags added by the "release type" setting in CMake to be shown
480 in the configuration summary.
483 Version 7.8 05-Sep-08
484 ---------------------
486 1. Replaced UCP searching code with optimized version as implemented for Ad
487 Muncher (http://www.admuncher.com/) by Peter Kankowski. This uses a two-
488 stage table and inline lookup instead of a function, giving speed ups of 2
489 to 5 times on some simple patterns that I tested. Permission was given to
490 distribute the MultiStage2.py script that generates the tables (it's not in
491 the tarball, but is in the Subversion repository).
493 2. Updated the Unicode datatables to Unicode 5.1.0. This adds yet more
494 scripts.
496 3. Change 12 for 7.7 introduced a bug in pcre_study() when a pattern contained
497 a group with a zero qualifier. The result of the study could be incorrect,
498 or the function might crash, depending on the pattern.
500 4. Caseless matching was not working for non-ASCII characters in back
501 references. For example, /(\x{de})\1/8i was not matching \x{de}\x{fe}.
502 It now works when Unicode Property Support is available.
504 5. In pcretest, an escape such as \x{de} in the data was always generating
505 a UTF-8 string, even in non-UTF-8 mode. Now it generates a single byte in
506 non-UTF-8 mode. If the value is greater than 255, it gives a warning about
507 truncation.
509 6. Minor bugfix in pcrecpp.cc (change "" == ... to NULL == ...).
511 7. Added two (int) casts to pcregrep when printing the difference of two
512 pointers, in case they are 64-bit values.
514 8. Added comments about Mac OS X stack usage to the pcrestack man page and to
515 test 2 if it fails.
517 9. Added PCRE_CALL_CONVENTION just before the names of all exported functions,
518 and a #define of that name to empty if it is not externally set. This is to
519 allow users of MSVC to set it if necessary.
521 10. The PCRE_EXP_DEFN macro which precedes exported functions was missing from
522 the convenience functions in the pcre_get.c source file.
524 11. An option change at the start of a pattern that had top-level alternatives
525 could cause overwriting and/or a crash. This command provoked a crash in
526 some environments:
528 printf "/(?i)[\xc3\xa9\xc3\xbd]|[\xc3\xa9\xc3\xbdA]/8\n" | pcretest
530 This potential security problem was recorded as CVE-2008-2371.
532 12. For a pattern where the match had to start at the beginning or immediately
533 after a newline (e.g /.*anything/ without the DOTALL flag), pcre_exec() and
534 pcre_dfa_exec() could read past the end of the passed subject if there was
535 no match. To help with detecting such bugs (e.g. with valgrind), I modified
536 pcretest so that it places the subject at the end of its malloc-ed buffer.
538 13. The change to pcretest in 12 above threw up a couple more cases when pcre_
539 exec() might read past the end of the data buffer in UTF-8 mode.
541 14. A similar bug to 7.3/2 existed when the PCRE_FIRSTLINE option was set and
542 the data contained the byte 0x85 as part of a UTF-8 character within its
543 first line. This applied both to normal and DFA matching.
545 15. Lazy qualifiers were not working in some cases in UTF-8 mode. For example,
546 /^[^d]*?$/8 failed to match "abc".
548 16. Added a missing copyright notice to pcrecpp_internal.h.
550 17. Make it more clear in the documentation that values returned from
551 pcre_exec() in ovector are byte offsets, not character counts.
553 18. Tidied a few places to stop certain compilers from issuing warnings.
555 19. Updated the Virtual Pascal + BCC files to compile the latest v7.7, as
556 supplied by Stefan Weber. I made a further small update for 7.8 because
557 there is a change of source arrangements: the pcre_searchfuncs.c module is
558 replaced by pcre_ucd.c.
561 Version 7.7 07-May-08
562 ---------------------
564 1. Applied Craig's patch to sort out a long long problem: "If we can't convert
565 a string to a long long, pretend we don't even have a long long." This is
566 done by checking for the strtoq, strtoll, and _strtoi64 functions.
568 2. Applied Craig's patch to pcrecpp.cc to restore ABI compatibility with
569 pre-7.6 versions, which defined a global no_arg variable instead of putting
570 it in the RE class. (See also #8 below.)
572 3. Remove a line of dead code, identified by coverity and reported by Nuno
573 Lopes.
575 4. Fixed two related pcregrep bugs involving -r with --include or --exclude:
577 (1) The include/exclude patterns were being applied to the whole pathnames
578 of files, instead of just to the final components.
580 (2) If there was more than one level of directory, the subdirectories were
581 skipped unless they satisfied the include/exclude conditions. This is
582 inconsistent with GNU grep (and could even be seen as contrary to the
583 pcregrep specification - which I improved to make it absolutely clear).
584 The action now is always to scan all levels of directory, and just
585 apply the include/exclude patterns to regular files.
587 5. Added the --include_dir and --exclude_dir patterns to pcregrep, and used
588 --exclude_dir in the tests to avoid scanning .svn directories.
590 6. Applied Craig's patch to the QuoteMeta function so that it escapes the
591 NUL character as backslash + 0 rather than backslash + NUL, because PCRE
592 doesn't support NULs in patterns.
594 7. Added some missing "const"s to declarations of static tables in
595 pcre_compile.c and pcre_dfa_exec.c.
597 8. Applied Craig's patch to pcrecpp.cc to fix a problem in OS X that was
598 caused by fix #2 above. (Subsequently also a second patch to fix the
599 first patch. And a third patch - this was a messy problem.)
601 9. Applied Craig's patch to remove the use of push_back().
603 10. Applied Alan Lehotsky's patch to add REG_STARTEND support to the POSIX
604 matching function regexec().
606 11. Added support for the Oniguruma syntax \g<name>, \g<n>, \g'name', \g'n',
607 which, however, unlike Perl's \g{...}, are subroutine calls, not back
608 references. PCRE supports relative numbers with this syntax (I don't think
609 Oniguruma does).
611 12. Previously, a group with a zero repeat such as (...){0} was completely
612 omitted from the compiled regex. However, this means that if the group
613 was called as a subroutine from elsewhere in the pattern, things went wrong
614 (an internal error was given). Such groups are now left in the compiled
615 pattern, with a new opcode that causes them to be skipped at execution
616 time.
618 13. Added the PCRE_JAVASCRIPT_COMPAT option. This makes the following changes
619 to the way PCRE behaves:
621 (a) A lone ] character is dis-allowed (Perl treats it as data).
623 (b) A back reference to an unmatched subpattern matches an empty string
624 (Perl fails the current match path).
626 (c) A data ] in a character class must be notated as \] because if the
627 first data character in a class is ], it defines an empty class. (In
628 Perl it is not possible to have an empty class.) The empty class []
629 never matches; it forces failure and is equivalent to (*FAIL) or (?!).
630 The negative empty class [^] matches any one character, independently
631 of the DOTALL setting.
633 14. A pattern such as /(?2)[]a()b](abc)/ which had a forward reference to a
634 non-existent subpattern following a character class starting with ']' and
635 containing () gave an internal compiling error instead of "reference to
636 non-existent subpattern". Fortunately, when the pattern did exist, the
637 compiled code was correct. (When scanning forwards to check for the
638 existencd of the subpattern, it was treating the data ']' as terminating
639 the class, so got the count wrong. When actually compiling, the reference
640 was subsequently set up correctly.)
642 15. The "always fail" assertion (?!) is optimzed to (*FAIL) by pcre_compile;
643 it was being rejected as not supported by pcre_dfa_exec(), even though
644 other assertions are supported. I have made pcre_dfa_exec() support
645 (*FAIL).
647 16. The implementation of 13c above involved the invention of a new opcode,
648 OP_ALLANY, which is like OP_ANY but doesn't check the /s flag. Since /s
649 cannot be changed at match time, I realized I could make a small
650 improvement to matching performance by compiling OP_ALLANY instead of
651 OP_ANY for "." when DOTALL was set, and then removing the runtime tests
652 on the OP_ANY path.
654 17. Compiling pcretest on Windows with readline support failed without the
655 following two fixes: (1) Make the unistd.h include conditional on
656 HAVE_UNISTD_H; (2) #define isatty and fileno as _isatty and _fileno.
658 18. Changed CMakeLists.txt and cmake/FindReadline.cmake to arrange for the
659 ncurses library to be included for pcretest when ReadLine support is
660 requested, but also to allow for it to be overridden. This patch came from
661 Daniel Bergström.
663 19. There was a typo in the file ucpinternal.h where f0_rangeflag was defined
664 as 0x00f00000 instead of 0x00800000. Luckily, this would not have caused
665 any errors with the current Unicode tables. Thanks to Peter Kankowski for
666 spotting this.
669 Version 7.6 28-Jan-08
670 ---------------------
672 1. A character class containing a very large number of characters with
673 codepoints greater than 255 (in UTF-8 mode, of course) caused a buffer
674 overflow.
676 2. Patch to cut out the "long long" test in pcrecpp_unittest when
677 HAVE_LONG_LONG is not defined.
679 3. Applied Christian Ehrlicher's patch to update the CMake build files to
680 bring them up to date and include new features. This patch includes:
682 - Fixed PH's badly added libz and libbz2 support.
683 - Fixed a problem with static linking.
684 - Added pcredemo. [But later removed - see 7 below.]
685 - Fixed dftables problem and added an option.
686 - Added a number of HAVE_XXX tests, including HAVE_WINDOWS_H and
688 - Added readline support for pcretest.
689 - Added an listing of the option settings after cmake has run.
691 4. A user submitted a patch to Makefile that makes it easy to create
692 "pcre.dll" under mingw when using Configure/Make. I added stuff to
693 Makefile.am that cause it to include this special target, without
694 affecting anything else. Note that the same mingw target plus all
695 the other distribution libraries and programs are now supported
696 when configuring with CMake (see 6 below) instead of with
697 Configure/Make.
699 5. Applied Craig's patch that moves no_arg into the RE class in the C++ code.
700 This is an attempt to solve the reported problem "pcrecpp::no_arg is not
701 exported in the Windows port". It has not yet been confirmed that the patch
702 solves the problem, but it does no harm.
704 6. Applied Sheri's patch to CMakeLists.txt to add NON_STANDARD_LIB_PREFIX and
705 NON_STANDARD_LIB_SUFFIX for dll names built with mingw when configured
706 with CMake, and also correct the comment about stack recursion.
708 7. Remove the automatic building of pcredemo from the ./configure system and
709 from CMakeLists.txt. The whole idea of pcredemo.c is that it is an example
710 of a program that users should build themselves after PCRE is installed, so
711 building it automatically is not really right. What is more, it gave
712 trouble in some build environments.
714 8. Further tidies to CMakeLists.txt from Sheri and Christian.
717 Version 7.5 10-Jan-08
718 ---------------------
720 1. Applied a patch from Craig: "This patch makes it possible to 'ignore'
721 values in parens when parsing an RE using the C++ wrapper."
723 2. Negative specials like \S did not work in character classes in UTF-8 mode.
724 Characters greater than 255 were excluded from the class instead of being
725 included.
727 3. The same bug as (2) above applied to negated POSIX classes such as
728 [:^space:].
730 4. PCRECPP_STATIC was referenced in pcrecpp_internal.h, but nowhere was it
731 defined or documented. It seems to have been a typo for PCRE_STATIC, so
732 I have changed it.
734 5. The construct (?&) was not diagnosed as a syntax error (it referenced the
735 first named subpattern) and a construct such as (?&a) would reference the
736 first named subpattern whose name started with "a" (in other words, the
737 length check was missing). Both these problems are fixed. "Subpattern name
738 expected" is now given for (?&) (a zero-length name), and this patch also
739 makes it give the same error for \k'' (previously it complained that that
740 was a reference to a non-existent subpattern).
742 6. The erroneous patterns (?+-a) and (?-+a) give different error messages;
743 this is right because (?- can be followed by option settings as well as by
744 digits. I have, however, made the messages clearer.
746 7. Patterns such as (?(1)a|b) (a pattern that contains fewer subpatterns
747 than the number used in the conditional) now cause a compile-time error.
748 This is actually not compatible with Perl, which accepts such patterns, but
749 treats the conditional as always being FALSE (as PCRE used to), but it
750 seems to me that giving a diagnostic is better.
752 8. Change "alphameric" to the more common word "alphanumeric" in comments
753 and messages.
755 9. Fix two occurrences of "backslash" in comments that should have been
756 "backspace".
758 10. Remove two redundant lines of code that can never be obeyed (their function
759 was moved elsewhere).
761 11. The program that makes PCRE's Unicode character property table had a bug
762 which caused it to generate incorrect table entries for sequences of
763 characters that have the same character type, but are in different scripts.
764 It amalgamated them into a single range, with the script of the first of
765 them. In other words, some characters were in the wrong script. There were
766 thirteen such cases, affecting characters in the following ranges:
768 U+002b0 - U+002c1
769 U+0060c - U+0060d
770 U+0061e - U+00612
771 U+0064b - U+0065e
772 U+0074d - U+0076d
773 U+01800 - U+01805
774 U+01d00 - U+01d77
775 U+01d9b - U+01dbf
776 U+0200b - U+0200f
777 U+030fc - U+030fe
778 U+03260 - U+0327f
779 U+0fb46 - U+0fbb1
780 U+10450 - U+1049d
782 12. The -o option (show only the matching part of a line) for pcregrep was not
783 compatible with GNU grep in that, if there was more than one match in a
784 line, it showed only the first of them. It now behaves in the same way as
785 GNU grep.
787 13. If the -o and -v options were combined for pcregrep, it printed a blank
788 line for every non-matching line. GNU grep prints nothing, and pcregrep now
789 does the same. The return code can be used to tell if there were any
790 non-matching lines.
792 14. Added --file-offsets and --line-offsets to pcregrep.
794 15. The pattern (?=something)(?R) was not being diagnosed as a potentially
795 infinitely looping recursion. The bug was that positive lookaheads were not
796 being skipped when checking for a possible empty match (negative lookaheads
797 and both kinds of lookbehind were skipped).
799 16. Fixed two typos in the Windows-only code in pcregrep.c, and moved the
800 inclusion of <windows.h> to before rather than after the definition of
801 INVALID_FILE_ATTRIBUTES (patch from David Byron).
803 17. Specifying a possessive quantifier with a specific limit for a Unicode
804 character property caused pcre_compile() to compile bad code, which led at
805 runtime to PCRE_ERROR_INTERNAL (-14). Examples of patterns that caused this
806 are: /\p{Zl}{2,3}+/8 and /\p{Cc}{2}+/8. It was the possessive "+" that
807 caused the error; without that there was no problem.
809 18. Added --enable-pcregrep-libz and --enable-pcregrep-libbz2.
811 19. Added --enable-pcretest-libreadline.
813 20. In pcrecpp.cc, the variable 'count' was incremented twice in
814 RE::GlobalReplace(). As a result, the number of replacements returned was
815 double what it should be. I removed one of the increments, but Craig sent a
816 later patch that removed the other one (the right fix) and added unit tests
817 that check the return values (which was not done before).
819 21. Several CMake things:
821 (1) Arranged that, when cmake is used on Unix, the libraries end up with
822 the names libpcre and libpcreposix, not just pcre and pcreposix.
824 (2) The above change means that pcretest and pcregrep are now correctly
825 linked with the newly-built libraries, not previously installed ones.
829 22. In UTF-8 mode, with newline set to "any", a pattern such as .*a.*=.b.*
830 crashed when matching a string such as a\x{2029}b (note that \x{2029} is a
831 UTF-8 newline character). The key issue is that the pattern starts .*;
832 this means that the match must be either at the beginning, or after a
833 newline. The bug was in the code for advancing after a failed match and
834 checking that the new position followed a newline. It was not taking
835 account of UTF-8 characters correctly.
837 23. PCRE was behaving differently from Perl in the way it recognized POSIX
838 character classes. PCRE was not treating the sequence [:...:] as a
839 character class unless the ... were all letters. Perl, however, seems to
840 allow any characters between [: and :], though of course it rejects as
841 unknown any "names" that contain non-letters, because all the known class
842 names consist only of letters. Thus, Perl gives an error for [[:1234:]],
843 for example, whereas PCRE did not - it did not recognize a POSIX character
844 class. This seemed a bit dangerous, so the code has been changed to be
845 closer to Perl. The behaviour is not identical to Perl, because PCRE will
846 diagnose an unknown class for, for example, [[:l\ower:]] where Perl will
847 treat it as [[:lower:]]. However, PCRE does now give "unknown" errors where
848 Perl does, and where it didn't before.
850 24. Rewrite so as to remove the single use of %n from pcregrep because in some
851 Windows environments %n is disabled by default.
854 Version 7.4 21-Sep-07
855 ---------------------
857 1. Change 7.3/28 was implemented for classes by looking at the bitmap. This
858 means that a class such as [\s] counted as "explicit reference to CR or
859 LF". That isn't really right - the whole point of the change was to try to
860 help when there was an actual mention of one of the two characters. So now
861 the change happens only if \r or \n (or a literal CR or LF) character is
862 encountered.
864 2. The 32-bit options word was also used for 6 internal flags, but the numbers
865 of both had grown to the point where there were only 3 bits left.
866 Fortunately, there was spare space in the data structure, and so I have
867 moved the internal flags into a new 16-bit field to free up more option
868 bits.
870 3. The appearance of (?J) at the start of a pattern set the DUPNAMES option,
871 but did not set the internal JCHANGED flag - either of these is enough to
872 control the way the "get" function works - but the PCRE_INFO_JCHANGED
873 facility is supposed to tell if (?J) was ever used, so now (?J) at the
874 start sets both bits.
876 4. Added options (at build time, compile time, exec time) to change \R from
877 matching any Unicode line ending sequence to just matching CR, LF, or CRLF.
879 5. doc/pcresyntax.html was missing from the distribution.
881 6. Put back the definition of PCRE_ERROR_NULLWSLIMIT, for backward
882 compatibility, even though it is no longer used.
884 7. Added macro for snprintf to pcrecpp_unittest.cc and also for strtoll and
885 strtoull to pcrecpp.cc to select the available functions in WIN32 when the
886 windows.h file is present (where different names are used). [This was
887 reversed later after testing - see 16 below.]
889 8. Changed all #include <config.h> to #include "config.h". There were also
890 some further <pcre.h> cases that I changed to "pcre.h".
892 9. When pcregrep was used with the --colour option, it missed the line ending
893 sequence off the lines that it output.
895 10. It was pointed out to me that arrays of string pointers cause lots of
896 relocations when a shared library is dynamically loaded. A technique of
897 using a single long string with a table of offsets can drastically reduce
898 these. I have refactored PCRE in four places to do this. The result is
899 dramatic:
901 Originally: 290
902 After changing UCP table: 187
903 After changing error message table: 43
904 After changing table of "verbs" 36
905 After changing table of Posix names 22
907 Thanks to the folks working on Gregex for glib for this insight.
909 11. --disable-stack-for-recursion caused compiling to fail unless -enable-
910 unicode-properties was also set.
912 12. Updated the tests so that they work when \R is defaulted to ANYCRLF.
914 13. Added checks for ANY and ANYCRLF to pcrecpp.cc where it previously
915 checked only for CRLF.
917 14. Added casts to pcretest.c to avoid compiler warnings.
919 15. Added Craig's patch to various pcrecpp modules to avoid compiler warnings.
921 16. Added Craig's patch to remove the WINDOWS_H tests, that were not working,
922 and instead check for _strtoi64 explicitly, and avoid the use of snprintf()
923 entirely. This removes changes made in 7 above.
925 17. The CMake files have been updated, and there is now more information about
926 building with CMake in the NON-UNIX-USE document.
929 Version 7.3 28-Aug-07
930 ---------------------
932 1. In the rejigging of the build system that eventually resulted in 7.1, the
933 line "#include <pcre.h>" was included in pcre_internal.h. The use of angle
934 brackets there is not right, since it causes compilers to look for an
935 installed pcre.h, not the version that is in the source that is being
936 compiled (which of course may be different). I have changed it back to:
938 #include "pcre.h"
940 I have a vague recollection that the change was concerned with compiling in
941 different directories, but in the new build system, that is taken care of
942 by the VPATH setting the Makefile.
944 2. The pattern .*$ when run in not-DOTALL UTF-8 mode with newline=any failed
945 when the subject happened to end in the byte 0x85 (e.g. if the last
946 character was \x{1ec5}). *Character* 0x85 is one of the "any" newline
947 characters but of course it shouldn't be taken as a newline when it is part
948 of another character. The bug was that, for an unlimited repeat of . in
949 not-DOTALL UTF-8 mode, PCRE was advancing by bytes rather than by
950 characters when looking for a newline.
952 3. A small performance improvement in the DOTALL UTF-8 mode .* case.
954 4. Debugging: adjusted the names of opcodes for different kinds of parentheses
955 in debug output.
957 5. Arrange to use "%I64d" instead of "%lld" and "%I64u" instead of "%llu" for
958 long printing in the pcrecpp unittest when running under MinGW.
960 6. ESC_K was left out of the EBCDIC table.
962 7. Change 7.0/38 introduced a new limit on the number of nested non-capturing
963 parentheses; I made it 1000, which seemed large enough. Unfortunately, the
964 limit also applies to "virtual nesting" when a pattern is recursive, and in
965 this case 1000 isn't so big. I have been able to remove this limit at the
966 expense of backing off one optimization in certain circumstances. Normally,
967 when pcre_exec() would call its internal match() function recursively and
968 immediately return the result unconditionally, it uses a "tail recursion"
969 feature to save stack. However, when a subpattern that can match an empty
970 string has an unlimited repetition quantifier, it no longer makes this
971 optimization. That gives it a stack frame in which to save the data for
972 checking that an empty string has been matched. Previously this was taken
973 from the 1000-entry workspace that had been reserved. So now there is no
974 explicit limit, but more stack is used.
976 8. Applied Daniel's patches to solve problems with the import/export magic
977 syntax that is required for Windows, and which was going wrong for the
978 pcreposix and pcrecpp parts of the library. These were overlooked when this
979 problem was solved for the main library.
981 9. There were some crude static tests to avoid integer overflow when computing
982 the size of patterns that contain repeated groups with explicit upper
983 limits. As the maximum quantifier is 65535, the maximum group length was
984 set at 30,000 so that the product of these two numbers did not overflow a
985 32-bit integer. However, it turns out that people want to use groups that
986 are longer than 30,000 bytes (though not repeat them that many times).
987 Change 7.0/17 (the refactoring of the way the pattern size is computed) has
988 made it possible to implement the integer overflow checks in a much more
989 dynamic way, which I have now done. The artificial limitation on group
990 length has been removed - we now have only the limit on the total length of
991 the compiled pattern, which depends on the LINK_SIZE setting.
993 10. Fixed a bug in the documentation for get/copy named substring when
994 duplicate names are permitted. If none of the named substrings are set, the
995 functions return PCRE_ERROR_NOSUBSTRING (7); the doc said they returned an
996 empty string.
998 11. Because Perl interprets \Q...\E at a high level, and ignores orphan \E
999 instances, patterns such as [\Q\E] or [\E] or even [^\E] cause an error,
1000 because the ] is interpreted as the first data character and the
1001 terminating ] is not found. PCRE has been made compatible with Perl in this
1002 regard. Previously, it interpreted [\Q\E] as an empty class, and [\E] could
1003 cause memory overwriting.
1005 10. Like Perl, PCRE automatically breaks an unlimited repeat after an empty
1006 string has been matched (to stop an infinite loop). It was not recognizing
1007 a conditional subpattern that could match an empty string if that
1008 subpattern was within another subpattern. For example, it looped when
1009 trying to match (((?(1)X|))*) but it was OK with ((?(1)X|)*) where the
1010 condition was not nested. This bug has been fixed.
1012 12. A pattern like \X?\d or \P{L}?\d in non-UTF-8 mode could cause a backtrack
1013 past the start of the subject in the presence of bytes with the top bit
1014 set, for example "\x8aBCD".
1016 13. Added Perl 5.10 experimental backtracking controls (*FAIL), (*F), (*PRUNE),
1017 (*SKIP), (*THEN), (*COMMIT), and (*ACCEPT).
1019 14. Optimized (?!) to (*FAIL).
1021 15. Updated the test for a valid UTF-8 string to conform to the later RFC 3629.
1022 This restricts code points to be within the range 0 to 0x10FFFF, excluding
1023 the "low surrogate" sequence 0xD800 to 0xDFFF. Previously, PCRE allowed the
1024 full range 0 to 0x7FFFFFFF, as defined by RFC 2279. Internally, it still
1025 does: it's just the validity check that is more restrictive.
1027 16. Inserted checks for integer overflows during escape sequence (backslash)
1028 processing, and also fixed erroneous offset values for syntax errors during
1029 backslash processing.
1031 17. Fixed another case of looking too far back in non-UTF-8 mode (cf 12 above)
1032 for patterns like [\PPP\x8a]{1,}\x80 with the subject "A\x80".
1034 18. An unterminated class in a pattern like (?1)\c[ with a "forward reference"
1035 caused an overrun.
1037 19. A pattern like (?:[\PPa*]*){8,} which had an "extended class" (one with
1038 something other than just ASCII characters) inside a group that had an
1039 unlimited repeat caused a loop at compile time (while checking to see
1040 whether the group could match an empty string).
1042 20. Debugging a pattern containing \p or \P could cause a crash. For example,
1043 [\P{Any}] did so. (Error in the code for printing property names.)
1045 21. An orphan \E inside a character class could cause a crash.
1047 22. A repeated capturing bracket such as (A)? could cause a wild memory
1048 reference during compilation.
1050 23. There are several functions in pcre_compile() that scan along a compiled
1051 expression for various reasons (e.g. to see if it's fixed length for look
1052 behind). There were bugs in these functions when a repeated \p or \P was
1053 present in the pattern. These operators have additional parameters compared
1054 with \d, etc, and these were not being taken into account when moving along
1055 the compiled data. Specifically:
1057 (a) A item such as \p{Yi}{3} in a lookbehind was not treated as fixed
1058 length.
1060 (b) An item such as \pL+ within a repeated group could cause crashes or
1061 loops.
1063 (c) A pattern such as \p{Yi}+(\P{Yi}+)(?1) could give an incorrect
1064 "reference to non-existent subpattern" error.
1066 (d) A pattern like (\P{Yi}{2}\277)? could loop at compile time.
1068 24. A repeated \S or \W in UTF-8 mode could give wrong answers when multibyte
1069 characters were involved (for example /\S{2}/8g with "A\x{a3}BC").
1071 25. Using pcregrep in multiline, inverted mode (-Mv) caused it to loop.
1073 26. Patterns such as [\P{Yi}A] which include \p or \P and just one other
1074 character were causing crashes (broken optimization).
1076 27. Patterns such as (\P{Yi}*\277)* (group with possible zero repeat containing
1077 \p or \P) caused a compile-time loop.
1079 28. More problems have arisen in unanchored patterns when CRLF is a valid line
1080 break. For example, the unstudied pattern [\r\n]A does not match the string
1081 "\r\nA" because change 7.0/46 below moves the current point on by two
1082 characters after failing to match at the start. However, the pattern \nA
1083 *does* match, because it doesn't start till \n, and if [\r\n]A is studied,
1084 the same is true. There doesn't seem any very clean way out of this, but
1085 what I have chosen to do makes the common cases work: PCRE now takes note
1086 of whether there can be an explicit match for \r or \n anywhere in the
1087 pattern, and if so, 7.0/46 no longer applies. As part of this change,
1088 there's a new PCRE_INFO_HASCRORLF option for finding out whether a compiled
1089 pattern has explicit CR or LF references.
1091 29. Added (*CR) etc for changing newline setting at start of pattern.
1094 Version 7.2 19-Jun-07
1095 ---------------------
1097 1. If the fr_FR locale cannot be found for test 3, try the "french" locale,
1098 which is apparently normally available under Windows.
1100 2. Re-jig the pcregrep tests with different newline settings in an attempt
1101 to make them independent of the local environment's newline setting.
1103 3. Add code to configure.ac to remove -g from the CFLAGS default settings.
1105 4. Some of the "internals" tests were previously cut out when the link size
1106 was not 2, because the output contained actual offsets. The recent new
1107 "Z" feature of pcretest means that these can be cut out, making the tests
1108 usable with all link sizes.
1110 5. Implemented Stan Switzer's goto replacement for longjmp() when not using
1111 stack recursion. This gives a massive performance boost under BSD, but just
1112 a small improvement under Linux. However, it saves one field in the frame
1113 in all cases.
1115 6. Added more features from the forthcoming Perl 5.10:
1117 (a) (?-n) (where n is a string of digits) is a relative subroutine or
1118 recursion call. It refers to the nth most recently opened parentheses.
1120 (b) (?+n) is also a relative subroutine call; it refers to the nth next
1121 to be opened parentheses.
1123 (c) Conditions that refer to capturing parentheses can be specified
1124 relatively, for example, (?(-2)... or (?(+3)...
1126 (d) \K resets the start of the current match so that everything before
1127 is not part of it.
1129 (e) \k{name} is synonymous with \k<name> and \k'name' (.NET compatible).
1131 (f) \g{name} is another synonym - part of Perl 5.10's unification of
1132 reference syntax.
1134 (g) (?| introduces a group in which the numbering of parentheses in each
1135 alternative starts with the same number.
1137 (h) \h, \H, \v, and \V match horizontal and vertical whitespace.
1139 7. Added two new calls to pcre_fullinfo(): PCRE_INFO_OKPARTIAL and
1142 8. A pattern such as (.*(.)?)* caused pcre_exec() to fail by either not
1143 terminating or by crashing. Diagnosed by Viktor Griph; it was in the code
1144 for detecting groups that can match an empty string.
1146 9. A pattern with a very large number of alternatives (more than several
1147 hundred) was running out of internal workspace during the pre-compile
1148 phase, where pcre_compile() figures out how much memory will be needed. A
1149 bit of new cunning has reduced the workspace needed for groups with
1150 alternatives. The 1000-alternative test pattern now uses 12 bytes of
1151 workspace instead of running out of the 4096 that are available.
1153 10. Inserted some missing (unsigned int) casts to get rid of compiler warnings.
1155 11. Applied patch from Google to remove an optimization that didn't quite work.
1156 The report of the bug said:
1158 pcrecpp::RE("a*").FullMatch("aaa") matches, while
1159 pcrecpp::RE("a*?").FullMatch("aaa") does not, and
1160 pcrecpp::RE("a*?\\z").FullMatch("aaa") does again.
1162 12. If \p or \P was used in non-UTF-8 mode on a character greater than 127
1163 it matched the wrong number of bytes.
1166 Version 7.1 24-Apr-07
1167 ---------------------
1169 1. Applied Bob Rossi and Daniel G's patches to convert the build system to one
1170 that is more "standard", making use of automake and other Autotools. There
1171 is some re-arrangement of the files and adjustment of comments consequent
1172 on this.
1174 2. Part of the patch fixed a problem with the pcregrep tests. The test of -r
1175 for recursive directory scanning broke on some systems because the files
1176 are not scanned in any specific order and on different systems the order
1177 was different. A call to "sort" has been inserted into RunGrepTest for the
1178 approprate test as a short-term fix. In the longer term there may be an
1179 alternative.
1181 3. I had an email from Eric Raymond about problems translating some of PCRE's
1182 man pages to HTML (despite the fact that I distribute HTML pages, some
1183 people do their own conversions for various reasons). The problems
1184 concerned the use of low-level troff macros .br and .in. I have therefore
1185 removed all such uses from the man pages (some were redundant, some could
1186 be replaced by .nf/.fi pairs). The 132html script that I use to generate
1187 HTML has been updated to handle .nf/.fi and to complain if it encounters
1188 .br or .in.
1190 4. Updated comments in configure.ac that get placed in config.h.in and also
1191 arranged for config.h to be included in the distribution, with the name
1192 config.h.generic, for the benefit of those who have to compile without
1193 Autotools (compare pcre.h, which is now distributed as pcre.h.generic).
1195 5. Updated the support (such as it is) for Virtual Pascal, thanks to Stefan
1196 Weber: (1) pcre_internal.h was missing some function renames; (2) updated
1197 makevp.bat for the current PCRE, using the additional files
1198 makevp_c.txt, makevp_l.txt, and pcregexp.pas.
1200 6. A Windows user reported a minor discrepancy with test 2, which turned out
1201 to be caused by a trailing space on an input line that had got lost in his
1202 copy. The trailing space was an accident, so I've just removed it.
1204 7. Add -Wl,-R... flags in pcre-config.in for *BSD* systems, as I'm told
1205 that is needed.
1207 8. Mark ucp_table (in ucptable.h) and ucp_gentype (in pcre_ucp_searchfuncs.c)
1208 as "const" (a) because they are and (b) because it helps the PHP
1209 maintainers who have recently made a script to detect big data structures
1210 in the php code that should be moved to the .rodata section. I remembered
1211 to update Builducptable as well, so it won't revert if ucptable.h is ever
1212 re-created.
1214 9. Added some extra #ifdef SUPPORT_UTF8 conditionals into pcretest.c,
1215 pcre_printint.src, pcre_compile.c, pcre_study.c, and pcre_tables.c, in
1216 order to be able to cut out the UTF-8 tables in the latter when UTF-8
1217 support is not required. This saves 1.5-2K of code, which is important in
1218 some applications.
1220 Later: more #ifdefs are needed in pcre_ord2utf8.c and pcre_valid_utf8.c
1221 so as not to refer to the tables, even though these functions will never be
1222 called when UTF-8 support is disabled. Otherwise there are problems with a
1223 shared library.
1225 10. Fixed two bugs in the emulated memmove() function in pcre_internal.h:
1227 (a) It was defining its arguments as char * instead of void *.
1229 (b) It was assuming that all moves were upwards in memory; this was true
1230 a long time ago when I wrote it, but is no longer the case.
1232 The emulated memove() is provided for those environments that have neither
1233 memmove() nor bcopy(). I didn't think anyone used it these days, but that
1234 is clearly not the case, as these two bugs were recently reported.
1236 11. The script PrepareRelease is now distributed: it calls 132html, CleanTxt,
1237 and Detrail to create the HTML documentation, the .txt form of the man
1238 pages, and it removes trailing spaces from listed files. It also creates
1239 pcre.h.generic and config.h.generic from pcre.h and config.h. In the latter
1240 case, it wraps all the #defines with #ifndefs. This script should be run
1241 before "make dist".
1243 12. Fixed two fairly obscure bugs concerned with quantified caseless matching
1244 with Unicode property support.
1246 (a) For a maximizing quantifier, if the two different cases of the
1247 character were of different lengths in their UTF-8 codings (there are
1248 some cases like this - I found 11), and the matching function had to
1249 back up over a mixture of the two cases, it incorrectly assumed they
1250 were both the same length.
1252 (b) When PCRE was configured to use the heap rather than the stack for
1253 recursion during matching, it was not correctly preserving the data for
1254 the other case of a UTF-8 character when checking ahead for a match
1255 while processing a minimizing repeat. If the check also involved
1256 matching a wide character, but failed, corruption could cause an
1257 erroneous result when trying to check for a repeat of the original
1258 character.
1260 13. Some tidying changes to the testing mechanism:
1262 (a) The RunTest script now detects the internal link size and whether there
1263 is UTF-8 and UCP support by running ./pcretest -C instead of relying on
1264 values substituted by "configure". (The RunGrepTest script already did
1265 this for UTF-8.) The configure.ac script no longer substitutes the
1266 relevant variables.
1268 (b) The debugging options /B and /D in pcretest show the compiled bytecode
1269 with length and offset values. This means that the output is different
1270 for different internal link sizes. Test 2 is skipped for link sizes
1271 other than 2 because of this, bypassing the problem. Unfortunately,
1272 there was also a test in test 3 (the locale tests) that used /B and
1273 failed for link sizes other than 2. Rather than cut the whole test out,
1274 I have added a new /Z option to pcretest that replaces the length and
1275 offset values with spaces. This is now used to make test 3 independent
1276 of link size. (Test 2 will be tidied up later.)
1278 14. If erroroffset was passed as NULL to pcre_compile, it provoked a
1279 segmentation fault instead of returning the appropriate error message.
1281 15. In multiline mode when the newline sequence was set to "any", the pattern
1282 ^$ would give a match between the \r and \n of a subject such as "A\r\nB".
1283 This doesn't seem right; it now treats the CRLF combination as the line
1284 ending, and so does not match in that case. It's only a pattern such as ^$
1285 that would hit this one: something like ^ABC$ would have failed after \r
1286 and then tried again after \r\n.
1288 16. Changed the comparison command for RunGrepTest from "diff -u" to "diff -ub"
1289 in an attempt to make files that differ only in their line terminators
1290 compare equal. This works on Linux.
1292 17. Under certain error circumstances pcregrep might try to free random memory
1293 as it exited. This is now fixed, thanks to valgrind.
1295 19. In pcretest, if the pattern /(?m)^$/g<any> was matched against the string
1296 "abc\r\n\r\n", it found an unwanted second match after the second \r. This
1297 was because its rules for how to advance for /g after matching an empty
1298 string at the end of a line did not allow for this case. They now check for
1299 it specially.
1301 20. pcretest is supposed to handle patterns and data of any length, by
1302 extending its buffers when necessary. It was getting this wrong when the
1303 buffer for a data line had to be extended.
1305 21. Added PCRE_NEWLINE_ANYCRLF which is like ANY, but matches only CR, LF, or
1306 CRLF as a newline sequence.
1308 22. Code for handling Unicode properties in pcre_dfa_exec() wasn't being cut
1309 out by #ifdef SUPPORT_UCP. This did no harm, as it could never be used, but
1310 I have nevertheless tidied it up.
1312 23. Added some casts to kill warnings from HP-UX ia64 compiler.
1314 24. Added a man page for pcre-config.
1317 Version 7.0 19-Dec-06
1318 ---------------------
1320 1. Fixed a signed/unsigned compiler warning in pcre_compile.c, shown up by
1321 moving to gcc 4.1.1.
1323 2. The -S option for pcretest uses setrlimit(); I had omitted to #include
1324 sys/time.h, which is documented as needed for this function. It doesn't
1325 seem to matter on Linux, but it showed up on some releases of OS X.
1327 3. It seems that there are systems where bytes whose values are greater than
1328 127 match isprint() in the "C" locale. The "C" locale should be the
1329 default when a C program starts up. In most systems, only ASCII printing
1330 characters match isprint(). This difference caused the output from pcretest
1331 to vary, making some of the tests fail. I have changed pcretest so that:
1333 (a) When it is outputting text in the compiled version of a pattern, bytes
1334 other than 32-126 are always shown as hex escapes.
1336 (b) When it is outputting text that is a matched part of a subject string,
1337 it does the same, unless a different locale has been set for the match
1338 (using the /L modifier). In this case, it uses isprint() to decide.
1340 4. Fixed a major bug that caused incorrect computation of the amount of memory
1341 required for a compiled pattern when options that changed within the
1342 pattern affected the logic of the preliminary scan that determines the
1343 length. The relevant options are -x, and -i in UTF-8 mode. The result was
1344 that the computed length was too small. The symptoms of this bug were
1345 either the PCRE error "internal error: code overflow" from pcre_compile(),
1346 or a glibc crash with a message such as "pcretest: free(): invalid next
1347 size (fast)". Examples of patterns that provoked this bug (shown in
1348 pcretest format) are:
1350 /(?-x: )/x
1351 /(?x)(?-x: \s*#\s*)/
1352 /((?i)[\x{c0}])/8
1353 /(?i:[\x{c0}])/8
1355 HOWEVER: Change 17 below makes this fix obsolete as the memory computation
1356 is now done differently.
1358 5. Applied patches from Google to: (a) add a QuoteMeta function to the C++
1359 wrapper classes; (b) implement a new function in the C++ scanner that is
1360 more efficient than the old way of doing things because it avoids levels of
1361 recursion in the regex matching; (c) add a paragraph to the documentation
1362 for the FullMatch() function.
1364 6. The escape sequence \n was being treated as whatever was defined as
1365 "newline". Not only was this contrary to the documentation, which states
1366 that \n is character 10 (hex 0A), but it also went horribly wrong when
1367 "newline" was defined as CRLF. This has been fixed.
1369 7. In pcre_dfa_exec.c the value of an unsigned integer (the variable called c)
1370 was being set to -1 for the "end of line" case (supposedly a value that no
1371 character can have). Though this value is never used (the check for end of
1372 line is "zero bytes in current character"), it caused compiler complaints.
1373 I've changed it to 0xffffffff.
1375 8. In pcre_version.c, the version string was being built by a sequence of
1376 C macros that, in the event of PCRE_PRERELEASE being defined as an empty
1377 string (as it is for production releases) called a macro with an empty
1378 argument. The C standard says the result of this is undefined. The gcc
1379 compiler treats it as an empty string (which was what was wanted) but it is
1380 reported that Visual C gives an error. The source has been hacked around to
1381 avoid this problem.
1383 9. On the advice of a Windows user, included <io.h> and <fcntl.h> in Windows
1384 builds of pcretest, and changed the call to _setmode() to use _O_BINARY
1385 instead of 0x8000. Made all the #ifdefs test both _WIN32 and WIN32 (not all
1386 of them did).
1388 10. Originally, pcretest opened its input and output without "b"; then I was
1389 told that "b" was needed in some environments, so it was added for release
1390 5.0 to both the input and output. (It makes no difference on Unix-like
1391 systems.) Later I was told that it is wrong for the input on Windows. I've
1392 now abstracted the modes into two macros, to make it easier to fiddle with
1393 them, and removed "b" from the input mode under Windows.
1395 11. Added pkgconfig support for the C++ wrapper library, libpcrecpp.
1397 12. Added -help and --help to pcretest as an official way of being reminded
1398 of the options.
1400 13. Removed some redundant semicolons after macro calls in pcrecpparg.h.in
1401 and pcrecpp.cc because they annoy compilers at high warning levels.
1403 14. A bit of tidying/refactoring in pcre_exec.c in the main bumpalong loop.
1405 15. Fixed an occurrence of == in configure.ac that should have been = (shell
1406 scripts are not C programs :-) and which was not noticed because it works
1407 on Linux.
1409 16. pcretest is supposed to handle any length of pattern and data line (as one
1410 line or as a continued sequence of lines) by extending its input buffer if
1411 necessary. This feature was broken for very long pattern lines, leading to
1412 a string of junk being passed to pcre_compile() if the pattern was longer
1413 than about 50K.
1415 17. I have done a major re-factoring of the way pcre_compile() computes the
1416 amount of memory needed for a compiled pattern. Previously, there was code
1417 that made a preliminary scan of the pattern in order to do this. That was
1418 OK when PCRE was new, but as the facilities have expanded, it has become
1419 harder and harder to keep it in step with the real compile phase, and there
1420 have been a number of bugs (see for example, 4 above). I have now found a
1421 cunning way of running the real compile function in a "fake" mode that
1422 enables it to compute how much memory it would need, while actually only
1423 ever using a few hundred bytes of working memory and without too many
1424 tests of the mode. This should make future maintenance and development
1425 easier. A side effect of this work is that the limit of 200 on the nesting
1426 depth of parentheses has been removed (though this was never a serious
1427 limitation, I suspect). However, there is a downside: pcre_compile() now
1428 runs more slowly than before (30% or more, depending on the pattern). I
1429 hope this isn't a big issue. There is no effect on runtime performance.
1431 18. Fixed a minor bug in pcretest: if a pattern line was not terminated by a
1432 newline (only possible for the last line of a file) and it was a
1433 pattern that set a locale (followed by /Lsomething), pcretest crashed.
1435 19. Added additional timing features to pcretest. (1) The -tm option now times
1436 matching only, not compiling. (2) Both -t and -tm can be followed, as a
1437 separate command line item, by a number that specifies the number of
1438 repeats to use when timing. The default is 50000; this gives better
1439 precision, but takes uncomfortably long for very large patterns.
1441 20. Extended pcre_study() to be more clever in cases where a branch of a
1442 subpattern has no definite first character. For example, (a*|b*)[cd] would
1443 previously give no result from pcre_study(). Now it recognizes that the
1444 first character must be a, b, c, or d.
1446 21. There was an incorrect error "recursive call could loop indefinitely" if
1447 a subpattern (or the entire pattern) that was being tested for matching an
1448 empty string contained only one non-empty item after a nested subpattern.
1449 For example, the pattern (?>\x{100}*)\d(?R) provoked this error
1450 incorrectly, because the \d was being skipped in the check.
1452 22. The pcretest program now has a new pattern option /B and a command line
1453 option -b, which is equivalent to adding /B to every pattern. This causes
1454 it to show the compiled bytecode, without the additional information that
1455 -d shows. The effect of -d is now the same as -b with -i (and similarly, /D
1456 is the same as /B/I).
1458 23. A new optimization is now able automatically to treat some sequences such
1459 as a*b as a*+b. More specifically, if something simple (such as a character
1460 or a simple class like \d) has an unlimited quantifier, and is followed by
1461 something that cannot possibly match the quantified thing, the quantifier
1462 is automatically "possessified".
1464 24. A recursive reference to a subpattern whose number was greater than 39
1465 went wrong under certain circumstances in UTF-8 mode. This bug could also
1466 have affected the operation of pcre_study().
1468 25. Realized that a little bit of performance could be had by replacing
1469 (c & 0xc0) == 0xc0 with c >= 0xc0 when processing UTF-8 characters.
1471 26. Timing data from pcretest is now shown to 4 decimal places instead of 3.
1473 27. Possessive quantifiers such as a++ were previously implemented by turning
1474 them into atomic groups such as ($>a+). Now they have their own opcodes,
1475 which improves performance. This includes the automatically created ones
1476 from 23 above.
1478 28. A pattern such as (?=(\w+))\1: which simulates an atomic group using a
1479 lookahead was broken if it was not anchored. PCRE was mistakenly expecting
1480 the first matched character to be a colon. This applied both to named and
1481 numbered groups.
1483 29. The ucpinternal.h header file was missing its idempotency #ifdef.
1485 30. I was sent a "project" file called libpcre.a.dev which I understand makes
1486 building PCRE on Windows easier, so I have included it in the distribution.
1488 31. There is now a check in pcretest against a ridiculously large number being
1489 returned by pcre_exec() or pcre_dfa_exec(). If this happens in a /g or /G
1490 loop, the loop is abandoned.
1492 32. Forward references to subpatterns in conditions such as (?(2)...) where
1493 subpattern 2 is defined later cause pcre_compile() to search forwards in
1494 the pattern for the relevant set of parentheses. This search went wrong
1495 when there were unescaped parentheses in a character class, parentheses
1496 escaped with \Q...\E, or parentheses in a #-comment in /x mode.
1498 33. "Subroutine" calls and backreferences were previously restricted to
1499 referencing subpatterns earlier in the regex. This restriction has now
1500 been removed.
1502 34. Added a number of extra features that are going to be in Perl 5.10. On the
1503 whole, these are just syntactic alternatives for features that PCRE had
1504 previously implemented using the Python syntax or my own invention. The
1505 other formats are all retained for compatibility.
1507 (a) Named groups can now be defined as (?<name>...) or (?'name'...) as well
1508 as (?P<name>...). The new forms, as well as being in Perl 5.10, are
1509 also .NET compatible.
1511 (b) A recursion or subroutine call to a named group can now be defined as
1512 (?&name) as well as (?P>name).
1514 (c) A backreference to a named group can now be defined as \k<name> or
1515 \k'name' as well as (?P=name). The new forms, as well as being in Perl
1516 5.10, are also .NET compatible.
1518 (d) A conditional reference to a named group can now use the syntax
1519 (?(<name>) or (?('name') as well as (?(name).
1521 (e) A "conditional group" of the form (?(DEFINE)...) can be used to define
1522 groups (named and numbered) that are never evaluated inline, but can be
1523 called as "subroutines" from elsewhere. In effect, the DEFINE condition
1524 is always false. There may be only one alternative in such a group.
1526 (f) A test for recursion can be given as (?(R1).. or (?(R&name)... as well
1527 as the simple (?(R). The condition is true only if the most recent
1528 recursion is that of the given number or name. It does not search out
1529 through the entire recursion stack.
1531 (g) The escape \gN or \g{N} has been added, where N is a positive or
1532 negative number, specifying an absolute or relative reference.
1534 35. Tidied to get rid of some further signed/unsigned compiler warnings and
1535 some "unreachable code" warnings.
1537 36. Updated the Unicode property tables to Unicode version 5.0.0. Amongst other
1538 things, this adds five new scripts.
1540 37. Perl ignores orphaned \E escapes completely. PCRE now does the same.
1541 There were also incompatibilities regarding the handling of \Q..\E inside
1542 character classes, for example with patterns like [\Qa\E-\Qz\E] where the
1543 hyphen was adjacent to \Q or \E. I hope I've cleared all this up now.
1545 38. Like Perl, PCRE detects when an indefinitely repeated parenthesized group
1546 matches an empty string, and forcibly breaks the loop. There were bugs in
1547 this code in non-simple cases. For a pattern such as ^(a()*)* matched
1548 against aaaa the result was just "a" rather than "aaaa", for example. Two
1549 separate and independent bugs (that affected different cases) have been
1550 fixed.
1552 39. Refactored the code to abolish the use of different opcodes for small
1553 capturing bracket numbers. This is a tidy that I avoided doing when I
1554 removed the limit on the number of capturing brackets for 3.5 back in 2001.
1555 The new approach is not only tidier, it makes it possible to reduce the
1556 memory needed to fix the previous bug (38).
1558 40. Implemented PCRE_NEWLINE_ANY to recognize any of the Unicode newline
1559 sequences (http://unicode.org/unicode/reports/tr18/) as "newline" when
1560 processing dot, circumflex, or dollar metacharacters, or #-comments in /x
1561 mode.
1563 41. Add \R to match any Unicode newline sequence, as suggested in the Unicode
1564 report.
1566 42. Applied patch, originally from Ari Pollak, modified by Google, to allow
1567 copy construction and assignment in the C++ wrapper.
1569 43. Updated pcregrep to support "--newline=any". In the process, I fixed a
1570 couple of bugs that could have given wrong results in the "--newline=crlf"
1571 case.
1573 44. Added a number of casts and did some reorganization of signed/unsigned int
1574 variables following suggestions from Dair Grant. Also renamed the variable
1575 "this" as "item" because it is a C++ keyword.
1577 45. Arranged for dftables to add
1579 #include "pcre_internal.h"
1581 to pcre_chartables.c because without it, gcc 4.x may remove the array
1582 definition from the final binary if PCRE is built into a static library and
1583 dead code stripping is activated.
1585 46. For an unanchored pattern, if a match attempt fails at the start of a
1586 newline sequence, and the newline setting is CRLF or ANY, and the next two
1587 characters are CRLF, advance by two characters instead of one.
1590 Version 6.7 04-Jul-06
1591 ---------------------
1593 1. In order to handle tests when input lines are enormously long, pcretest has
1594 been re-factored so that it automatically extends its buffers when
1595 necessary. The code is crude, but this _is_ just a test program. The
1596 default size has been increased from 32K to 50K.
1598 2. The code in pcre_study() was using the value of the re argument before
1599 testing it for NULL. (Of course, in any sensible call of the function, it
1600 won't be NULL.)
1602 3. The memmove() emulation function in pcre_internal.h, which is used on
1603 systems that lack both memmove() and bcopy() - that is, hardly ever -
1604 was missing a "static" storage class specifier.
1606 4. When UTF-8 mode was not set, PCRE looped when compiling certain patterns
1607 containing an extended class (one that cannot be represented by a bitmap
1608 because it contains high-valued characters or Unicode property items, e.g.
1609 [\pZ]). Almost always one would set UTF-8 mode when processing such a
1610 pattern, but PCRE should not loop if you do not (it no longer does).
1611 [Detail: two cases were found: (a) a repeated subpattern containing an
1612 extended class; (b) a recursive reference to a subpattern that followed a
1613 previous extended class. It wasn't skipping over the extended class
1614 correctly when UTF-8 mode was not set.]
1616 5. A negated single-character class was not being recognized as fixed-length
1617 in lookbehind assertions such as (?<=[^f]), leading to an incorrect
1618 compile error "lookbehind assertion is not fixed length".
1620 6. The RunPerlTest auxiliary script was showing an unexpected difference
1621 between PCRE and Perl for UTF-8 tests. It turns out that it is hard to
1622 write a Perl script that can interpret lines of an input file either as
1623 byte characters or as UTF-8, which is what "perltest" was being required to
1624 do for the non-UTF-8 and UTF-8 tests, respectively. Essentially what you
1625 can't do is switch easily at run time between having the "use utf8;" pragma
1626 or not. In the end, I fudged it by using the RunPerlTest script to insert
1627 "use utf8;" explicitly for the UTF-8 tests.
1629 7. In multiline (/m) mode, PCRE was matching ^ after a terminating newline at
1630 the end of the subject string, contrary to the documentation and to what
1631 Perl does. This was true of both matching functions. Now it matches only at
1632 the start of the subject and immediately after *internal* newlines.
1634 8. A call of pcre_fullinfo() from pcretest to get the option bits was passing
1635 a pointer to an int instead of a pointer to an unsigned long int. This
1636 caused problems on 64-bit systems.
1638 9. Applied a patch from the folks at Google to pcrecpp.cc, to fix "another
1639 instance of the 'standard' template library not being so standard".
1641 10. There was no check on the number of named subpatterns nor the maximum
1642 length of a subpattern name. The product of these values is used to compute
1643 the size of the memory block for a compiled pattern. By supplying a very
1644 long subpattern name and a large number of named subpatterns, the size
1645 computation could be caused to overflow. This is now prevented by limiting
1646 the length of names to 32 characters, and the number of named subpatterns
1647 to 10,000.
1649 11. Subpatterns that are repeated with specific counts have to be replicated in
1650 the compiled pattern. The size of memory for this was computed from the
1651 length of the subpattern and the repeat count. The latter is limited to
1652 65535, but there was no limit on the former, meaning that integer overflow
1653 could in principle occur. The compiled length of a repeated subpattern is
1654 now limited to 30,000 bytes in order to prevent this.
1656 12. Added the optional facility to have named substrings with the same name.
1658 13. Added the ability to use a named substring as a condition, using the
1659 Python syntax: (?(name)yes|no). This overloads (?(R)... and names that
1660 are numbers (not recommended). Forward references are permitted.
1662 14. Added forward references in named backreferences (if you see what I mean).
1664 15. In UTF-8 mode, with the PCRE_DOTALL option set, a quantified dot in the
1665 pattern could run off the end of the subject. For example, the pattern
1666 "(?s)(.{1,5})"8 did this with the subject "ab".
1668 16. If PCRE_DOTALL or PCRE_MULTILINE were set, pcre_dfa_exec() behaved as if
1669 PCRE_CASELESS was set when matching characters that were quantified with ?
1670 or *.
1672 17. A character class other than a single negated character that had a minimum
1673 but no maximum quantifier - for example [ab]{6,} - was not handled
1674 correctly by pce_dfa_exec(). It would match only one character.
1676 18. A valid (though odd) pattern that looked like a POSIX character
1677 class but used an invalid character after [ (for example [[,abc,]]) caused
1678 pcre_compile() to give the error "Failed: internal error: code overflow" or
1679 in some cases to crash with a glibc free() error. This could even happen if
1680 the pattern terminated after [[ but there just happened to be a sequence of
1681 letters, a binary zero, and a closing ] in the memory that followed.
1683 19. Perl's treatment of octal escapes in the range \400 to \777 has changed
1684 over the years. Originally (before any Unicode support), just the bottom 8
1685 bits were taken. Thus, for example, \500 really meant \100. Nowadays the
1686 output from "man perlunicode" includes this:
1688 The regular expression compiler produces polymorphic opcodes. That
1689 is, the pattern adapts to the data and automatically switches to
1690 the Unicode character scheme when presented with Unicode data--or
1691 instead uses a traditional byte scheme when presented with byte
1692 data.
1694 Sadly, a wide octal escape does not cause a switch, and in a string with
1695 no other multibyte characters, these octal escapes are treated as before.
1696 Thus, in Perl, the pattern /\500/ actually matches \100 but the pattern
1697 /\500|\x{1ff}/ matches \500 or \777 because the whole thing is treated as a
1698 Unicode string.
1700 I have not perpetrated such confusion in PCRE. Up till now, it took just
1701 the bottom 8 bits, as in old Perl. I have now made octal escapes with
1702 values greater than \377 illegal in non-UTF-8 mode. In UTF-8 mode they
1703 translate to the appropriate multibyte character.
1705 29. Applied some refactoring to reduce the number of warnings from Microsoft
1706 and Borland compilers. This has included removing the fudge introduced
1707 seven years ago for the OS/2 compiler (see 2.02/2 below) because it caused
1708 a warning about an unused variable.
1710 21. PCRE has not included VT (character 0x0b) in the set of whitespace
1711 characters since release 4.0, because Perl (from release 5.004) does not.
1712 [Or at least, is documented not to: some releases seem to be in conflict
1713 with the documentation.] However, when a pattern was studied with
1714 pcre_study() and all its branches started with \s, PCRE still included VT
1715 as a possible starting character. Of course, this did no harm; it just
1716 caused an unnecessary match attempt.
1718 22. Removed a now-redundant internal flag bit that recorded the fact that case
1719 dependency changed within the pattern. This was once needed for "required
1720 byte" processing, but is no longer used. This recovers a now-scarce options
1721 bit. Also moved the least significant internal flag bit to the most-
1722 significant bit of the word, which was not previously used (hangover from
1723 the days when it was an int rather than a uint) to free up another bit for
1724 the future.
1726 23. Added support for CRLF line endings as well as CR and LF. As well as the
1727 default being selectable at build time, it can now be changed at runtime
1728 via the PCRE_NEWLINE_xxx flags. There are now options for pcregrep to
1729 specify that it is scanning data with non-default line endings.
1731 24. Changed the definition of CXXLINK to make it agree with the definition of
1732 LINK in the Makefile, by replacing LDFLAGS to CXXFLAGS.
1734 25. Applied Ian Taylor's patches to avoid using another stack frame for tail
1735 recursions. This makes a big different to stack usage for some patterns.
1737 26. If a subpattern containing a named recursion or subroutine reference such
1738 as (?P>B) was quantified, for example (xxx(?P>B)){3}, the calculation of
1739 the space required for the compiled pattern went wrong and gave too small a
1740 value. Depending on the environment, this could lead to "Failed: internal
1741 error: code overflow at offset 49" or "glibc detected double free or
1742 corruption" errors.
1744 27. Applied patches from Google (a) to support the new newline modes and (b) to
1745 advance over multibyte UTF-8 characters in GlobalReplace.
1747 28. Change free() to pcre_free() in pcredemo.c. Apparently this makes a
1748 difference for some implementation of PCRE in some Windows version.
1750 29. Added some extra testing facilities to pcretest:
1752 \q<number> in a data line sets the "match limit" value
1753 \Q<number> in a data line sets the "match recursion limt" value
1754 -S <number> sets the stack size, where <number> is in megabytes
1756 The -S option isn't available for Windows.
1759 Version 6.6 06-Feb-06
1760 ---------------------
1762 1. Change 16(a) for 6.5 broke things, because PCRE_DATA_SCOPE was not defined
1763 in pcreposix.h. I have copied the definition from pcre.h.
1765 2. Change 25 for 6.5 broke compilation in a build directory out-of-tree
1766 because pcre.h is no longer a built file.
1768 3. Added Jeff Friedl's additional debugging patches to pcregrep. These are
1769 not normally included in the compiled code.
1772 Version 6.5 01-Feb-06
1773 ---------------------
1775 1. When using the partial match feature with pcre_dfa_exec(), it was not
1776 anchoring the second and subsequent partial matches at the new starting
1777 point. This could lead to incorrect results. For example, with the pattern
1778 /1234/, partially matching against "123" and then "a4" gave a match.
1780 2. Changes to pcregrep:
1782 (a) All non-match returns from pcre_exec() were being treated as failures
1783 to match the line. Now, unless the error is PCRE_ERROR_NOMATCH, an
1784 error message is output. Some extra information is given for the
1786 probably the only errors that are likely to be caused by users (by
1787 specifying a regex that has nested indefinite repeats, for instance).
1788 If there are more than 20 of these errors, pcregrep is abandoned.
1790 (b) A binary zero was treated as data while matching, but terminated the
1791 output line if it was written out. This has been fixed: binary zeroes
1792 are now no different to any other data bytes.
1794 (c) Whichever of the LC_ALL or LC_CTYPE environment variables is set is
1795 used to set a locale for matching. The --locale=xxxx long option has
1796 been added (no short equivalent) to specify a locale explicitly on the
1797 pcregrep command, overriding the environment variables.
1799 (d) When -B was used with -n, some line numbers in the output were one less
1800 than they should have been.
1802 (e) Added the -o (--only-matching) option.
1804 (f) If -A or -C was used with -c (count only), some lines of context were
1805 accidentally printed for the final match.
1807 (g) Added the -H (--with-filename) option.
1809 (h) The combination of options -rh failed to suppress file names for files
1810 that were found from directory arguments.
1812 (i) Added the -D (--devices) and -d (--directories) options.
1814 (j) Added the -F (--fixed-strings) option.
1816 (k) Allow "-" to be used as a file name for -f as well as for a data file.
1818 (l) Added the --colo(u)r option.
1820 (m) Added Jeffrey Friedl's -S testing option, but within #ifdefs so that it
1821 is not present by default.
1823 3. A nasty bug was discovered in the handling of recursive patterns, that is,
1824 items such as (?R) or (?1), when the recursion could match a number of
1825 alternatives. If it matched one of the alternatives, but subsequently,
1826 outside the recursion, there was a failure, the code tried to back up into
1827 the recursion. However, because of the way PCRE is implemented, this is not
1828 possible, and the result was an incorrect result from the match.
1830 In order to prevent this happening, the specification of recursion has
1831 been changed so that all such subpatterns are automatically treated as
1832 atomic groups. Thus, for example, (?R) is treated as if it were (?>(?R)).
1834 4. I had overlooked the fact that, in some locales, there are characters for
1835 which isalpha() is true but neither isupper() nor islower() are true. In
1836 the fr_FR locale, for instance, the \xAA and \xBA characters (ordmasculine
1837 and ordfeminine) are like this. This affected the treatment of \w and \W
1838 when they appeared in character classes, but not when they appeared outside
1839 a character class. The bit map for "word" characters is now created
1840 separately from the results of isalnum() instead of just taking it from the
1841 upper, lower, and digit maps. (Plus the underscore character, of course.)
1843 5. The above bug also affected the handling of POSIX character classes such as
1844 [[:alpha:]] and [[:alnum:]]. These do not have their own bit maps in PCRE's
1845 permanent tables. Instead, the bit maps for such a class were previously
1846 created as the appropriate unions of the upper, lower, and digit bitmaps.
1847 Now they are created by subtraction from the [[:word:]] class, which has
1848 its own bitmap.
1850 6. The [[:blank:]] character class matches horizontal, but not vertical space.
1851 It is created by subtracting the vertical space characters (\x09, \x0a,
1852 \x0b, \x0c) from the [[:space:]] bitmap. Previously, however, the
1853 subtraction was done in the overall bitmap for a character class, meaning
1854 that a class such as [\x0c[:blank:]] was incorrect because \x0c would not
1855 be recognized. This bug has been fixed.
1857 7. Patches from the folks at Google:
1859 (a) pcrecpp.cc: "to handle a corner case that may or may not happen in
1860 real life, but is still worth protecting against".
1862 (b) pcrecpp.cc: "corrects a bug when negative radixes are used with
1863 regular expressions".
1865 (c) pcre_scanner.cc: avoid use of std::count() because not all systems
1866 have it.
1868 (d) Split off pcrecpparg.h from pcrecpp.h and had the former built by
1869 "configure" and the latter not, in order to fix a problem somebody had
1870 with compiling the Arg class on HP-UX.
1872 (e) Improve the error-handling of the C++ wrapper a little bit.
1874 (f) New tests for checking recursion limiting.
1876 8. The pcre_memmove() function, which is used only if the environment does not
1877 have a standard memmove() function (and is therefore rarely compiled),
1878 contained two bugs: (a) use of int instead of size_t, and (b) it was not
1879 returning a result (though PCRE never actually uses the result).
1881 9. In the POSIX regexec() interface, if nmatch is specified as a ridiculously
1882 large number - greater than INT_MAX/(3*sizeof(int)) - REG_ESPACE is
1883 returned instead of calling malloc() with an overflowing number that would
1884 most likely cause subsequent chaos.
1886 10. The debugging option of pcretest was not showing the NO_AUTO_CAPTURE flag.
1888 11. The POSIX flag REG_NOSUB is now supported. When a pattern that was compiled
1889 with this option is matched, the nmatch and pmatch options of regexec() are
1890 ignored.
1892 12. Added REG_UTF8 to the POSIX interface. This is not defined by POSIX, but is
1893 provided in case anyone wants to the the POSIX interface with UTF-8
1894 strings.
1896 13. Added CXXLDFLAGS to the Makefile parameters to provide settings only on the
1897 C++ linking (needed for some HP-UX environments).
1899 14. Avoid compiler warnings in get_ucpname() when compiled without UCP support
1900 (unused parameter) and in the pcre_printint() function (omitted "default"
1901 switch label when the default is to do nothing).
1903 15. Added some code to make it possible, when PCRE is compiled as a C++
1904 library, to replace subject pointers for pcre_exec() with a smart pointer
1905 class, thus making it possible to process discontinuous strings.
1907 16. The two macros PCRE_EXPORT and PCRE_DATA_SCOPE are confusing, and perform
1908 much the same function. They were added by different people who were trying
1909 to make PCRE easy to compile on non-Unix systems. It has been suggested
1910 that PCRE_EXPORT be abolished now that there is more automatic apparatus
1911 for compiling on Windows systems. I have therefore replaced it with
1912 PCRE_DATA_SCOPE. This is set automatically for Windows; if not set it
1913 defaults to "extern" for C or "extern C" for C++, which works fine on
1914 Unix-like systems. It is now possible to override the value of PCRE_DATA_
1915 SCOPE with something explicit in config.h. In addition:
1917 (a) pcreposix.h still had just "extern" instead of either of these macros;
1918 I have replaced it with PCRE_DATA_SCOPE.
1920 (b) Functions such as _pcre_xclass(), which are internal to the library,
1921 but external in the C sense, all had PCRE_EXPORT in their definitions.
1922 This is apparently wrong for the Windows case, so I have removed it.
1923 (It makes no difference on Unix-like systems.)
1925 17. Added a new limit, MATCH_LIMIT_RECURSION, which limits the depth of nesting
1926 of recursive calls to match(). This is different to MATCH_LIMIT because
1927 that limits the total number of calls to match(), not all of which increase
1928 the depth of recursion. Limiting the recursion depth limits the amount of
1929 stack (or heap if NO_RECURSE is set) that is used. The default can be set
1930 when PCRE is compiled, and changed at run time. A patch from Google adds
1931 this functionality to the C++ interface.
1933 18. Changes to the handling of Unicode character properties:
1935 (a) Updated the table to Unicode 4.1.0.
1937 (b) Recognize characters that are not in the table as "Cn" (undefined).
1939 (c) I revised the way the table is implemented to a much improved format
1940 which includes recognition of ranges. It now supports the ranges that
1941 are defined in UnicodeData.txt, and it also amalgamates other
1942 characters into ranges. This has reduced the number of entries in the
1943 table from around 16,000 to around 3,000, thus reducing its size
1944 considerably. I realized I did not need to use a tree structure after
1945 all - a binary chop search is just as efficient. Having reduced the
1946 number of entries, I extended their size from 6 bytes to 8 bytes to
1947 allow for more data.
1949 (d) Added support for Unicode script names via properties such as \p{Han}.
1951 19. In UTF-8 mode, a backslash followed by a non-Ascii character was not
1952 matching that character.
1954 20. When matching a repeated Unicode property with a minimum greater than zero,
1955 (for example \pL{2,}), PCRE could look past the end of the subject if it
1956 reached it while seeking the minimum number of characters. This could
1957 happen only if some of the characters were more than one byte long, because
1958 there is a check for at least the minimum number of bytes.
1960 21. Refactored the implementation of \p and \P so as to be more general, to
1961 allow for more different types of property in future. This has changed the
1962 compiled form incompatibly. Anybody with saved compiled patterns that use
1963 \p or \P will have to recompile them.
1965 22. Added "Any" and "L&" to the supported property types.
1967 23. Recognize \x{...} as a code point specifier, even when not in UTF-8 mode,
1968 but give a compile time error if the value is greater than 0xff.
1970 24. The man pages for pcrepartial, pcreprecompile, and pcre_compile2 were
1971 accidentally not being installed or uninstalled.
1973 25. The pcre.h file was built from pcre.h.in, but the only changes that were
1974 made were to insert the current release number. This seemed silly, because
1975 it made things harder for people building PCRE on systems that don't run
1976 "configure". I have turned pcre.h into a distributed file, no longer built
1977 by "configure", with the version identification directly included. There is
1978 no longer a pcre.h.in file.
1980 However, this change necessitated a change to the pcre-config script as
1981 well. It is built from pcre-config.in, and one of the substitutions was the
1982 release number. I have updated configure.ac so that ./configure now finds
1983 the release number by grepping pcre.h.
1985 26. Added the ability to run the tests under valgrind.
1988 Version 6.4 05-Sep-05
1989 ---------------------
1991 1. Change 6.0/10/(l) to pcregrep introduced a bug that caused separator lines
1992 "--" to be printed when multiple files were scanned, even when none of the
1993 -A, -B, or -C options were used. This is not compatible with Gnu grep, so I
1994 consider it to be a bug, and have restored the previous behaviour.
1996 2. A couple of code tidies to get rid of compiler warnings.
1998 3. The pcretest program used to cheat by referring to symbols in the library
1999 whose names begin with _pcre_. These are internal symbols that are not
2000 really supposed to be visible externally, and in some environments it is
2001 possible to suppress them. The cheating is now confined to including
2002 certain files from the library's source, which is a bit cleaner.
2004 4. Renamed pcre.in as pcre.h.in to go with pcrecpp.h.in; it also makes the
2005 file's purpose clearer.
2007 5. Reorganized pcre_ucp_findchar().
2010 Version 6.3 15-Aug-05
2011 ---------------------
2013 1. The file libpcre.pc.in did not have general read permission in the tarball.
2015 2. There were some problems when building without C++ support:
2017 (a) If C++ support was not built, "make install" and "make test" still
2018 tried to test it.
2020 (b) There were problems when the value of CXX was explicitly set. Some
2021 changes have been made to try to fix these, and ...
2023 (c) --disable-cpp can now be used to explicitly disable C++ support.
2025 (d) The use of @CPP_OBJ@ directly caused a blank line preceded by a
2026 backslash in a target when C++ was disabled. This confuses some
2027 versions of "make", apparently. Using an intermediate variable solves
2028 this. (Same for CPP_LOBJ.)
2030 3. $(LINK_FOR_BUILD) now includes $(CFLAGS_FOR_BUILD) and $(LINK)
2031 (non-Windows) now includes $(CFLAGS) because these flags are sometimes
2032 necessary on certain architectures.
2034 4. Added a setting of -export-symbols-regex to the link command to remove
2035 those symbols that are exported in the C sense, but actually are local
2036 within the library, and not documented. Their names all begin with
2037 "_pcre_". This is not a perfect job, because (a) we have to except some
2038 symbols that pcretest ("illegally") uses, and (b) the facility isn't always
2039 available (and never for static libraries). I have made a note to try to
2040 find a way round (a) in the future.
2043 Version 6.2 01-Aug-05
2044 ---------------------
2046 1. There was no test for integer overflow of quantifier values. A construction
2047 such as {1111111111111111} would give undefined results. What is worse, if
2048 a minimum quantifier for a parenthesized subpattern overflowed and became
2049 negative, the calculation of the memory size went wrong. This could have
2050 led to memory overwriting.
2052 2. Building PCRE using VPATH was broken. Hopefully it is now fixed.
2054 3. Added "b" to the 2nd argument of fopen() in dftables.c, for non-Unix-like
2055 operating environments where this matters.
2057 4. Applied Giuseppe Maxia's patch to add additional features for controlling
2058 PCRE options from within the C++ wrapper.
2060 5. Named capturing subpatterns were not being correctly counted when a pattern
2061 was compiled. This caused two problems: (a) If there were more than 100
2062 such subpatterns, the calculation of the memory needed for the whole
2063 compiled pattern went wrong, leading to an overflow error. (b) Numerical
2064 back references of the form \12, where the number was greater than 9, were
2065 not recognized as back references, even though there were sufficient
2066 previous subpatterns.
2068 6. Two minor patches to pcrecpp.cc in order to allow it to compile on older
2069 versions of gcc, e.g. 2.95.4.
2072 Version 6.1 21-Jun-05
2073 ---------------------
2075 1. There was one reference to the variable "posix" in pcretest.c that was not
2076 surrounded by "#if !defined NOPOSIX".
2078 2. Make it possible to compile pcretest without DFA support, UTF8 support, or
2079 the cross-check on the old pcre_info() function, for the benefit of the
2080 cut-down version of PCRE that is currently imported into Exim.
2082 3. A (silly) pattern starting with (?i)(?-i) caused an internal space
2083 allocation error. I've done the easy fix, which wastes 2 bytes for sensible
2084 patterns that start (?i) but I don't think that matters. The use of (?i) is
2085 just an example; this all applies to the other options as well.
2087 4. Since libtool seems to echo the compile commands it is issuing, the output
2088 from "make" can be reduced a bit by putting "@" in front of each libtool
2089 compile command.
2091 5. Patch from the folks at Google for configure.in to be a bit more thorough
2092 in checking for a suitable C++ installation before trying to compile the
2093 C++ stuff. This should fix a reported problem when a compiler was present,
2094 but no suitable headers.
2096 6. The man pages all had just "PCRE" as their title. I have changed them to
2097 be the relevant file name. I have also arranged that these names are
2098 retained in the file doc/pcre.txt, which is a concatenation in text format
2099 of all the man pages except the little individual ones for each function.
2101 7. The NON-UNIX-USE file had not been updated for the different set of source
2102 files that come with release 6. I also added a few comments about the C++
2103 wrapper.
2106 Version 6.0 07-Jun-05
2107 ---------------------
2109 1. Some minor internal re-organization to help with my DFA experiments.
2111 2. Some missing #ifdef SUPPORT_UCP conditionals in pcretest and printint that
2112 didn't matter for the library itself when fully configured, but did matter
2113 when compiling without UCP support, or within Exim, where the ucp files are
2114 not imported.
2116 3. Refactoring of the library code to split up the various functions into
2117 different source modules. The addition of the new DFA matching code (see
2118 below) to a single monolithic source would have made it really too
2119 unwieldy, quite apart from causing all the code to be include in a
2120 statically linked application, when only some functions are used. This is
2121 relevant even without the DFA addition now that patterns can be compiled in
2122 one application and matched in another.
2124 The downside of splitting up is that there have to be some external
2125 functions and data tables that are used internally in different modules of
2126 the library but which are not part of the API. These have all had their
2127 names changed to start with "_pcre_" so that they are unlikely to clash
2128 with other external names.
2130 4. Added an alternate matching function, pcre_dfa_exec(), which matches using
2131 a different (DFA) algorithm. Although it is slower than the original
2132 function, it does have some advantages for certain types of matching
2133 problem.
2135 5. Upgrades to pcretest in order to test the features of pcre_dfa_exec(),
2136 including restarting after a partial match.
2138 6. A patch for pcregrep that defines INVALID_FILE_ATTRIBUTES if it is not
2139 defined when compiling for Windows was sent to me. I have put it into the
2140 code, though I have no means of testing or verifying it.
2142 7. Added the pcre_refcount() auxiliary function.
2144 8. Added the PCRE_FIRSTLINE option. This constrains an unanchored pattern to
2145 match before or at the first newline in the subject string. In pcretest,
2146 the /f option on a pattern can be used to set this.
2148 9. A repeated \w when used in UTF-8 mode with characters greater than 256
2149 would behave wrongly. This has been present in PCRE since release 4.0.
2151 10. A number of changes to the pcregrep command:
2153 (a) Refactored how -x works; insert ^(...)$ instead of setting
2154 PCRE_ANCHORED and checking the length, in preparation for adding
2155 something similar for -w.
2157 (b) Added the -w (match as a word) option.
2159 (c) Refactored the way lines are read and buffered so as to have more
2160 than one at a time available.
2162 (d) Implemented a pcregrep test script.
2164 (e) Added the -M (multiline match) option. This allows patterns to match
2165 over several lines of the subject. The buffering ensures that at least
2166 8K, or the rest of the document (whichever is the shorter) is available
2167 for matching (and similarly the previous 8K for lookbehind assertions).
2169 (f) Changed the --help output so that it now says
2171 -w, --word-regex(p)
2173 instead of two lines, one with "regex" and the other with "regexp"
2174 because that confused at least one person since the short forms are the
2175 same. (This required a bit of code, as the output is generated
2176 automatically from a table. It wasn't just a text change.)
2178 (g) -- can be used to terminate pcregrep options if the next thing isn't an
2179 option but starts with a hyphen. Could be a pattern or a path name
2180 starting with a hyphen, for instance.
2182 (h) "-" can be given as a file name to represent stdin.
2184 (i) When file names are being printed, "(standard input)" is used for
2185 the standard input, for compatibility with GNU grep. Previously
2186 "<stdin>" was used.
2188 (j) The option --label=xxx can be used to supply a name to be used for
2189 stdin when file names are being printed. There is no short form.
2191 (k) Re-factored the options decoding logic because we are going to add
2192 two more options that take data. Such options can now be given in four
2193 different ways, e.g. "-fname", "-f name", "--file=name", "--file name".
2195 (l) Added the -A, -B, and -C options for requesting that lines of context
2196 around matches be printed.
2198 (m) Added the -L option to print the names of files that do not contain
2199 any matching lines, that is, the complement of -l.
2201 (n) The return code is 2 if any file cannot be opened, but pcregrep does
2202 continue to scan other files.
2204 (o) The -s option was incorrectly implemented. For compatibility with other
2205 greps, it now suppresses the error message for a non-existent or non-
2206 accessible file (but not the return code). There is a new option called
2207 -q that suppresses the output of matching lines, which was what -s was
2208 previously doing.
2210 (p) Added --include and --exclude options to specify files for inclusion
2211 and exclusion when recursing.
2213 11. The Makefile was not using the Autoconf-supported LDFLAGS macro properly.
2214 Hopefully, it now does.
2216 12. Missing cast in pcre_study().
2218 13. Added an "uninstall" target to the makefile.
2220 14. Replaced "extern" in the function prototypes in Makefile.in with
2221 "PCRE_DATA_SCOPE", which defaults to 'extern' or 'extern "C"' in the Unix
2222 world, but is set differently for Windows.
2224 15. Added a second compiling function called pcre_compile2(). The only
2225 difference is that it has an extra argument, which is a pointer to an
2226 integer error code. When there is a compile-time failure, this is set
2227 non-zero, in addition to the error test pointer being set to point to an
2228 error message. The new argument may be NULL if no error number is required
2229 (but then you may as well call pcre_compile(), which is now just a
2230 wrapper). This facility is provided because some applications need a
2231 numeric error indication, but it has also enabled me to tidy up the way
2232 compile-time errors are handled in the POSIX wrapper.
2234 16. Added VPATH=.libs to the makefile; this should help when building with one
2235 prefix path and installing with another. (Or so I'm told by someone who
2236 knows more about this stuff than I do.)
2238 17. Added a new option, REG_DOTALL, to the POSIX function regcomp(). This
2239 passes PCRE_DOTALL to the pcre_compile() function, making the "." character
2240 match everything, including newlines. This is not POSIX-compatible, but
2241 somebody wanted the feature. From pcretest it can be activated by using
2242 both the P and the s flags.
2244 18. AC_PROG_LIBTOOL appeared twice in Makefile.in. Removed one.
2246 19. libpcre.pc was being incorrectly installed as executable.
2248 20. A couple of places in pcretest check for end-of-line by looking for '\n';
2249 it now also looks for '\r' so that it will work unmodified on Windows.
2251 21. Added Google's contributed C++ wrapper to the distribution.
2253 22. Added some untidy missing memory free() calls in pcretest, to keep
2254 Electric Fence happy when testing.
2258 Version 5.0 13-Sep-04
2259 ---------------------
2261 1. Internal change: literal characters are no longer packed up into items
2262 containing multiple characters in a single byte-string. Each character
2263 is now matched using a separate opcode. However, there may be more than one
2264 byte in the character in UTF-8 mode.
2266 2. The pcre_callout_block structure has two new fields: pattern_position and
2267 next_item_length. These contain the offset in the pattern to the next match
2268 item, and its length, respectively.
2270 3. The PCRE_AUTO_CALLOUT option for pcre_compile() requests the automatic
2271 insertion of callouts before each pattern item. Added the /C option to
2272 pcretest to make use of this.
2274 4. On the advice of a Windows user, the lines
2276 #if defined(_WIN32) || defined(WIN32)
2277 _setmode( _fileno( stdout ), 0x8000 );
2278 #endif /* defined(_WIN32) || defined(WIN32) */
2280 have been added to the source of pcretest. This apparently does useful
2281 magic in relation to line terminators.
2283 5. Changed "r" and "w" in the calls to fopen() in pcretest to "rb" and "wb"
2284 for the benefit of those environments where the "b" makes a difference.
2286 6. The icc compiler has the same options as gcc, but "configure" doesn't seem
2287 to know about it. I have put a hack into configure.in that adds in code
2288 to set GCC=yes if CC=icc. This seems to end up at a point in the
2289 generated configure script that is early enough to affect the setting of
2290 compiler options, which is what is needed, but I have no means of testing
2291 whether it really works. (The user who reported this had patched the
2292 generated configure script, which of course I cannot do.)
2294 LATER: After change 22 below (new libtool files), the configure script
2295 seems to know about icc (and also ecc). Therefore, I have commented out
2296 this hack in configure.in.
2298 7. Added support for pkg-config (2 patches were sent in).
2300 8. Negated POSIX character classes that used a combination of internal tables
2301 were completely broken. These were [[:^alpha:]], [[:^alnum:]], and
2302 [[:^ascii]]. Typically, they would match almost any characters. The other
2303 POSIX classes were not broken in this way.
2305 9. Matching the pattern "\b.*?" against "ab cd", starting at offset 1, failed
2306 to find the match, as PCRE was deluded into thinking that the match had to
2307 start at the start point or following a newline. The same bug applied to
2308 patterns with negative forward assertions or any backward assertions
2309 preceding ".*" at the start, unless the pattern required a fixed first
2310 character. This was a failing pattern: "(?!.bcd).*". The bug is now fixed.
2312 10. In UTF-8 mode, when moving forwards in the subject after a failed match
2313 starting at the last subject character, bytes beyond the end of the subject
2314 string were read.
2316 11. Renamed the variable "class" as "classbits" to make life easier for C++
2317 users. (Previously there was a macro definition, but it apparently wasn't
2318 enough.)
2320 12. Added the new field "tables" to the extra data so that tables can be passed
2321 in at exec time, or the internal tables can be re-selected. This allows
2322 a compiled regex to be saved and re-used at a later time by a different
2323 program that might have everything at different addresses.
2325 13. Modified the pcre-config script so that, when run on Solaris, it shows a
2326 -R library as well as a -L library.
2328 14. The debugging options of pcretest (-d on the command line or D on a
2329 pattern) showed incorrect output for anything following an extended class
2330 that contained multibyte characters and which was followed by a quantifier.
2332 15. Added optional support for general category Unicode character properties
2333 via the \p, \P, and \X escapes. Unicode property support implies UTF-8
2334 support. It adds about 90K to the size of the library. The meanings of the
2335 inbuilt class escapes such as \d and \s have NOT been changed.
2337 16. Updated pcredemo.c to include calls to free() to release the memory for the
2338 compiled pattern.
2340 17. The generated file chartables.c was being created in the source directory
2341 instead of in the building directory. This caused the build to fail if the
2342 source directory was different from the building directory, and was
2343 read-only.
2345 18. Added some sample Win commands from Mark Tetrode into the NON-UNIX-USE
2346 file. No doubt somebody will tell me if they don't make sense... Also added
2347 Dan Mooney's comments about building on OpenVMS.
2349 19. Added support for partial matching via the PCRE_PARTIAL option for
2350 pcre_exec() and the \P data escape in pcretest.
2352 20. Extended pcretest with 3 new pattern features:
2354 (i) A pattern option of the form ">rest-of-line" causes pcretest to
2355 write the compiled pattern to the file whose name is "rest-of-line".
2356 This is a straight binary dump of the data, with the saved pointer to
2357 the character tables forced to be NULL. The study data, if any, is
2358 written too. After writing, pcretest reads a new pattern.
2360 (ii) If, instead of a pattern, "<rest-of-line" is given, pcretest reads a
2361 compiled pattern from the given file. There must not be any
2362 occurrences of "<" in the file name (pretty unlikely); if there are,
2363 pcretest will instead treat the initial "<" as a pattern delimiter.
2364 After reading in the pattern, pcretest goes on to read data lines as
2365 usual.
2367 (iii) The F pattern option causes pcretest to flip the bytes in the 32-bit
2368 and 16-bit fields in a compiled pattern, to simulate a pattern that
2369 was compiled on a host of opposite endianness.
2371 21. The pcre-exec() function can now cope with patterns that were compiled on
2372 hosts of opposite endianness, with this restriction:
2374 As for any compiled expression that is saved and used later, the tables
2375 pointer field cannot be preserved; the extra_data field in the arguments
2376 to pcre_exec() should be used to pass in a tables address if a value
2377 other than the default internal tables were used at compile time.
2379 22. Calling pcre_exec() with a negative value of the "ovecsize" parameter is
2380 now diagnosed as an error. Previously, most of the time, a negative number
2381 would have been treated as zero, but if in addition "ovector" was passed as
2382 NULL, a crash could occur.
2384 23. Updated the files ltmain.sh, config.sub, config.guess, and aclocal.m4 with
2385 new versions from the libtool 1.5 distribution (the last one is a copy of
2386 a file called libtool.m4). This seems to have fixed the need to patch
2387 "configure" to support Darwin 1.3 (which I used to do). However, I still
2388 had to patch ltmain.sh to ensure that ${SED} is set (it isn't on my
2389 workstation).
2391 24. Changed the PCRE licence to be the more standard "BSD" licence.
2394 Version 4.5 01-Dec-03
2395 ---------------------
2397 1. There has been some re-arrangement of the code for the match() function so
2398 that it can be compiled in a version that does not call itself recursively.
2399 Instead, it keeps those local variables that need separate instances for
2400 each "recursion" in a frame on the heap, and gets/frees frames whenever it
2401 needs to "recurse". Keeping track of where control must go is done by means
2402 of setjmp/longjmp. The whole thing is implemented by a set of macros that
2403 hide most of the details from the main code, and operates only if
2404 NO_RECURSE is defined while compiling pcre.c. If PCRE is built using the
2405 "configure" mechanism, "--disable-stack-for-recursion" turns on this way of
2406 operating.
2408 To make it easier for callers to provide specially tailored get/free
2409 functions for this usage, two new functions, pcre_stack_malloc, and
2410 pcre_stack_free, are used. They are always called in strict stacking order,
2411 and the size of block requested is always the same.
2413 The PCRE_CONFIG_STACKRECURSE info parameter can be used to find out whether
2414 PCRE has been compiled to use the stack or the heap for recursion. The
2415 -C option of pcretest uses this to show which version is compiled.
2417 A new data escape \S, is added to pcretest; it causes the amounts of store
2418 obtained and freed by both kinds of malloc/free at match time to be added
2419 to the output.
2421 2. Changed the locale test to use "fr_FR" instead of "fr" because that's
2422 what's available on my current Linux desktop machine.
2424 3. When matching a UTF-8 string, the test for a valid string at the start has
2425 been extended. If start_offset is not zero, PCRE now checks that it points
2426 to a byte that is the start of a UTF-8 character. If not, it returns
2427 PCRE_ERROR_BADUTF8_OFFSET (-11). Note: the whole string is still checked;
2428 this is necessary because there may be backward assertions in the pattern.
2429 When matching the same subject several times, it may save resources to use
2430 PCRE_NO_UTF8_CHECK on all but the first call if the string is long.
2432 4. The code for checking the validity of UTF-8 strings has been tightened so
2433 that it rejects (a) strings containing 0xfe or 0xff bytes and (b) strings
2434 containing "overlong sequences".
2436 5. Fixed a bug (appearing twice) that I could not find any way of exploiting!
2437 I had written "if ((digitab[*p++] && chtab_digit) == 0)" where the "&&"
2438 should have been "&", but it just so happened that all the cases this let
2439 through by mistake were picked up later in the function.
2441 6. I had used a variable called "isblank" - this is a C99 function, causing
2442 some compilers to warn. To avoid this, I renamed it (as "blankclass").
2444 7. Cosmetic: (a) only output another newline at the end of pcretest if it is
2445 prompting; (b) run "./pcretest /dev/null" at the start of the test script
2446 so the version is shown; (c) stop "make test" echoing "./RunTest".
2448 8. Added patches from David Burgess to enable PCRE to run on EBCDIC systems.
2450 9. The prototype for memmove() for systems that don't have it was using
2451 size_t, but the inclusion of the header that defines size_t was later. I've
2452 moved the #includes for the C headers earlier to avoid this.
2454 10. Added some adjustments to the code to make it easier to compiler on certain
2455 special systems:
2457 (a) Some "const" qualifiers were missing.
2458 (b) Added the macro EXPORT before all exported functions; by default this
2459 is defined to be empty.
2460 (c) Changed the dftables auxiliary program (that builds chartables.c) so
2461 that it reads its output file name as an argument instead of writing
2462 to the standard output and assuming this can be redirected.
2464 11. In UTF-8 mode, if a recursive reference (e.g. (?1)) followed a character
2465 class containing characters with values greater than 255, PCRE compilation
2466 went into a loop.
2468 12. A recursive reference to a subpattern that was within another subpattern
2469 that had a minimum quantifier of zero caused PCRE to crash. For example,
2470 (x(y(?2))z)? provoked this bug with a subject that got as far as the
2471 recursion. If the recursively-called subpattern itself had a zero repeat,
2472 that was OK.
2474 13. In pcretest, the buffer for reading a data line was set at 30K, but the
2475 buffer into which it was copied (for escape processing) was still set at
2476 1024, so long lines caused crashes.
2478 14. A pattern such as /[ab]{1,3}+/ failed to compile, giving the error
2479 "internal error: code overflow...". This applied to any character class
2480 that was followed by a possessive quantifier.
2482 15. Modified the Makefile to add libpcre.la as a prerequisite for
2483 libpcreposix.la because I was told this is needed for a parallel build to
2484 work.
2486 16. If a pattern that contained .* following optional items at the start was
2487 studied, the wrong optimizing data was generated, leading to matching
2488 errors. For example, studying /[ab]*.*c/ concluded, erroneously, that any
2489 matching string must start with a or b or c. The correct conclusion for
2490 this pattern is that a match can start with any character.
2493 Version 4.4 13-Aug-03
2494 ---------------------
2496 1. In UTF-8 mode, a character class containing characters with values between
2497 127 and 255 was not handled correctly if the compiled pattern was studied.
2498 In fixing this, I have also improved the studying algorithm for such
2499 classes (slightly).
2501 2. Three internal functions had redundant arguments passed to them. Removal
2502 might give a very teeny performance improvement.
2504 3. Documentation bug: the value of the capture_top field in a callout is *one
2505 more than* the number of the hightest numbered captured substring.
2507 4. The Makefile linked pcretest and pcregrep with -lpcre, which could result
2508 in incorrectly linking with a previously installed version. They now link
2509 explicitly with libpcre.la.
2511 5. configure.in no longer needs to recognize Cygwin specially.
2513 6. A problem in pcre.in for Windows platforms is fixed.
2515 7. If a pattern was successfully studied, and the -d (or /D) flag was given to
2516 pcretest, it used to include the size of the study block as part of its
2517 output. Unfortunately, the structure contains a field that has a different
2518 size on different hardware architectures. This meant that the tests that
2519 showed this size failed. As the block is currently always of a fixed size,
2520 this information isn't actually particularly useful in pcretest output, so
2521 I have just removed it.
2523 8. Three pre-processor statements accidentally did not start in column 1.
2524 Sadly, there are *still* compilers around that complain, even though
2525 standard C has not required this for well over a decade. Sigh.
2527 9. In pcretest, the code for checking callouts passed small integers in the
2528 callout_data field, which is a void * field. However, some picky compilers
2529 complained about the casts involved for this on 64-bit systems. Now
2530 pcretest passes the address of the small integer instead, which should get
2531 rid of the warnings.
2533 10. By default, when in UTF-8 mode, PCRE now checks for valid UTF-8 strings at
2534 both compile and run time, and gives an error if an invalid UTF-8 sequence
2535 is found. There is a option for disabling this check in cases where the
2536 string is known to be correct and/or the maximum performance is wanted.
2538 11. In response to a bug report, I changed one line in Makefile.in from
2540 -Wl,--out-implib,.libs/lib@WIN_PREFIX@pcreposix.dll.a \
2541 to
2542 -Wl,--out-implib,.libs/@WIN_PREFIX@libpcreposix.dll.a \
2544 to look similar to other lines, but I have no way of telling whether this
2545 is the right thing to do, as I do not use Windows. No doubt I'll get told
2546 if it's wrong...
2549 Version 4.3 21-May-03
2550 ---------------------
2552 1. Two instances of @WIN_PREFIX@ omitted from the Windows targets in the
2553 Makefile.
2555 2. Some refactoring to improve the quality of the code:
2557 (i) The utf8_table... variables are now declared "const".
2559 (ii) The code for \cx, which used the "case flipping" table to upper case
2560 lower case letters, now just substracts 32. This is ASCII-specific,
2561 but the whole concept of \cx is ASCII-specific, so it seems
2562 reasonable.
2564 (iii) PCRE was using its character types table to recognize decimal and
2565 hexadecimal digits in the pattern. This is silly, because it handles
2566 only 0-9, a-f, and A-F, but the character types table is locale-
2567 specific, which means strange things might happen. A private
2568 table is now used for this - though it costs 256 bytes, a table is
2569 much faster than multiple explicit tests. Of course, the standard
2570 character types table is still used for matching digits in subject
2571 strings against \d.
2573 (iv) Strictly, the identifier ESC_t is reserved by POSIX (all identifiers
2574 ending in _t are). So I've renamed it as ESC_tee.
2576 3. The first argument for regexec() in the POSIX wrapper should have been
2577 defined as "const".
2579 4. Changed pcretest to use malloc() for its buffers so that they can be
2580 Electric Fenced for debugging.
2582 5. There were several places in the code where, in UTF-8 mode, PCRE would try
2583 to read one or more bytes before the start of the subject string. Often this
2584 had no effect on PCRE's behaviour, but in some circumstances it could
2585 provoke a segmentation fault.
2587 6. A lookbehind at the start of a pattern in UTF-8 mode could also cause PCRE
2588 to try to read one or more bytes before the start of the subject string.
2590 7. A lookbehind in a pattern matched in non-UTF-8 mode on a PCRE compiled with
2591 UTF-8 support could misbehave in various ways if the subject string
2592 contained bytes with the 0x80 bit set and the 0x40 bit unset in a lookbehind
2593 area. (PCRE was not checking for the UTF-8 mode flag, and trying to move
2594 back over UTF-8 characters.)
2597 Version 4.2 14-Apr-03
2598 ---------------------
2600 1. Typo "#if SUPPORT_UTF8" instead of "#ifdef SUPPORT_UTF8" fixed.
2602 2. Changes to the building process, supplied by Ronald Landheer-Cieslak
2603 [ON_WINDOWS]: new variable, "#" on non-Windows platforms
2604 [NOT_ON_WINDOWS]: new variable, "#" on Windows platforms
2605 [WIN_PREFIX]: new variable, "cyg" for Cygwin
2606 * Makefile.in: use autoconf substitution for OBJEXT, EXEEXT, BUILD_OBJEXT
2608 Note: automatic setting of the BUILD variables is not yet working
2609 set CPPFLAGS and BUILD_CPPFLAGS (but don't use yet) - should be used at
2610 compile-time but not at link-time
2611 [LINK]: use for linking executables only
2612 make different versions for Windows and non-Windows
2613 [LINKLIB]: new variable, copy of UNIX-style LINK, used for linking
2614 libraries
2615 [LINK_FOR_BUILD]: new variable
2616 [OBJEXT]: use throughout
2617 [EXEEXT]: use throughout
2618 <winshared>: new target
2619 <wininstall>: new target
2620 <dftables.o>: use native compiler
2621 <dftables>: use native linker
2622 <install>: handle Windows platform correctly
2623 <clean>: ditto
2624 <check>: ditto
2625 copy DLL to top builddir before testing
2627 As part of these changes, -no-undefined was removed again. This was reported
2628 to give trouble on HP-UX 11.0, so getting rid of it seems like a good idea
2629 in any case.
2631 3. Some tidies to get rid of compiler warnings:
2633 . In the match_data structure, match_limit was an unsigned long int, whereas
2634 match_call_count was an int. I've made them both unsigned long ints.
2636 . In pcretest the fact that a const uschar * doesn't automatically cast to
2637 a void * provoked a warning.
2639 . Turning on some more compiler warnings threw up some "shadow" variables
2640 and a few more missing casts.
2642 4. If PCRE was complied with UTF-8 support, but called without the PCRE_UTF8
2643 option, a class that contained a single character with a value between 128
2644 and 255 (e.g. /[\xFF]/) caused PCRE to crash.
2646 5. If PCRE was compiled with UTF-8 support, but called without the PCRE_UTF8
2647 option, a class that contained several characters, but with at least one
2648 whose value was between 128 and 255 caused PCRE to crash.
2651 Version 4.1 12-Mar-03
2652 ---------------------
2654 1. Compiling with gcc -pedantic found a couple of places where casts were
2655 needed, and a string in dftables.c that was longer than standard compilers are
2656 required to support.
2658 2. Compiling with Sun's compiler found a few more places where the code could
2659 be tidied up in order to avoid warnings.
2661 3. The variables for cross-compiling were called HOST_CC and HOST_CFLAGS; the
2662 first of these names is deprecated in the latest Autoconf in favour of the name
2663 CC_FOR_BUILD, because "host" is typically used to mean the system on which the
2664 compiled code will be run. I can't find a reference for HOST_CFLAGS, but by
2665 analogy I have changed it to CFLAGS_FOR_BUILD.
2667 4. Added -no-undefined to the linking command in the Makefile, because this is
2668 apparently helpful for Windows. To make it work, also added "-L. -lpcre" to the
2669 linking step for the pcreposix library.
2671 5. PCRE was failing to diagnose the case of two named groups with the same
2672 name.
2674 6. A problem with one of PCRE's optimizations was discovered. PCRE remembers a
2675 literal character that is needed in the subject for a match, and scans along to
2676 ensure that it is present before embarking on the full matching process. This
2677 saves time in cases of nested unlimited repeats that are never going to match.
2678 Problem: the scan can take a lot of time if the subject is very long (e.g.
2679 megabytes), thus penalizing straightforward matches. It is now done only if the
2680 amount of subject to be scanned is less than 1000 bytes.
2682 7. A lesser problem with the same optimization is that it was recording the
2683 first character of an anchored pattern as "needed", thus provoking a search
2684 right along the subject, even when the first match of the pattern was going to
2685 fail. The "needed" character is now not set for anchored patterns, unless it
2686 follows something in the pattern that is of non-fixed length. Thus, it still
2687 fulfils its original purpose of finding quick non-matches in cases of nested
2688 unlimited repeats, but isn't used for simple anchored patterns such as /^abc/.
2691 Version 4.0 17-Feb-03
2692 ---------------------
2694 1. If a comment in an extended regex that started immediately after a meta-item
2695 extended to the end of string, PCRE compiled incorrect data. This could lead to
2696 all kinds of weird effects. Example: /#/ was bad; /()#/ was bad; /a#/ was not.
2698 2. Moved to autoconf 2.53 and libtool 1.4.2.
2700 3. Perl 5.8 no longer needs "use utf8" for doing UTF-8 things. Consequently,
2701 the special perltest8 script is no longer needed - all the tests can be run
2702 from a single perltest script.
2704 4. From 5.004, Perl has not included the VT character (0x0b) in the set defined
2705 by \s. It has now been removed in PCRE. This means it isn't recognized as
2706 whitespace in /x regexes too, which is the same as Perl. Note that the POSIX
2707 class [:space:] *does* include VT, thereby creating a mess.
2709 5. Added the class [:blank:] (a GNU extension from Perl 5.8) to match only
2710 space and tab.
2712 6. Perl 5.005 was a long time ago. It's time to amalgamate the tests that use
2713 its new features into the main test script, reducing the number of scripts.
2715 7. Perl 5.8 has changed the meaning of patterns like /a(?i)b/. Earlier versions
2716 were backward compatible, and made the (?i) apply to the whole pattern, as if
2717 /i were given. Now it behaves more logically, and applies the option setting
2718 only to what follows. PCRE has been changed to follow suit. However, if it
2719 finds options settings right at the start of the pattern, it extracts them into
2720 the global options, as before. Thus, they show up in the info data.
2722 8. Added support for the \Q...\E escape sequence. Characters in between are
2723 treated as literals. This is slightly different from Perl in that $ and @ are
2724 also handled as literals inside the quotes. In Perl, they will cause variable
2725 interpolation. Note the following examples:
2727 Pattern PCRE matches Perl matches
2729 \Qabc$xyz\E abc$xyz abc followed by the contents of $xyz
2730 \Qabc\$xyz\E abc\$xyz abc\$xyz
2731 \Qabc\E\$\Qxyz\E abc$xyz abc$xyz
2733 For compatibility with Perl, \Q...\E sequences are recognized inside character
2734 classes as well as outside them.
2736 9. Re-organized 3 code statements in pcretest to avoid "overflow in
2737 floating-point constant arithmetic" warnings from a Microsoft compiler. Added a
2738 (size_t) cast to one statement in pcretest and one in pcreposix to avoid
2739 signed/unsigned warnings.
2741 10. SunOS4 doesn't have strtoul(). This was used only for unpicking the -o
2742 option for pcretest, so I've replaced it by a simple function that does just
2743 that job.
2745 11. pcregrep was ending with code 0 instead of 2 for the commands "pcregrep" or
2746 "pcregrep -".
2748 12. Added "possessive quantifiers" ?+, *+, ++, and {,}+ which come from Sun's
2749 Java package. This provides some syntactic sugar for simple cases of what my
2750 documentation calls "once-only subpatterns". A pattern such as x*+ is the same
2751 as (?>x*). In other words, if what is inside (?>...) is just a single repeated
2752 item, you can use this simplified notation. Note that only makes sense with
2753 greedy quantifiers. Consequently, the use of the possessive quantifier forces
2754 greediness, whatever the setting of the PCRE_UNGREEDY option.
2756 13. A change of greediness default within a pattern was not taking effect at
2757 the current level for patterns like /(b+(?U)a+)/. It did apply to parenthesized
2758 subpatterns that followed. Patterns like /b+(?U)a+/ worked because the option
2759 was abstracted outside.
2761 14. PCRE now supports the \G assertion. It is true when the current matching
2762 position is at the start point of the match. This differs from \A when the
2763 starting offset is non-zero. Used with the /g option of pcretest (or similar
2764 code), it works in the same way as it does for Perl's /g option. If all
2765 alternatives of a regex begin with \G, the expression is anchored to the start
2766 match position, and the "anchored" flag is set in the compiled expression.
2768 15. Some bugs concerning the handling of certain option changes within patterns
2769 have been fixed. These applied to options other than (?ims). For example,
2770 "a(?x: b c )d" did not match "XabcdY" but did match "Xa b c dY". It should have
2771 been the other way round. Some of this was related to change 7 above.
2773 16. PCRE now gives errors for /[.x.]/ and /[=x=]/ as unsupported POSIX
2774 features, as Perl does. Previously, PCRE gave the warnings only for /[[.x.]]/
2775 and /[[=x=]]/. PCRE now also gives an error for /[:name:]/ because it supports
2776 POSIX classes only within a class (e.g. /[[:alpha:]]/).
2778 17. Added support for Perl's \C escape. This matches one byte, even in UTF8
2779 mode. Unlike ".", it always matches newline, whatever the setting of
2780 PCRE_DOTALL. However, PCRE does not permit \C to appear in lookbehind
2781 assertions. Perl allows it, but it doesn't (in general) work because it can't
2782 calculate the length of the lookbehind. At least, that's the case for Perl
2783 5.8.0 - I've been told they are going to document that it doesn't work in
2784 future.
2786 18. Added an error diagnosis for escapes that PCRE does not support: these are
2787 \L, \l, \N, \P, \p, \U, \u, and \X.
2789 19. Although correctly diagnosing a missing ']' in a character class, PCRE was
2790 reading past the end of the pattern in cases such as /[abcd/.
2792 20. PCRE was getting more memory than necessary for patterns with classes that
2793 contained both POSIX named classes and other characters, e.g. /[[:space:]abc/.
2795 21. Added some code, conditional on #ifdef VPCOMPAT, to make life easier for
2796 compiling PCRE for use with Virtual Pascal.
2798 22. Small fix to the Makefile to make it work properly if the build is done
2799 outside the source tree.
2801 23. Added a new extension: a condition to go with recursion. If a conditional
2802 subpattern starts with (?(R) the "true" branch is used if recursion has
2803 happened, whereas the "false" branch is used only at the top level.
2805 24. When there was a very long string of literal characters (over 255 bytes
2806 without UTF support, over 250 bytes with UTF support), the computation of how
2807 much memory was required could be incorrect, leading to segfaults or other
2808 strange effects.
2810 25. PCRE was incorrectly assuming anchoring (either to start of subject or to
2811 start of line for a non-DOTALL pattern) when a pattern started with (.*) and
2812 there was a subsequent back reference to those brackets. This meant that, for
2813 example, /(.*)\d+\1/ failed to match "abc123bc". Unfortunately, it isn't
2814 possible to check for precisely this case. All we can do is abandon the
2815 optimization if .* occurs inside capturing brackets when there are any back
2816 references whatsoever. (See below for a better fix that came later.)
2818 26. The handling of the optimization for finding the first character of a
2819 non-anchored pattern, and for finding a character that is required later in the
2820 match were failing in some cases. This didn't break the matching; it just
2821 failed to optimize when it could. The way this is done has been re-implemented.
2823 27. Fixed typo in error message for invalid (?R item (it said "(?p").
2825 28. Added a new feature that provides some of the functionality that Perl
2826 provides with (?{...}). The facility is termed a "callout". The way it is done
2827 in PCRE is for the caller to provide an optional function, by setting
2828 pcre_callout to its entry point. Like pcre_malloc and pcre_free, this is a
2829 global variable. By default it is unset, which disables all calling out. To get
2830 the function called, the regex must include (?C) at appropriate points. This
2831 is, in fact, equivalent to (?C0), and any number <= 255 may be given with (?C).
2832 This provides a means of identifying different callout points. When PCRE
2833 reaches such a point in the regex, if pcre_callout has been set, the external
2834 function is called. It is provided with data in a structure called
2835 pcre_callout_block, which is defined in pcre.h. If the function returns 0,
2836 matching continues; if it returns a non-zero value, the match at the current
2837 point fails. However, backtracking will occur if possible. [This was changed
2838 later and other features added - see item 49 below.]
2840 29. pcretest is upgraded to test the callout functionality. It provides a
2841 callout function that displays information. By default, it shows the start of
2842 the match and the current position in the text. There are some new data escapes
2843 to vary what happens:
2845 \C+ in addition, show current contents of captured substrings
2846 \C- do not supply a callout function
2847 \C!n return 1 when callout number n is reached
2848 \C!n!m return 1 when callout number n is reached for the mth time
2850 30. If pcregrep was called with the -l option and just a single file name, it
2851 output "<stdin>" if a match was found, instead of the file name.
2853 31. Improve the efficiency of the POSIX API to PCRE. If the number of capturing
2854 slots is less than POSIX_MALLOC_THRESHOLD, use a block on the stack to pass to
2855 pcre_exec(). This saves a malloc/free per call. The default value of
2856 POSIX_MALLOC_THRESHOLD is 10; it can be changed by --with-posix-malloc-threshold
2857 when configuring.
2859 32. The default maximum size of a compiled pattern is 64K. There have been a
2860 few cases of people hitting this limit. The code now uses macros to handle the
2861 storing of links as offsets within the compiled pattern. It defaults to 2-byte
2862 links, but this can be changed to 3 or 4 bytes by --with-link-size when
2863 configuring. Tests 2 and 5 work only with 2-byte links because they output
2864 debugging information about compiled patterns.
2866 33. Internal code re-arrangements:
2868 (a) Moved the debugging function for printing out a compiled regex into
2869 its own source file (printint.c) and used #include to pull it into
2870 pcretest.c and, when DEBUG is defined, into pcre.c, instead of having two
2871 separate copies.
2873 (b) Defined the list of op-code names for debugging as a macro in
2874 internal.h so that it is next to the definition of the opcodes.
2876 (c) Defined a table of op-code lengths for simpler skipping along compiled
2877 code. This is again a macro in internal.h so that it is next to the
2878 definition of the opcodes.
2880 34. Added support for recursive calls to individual subpatterns, along the
2881 lines of Robin Houston's patch (but implemented somewhat differently).
2883 35. Further mods to the Makefile to help Win32. Also, added code to pcregrep to
2884 allow it to read and process whole directories in Win32. This code was
2885 contributed by Lionel Fourquaux; it has not been tested by me.
2887 36. Added support for named subpatterns. The Python syntax (?P<name>...) is
2888 used to name a group. Names consist of alphanumerics and underscores, and must
2889 be unique. Back references use the syntax (?P=name) and recursive calls use
2890 (?P>name) which is a PCRE extension to the Python extension. Groups still have
2891 numbers. The function pcre_fullinfo() can be used after compilation to extract
2892 a name/number map. There are three relevant calls:
2894 PCRE_INFO_NAMEENTRYSIZE yields the size of each entry in the map
2895 PCRE_INFO_NAMECOUNT yields the number of entries
2896 PCRE_INFO_NAMETABLE yields a pointer to the map.
2898 The map is a vector of fixed-size entries. The size of each entry depends on
2899 the length of the longest name used. The first two bytes of each entry are the
2900 group number, most significant byte first. There follows the corresponding
2901 name, zero terminated. The names are in alphabetical order.
2903 37. Make the maximum literal string in the compiled code 250 for the non-UTF-8
2904 case instead of 255. Making it the same both with and without UTF-8 support
2905 means that the same test output works with both.
2907 38. There was a case of malloc(0) in the POSIX testing code in pcretest. Avoid
2908 calling malloc() with a zero argument.
2910 39. Change 25 above had to resort to a heavy-handed test for the .* anchoring
2911 optimization. I've improved things by keeping a bitmap of backreferences with
2912 numbers 1-31 so that if .* occurs inside capturing brackets that are not in
2913 fact referenced, the optimization can be applied. It is unlikely that a
2914 relevant occurrence of .* (i.e. one which might indicate anchoring or forcing
2915 the match to follow \n) will appear inside brackets with a number greater than
2916 31, but if it does, any back reference > 31 suppresses the optimization.
2918 40. Added a new compile-time option PCRE_NO_AUTO_CAPTURE. This has the effect
2919 of disabling numbered capturing parentheses. Any opening parenthesis that is
2920 not followed by ? behaves as if it were followed by ?: but named parentheses
2921 can still be used for capturing (and they will acquire numbers in the usual
2922 way).
2924 41. Redesigned the return codes from the match() function into yes/no/error so
2925 that errors can be passed back from deep inside the nested calls. A malloc
2926 failure while inside a recursive subpattern call now causes the
2927 PCRE_ERROR_NOMEMORY return instead of quietly going wrong.
2929 42. It is now possible to set a limit on the number of times the match()
2930 function is called in a call to pcre_exec(). This facility makes it possible to
2931 limit the amount of recursion and backtracking, though not in a directly
2932 obvious way, because the match() function is used in a number of different
2933 circumstances. The count starts from zero for each position in the subject
2934 string (for non-anchored patterns). The default limit is, for compatibility, a
2935 large number, namely 10 000 000. You can change this in two ways:
2937 (a) When configuring PCRE before making, you can use --with-match-limit=n
2938 to set a default value for the compiled library.
2940 (b) For each call to pcre_exec(), you can pass a pcre_extra block in which
2941 a different value is set. See 45 below.
2943 If the limit is exceeded, pcre_exec() returns PCRE_ERROR_MATCHLIMIT.
2945 43. Added a new function pcre_config(int, void *) to enable run-time extraction
2946 of things that can be changed at compile time. The first argument specifies
2947 what is wanted and the second points to where the information is to be placed.
2948 The current list of available information is:
2952 The output is an integer that is set to one if UTF-8 support is available;
2953 otherwise it is set to zero.
2957 The output is an integer that it set to the value of the code that is used for
2958 newline. It is either LF (10) or CR (13).
2962 The output is an integer that contains the number of bytes used for internal
2963 linkage in compiled expressions. The value is 2, 3, or 4. See item 32 above.
2967 The output is an integer that contains the threshold above which the POSIX
2968 interface uses malloc() for output vectors. See item 31 above.
2972 The output is an unsigned integer that contains the default limit of the number
2973 of match() calls in a pcre_exec() execution. See 42 above.
2975 44. pcretest has been upgraded by the addition of the -C option. This causes it
2976 to extract all the available output from the new pcre_config() function, and to
2977 output it. The program then exits immediately.
2979 45. A need has arisen to pass over additional data with calls to pcre_exec() in
2980 order to support additional features. One way would have been to define
2981 pcre_exec2() (for example) with extra arguments, but this would not have been
2982 extensible, and would also have required all calls to the original function to
2983 be mapped to the new one. Instead, I have chosen to extend the mechanism that
2984 is used for passing in "extra" data from pcre_study().
2986 The pcre_extra structure is now exposed and defined in pcre.h. It currently
2987 contains the following fields:
2989 flags a bitmap indicating which of the following fields are set
2990 study_data opaque data from pcre_study()
2991 match_limit a way of specifying a limit on match() calls for a specific
2992 call to pcre_exec()
2993 callout_data data for callouts (see 49 below)
2995 The flag bits are also defined in pcre.h, and are
3001 The pcre_study() function now returns one of these new pcre_extra blocks, with
3002 the actual study data pointed to by the study_data field, and the
3003 PCRE_EXTRA_STUDY_DATA flag set. This can be passed directly to pcre_exec() as
3004 before. That is, this change is entirely upwards-compatible and requires no
3005 change to existing code.
3007 If you want to pass in additional data to pcre_exec(), you can either place it
3008 in a pcre_extra block provided by pcre_study(), or create your own pcre_extra
3009 block.
3011 46. pcretest has been extended to test the PCRE_EXTRA_MATCH_LIMIT feature. If a
3012 data string contains the escape sequence \M, pcretest calls pcre_exec() several
3013 times with different match limits, until it finds the minimum value needed for
3014 pcre_exec() to complete. The value is then output. This can be instructive; for
3015 most simple matches the number is quite small, but for pathological cases it
3016 gets very large very quickly.
3018 47. There's a new option for pcre_fullinfo() called PCRE_INFO_STUDYSIZE. It
3019 returns the size of the data block pointed to by the study_data field in a
3020 pcre_extra block, that is, the value that was passed as the argument to
3021 pcre_malloc() when PCRE was getting memory in which to place the information
3022 created by pcre_study(). The fourth argument should point to a size_t variable.
3023 pcretest has been extended so that this information is shown after a successful
3024 pcre_study() call when information about the compiled regex is being displayed.
3026 48. Cosmetic change to Makefile: there's no need to have / after $(DESTDIR)
3027 because what follows is always an absolute path. (Later: it turns out that this
3028 is more than cosmetic for MinGW, because it doesn't like empty path
3029 components.)
3031 49. Some changes have been made to the callout feature (see 28 above):
3033 (i) A callout function now has three choices for what it returns:
3035 0 => success, carry on matching
3036 > 0 => failure at this point, but backtrack if possible
3037 < 0 => serious error, return this value from pcre_exec()
3039 Negative values should normally be chosen from the set of PCRE_ERROR_xxx
3040 values. In particular, returning PCRE_ERROR_NOMATCH forces a standard
3041 "match failed" error. The error number PCRE_ERROR_CALLOUT is reserved for
3042 use by callout functions. It will never be used by PCRE itself.
3044 (ii) The pcre_extra structure (see 45 above) has a void * field called
3045 callout_data, with corresponding flag bit PCRE_EXTRA_CALLOUT_DATA. The
3046 pcre_callout_block structure has a field of the same name. The contents of
3047 the field passed in the pcre_extra structure are passed to the callout
3048 function in the corresponding field in the callout block. This makes it
3049 easier to use the same callout-containing regex from multiple threads. For
3050 testing, the pcretest program has a new data escape
3052 \C*n pass the number n (may be negative) as callout_data
3054 If the callout function in pcretest receives a non-zero value as
3055 callout_data, it returns that value.
3057 50. Makefile wasn't handling CFLAGS properly when compiling dftables. Also,
3058 there were some redundant $(CFLAGS) in commands that are now specified as
3059 $(LINK), which already includes $(CFLAGS).
3061 51. Extensions to UTF-8 support are listed below. These all apply when (a) PCRE
3062 has been compiled with UTF-8 support *and* pcre_compile() has been compiled
3063 with the PCRE_UTF8 flag. Patterns that are compiled without that flag assume
3064 one-byte characters throughout. Note that case-insensitive matching applies
3065 only to characters whose values are less than 256. PCRE doesn't support the
3066 notion of cases for higher-valued characters.
3068 (i) A character class whose characters are all within 0-255 is handled as
3069 a bit map, and the map is inverted for negative classes. Previously, a
3070 character > 255 always failed to match such a class; however it should
3071 match if the class was a negative one (e.g. [^ab]). This has been fixed.
3073 (ii) A negated character class with a single character < 255 is coded as
3074 "not this character" (OP_NOT). This wasn't working properly when the test
3075 character was multibyte, either singly or repeated.
3077 (iii) Repeats of multibyte characters are now handled correctly in UTF-8
3078 mode, for example: \x{100}{2,3}.
3080 (iv) The character escapes \b, \B, \d, \D, \s, \S, \w, and \W (either
3081 singly or repeated) now correctly test multibyte characters. However,
3082 PCRE doesn't recognize any characters with values greater than 255 as
3083 digits, spaces, or word characters. Such characters always match \D, \S,
3084 and \W, and never match \d, \s, or \w.
3086 (v) Classes may now contain characters and character ranges with values
3087 greater than 255. For example: [ab\x{100}-\x{400}].
3089 (vi) pcregrep now has a --utf-8 option (synonym -u) which makes it call
3090 PCRE in UTF-8 mode.
3092 52. The info request value PCRE_INFO_FIRSTCHAR has been renamed
3093 PCRE_INFO_FIRSTBYTE because it is a byte value. However, the old name is
3094 retained for backwards compatibility. (Note that LASTLITERAL is also a byte
3095 value.)
3097 53. The single man page has become too large. I have therefore split it up into
3098 a number of separate man pages. These also give rise to individual HTML pages;
3099 these are now put in a separate directory, and there is an index.html page that
3100 lists them all. Some hyperlinking between the pages has been installed.
3102 54. Added convenience functions for handling named capturing parentheses.
3104 55. Unknown escapes inside character classes (e.g. [\M]) and escapes that
3105 aren't interpreted therein (e.g. [\C]) are literals in Perl. This is now also
3106 true in PCRE, except when the PCRE_EXTENDED option is set, in which case they
3107 are faulted.
3109 56. Introduced HOST_CC and HOST_CFLAGS which can be set in the environment when
3110 calling configure. These values are used when compiling the dftables.c program
3111 which is run to generate the source of the default character tables. They
3112 default to the values of CC and CFLAGS. If you are cross-compiling PCRE,
3113 you will need to set these values.
3115 57. Updated the building process for Windows DLL, as provided by Fred Cox.
3118 Version 3.9 02-Jan-02
3119 ---------------------
3121 1. A bit of extraneous text had somehow crept into the pcregrep documentation.
3123 2. If --disable-static was given, the building process failed when trying to
3124 build pcretest and pcregrep. (For some reason it was using libtool to compile
3125 them, which is not right, as they aren't part of the library.)
3128 Version 3.8 18-Dec-01
3129 ---------------------
3131 1. The experimental UTF-8 code was completely screwed up. It was packing the
3132 bytes in the wrong order. How dumb can you get?
3135 Version 3.7 29-Oct-01
3136 ---------------------
3138 1. In updating pcretest to check change 1 of version 3.6, I screwed up.
3139 This caused pcretest, when used on the test data, to segfault. Unfortunately,
3140 this didn't happen under Solaris 8, where I normally test things.
3142 2. The Makefile had to be changed to make it work on BSD systems, where 'make'
3143 doesn't seem to recognize that ./xxx and xxx are the same file. (This entry
3144 isn't in ChangeLog distributed with 3.7 because I forgot when I hastily made
3145 this fix an hour or so after the initial 3.7 release.)
3148 Version 3.6 23-Oct-01
3149 ---------------------
3151 1. Crashed with /(sens|respons)e and \1ibility/ and "sense and sensibility" if
3152 offsets passed as NULL with zero offset count.
3154 2. The config.guess and config.sub files had not been updated when I moved to
3155 the latest autoconf.
3158 Version 3.5 15-Aug-01
3159 ---------------------
3161 1. Added some missing #if !defined NOPOSIX conditionals in pcretest.c that
3162 had been forgotten.
3164 2. By using declared but undefined structures, we can avoid using "void"
3165 definitions in pcre.h while keeping the internal definitions of the structures
3166 private.
3168 3. The distribution is now built using autoconf 2.50 and libtool 1.4. From a
3169 user point of view, this means that both static and shared libraries are built
3170 by default, but this can be individually controlled. More of the work of
3171 handling this static/shared cases is now inside libtool instead of PCRE's make
3172 file.
3174 4. The pcretest utility is now installed along with pcregrep because it is
3175 useful for users (to test regexs) and by doing this, it automatically gets
3176 relinked by libtool. The documentation has been turned into a man page, so
3177 there are now .1, .txt, and .html versions in /doc.
3179 5. Upgrades to pcregrep:
3180 (i) Added long-form option names like gnu grep.
3181 (ii) Added --help to list all options with an explanatory phrase.
3182 (iii) Added -r, --recursive to recurse into sub-directories.
3183 (iv) Added -f, --file to read patterns from a file.
3185 6. pcre_exec() was referring to its "code" argument before testing that
3186 argument for NULL (and giving an error if it was NULL).
3188 7. Upgraded Makefile.in to allow for compiling in a different directory from
3189 the source directory.
3191 8. Tiny buglet in pcretest: when pcre_fullinfo() was called to retrieve the
3192 options bits, the pointer it was passed was to an int instead of to an unsigned
3193 long int. This mattered only on 64-bit systems.
3195 9. Fixed typo (3.4/1) in pcre.h again. Sigh. I had changed pcre.h (which is
3196 generated) instead of pcre.in, which it its source. Also made the same change
3197 in several of the .c files.
3199 10. A new release of gcc defines printf() as a macro, which broke pcretest
3200 because it had an ifdef in the middle of a string argument for printf(). Fixed
3201 by using separate calls to printf().
3203 11. Added --enable-newline-is-cr and --enable-newline-is-lf to the configure
3204 script, to force use of CR or LF instead of \n in the source. On non-Unix
3205 systems, the value can be set in config.h.
3207 12. The limit of 200 on non-capturing parentheses is a _nesting_ limit, not an
3208 absolute limit. Changed the text of the error message to make this clear, and
3209 likewise updated the man page.
3211 13. The limit of 99 on the number of capturing subpatterns has been removed.
3212 The new limit is 65535, which I hope will not be a "real" limit.
3215 Version 3.4 22-Aug-00
3216 ---------------------
3218 1. Fixed typo in pcre.h: unsigned const char * changed to const unsigned char *.
3220 2. Diagnose condition (?(0) as an error instead of crashing on matching.
3223 Version 3.3 01-Aug-00
3224 ---------------------
3226 1. If an octal character was given, but the value was greater than \377, it
3227 was not getting masked to the least significant bits, as documented. This could
3228 lead to crashes in some systems.
3230 2. Perl 5.6 (if not earlier versions) accepts classes like [a-\d] and treats
3231 the hyphen as a literal. PCRE used to give an error; it now behaves like Perl.
3233 3. Added the functions pcre_free_substring() and pcre_free_substring_list().
3234 These just pass their arguments on to (pcre_free)(), but they are provided
3235 because some uses of PCRE bind it to non-C systems that can call its functions,
3236 but cannot call free() or pcre_free() directly.
3238 4. Add "make test" as a synonym for "make check". Corrected some comments in
3239 the Makefile.
3241 5. Add $(DESTDIR)/ in front of all the paths in the "install" target in the
3242 Makefile.
3244 6. Changed the name of pgrep to pcregrep, because Solaris has introduced a
3245 command called pgrep for grepping around the active processes.
3247 7. Added the beginnings of support for UTF-8 character strings.
3249 8. Arranged for the Makefile to pass over the settings of CC, CFLAGS, and
3250 RANLIB to ./ltconfig so that they are used by libtool. I think these are all
3251 the relevant ones. (AR is not passed because ./ltconfig does its own figuring
3252 out for the ar command.)
3255 Version 3.2 12-May-00
3256 ---------------------
3258 This is purely a bug fixing release.
3260 1. If the pattern /((Z)+|A)*/ was matched agained ZABCDEFG it matched Z instead
3261 of ZA. This was just one example of several cases that could provoke this bug,
3262 which was introduced by change 9 of version 2.00. The code for breaking
3263 infinite loops after an iteration that matches an empty string was't working
3264 correctly.
3266 2. The pcretest program was not imitating Perl correctly for the pattern /a*/g
3267 when matched against abbab (for example). After matching an empty string, it
3268 wasn't forcing anchoring when setting PCRE_NOTEMPTY for the next attempt; this
3269 caused it to match further down the string than it should.
3271 3. The code contained an inclusion of sys/types.h. It isn't clear why this
3272 was there because it doesn't seem to be needed, and it causes trouble on some
3273 systems, as it is not a Standard C header. It has been removed.
3275 4. Made 4 silly changes to the source to avoid stupid compiler warnings that
3276 were reported on the Macintosh. The changes were from
3278 while ((c = *(++ptr)) != 0 && c != '\n');
3279 to
3280 while ((c = *(++ptr)) != 0 && c != '\n') ;
3282 Totally extraordinary, but if that's what it takes...
3284 5. PCRE is being used in one environment where neither memmove() nor bcopy() is
3285 available. Added HAVE_BCOPY and an autoconf test for it; if neither
3286 HAVE_MEMMOVE nor HAVE_BCOPY is set, use a built-in emulation function which
3287 assumes the way PCRE uses memmove() (always moving upwards).
3289 6. PCRE is being used in one environment where strchr() is not available. There
3290 was only one use in pcre.c, and writing it out to avoid strchr() probably gives
3291 faster code anyway.
3294 Version 3.1 09-Feb-00
3295 ---------------------
3297 The only change in this release is the fixing of some bugs in Makefile.in for
3298 the "install" target:
3300 (1) It was failing to install pcreposix.h.
3302 (2) It was overwriting the pcre.3 man page with the pcreposix.3 man page.
3305 Version 3.0 01-Feb-00
3306 ---------------------
3308 1. Add support for the /+ modifier to perltest (to output $` like it does in
3309 pcretest).
3311 2. Add support for the /g modifier to perltest.
3313 3. Fix pcretest so that it behaves even more like Perl for /g when the pattern
3314 matches null strings.
3316 4. Fix perltest so that it doesn't do unwanted things when fed an empty
3317 pattern. Perl treats empty patterns specially - it reuses the most recent
3318 pattern, which is not what we want. Replace // by /(?#)/ in order to avoid this
3319 effect.
3321 5. The POSIX interface was broken in that it was just handing over the POSIX
3322 captured string vector to pcre_exec(), but (since release 2.00) PCRE has
3323 required a bigger vector, with some working space on the end. This means that
3324 the POSIX wrapper now has to get and free some memory, and copy the results.
3326 6. Added some simple autoconf support, placing the test data and the
3327 documentation in separate directories, re-organizing some of the
3328 information files, and making it build pcre-config (a GNU standard). Also added
3329 libtool support for building PCRE as a shared library, which is now the
3330 default.
3332 7. Got rid of the leading zero in the definition of PCRE_MINOR because 08 and
3333 09 are not valid octal constants. Single digits will be used for minor values
3334 less than 10.
3336 8. Defined REG_EXTENDED and REG_NOSUB as zero in the POSIX header, so that
3337 existing programs that set these in the POSIX interface can use PCRE without
3338 modification.
3340 9. Added a new function, pcre_fullinfo() with an extensible interface. It can
3341 return all that pcre_info() returns, plus additional data. The pcre_info()
3342 function is retained for compatibility, but is considered to be obsolete.
3344 10. Added experimental recursion feature (?R) to handle one common case that
3345 Perl 5.6 will be able to do with (?p{...}).
3347 11. Added support for POSIX character classes like [:alpha:], which Perl is
3348 adopting.
3351 Version 2.08 31-Aug-99
3352 ----------------------
3354 1. When startoffset was not zero and the pattern began with ".*", PCRE was not
3355 trying to match at the startoffset position, but instead was moving forward to
3356 the next newline as if a previous match had failed.
3358 2. pcretest was not making use of PCRE_NOTEMPTY when repeating for /g and /G,
3359 and could get into a loop if a null string was matched other than at the start
3360 of the subject.
3362 3. Added definitions of PCRE_MAJOR and PCRE_MINOR to pcre.h so the version can
3363 be distinguished at compile time, and for completeness also added PCRE_DATE.
3365 5. Added Paul Sokolovsky's minor changes to make it easy to compile a Win32 DLL
3366 in GnuWin32 environments.
3369 Version 2.07 29-Jul-99
3370 ----------------------
3372 1. The documentation is now supplied in plain text form and HTML as well as in
3373 the form of man page sources.
3375 2. C++ compilers don't like assigning (void *) values to other pointer types.
3376 In particular this affects malloc(). Although there is no problem in Standard
3377 C, I've put in casts to keep C++ compilers happy.
3379 3. Typo on pcretest.c; a cast of (unsigned char *) in the POSIX regexec() call
3380 should be (const char *).
3382 4. If NOPOSIX is defined, pcretest.c compiles without POSIX support. This may
3383 be useful for non-Unix systems who don't want to bother with the POSIX stuff.
3384 However, I haven't made this a standard facility. The documentation doesn't
3385 mention it, and the Makefile doesn't support it.
3387 5. The Makefile now contains an "install" target, with editable destinations at
3388 the top of the file. The pcretest program is not installed.
3390 6. pgrep -V now gives the PCRE version number and date.
3392 7. Fixed bug: a zero repetition after a literal string (e.g. /abcde{0}/) was
3393 causing the entire string to be ignored, instead of just the last character.
3395 8. If a pattern like /"([^\\"]+|\\.)*"/ is applied in the normal way to a
3396 non-matching string, it can take a very, very long time, even for strings of
3397 quite modest length, because of the nested recursion. PCRE now does better in
3398 some of these cases. It does this by remembering the last required literal
3399 character in the pattern, and pre-searching the subject to ensure it is present
3400 before running the real match. In other words, it applies a heuristic to detect
3401 some types of certain failure quickly, and in the above example, if presented
3402 with a string that has no trailing " it gives "no match" very quickly.
3404 9. A new runtime option PCRE_NOTEMPTY causes null string matches to be ignored;
3405 other alternatives are tried instead.
3408 Version 2.06 09-Jun-99
3409 ----------------------
3411 1. Change pcretest's output for amount of store used to show just the code
3412 space, because the remainder (the data block) varies in size between 32-bit and
3413 64-bit systems.
3415 2. Added an extra argument to pcre_exec() to supply an offset in the subject to
3416 start matching at. This allows lookbehinds to work when searching for multiple
3417 occurrences in a string.
3419 3. Added additional options to pcretest for testing multiple occurrences:
3421 /+ outputs the rest of the string that follows a match
3422 /g loops for multiple occurrences, using the new startoffset argument
3423 /G loops for multiple occurrences by passing an incremented pointer
3425 4. PCRE wasn't doing the "first character" optimization for patterns starting
3426 with \b or \B, though it was doing it for other lookbehind assertions. That is,
3427 it wasn't noticing that a match for a pattern such as /\bxyz/ has to start with
3428 the letter 'x'. On long subject strings, this gives a significant speed-up.
3431 Version 2.05 21-Apr-99
3432 ----------------------
3434 1. Changed the type of magic_number from int to long int so that it works
3435 properly on 16-bit systems.
3437 2. Fixed a bug which caused patterns starting with .* not to work correctly
3438 when the subject string contained newline characters. PCRE was assuming
3439 anchoring for such patterns in all cases, which is not correct because .* will
3440 not pass a newline unless PCRE_DOTALL is set. It now assumes anchoring only if
3441 DOTALL is set at top level; otherwise it knows that patterns starting with .*
3442 must be retried after every newline in the subject.
3445 Version 2.04 18-Feb-99
3446 ----------------------
3448 1. For parenthesized subpatterns with repeats whose minimum was zero, the
3449 computation of the store needed to hold the pattern was incorrect (too large).
3450 If such patterns were nested a few deep, this could multiply and become a real
3451 problem.
3453 2. Added /M option to pcretest to show the memory requirement of a specific
3454 pattern. Made -m a synonym of -s (which does this globally) for compatibility.
3456 3. Subpatterns of the form (regex){n,m} (i.e. limited maximum) were being
3457 compiled in such a way that the backtracking after subsequent failure was
3458 pessimal. Something like (a){0,3} was compiled as (a)?(a)?(a)? instead of
3459 ((a)((a)(a)?)?)? with disastrous performance if the maximum was of any size.
3462 Version 2.03 02-Feb-99
3463 ----------------------
3465 1. Fixed typo and small mistake in man page.
3467 2. Added 4th condition (GPL supersedes if conflict) and created separate
3468 LICENCE file containing the conditions.
3470 3. Updated pcretest so that patterns such as /abc\/def/ work like they do in
3471 Perl, that is the internal \ allows the delimiter to be included in the
3472 pattern. Locked out the use of \ as a delimiter. If \ immediately follows
3473 the final delimiter, add \ to the end of the pattern (to test the error).
3475 4. Added the convenience functions for extracting substrings after a successful
3476 match. Updated pcretest to make it able to test these functions.
3479 Version 2.02 14-Jan-99
3480 ----------------------
3482 1. Initialized the working variables associated with each extraction so that
3483 their saving and restoring doesn't refer to uninitialized store.
3485 2. Put dummy code into study.c in order to trick the optimizer of the IBM C
3486 compiler for OS/2 into generating correct code. Apparently IBM isn't going to
3487 fix the problem.
3489 3. Pcretest: the timing code wasn't using LOOPREPEAT for timing execution
3490 calls, and wasn't printing the correct value for compiling calls. Increased the
3491 default value of LOOPREPEAT, and the number of significant figures in the
3492 times.
3494 4. Changed "/bin/rm" in the Makefile to "-rm" so it works on Windows NT.
3496 5. Renamed "deftables" as "dftables" to get it down to 8 characters, to avoid
3497 a building problem on Windows NT with a FAT file system.
3500 Version 2.01 21-Oct-98
3501 ----------------------
3503 1. Changed the API for pcre_compile() to allow for the provision of a pointer
3504 to character tables built by pcre_maketables() in the current locale. If NULL
3505 is passed, the default tables are used.
3508 Version 2.00 24-Sep-98
3509 ----------------------
3511 1. Since the (>?) facility is in Perl 5.005, don't require PCRE_EXTRA to enable
3512 it any more.
3514 2. Allow quantification of (?>) groups, and make it work correctly.
3516 3. The first character computation wasn't working for (?>) groups.
3518 4. Correct the implementation of \Z (it is permitted to match on the \n at the
3519 end of the subject) and add 5.005's \z, which really does match only at the
3520 very end of the subject.
3522 5. Remove the \X "cut" facility; Perl doesn't have it, and (?> is neater.
3524 6. Remove the ability to specify CASELESS, MULTILINE, DOTALL, and
3525 DOLLAR_END_ONLY at runtime, to make it possible to implement the Perl 5.005
3526 localized options. All options to pcre_study() were also removed.
3528 7. Add other new features from 5.005:
3530 $(?<= positive lookbehind
3531 $(?<! negative lookbehind
3532 (?imsx-imsx) added the unsetting capability
3533 such a setting is global if at outer level; local otherwise
3534 (?imsx-imsx:) non-capturing groups with option setting
3535 (?(cond)re|re) conditional pattern matching
3537 A backreference to itself in a repeated group matches the previous
3538 captured string.
3540 8. General tidying up of studying (both automatic and via "study")
3541 consequential on the addition of new assertions.
3543 9. As in 5.005, unlimited repeated groups that could match an empty substring
3544 are no longer faulted at compile time. Instead, the loop is forcibly broken at
3545 runtime if any iteration does actually match an empty substring.
3547 10. Include the RunTest script in the distribution.
3549 11. Added tests from the Perl 5.005_02 distribution. This showed up a few
3550 discrepancies, some of which were old and were also with respect to 5.004. They
3551 have now been fixed.
3554 Version 1.09 28-Apr-98
3555 ----------------------
3557 1. A negated single character class followed by a quantifier with a minimum
3558 value of one (e.g. [^x]{1,6} ) was not compiled correctly. This could lead to
3559 program crashes, or just wrong answers. This did not apply to negated classes
3560 containing more than one character, or to minima other than one.
3563 Version 1.08 27-Mar-98
3564 ----------------------
3566 1. Add PCRE_UNGREEDY to invert the greediness of quantifiers.
3568 2. Add (?U) and (?X) to set PCRE_UNGREEDY and PCRE_EXTRA respectively. The
3569 latter must appear before anything that relies on it in the pattern.
3572 Version 1.07 16-Feb-98
3573 ----------------------
3575 1. A pattern such as /((a)*)*/ was not being diagnosed as in error (unlimited
3576 repeat of a potentially empty string).
3579 Version 1.06 23-Jan-98
3580 ----------------------
3582 1. Added Markus Oberhumer's little patches for C++.
3584 2. Literal strings longer than 255 characters were broken.
3587 Version 1.05 23-Dec-97
3588 ----------------------
3590 1. Negated character classes containing more than one character were failing if
3591 PCRE_CASELESS was set at run time.
3594 Version 1.04 19-Dec-97
3595 ----------------------
3597 1. Corrected the man page, where some "const" qualifiers had been omitted.
3599 2. Made debugging output print "{0,xxx}" instead of just "{,xxx}" to agree with
3600 input syntax.
3602 3. Fixed memory leak which occurred when a regex with back references was
3603 matched with an offsets vector that wasn't big enough. The temporary memory
3604 that is used in this case wasn't being freed if the match failed.
3606 4. Tidied pcretest to ensure it frees memory that it gets.
3608 5. Temporary memory was being obtained in the case where the passed offsets
3609 vector was exactly big enough.
3611 6. Corrected definition of offsetof() from change 5 below.
3613 7. I had screwed up change 6 below and broken the rules for the use of
3614 setjmp(). Now fixed.
3617 Version 1.03 18-Dec-97
3618 ----------------------
3620 1. A erroneous regex with a missing opening parenthesis was correctly
3621 diagnosed, but PCRE attempted to access brastack[-1], which could cause crashes
3622 on some systems.
3624 2. Replaced offsetof(real_pcre, code) by offsetof(real_pcre, code[0]) because
3625 it was reported that one broken compiler failed on the former because "code" is
3626 also an independent variable.
3628 3. The erroneous regex a[]b caused an array overrun reference.
3630 4. A regex ending with a one-character negative class (e.g. /[^k]$/) did not
3631 fail on data ending with that character. (It was going on too far, and checking
3632 the next character, typically a binary zero.) This was specific to the
3633 optimized code for single-character negative classes.
3635 5. Added a contributed patch from the TIN world which does the following:
3637 + Add an undef for memmove, in case the the system defines a macro for it.
3639 + Add a definition of offsetof(), in case there isn't one. (I don't know
3640 the reason behind this - offsetof() is part of the ANSI standard - but
3641 it does no harm).
3643 + Reduce the ifdef's in pcre.c using macro DPRINTF, thereby eliminating
3644 most of the places where whitespace preceded '#'. I have given up and
3645 allowed the remaining 2 cases to be at the margin.
3647 + Rename some variables in pcre to eliminate shadowing. This seems very
3648 pedantic, but does no harm, of course.
3650 6. Moved the call to setjmp() into its own function, to get rid of warnings
3651 from gcc -Wall, and avoided calling it at all unless PCRE_EXTRA is used.
3653 7. Constructs such as \d{8,} were compiling into the equivalent of
3654 \d{8}\d{0,65527} instead of \d{8}\d* which didn't make much difference to the
3655 outcome, but in this particular case used more store than had been allocated,
3656 which caused the bug to be discovered because it threw up an internal error.
3658 8. The debugging code in both pcre and pcretest for outputting the compiled
3659 form of a regex was going wrong in the case of back references followed by
3660 curly-bracketed repeats.
3663 Version 1.02 12-Dec-97
3664 ----------------------
3666 1. Typos in pcre.3 and comments in the source fixed.
3668 2. Applied a contributed patch to get rid of places where it used to remove
3669 'const' from variables, and fixed some signed/unsigned and uninitialized
3670 variable warnings.
3672 3. Added the "runtest" target to Makefile.
3674 4. Set default compiler flag to -O2 rather than just -O.
3677 Version 1.01 19-Nov-97
3678 ----------------------
3680 1. PCRE was failing to diagnose unlimited repeat of empty string for patterns
3681 like /([ab]*)*/, that is, for classes with more than one character in them.
3683 2. Likewise, it wasn't diagnosing patterns with "once-only" subpatterns, such
3684 as /((?>a*))*/ (a PCRE_EXTRA facility).
3687 Version 1.00 18-Nov-97
3688 ----------------------
3690 1. Added compile-time macros to support systems such as SunOS4 which don't have
3691 memmove() or strerror() but have other things that can be used instead.
3693 2. Arranged that "make clean" removes the executables.
3696 Version 0.99 27-Oct-97
3697 ----------------------
3699 1. Fixed bug in code for optimizing classes with only one character. It was
3700 initializing a 32-byte map regardless, which could cause it to run off the end
3701 of the memory it had got.
3703 2. Added, conditional on PCRE_EXTRA, the proposed (?>REGEX) construction.
3706 Version 0.98 22-Oct-97
3707 ----------------------
3709 1. Fixed bug in code for handling temporary memory usage when there are more
3710 back references than supplied space in the ovector. This could cause segfaults.
3713 Version 0.97 21-Oct-97
3714 ----------------------
3716 1. Added the \X "cut" facility, conditional on PCRE_EXTRA.
3718 2. Optimized negated single characters not to use a bit map.
3720 3. Brought error texts together as macro definitions; clarified some of them;
3721 fixed one that was wrong - it said "range out of order" when it meant "invalid
3722 escape sequence".
3724 4. Changed some char * arguments to const char *.
3726 5. Added PCRE_NOTBOL and PCRE_NOTEOL (from POSIX).
3728 6. Added the POSIX-style API wrapper in pcreposix.a and testing facilities in
3729 pcretest.
3732 Version 0.96 16-Oct-97
3733 ----------------------
3735 1. Added a simple "pgrep" utility to the distribution.
3737 2. Fixed an incompatibility with Perl: "{" is now treated as a normal character
3738 unless it appears in one of the precise forms "{ddd}", "{ddd,}", or "{ddd,ddd}"
3739 where "ddd" means "one or more decimal digits".
3741 3. Fixed serious bug. If a pattern had a back reference, but the call to
3742 pcre_exec() didn't supply a large enough ovector to record the related
3743 identifying subpattern, the match always failed. PCRE now remembers the number
3744 of the largest back reference, and gets some temporary memory in which to save
3745 the offsets during matching if necessary, in order to ensure that
3746 backreferences always work.
3748 4. Increased the compatibility with Perl in a number of ways:
3750 (a) . no longer matches \n by default; an option PCRE_DOTALL is provided
3751 to request this handling. The option can be set at compile or exec time.
3753 (b) $ matches before a terminating newline by default; an option
3754 PCRE_DOLLAR_ENDONLY is provided to override this (but not in multiline
3755 mode). The option can be set at compile or exec time.
3757 (c) The handling of \ followed by a digit other than 0 is now supposed to be
3758 the same as Perl's. If the decimal number it represents is less than 10
3759 or there aren't that many previous left capturing parentheses, an octal
3760 escape is read. Inside a character class, it's always an octal escape,
3761 even if it is a single digit.
3763 (d) An escaped but undefined alphabetic character is taken as a literal,
3764 unless PCRE_EXTRA is set. Currently this just reserves the remaining
3765 escapes.
3767 (e) {0} is now permitted. (The previous item is removed from the compiled
3768 pattern).
3770 5. Changed all the names of code files so that the basic parts are no longer
3771 than 10 characters, and abolished the teeny "globals.c" file.
3773 6. Changed the handling of character classes; they are now done with a 32-byte
3774 bit map always.
3776 7. Added the -d and /D options to pcretest to make it possible to look at the
3777 internals of compilation without having to recompile pcre.
3780 Version 0.95 23-Sep-97
3781 ----------------------
3783 1. Fixed bug in pre-pass concerning escaped "normal" characters such as \x5c or
3784 \x20 at the start of a run of normal characters. These were being treated as
3785 real characters, instead of the source characters being re-checked.
3788 Version 0.94 18-Sep-97
3789 ----------------------
3791 1. The functions are now thread-safe, with the caveat that the global variables
3792 containing pointers to malloc() and free() or alternative functions are the
3793 same for all threads.
3795 2. Get pcre_study() to generate a bitmap of initial characters for non-
3796 anchored patterns when this is possible, and use it if passed to pcre_exec().
3799 Version 0.93 15-Sep-97
3800 ----------------------
3802 1. /(b)|(:+)/ was computing an incorrect first character.
3804 2. Add pcre_study() to the API and the passing of pcre_extra to pcre_exec(),
3805 but not actually doing anything yet.
3807 3. Treat "-" characters in classes that cannot be part of ranges as literals,
3808 as Perl does (e.g. [-az] or [az-]).
3810 4. Set the anchored flag if a branch starts with .* or .*? because that tests
3811 all possible positions.
3813 5. Split up into different modules to avoid including unneeded functions in a
3814 compiled binary. However, compile and exec are still in one module. The "study"
3815 function is split off.
3817 6. The character tables are now in a separate module whose source is generated
3818 by an auxiliary program - but can then be edited by hand if required. There are
3819 now no calls to isalnum(), isspace(), isdigit(), isxdigit(), tolower() or
3820 toupper() in the code.
3822 7. Turn the malloc/free funtions variables into pcre_malloc and pcre_free and
3823 make them global. Abolish the function for setting them, as the caller can now
3824 set them directly.
3827 Version 0.92 11-Sep-97
3828 ----------------------
3830 1. A repeat with a fixed maximum and a minimum of 1 for an ordinary character
3831 (e.g. /a{1,3}/) was broken (I mis-optimized it).
3833 2. Caseless matching was not working in character classes if the characters in
3834 the pattern were in upper case.
3836 3. Make ranges like [W-c] work in the same way as Perl for caseless matching.
3838 4. Make PCRE_ANCHORED public and accept as a compile option.
3840 5. Add an options word to pcre_exec() and accept PCRE_ANCHORED and
3841 PCRE_CASELESS at run time. Add escapes \A and \I to pcretest to cause it to
3842 pass them.
3844 6. Give an error if bad option bits passed at compile or run time.
3846 7. Add PCRE_MULTILINE at compile and exec time, and (?m) as well. Add \M to
3847 pcretest to cause it to pass that flag.
3849 8. Add pcre_info(), to get the number of identifying subpatterns, the stored
3850 options, and the first character, if set.
3852 9. Recognize C+ or C{n,m} where n >= 1 as providing a fixed starting character.
3855 Version 0.91 10-Sep-97
3856 ----------------------
3858 1. PCRE was failing to diagnose unlimited repeats of subpatterns that could
3859 match the empty string as in /(a*)*/. It was looping and ultimately crashing.
3861 2. PCRE was looping on encountering an indefinitely repeated back reference to
3862 a subpattern that had matched an empty string, e.g. /(a|)\1*/. It now does what
3863 Perl does - treats the match as successful.
3865 ****


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

  ViewVC Help
Powered by ViewVC 1.1.5