ViewVC logotype

Contents of /code/trunk/ChangeLog

Parent Directory Parent Directory | Revision Log Revision Log

Revision 498 - (show annotations)
Wed Mar 3 19:29:38 2010 UTC (10 years, 10 months ago) by ph10
File size: 182903 byte(s)
Fix missing data in coptable and poptable vectors; add compile-time checks for 
their lengths.
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).
34 Version 8.01 19-Jan-2010
35 ------------------------
37 1. If a pattern contained a conditional subpattern with only one branch (in
38 particular, this includes all (*DEFINE) patterns), a call to pcre_study()
39 computed the wrong minimum data length (which is of course zero for such
40 subpatterns). This could cause incorrect "no match" results.
42 2. For patterns such as (?i)a(?-i)b|c where an option setting at the start of
43 the pattern is reset in the first branch, pcre_compile() failed with
44 "internal error: code overflow at offset...". This happened only when
45 the reset was to the original external option setting. (An optimization
46 abstracts leading options settings into an external setting, which was the
47 cause of this.)
49 3. A pattern such as ^(?!a(*SKIP)b) where a negative assertion contained one
50 of the verbs SKIP, PRUNE, or COMMIT, did not work correctly. When the
51 assertion pattern did not match (meaning that the assertion was true), it
52 was incorrectly treated as false if the SKIP had been reached during the
53 matching. This also applied to assertions used as conditions.
55 4. If an item that is not supported by pcre_dfa_exec() was encountered in an
56 assertion subpattern, including such a pattern used as a condition,
57 unpredictable results occurred, instead of the error return
60 5. The C++ GlobalReplace function was not working like Perl for the special
61 situation when an empty string is matched. It now does the fancy magic
62 stuff that is necessary.
64 6. In pcre_internal.h, obsolete includes to setjmp.h and stdarg.h have been
65 removed. (These were left over from very, very early versions of PCRE.)
67 7. Some cosmetic changes to the code to make life easier when compiling it
68 as part of something else:
70 (a) Change DEBUG to PCRE_DEBUG.
72 (b) In pcre_compile(), rename the member of the "branch_chain" structure
73 called "current" as "current_branch", to prevent a collision with the
74 Linux macro when compiled as a kernel module.
76 (c) In pcre_study(), rename the function set_bit() as set_table_bit(), to
77 prevent a collision with the Linux macro when compiled as a kernel
78 module.
80 8. In pcre_compile() there are some checks for integer overflows that used to
81 cast potentially large values to (double). This has been changed to that
82 when building, a check for int64_t is made, and if it is found, it is used
83 instead, thus avoiding the use of floating point arithmetic. (There is no
84 other use of FP in PCRE.) If int64_t is not found, the fallback is to
85 double.
87 9. Added two casts to avoid signed/unsigned warnings from VS Studio Express
88 2005 (difference between two addresses compared to an unsigned value).
90 10. Change the standard AC_CHECK_LIB test for libbz2 in configure.ac to a
91 custom one, because of the following reported problem in Windows:
93 - libbz2 uses the Pascal calling convention (WINAPI) for the functions
94 under Win32.
95 - The standard autoconf AC_CHECK_LIB fails to include "bzlib.h",
96 therefore missing the function definition.
97 - The compiler thus generates a "C" signature for the test function.
98 - The linker fails to find the "C" function.
99 - PCRE fails to configure if asked to do so against libbz2.
101 11. When running libtoolize from libtool-2.2.6b as part of autogen.sh, these
102 messages were output:
104 Consider adding `AC_CONFIG_MACRO_DIR([m4])' to configure.ac and
105 rerunning libtoolize, to keep the correct libtool macros in-tree.
106 Consider adding `-I m4' to ACLOCAL_AMFLAGS in Makefile.am.
108 I have done both of these things.
110 12. Although pcre_dfa_exec() does not use nearly as much stack as pcre_exec()
111 most of the time, it *can* run out if it is given a pattern that contains a
112 runaway infinite recursion. I updated the discussion in the pcrestack man
113 page.
115 13. Now that we have gone to the x.xx style of version numbers, the minor
116 version may start with zero. Using 08 or 09 is a bad idea because users
117 might check the value of PCRE_MINOR in their code, and 08 or 09 may be
118 interpreted as invalid octal numbers. I've updated the previous comment in
119 configure.ac, and also added a check that gives an error if 08 or 09 are
120 used.
122 14. Change 8.00/11 was not quite complete: code had been accidentally omitted,
123 causing partial matching to fail when the end of the subject matched \W
124 in a UTF-8 pattern where \W was quantified with a minimum of 3.
126 15. There were some discrepancies between the declarations in pcre_internal.h
127 of _pcre_is_newline(), _pcre_was_newline(), and _pcre_valid_utf8() and
128 their definitions. The declarations used "const uschar *" and the
129 definitions used USPTR. Even though USPTR is normally defined as "const
130 unsigned char *" (and uschar is typedeffed as "unsigned char"), it was
131 reported that: "This difference in casting confuses some C++ compilers, for
132 example, SunCC recognizes above declarations as different functions and
133 generates broken code for hbpcre." I have changed the declarations to use
134 USPTR.
136 16. GNU libtool is named differently on some systems. The autogen.sh script now
137 tries several variants such as glibtoolize (MacOSX) and libtoolize1x
138 (FreeBSD).
140 17. Applied Craig's patch that fixes an HP aCC compile error in pcre 8.00
141 (strtoXX undefined when compiling pcrecpp.cc). The patch contains this
142 comment: "Figure out how to create a longlong from a string: strtoll and
143 equivalent. It's not enough to call AC_CHECK_FUNCS: hpux has a strtoll, for
144 instance, but it only takes 2 args instead of 3!"
146 18. A subtle bug concerned with back references has been fixed by a change of
147 specification, with a corresponding code fix. A pattern such as
148 ^(xa|=?\1a)+$ which contains a back reference inside the group to which it
149 refers, was giving matches when it shouldn't. For example, xa=xaaa would
150 match that pattern. Interestingly, Perl (at least up to 5.11.3) has the
151 same bug. Such groups have to be quantified to be useful, or contained
152 inside another quantified group. (If there's no repetition, the reference
153 can never match.) The problem arises because, having left the group and
154 moved on to the rest of the pattern, a later failure that backtracks into
155 the group uses the captured value from the final iteration of the group
156 rather than the correct earlier one. I have fixed this in PCRE by forcing
157 any group that contains a reference to itself to be an atomic group; that
158 is, there cannot be any backtracking into it once it has completed. This is
159 similar to recursive and subroutine calls.
162 Version 8.00 19-Oct-09
163 ----------------------
165 1. The table for translating pcre_compile() error codes into POSIX error codes
166 was out-of-date, and there was no check on the pcre_compile() error code
167 being within the table. This could lead to an OK return being given in
168 error.
170 2. Changed the call to open a subject file in pcregrep from fopen(pathname,
171 "r") to fopen(pathname, "rb"), which fixed a problem with some of the tests
172 in a Windows environment.
174 3. The pcregrep --count option prints the count for each file even when it is
175 zero, as does GNU grep. However, pcregrep was also printing all files when
176 --files-with-matches was added. Now, when both options are given, it prints
177 counts only for those files that have at least one match. (GNU grep just
178 prints the file name in this circumstance, but including the count seems
179 more useful - otherwise, why use --count?) Also ensured that the
180 combination -clh just lists non-zero counts, with no names.
182 4. The long form of the pcregrep -F option was incorrectly implemented as
183 --fixed_strings instead of --fixed-strings. This is an incompatible change,
184 but it seems right to fix it, and I didn't think it was worth preserving
185 the old behaviour.
187 5. The command line items --regex=pattern and --regexp=pattern were not
188 recognized by pcregrep, which required --regex pattern or --regexp pattern
189 (with a space rather than an '='). The man page documented the '=' forms,
190 which are compatible with GNU grep; these now work.
192 6. No libpcreposix.pc file was created for pkg-config; there was just
193 libpcre.pc and libpcrecpp.pc. The omission has been rectified.
195 7. Added #ifndef SUPPORT_UCP into the pcre_ucd.c module, to reduce its size
196 when UCP support is not needed, by modifying the Python script that
197 generates it from Unicode data files. This should not matter if the module
198 is correctly used as a library, but I received one complaint about 50K of
199 unwanted data. My guess is that the person linked everything into his
200 program rather than using a library. Anyway, it does no harm.
202 8. A pattern such as /\x{123}{2,2}+/8 was incorrectly compiled; the trigger
203 was a minimum greater than 1 for a wide character in a possessive
204 repetition. The same bug could also affect patterns like /(\x{ff}{0,2})*/8
205 which had an unlimited repeat of a nested, fixed maximum repeat of a wide
206 character. Chaos in the form of incorrect output or a compiling loop could
207 result.
209 9. The restrictions on what a pattern can contain when partial matching is
210 requested for pcre_exec() have been removed. All patterns can now be
211 partially matched by this function. In addition, if there are at least two
212 slots in the offset vector, the offset of the earliest inspected character
213 for the match and the offset of the end of the subject are set in them when
214 PCRE_ERROR_PARTIAL is returned.
216 10. Partial matching has been split into two forms: PCRE_PARTIAL_SOFT, which is
217 synonymous with PCRE_PARTIAL, for backwards compatibility, and
218 PCRE_PARTIAL_HARD, which causes a partial match to supersede a full match,
219 and may be more useful for multi-segment matching.
221 11. Partial matching with pcre_exec() is now more intuitive. A partial match
222 used to be given if ever the end of the subject was reached; now it is
223 given only if matching could not proceed because another character was
224 needed. This makes a difference in some odd cases such as Z(*FAIL) with the
225 string "Z", which now yields "no match" instead of "partial match". In the
226 case of pcre_dfa_exec(), "no match" is given if every matching path for the
227 final character ended with (*FAIL).
229 12. Restarting a match using pcre_dfa_exec() after a partial match did not work
230 if the pattern had a "must contain" character that was already found in the
231 earlier partial match, unless partial matching was again requested. For
232 example, with the pattern /dog.(body)?/, the "must contain" character is
233 "g". If the first part-match was for the string "dog", restarting with
234 "sbody" failed. This bug has been fixed.
236 13. The string returned by pcre_dfa_exec() after a partial match has been
237 changed so that it starts at the first inspected character rather than the
238 first character of the match. This makes a difference only if the pattern
239 starts with a lookbehind assertion or \b or \B (\K is not supported by
240 pcre_dfa_exec()). It's an incompatible change, but it makes the two
241 matching functions compatible, and I think it's the right thing to do.
243 14. Added a pcredemo man page, created automatically from the pcredemo.c file,
244 so that the demonstration program is easily available in environments where
245 PCRE has not been installed from source.
247 15. Arranged to add -DPCRE_STATIC to cflags in libpcre.pc, libpcreposix.cp,
248 libpcrecpp.pc and pcre-config when PCRE is not compiled as a shared
249 library.
251 16. Added REG_UNGREEDY to the pcreposix interface, at the request of a user.
252 It maps to PCRE_UNGREEDY. It is not, of course, POSIX-compatible, but it
253 is not the first non-POSIX option to be added. Clearly some people find
254 these options useful.
256 17. If a caller to the POSIX matching function regexec() passes a non-zero
257 value for nmatch with a NULL value for pmatch, the value of
258 nmatch is forced to zero.
260 18. RunGrepTest did not have a test for the availability of the -u option of
261 the diff command, as RunTest does. It now checks in the same way as
262 RunTest, and also checks for the -b option.
264 19. If an odd number of negated classes containing just a single character
265 interposed, within parentheses, between a forward reference to a named
266 subpattern and the definition of the subpattern, compilation crashed with
267 an internal error, complaining that it could not find the referenced
268 subpattern. An example of a crashing pattern is /(?&A)(([^m])(?<A>))/.
269 [The bug was that it was starting one character too far in when skipping
270 over the character class, thus treating the ] as data rather than
271 terminating the class. This meant it could skip too much.]
273 20. Added PCRE_NOTEMPTY_ATSTART in order to be able to correctly implement the
274 /g option in pcretest when the pattern contains \K, which makes it possible
275 to have an empty string match not at the start, even when the pattern is
276 anchored. Updated pcretest and pcredemo to use this option.
278 21. If the maximum number of capturing subpatterns in a recursion was greater
279 than the maximum at the outer level, the higher number was returned, but
280 with unset values at the outer level. The correct (outer level) value is
281 now given.
283 22. If (*ACCEPT) appeared inside capturing parentheses, previous releases of
284 PCRE did not set those parentheses (unlike Perl). I have now found a way to
285 make it do so. The string so far is captured, making this feature
286 compatible with Perl.
288 23. The tests have been re-organized, adding tests 11 and 12, to make it
289 possible to check the Perl 5.10 features against Perl 5.10.
291 24. Perl 5.10 allows subroutine calls in lookbehinds, as long as the subroutine
292 pattern matches a fixed length string. PCRE did not allow this; now it
293 does. Neither allows recursion.
295 25. I finally figured out how to implement a request to provide the minimum
296 length of subject string that was needed in order to match a given pattern.
297 (It was back references and recursion that I had previously got hung up
298 on.) This code has now been added to pcre_study(); it finds a lower bound
299 to the length of subject needed. It is not necessarily the greatest lower
300 bound, but using it to avoid searching strings that are too short does give
301 some useful speed-ups. The value is available to calling programs via
302 pcre_fullinfo().
304 26. While implementing 25, I discovered to my embarrassment that pcretest had
305 not been passing the result of pcre_study() to pcre_dfa_exec(), so the
306 study optimizations had never been tested with that matching function.
307 Oops. What is worse, even when it was passed study data, there was a bug in
308 pcre_dfa_exec() that meant it never actually used it. Double oops. There
309 were also very few tests of studied patterns with pcre_dfa_exec().
311 27. If (?| is used to create subpatterns with duplicate numbers, they are now
312 allowed to have the same name, even if PCRE_DUPNAMES is not set. However,
313 on the other side of the coin, they are no longer allowed to have different
314 names, because these cannot be distinguished in PCRE, and this has caused
315 confusion. (This is a difference from Perl.)
317 28. When duplicate subpattern names are present (necessarily with different
318 numbers, as required by 27 above), and a test is made by name in a
319 conditional pattern, either for a subpattern having been matched, or for
320 recursion in such a pattern, all the associated numbered subpatterns are
321 tested, and the overall condition is true if the condition is true for any
322 one of them. This is the way Perl works, and is also more like the way
323 testing by number works.
326 Version 7.9 11-Apr-09
327 ---------------------
329 1. When building with support for bzlib/zlib (pcregrep) and/or readline
330 (pcretest), all targets were linked against these libraries. This included
331 libpcre, libpcreposix, and libpcrecpp, even though they do not use these
332 libraries. This caused unwanted dependencies to be created. This problem
333 has been fixed, and now only pcregrep is linked with bzlib/zlib and only
334 pcretest is linked with readline.
336 2. The "typedef int BOOL" in pcre_internal.h that was included inside the
337 "#ifndef FALSE" condition by an earlier change (probably 7.8/18) has been
338 moved outside it again, because FALSE and TRUE are already defined in AIX,
339 but BOOL is not.
341 3. The pcre_config() function was treating the PCRE_MATCH_LIMIT and
342 PCRE_MATCH_LIMIT_RECURSION values as ints, when they should be long ints.
344 4. The pcregrep documentation said spaces were inserted as well as colons (or
345 hyphens) following file names and line numbers when outputting matching
346 lines. This is not true; no spaces are inserted. I have also clarified the
347 wording for the --colour (or --color) option.
349 5. In pcregrep, when --colour was used with -o, the list of matching strings
350 was not coloured; this is different to GNU grep, so I have changed it to be
351 the same.
353 6. When --colo(u)r was used in pcregrep, only the first matching substring in
354 each matching line was coloured. Now it goes on to look for further matches
355 of any of the test patterns, which is the same behaviour as GNU grep.
357 7. A pattern that could match an empty string could cause pcregrep to loop; it
358 doesn't make sense to accept an empty string match in pcregrep, so I have
359 locked it out (using PCRE's PCRE_NOTEMPTY option). By experiment, this
360 seems to be how GNU grep behaves.
362 8. The pattern (?(?=.*b)b|^) was incorrectly compiled as "match must be at
363 start or after a newline", because the conditional assertion was not being
364 correctly handled. The rule now is that both the assertion and what follows
365 in the first alternative must satisfy the test.
367 9. If auto-callout was enabled in a pattern with a conditional group whose
368 condition was an assertion, PCRE could crash during matching, both with
369 pcre_exec() and pcre_dfa_exec().
371 10. The PCRE_DOLLAR_ENDONLY option was not working when pcre_dfa_exec() was
372 used for matching.
374 11. Unicode property support in character classes was not working for
375 characters (bytes) greater than 127 when not in UTF-8 mode.
377 12. Added the -M command line option to pcretest.
379 14. Added the non-standard REG_NOTEMPTY option to the POSIX interface.
381 15. Added the PCRE_NO_START_OPTIMIZE match-time option.
383 16. Added comments and documentation about mis-use of no_arg in the C++
384 wrapper.
386 17. Implemented support for UTF-8 encoding in EBCDIC environments, a patch
387 from Martin Jerabek that uses macro names for all relevant character and
388 string constants.
390 18. Added to pcre_internal.h two configuration checks: (a) If both EBCDIC and
391 SUPPORT_UTF8 are set, give an error; (b) If SUPPORT_UCP is set without
392 SUPPORT_UTF8, define SUPPORT_UTF8. The "configure" script handles both of
393 these, but not everybody uses configure.
395 19. A conditional group that had only one branch was not being correctly
396 recognized as an item that could match an empty string. This meant that an
397 enclosing group might also not be so recognized, causing infinite looping
398 (and probably a segfault) for patterns such as ^"((?(?=[a])[^"])|b)*"$
399 with the subject "ab", where knowledge that the repeated group can match
400 nothing is needed in order to break the loop.
402 20. If a pattern that was compiled with callouts was matched using pcre_dfa_
403 exec(), but without supplying a callout function, matching went wrong.
405 21. If PCRE_ERROR_MATCHLIMIT occurred during a recursion, there was a memory
406 leak if the size of the offset vector was greater than 30. When the vector
407 is smaller, the saved offsets during recursion go onto a local stack
408 vector, but for larger vectors malloc() is used. It was failing to free
409 when the recursion yielded PCRE_ERROR_MATCH_LIMIT (or any other "abnormal"
410 error, in fact).
412 22. There was a missing #ifdef SUPPORT_UTF8 round one of the variables in the
413 heapframe that is used only when UTF-8 support is enabled. This caused no
414 problem, but was untidy.
416 23. Steven Van Ingelgem's patch to CMakeLists.txt to change the name
417 CMAKE_BINARY_DIR to PROJECT_BINARY_DIR so that it works when PCRE is
418 included within another project.
420 24. Steven Van Ingelgem's patches to add more options to the CMake support,
421 slightly modified by me:
423 (a) PCRE_BUILD_TESTS can be set OFF not to build the tests, including
424 not building pcregrep.
426 (b) PCRE_BUILD_PCREGREP can be see OFF not to build pcregrep, but only
427 if PCRE_BUILD_TESTS is also set OFF, because the tests use pcregrep.
429 25. Forward references, both numeric and by name, in patterns that made use of
430 duplicate group numbers, could behave incorrectly or give incorrect errors,
431 because when scanning forward to find the reference group, PCRE was not
432 taking into account the duplicate group numbers. A pattern such as
433 ^X(?3)(a)(?|(b)|(q))(Y) is an example.
435 26. Changed a few more instances of "const unsigned char *" to USPTR, making
436 the feature of a custom pointer more persuasive (as requested by a user).
438 27. Wrapped the definitions of fileno and isatty for Windows, which appear in
439 pcretest.c, inside #ifndefs, because it seems they are sometimes already
440 pre-defined.
442 28. Added support for (*UTF8) at the start of a pattern.
444 29. Arrange for flags added by the "release type" setting in CMake to be shown
445 in the configuration summary.
448 Version 7.8 05-Sep-08
449 ---------------------
451 1. Replaced UCP searching code with optimized version as implemented for Ad
452 Muncher (http://www.admuncher.com/) by Peter Kankowski. This uses a two-
453 stage table and inline lookup instead of a function, giving speed ups of 2
454 to 5 times on some simple patterns that I tested. Permission was given to
455 distribute the MultiStage2.py script that generates the tables (it's not in
456 the tarball, but is in the Subversion repository).
458 2. Updated the Unicode datatables to Unicode 5.1.0. This adds yet more
459 scripts.
461 3. Change 12 for 7.7 introduced a bug in pcre_study() when a pattern contained
462 a group with a zero qualifier. The result of the study could be incorrect,
463 or the function might crash, depending on the pattern.
465 4. Caseless matching was not working for non-ASCII characters in back
466 references. For example, /(\x{de})\1/8i was not matching \x{de}\x{fe}.
467 It now works when Unicode Property Support is available.
469 5. In pcretest, an escape such as \x{de} in the data was always generating
470 a UTF-8 string, even in non-UTF-8 mode. Now it generates a single byte in
471 non-UTF-8 mode. If the value is greater than 255, it gives a warning about
472 truncation.
474 6. Minor bugfix in pcrecpp.cc (change "" == ... to NULL == ...).
476 7. Added two (int) casts to pcregrep when printing the difference of two
477 pointers, in case they are 64-bit values.
479 8. Added comments about Mac OS X stack usage to the pcrestack man page and to
480 test 2 if it fails.
482 9. Added PCRE_CALL_CONVENTION just before the names of all exported functions,
483 and a #define of that name to empty if it is not externally set. This is to
484 allow users of MSVC to set it if necessary.
486 10. The PCRE_EXP_DEFN macro which precedes exported functions was missing from
487 the convenience functions in the pcre_get.c source file.
489 11. An option change at the start of a pattern that had top-level alternatives
490 could cause overwriting and/or a crash. This command provoked a crash in
491 some environments:
493 printf "/(?i)[\xc3\xa9\xc3\xbd]|[\xc3\xa9\xc3\xbdA]/8\n" | pcretest
495 This potential security problem was recorded as CVE-2008-2371.
497 12. For a pattern where the match had to start at the beginning or immediately
498 after a newline (e.g /.*anything/ without the DOTALL flag), pcre_exec() and
499 pcre_dfa_exec() could read past the end of the passed subject if there was
500 no match. To help with detecting such bugs (e.g. with valgrind), I modified
501 pcretest so that it places the subject at the end of its malloc-ed buffer.
503 13. The change to pcretest in 12 above threw up a couple more cases when pcre_
504 exec() might read past the end of the data buffer in UTF-8 mode.
506 14. A similar bug to 7.3/2 existed when the PCRE_FIRSTLINE option was set and
507 the data contained the byte 0x85 as part of a UTF-8 character within its
508 first line. This applied both to normal and DFA matching.
510 15. Lazy qualifiers were not working in some cases in UTF-8 mode. For example,
511 /^[^d]*?$/8 failed to match "abc".
513 16. Added a missing copyright notice to pcrecpp_internal.h.
515 17. Make it more clear in the documentation that values returned from
516 pcre_exec() in ovector are byte offsets, not character counts.
518 18. Tidied a few places to stop certain compilers from issuing warnings.
520 19. Updated the Virtual Pascal + BCC files to compile the latest v7.7, as
521 supplied by Stefan Weber. I made a further small update for 7.8 because
522 there is a change of source arrangements: the pcre_searchfuncs.c module is
523 replaced by pcre_ucd.c.
526 Version 7.7 07-May-08
527 ---------------------
529 1. Applied Craig's patch to sort out a long long problem: "If we can't convert
530 a string to a long long, pretend we don't even have a long long." This is
531 done by checking for the strtoq, strtoll, and _strtoi64 functions.
533 2. Applied Craig's patch to pcrecpp.cc to restore ABI compatibility with
534 pre-7.6 versions, which defined a global no_arg variable instead of putting
535 it in the RE class. (See also #8 below.)
537 3. Remove a line of dead code, identified by coverity and reported by Nuno
538 Lopes.
540 4. Fixed two related pcregrep bugs involving -r with --include or --exclude:
542 (1) The include/exclude patterns were being applied to the whole pathnames
543 of files, instead of just to the final components.
545 (2) If there was more than one level of directory, the subdirectories were
546 skipped unless they satisfied the include/exclude conditions. This is
547 inconsistent with GNU grep (and could even be seen as contrary to the
548 pcregrep specification - which I improved to make it absolutely clear).
549 The action now is always to scan all levels of directory, and just
550 apply the include/exclude patterns to regular files.
552 5. Added the --include_dir and --exclude_dir patterns to pcregrep, and used
553 --exclude_dir in the tests to avoid scanning .svn directories.
555 6. Applied Craig's patch to the QuoteMeta function so that it escapes the
556 NUL character as backslash + 0 rather than backslash + NUL, because PCRE
557 doesn't support NULs in patterns.
559 7. Added some missing "const"s to declarations of static tables in
560 pcre_compile.c and pcre_dfa_exec.c.
562 8. Applied Craig's patch to pcrecpp.cc to fix a problem in OS X that was
563 caused by fix #2 above. (Subsequently also a second patch to fix the
564 first patch. And a third patch - this was a messy problem.)
566 9. Applied Craig's patch to remove the use of push_back().
568 10. Applied Alan Lehotsky's patch to add REG_STARTEND support to the POSIX
569 matching function regexec().
571 11. Added support for the Oniguruma syntax \g<name>, \g<n>, \g'name', \g'n',
572 which, however, unlike Perl's \g{...}, are subroutine calls, not back
573 references. PCRE supports relative numbers with this syntax (I don't think
574 Oniguruma does).
576 12. Previously, a group with a zero repeat such as (...){0} was completely
577 omitted from the compiled regex. However, this means that if the group
578 was called as a subroutine from elsewhere in the pattern, things went wrong
579 (an internal error was given). Such groups are now left in the compiled
580 pattern, with a new opcode that causes them to be skipped at execution
581 time.
583 13. Added the PCRE_JAVASCRIPT_COMPAT option. This makes the following changes
584 to the way PCRE behaves:
586 (a) A lone ] character is dis-allowed (Perl treats it as data).
588 (b) A back reference to an unmatched subpattern matches an empty string
589 (Perl fails the current match path).
591 (c) A data ] in a character class must be notated as \] because if the
592 first data character in a class is ], it defines an empty class. (In
593 Perl it is not possible to have an empty class.) The empty class []
594 never matches; it forces failure and is equivalent to (*FAIL) or (?!).
595 The negative empty class [^] matches any one character, independently
596 of the DOTALL setting.
598 14. A pattern such as /(?2)[]a()b](abc)/ which had a forward reference to a
599 non-existent subpattern following a character class starting with ']' and
600 containing () gave an internal compiling error instead of "reference to
601 non-existent subpattern". Fortunately, when the pattern did exist, the
602 compiled code was correct. (When scanning forwards to check for the
603 existencd of the subpattern, it was treating the data ']' as terminating
604 the class, so got the count wrong. When actually compiling, the reference
605 was subsequently set up correctly.)
607 15. The "always fail" assertion (?!) is optimzed to (*FAIL) by pcre_compile;
608 it was being rejected as not supported by pcre_dfa_exec(), even though
609 other assertions are supported. I have made pcre_dfa_exec() support
610 (*FAIL).
612 16. The implementation of 13c above involved the invention of a new opcode,
613 OP_ALLANY, which is like OP_ANY but doesn't check the /s flag. Since /s
614 cannot be changed at match time, I realized I could make a small
615 improvement to matching performance by compiling OP_ALLANY instead of
616 OP_ANY for "." when DOTALL was set, and then removing the runtime tests
617 on the OP_ANY path.
619 17. Compiling pcretest on Windows with readline support failed without the
620 following two fixes: (1) Make the unistd.h include conditional on
621 HAVE_UNISTD_H; (2) #define isatty and fileno as _isatty and _fileno.
623 18. Changed CMakeLists.txt and cmake/FindReadline.cmake to arrange for the
624 ncurses library to be included for pcretest when ReadLine support is
625 requested, but also to allow for it to be overridden. This patch came from
626 Daniel Bergström.
628 19. There was a typo in the file ucpinternal.h where f0_rangeflag was defined
629 as 0x00f00000 instead of 0x00800000. Luckily, this would not have caused
630 any errors with the current Unicode tables. Thanks to Peter Kankowski for
631 spotting this.
634 Version 7.6 28-Jan-08
635 ---------------------
637 1. A character class containing a very large number of characters with
638 codepoints greater than 255 (in UTF-8 mode, of course) caused a buffer
639 overflow.
641 2. Patch to cut out the "long long" test in pcrecpp_unittest when
642 HAVE_LONG_LONG is not defined.
644 3. Applied Christian Ehrlicher's patch to update the CMake build files to
645 bring them up to date and include new features. This patch includes:
647 - Fixed PH's badly added libz and libbz2 support.
648 - Fixed a problem with static linking.
649 - Added pcredemo. [But later removed - see 7 below.]
650 - Fixed dftables problem and added an option.
651 - Added a number of HAVE_XXX tests, including HAVE_WINDOWS_H and
653 - Added readline support for pcretest.
654 - Added an listing of the option settings after cmake has run.
656 4. A user submitted a patch to Makefile that makes it easy to create
657 "pcre.dll" under mingw when using Configure/Make. I added stuff to
658 Makefile.am that cause it to include this special target, without
659 affecting anything else. Note that the same mingw target plus all
660 the other distribution libraries and programs are now supported
661 when configuring with CMake (see 6 below) instead of with
662 Configure/Make.
664 5. Applied Craig's patch that moves no_arg into the RE class in the C++ code.
665 This is an attempt to solve the reported problem "pcrecpp::no_arg is not
666 exported in the Windows port". It has not yet been confirmed that the patch
667 solves the problem, but it does no harm.
669 6. Applied Sheri's patch to CMakeLists.txt to add NON_STANDARD_LIB_PREFIX and
670 NON_STANDARD_LIB_SUFFIX for dll names built with mingw when configured
671 with CMake, and also correct the comment about stack recursion.
673 7. Remove the automatic building of pcredemo from the ./configure system and
674 from CMakeLists.txt. The whole idea of pcredemo.c is that it is an example
675 of a program that users should build themselves after PCRE is installed, so
676 building it automatically is not really right. What is more, it gave
677 trouble in some build environments.
679 8. Further tidies to CMakeLists.txt from Sheri and Christian.
682 Version 7.5 10-Jan-08
683 ---------------------
685 1. Applied a patch from Craig: "This patch makes it possible to 'ignore'
686 values in parens when parsing an RE using the C++ wrapper."
688 2. Negative specials like \S did not work in character classes in UTF-8 mode.
689 Characters greater than 255 were excluded from the class instead of being
690 included.
692 3. The same bug as (2) above applied to negated POSIX classes such as
693 [:^space:].
695 4. PCRECPP_STATIC was referenced in pcrecpp_internal.h, but nowhere was it
696 defined or documented. It seems to have been a typo for PCRE_STATIC, so
697 I have changed it.
699 5. The construct (?&) was not diagnosed as a syntax error (it referenced the
700 first named subpattern) and a construct such as (?&a) would reference the
701 first named subpattern whose name started with "a" (in other words, the
702 length check was missing). Both these problems are fixed. "Subpattern name
703 expected" is now given for (?&) (a zero-length name), and this patch also
704 makes it give the same error for \k'' (previously it complained that that
705 was a reference to a non-existent subpattern).
707 6. The erroneous patterns (?+-a) and (?-+a) give different error messages;
708 this is right because (?- can be followed by option settings as well as by
709 digits. I have, however, made the messages clearer.
711 7. Patterns such as (?(1)a|b) (a pattern that contains fewer subpatterns
712 than the number used in the conditional) now cause a compile-time error.
713 This is actually not compatible with Perl, which accepts such patterns, but
714 treats the conditional as always being FALSE (as PCRE used to), but it
715 seems to me that giving a diagnostic is better.
717 8. Change "alphameric" to the more common word "alphanumeric" in comments
718 and messages.
720 9. Fix two occurrences of "backslash" in comments that should have been
721 "backspace".
723 10. Remove two redundant lines of code that can never be obeyed (their function
724 was moved elsewhere).
726 11. The program that makes PCRE's Unicode character property table had a bug
727 which caused it to generate incorrect table entries for sequences of
728 characters that have the same character type, but are in different scripts.
729 It amalgamated them into a single range, with the script of the first of
730 them. In other words, some characters were in the wrong script. There were
731 thirteen such cases, affecting characters in the following ranges:
733 U+002b0 - U+002c1
734 U+0060c - U+0060d
735 U+0061e - U+00612
736 U+0064b - U+0065e
737 U+0074d - U+0076d
738 U+01800 - U+01805
739 U+01d00 - U+01d77
740 U+01d9b - U+01dbf
741 U+0200b - U+0200f
742 U+030fc - U+030fe
743 U+03260 - U+0327f
744 U+0fb46 - U+0fbb1
745 U+10450 - U+1049d
747 12. The -o option (show only the matching part of a line) for pcregrep was not
748 compatible with GNU grep in that, if there was more than one match in a
749 line, it showed only the first of them. It now behaves in the same way as
750 GNU grep.
752 13. If the -o and -v options were combined for pcregrep, it printed a blank
753 line for every non-matching line. GNU grep prints nothing, and pcregrep now
754 does the same. The return code can be used to tell if there were any
755 non-matching lines.
757 14. Added --file-offsets and --line-offsets to pcregrep.
759 15. The pattern (?=something)(?R) was not being diagnosed as a potentially
760 infinitely looping recursion. The bug was that positive lookaheads were not
761 being skipped when checking for a possible empty match (negative lookaheads
762 and both kinds of lookbehind were skipped).
764 16. Fixed two typos in the Windows-only code in pcregrep.c, and moved the
765 inclusion of <windows.h> to before rather than after the definition of
766 INVALID_FILE_ATTRIBUTES (patch from David Byron).
768 17. Specifying a possessive quantifier with a specific limit for a Unicode
769 character property caused pcre_compile() to compile bad code, which led at
770 runtime to PCRE_ERROR_INTERNAL (-14). Examples of patterns that caused this
771 are: /\p{Zl}{2,3}+/8 and /\p{Cc}{2}+/8. It was the possessive "+" that
772 caused the error; without that there was no problem.
774 18. Added --enable-pcregrep-libz and --enable-pcregrep-libbz2.
776 19. Added --enable-pcretest-libreadline.
778 20. In pcrecpp.cc, the variable 'count' was incremented twice in
779 RE::GlobalReplace(). As a result, the number of replacements returned was
780 double what it should be. I removed one of the increments, but Craig sent a
781 later patch that removed the other one (the right fix) and added unit tests
782 that check the return values (which was not done before).
784 21. Several CMake things:
786 (1) Arranged that, when cmake is used on Unix, the libraries end up with
787 the names libpcre and libpcreposix, not just pcre and pcreposix.
789 (2) The above change means that pcretest and pcregrep are now correctly
790 linked with the newly-built libraries, not previously installed ones.
794 22. In UTF-8 mode, with newline set to "any", a pattern such as .*a.*=.b.*
795 crashed when matching a string such as a\x{2029}b (note that \x{2029} is a
796 UTF-8 newline character). The key issue is that the pattern starts .*;
797 this means that the match must be either at the beginning, or after a
798 newline. The bug was in the code for advancing after a failed match and
799 checking that the new position followed a newline. It was not taking
800 account of UTF-8 characters correctly.
802 23. PCRE was behaving differently from Perl in the way it recognized POSIX
803 character classes. PCRE was not treating the sequence [:...:] as a
804 character class unless the ... were all letters. Perl, however, seems to
805 allow any characters between [: and :], though of course it rejects as
806 unknown any "names" that contain non-letters, because all the known class
807 names consist only of letters. Thus, Perl gives an error for [[:1234:]],
808 for example, whereas PCRE did not - it did not recognize a POSIX character
809 class. This seemed a bit dangerous, so the code has been changed to be
810 closer to Perl. The behaviour is not identical to Perl, because PCRE will
811 diagnose an unknown class for, for example, [[:l\ower:]] where Perl will
812 treat it as [[:lower:]]. However, PCRE does now give "unknown" errors where
813 Perl does, and where it didn't before.
815 24. Rewrite so as to remove the single use of %n from pcregrep because in some
816 Windows environments %n is disabled by default.
819 Version 7.4 21-Sep-07
820 ---------------------
822 1. Change 7.3/28 was implemented for classes by looking at the bitmap. This
823 means that a class such as [\s] counted as "explicit reference to CR or
824 LF". That isn't really right - the whole point of the change was to try to
825 help when there was an actual mention of one of the two characters. So now
826 the change happens only if \r or \n (or a literal CR or LF) character is
827 encountered.
829 2. The 32-bit options word was also used for 6 internal flags, but the numbers
830 of both had grown to the point where there were only 3 bits left.
831 Fortunately, there was spare space in the data structure, and so I have
832 moved the internal flags into a new 16-bit field to free up more option
833 bits.
835 3. The appearance of (?J) at the start of a pattern set the DUPNAMES option,
836 but did not set the internal JCHANGED flag - either of these is enough to
837 control the way the "get" function works - but the PCRE_INFO_JCHANGED
838 facility is supposed to tell if (?J) was ever used, so now (?J) at the
839 start sets both bits.
841 4. Added options (at build time, compile time, exec time) to change \R from
842 matching any Unicode line ending sequence to just matching CR, LF, or CRLF.
844 5. doc/pcresyntax.html was missing from the distribution.
846 6. Put back the definition of PCRE_ERROR_NULLWSLIMIT, for backward
847 compatibility, even though it is no longer used.
849 7. Added macro for snprintf to pcrecpp_unittest.cc and also for strtoll and
850 strtoull to pcrecpp.cc to select the available functions in WIN32 when the
851 windows.h file is present (where different names are used). [This was
852 reversed later after testing - see 16 below.]
854 8. Changed all #include <config.h> to #include "config.h". There were also
855 some further <pcre.h> cases that I changed to "pcre.h".
857 9. When pcregrep was used with the --colour option, it missed the line ending
858 sequence off the lines that it output.
860 10. It was pointed out to me that arrays of string pointers cause lots of
861 relocations when a shared library is dynamically loaded. A technique of
862 using a single long string with a table of offsets can drastically reduce
863 these. I have refactored PCRE in four places to do this. The result is
864 dramatic:
866 Originally: 290
867 After changing UCP table: 187
868 After changing error message table: 43
869 After changing table of "verbs" 36
870 After changing table of Posix names 22
872 Thanks to the folks working on Gregex for glib for this insight.
874 11. --disable-stack-for-recursion caused compiling to fail unless -enable-
875 unicode-properties was also set.
877 12. Updated the tests so that they work when \R is defaulted to ANYCRLF.
879 13. Added checks for ANY and ANYCRLF to pcrecpp.cc where it previously
880 checked only for CRLF.
882 14. Added casts to pcretest.c to avoid compiler warnings.
884 15. Added Craig's patch to various pcrecpp modules to avoid compiler warnings.
886 16. Added Craig's patch to remove the WINDOWS_H tests, that were not working,
887 and instead check for _strtoi64 explicitly, and avoid the use of snprintf()
888 entirely. This removes changes made in 7 above.
890 17. The CMake files have been updated, and there is now more information about
891 building with CMake in the NON-UNIX-USE document.
894 Version 7.3 28-Aug-07
895 ---------------------
897 1. In the rejigging of the build system that eventually resulted in 7.1, the
898 line "#include <pcre.h>" was included in pcre_internal.h. The use of angle
899 brackets there is not right, since it causes compilers to look for an
900 installed pcre.h, not the version that is in the source that is being
901 compiled (which of course may be different). I have changed it back to:
903 #include "pcre.h"
905 I have a vague recollection that the change was concerned with compiling in
906 different directories, but in the new build system, that is taken care of
907 by the VPATH setting the Makefile.
909 2. The pattern .*$ when run in not-DOTALL UTF-8 mode with newline=any failed
910 when the subject happened to end in the byte 0x85 (e.g. if the last
911 character was \x{1ec5}). *Character* 0x85 is one of the "any" newline
912 characters but of course it shouldn't be taken as a newline when it is part
913 of another character. The bug was that, for an unlimited repeat of . in
914 not-DOTALL UTF-8 mode, PCRE was advancing by bytes rather than by
915 characters when looking for a newline.
917 3. A small performance improvement in the DOTALL UTF-8 mode .* case.
919 4. Debugging: adjusted the names of opcodes for different kinds of parentheses
920 in debug output.
922 5. Arrange to use "%I64d" instead of "%lld" and "%I64u" instead of "%llu" for
923 long printing in the pcrecpp unittest when running under MinGW.
925 6. ESC_K was left out of the EBCDIC table.
927 7. Change 7.0/38 introduced a new limit on the number of nested non-capturing
928 parentheses; I made it 1000, which seemed large enough. Unfortunately, the
929 limit also applies to "virtual nesting" when a pattern is recursive, and in
930 this case 1000 isn't so big. I have been able to remove this limit at the
931 expense of backing off one optimization in certain circumstances. Normally,
932 when pcre_exec() would call its internal match() function recursively and
933 immediately return the result unconditionally, it uses a "tail recursion"
934 feature to save stack. However, when a subpattern that can match an empty
935 string has an unlimited repetition quantifier, it no longer makes this
936 optimization. That gives it a stack frame in which to save the data for
937 checking that an empty string has been matched. Previously this was taken
938 from the 1000-entry workspace that had been reserved. So now there is no
939 explicit limit, but more stack is used.
941 8. Applied Daniel's patches to solve problems with the import/export magic
942 syntax that is required for Windows, and which was going wrong for the
943 pcreposix and pcrecpp parts of the library. These were overlooked when this
944 problem was solved for the main library.
946 9. There were some crude static tests to avoid integer overflow when computing
947 the size of patterns that contain repeated groups with explicit upper
948 limits. As the maximum quantifier is 65535, the maximum group length was
949 set at 30,000 so that the product of these two numbers did not overflow a
950 32-bit integer. However, it turns out that people want to use groups that
951 are longer than 30,000 bytes (though not repeat them that many times).
952 Change 7.0/17 (the refactoring of the way the pattern size is computed) has
953 made it possible to implement the integer overflow checks in a much more
954 dynamic way, which I have now done. The artificial limitation on group
955 length has been removed - we now have only the limit on the total length of
956 the compiled pattern, which depends on the LINK_SIZE setting.
958 10. Fixed a bug in the documentation for get/copy named substring when
959 duplicate names are permitted. If none of the named substrings are set, the
960 functions return PCRE_ERROR_NOSUBSTRING (7); the doc said they returned an
961 empty string.
963 11. Because Perl interprets \Q...\E at a high level, and ignores orphan \E
964 instances, patterns such as [\Q\E] or [\E] or even [^\E] cause an error,
965 because the ] is interpreted as the first data character and the
966 terminating ] is not found. PCRE has been made compatible with Perl in this
967 regard. Previously, it interpreted [\Q\E] as an empty class, and [\E] could
968 cause memory overwriting.
970 10. Like Perl, PCRE automatically breaks an unlimited repeat after an empty
971 string has been matched (to stop an infinite loop). It was not recognizing
972 a conditional subpattern that could match an empty string if that
973 subpattern was within another subpattern. For example, it looped when
974 trying to match (((?(1)X|))*) but it was OK with ((?(1)X|)*) where the
975 condition was not nested. This bug has been fixed.
977 12. A pattern like \X?\d or \P{L}?\d in non-UTF-8 mode could cause a backtrack
978 past the start of the subject in the presence of bytes with the top bit
979 set, for example "\x8aBCD".
981 13. Added Perl 5.10 experimental backtracking controls (*FAIL), (*F), (*PRUNE),
982 (*SKIP), (*THEN), (*COMMIT), and (*ACCEPT).
984 14. Optimized (?!) to (*FAIL).
986 15. Updated the test for a valid UTF-8 string to conform to the later RFC 3629.
987 This restricts code points to be within the range 0 to 0x10FFFF, excluding
988 the "low surrogate" sequence 0xD800 to 0xDFFF. Previously, PCRE allowed the
989 full range 0 to 0x7FFFFFFF, as defined by RFC 2279. Internally, it still
990 does: it's just the validity check that is more restrictive.
992 16. Inserted checks for integer overflows during escape sequence (backslash)
993 processing, and also fixed erroneous offset values for syntax errors during
994 backslash processing.
996 17. Fixed another case of looking too far back in non-UTF-8 mode (cf 12 above)
997 for patterns like [\PPP\x8a]{1,}\x80 with the subject "A\x80".
999 18. An unterminated class in a pattern like (?1)\c[ with a "forward reference"
1000 caused an overrun.
1002 19. A pattern like (?:[\PPa*]*){8,} which had an "extended class" (one with
1003 something other than just ASCII characters) inside a group that had an
1004 unlimited repeat caused a loop at compile time (while checking to see
1005 whether the group could match an empty string).
1007 20. Debugging a pattern containing \p or \P could cause a crash. For example,
1008 [\P{Any}] did so. (Error in the code for printing property names.)
1010 21. An orphan \E inside a character class could cause a crash.
1012 22. A repeated capturing bracket such as (A)? could cause a wild memory
1013 reference during compilation.
1015 23. There are several functions in pcre_compile() that scan along a compiled
1016 expression for various reasons (e.g. to see if it's fixed length for look
1017 behind). There were bugs in these functions when a repeated \p or \P was
1018 present in the pattern. These operators have additional parameters compared
1019 with \d, etc, and these were not being taken into account when moving along
1020 the compiled data. Specifically:
1022 (a) A item such as \p{Yi}{3} in a lookbehind was not treated as fixed
1023 length.
1025 (b) An item such as \pL+ within a repeated group could cause crashes or
1026 loops.
1028 (c) A pattern such as \p{Yi}+(\P{Yi}+)(?1) could give an incorrect
1029 "reference to non-existent subpattern" error.
1031 (d) A pattern like (\P{Yi}{2}\277)? could loop at compile time.
1033 24. A repeated \S or \W in UTF-8 mode could give wrong answers when multibyte
1034 characters were involved (for example /\S{2}/8g with "A\x{a3}BC").
1036 25. Using pcregrep in multiline, inverted mode (-Mv) caused it to loop.
1038 26. Patterns such as [\P{Yi}A] which include \p or \P and just one other
1039 character were causing crashes (broken optimization).
1041 27. Patterns such as (\P{Yi}*\277)* (group with possible zero repeat containing
1042 \p or \P) caused a compile-time loop.
1044 28. More problems have arisen in unanchored patterns when CRLF is a valid line
1045 break. For example, the unstudied pattern [\r\n]A does not match the string
1046 "\r\nA" because change 7.0/46 below moves the current point on by two
1047 characters after failing to match at the start. However, the pattern \nA
1048 *does* match, because it doesn't start till \n, and if [\r\n]A is studied,
1049 the same is true. There doesn't seem any very clean way out of this, but
1050 what I have chosen to do makes the common cases work: PCRE now takes note
1051 of whether there can be an explicit match for \r or \n anywhere in the
1052 pattern, and if so, 7.0/46 no longer applies. As part of this change,
1053 there's a new PCRE_INFO_HASCRORLF option for finding out whether a compiled
1054 pattern has explicit CR or LF references.
1056 29. Added (*CR) etc for changing newline setting at start of pattern.
1059 Version 7.2 19-Jun-07
1060 ---------------------
1062 1. If the fr_FR locale cannot be found for test 3, try the "french" locale,
1063 which is apparently normally available under Windows.
1065 2. Re-jig the pcregrep tests with different newline settings in an attempt
1066 to make them independent of the local environment's newline setting.
1068 3. Add code to configure.ac to remove -g from the CFLAGS default settings.
1070 4. Some of the "internals" tests were previously cut out when the link size
1071 was not 2, because the output contained actual offsets. The recent new
1072 "Z" feature of pcretest means that these can be cut out, making the tests
1073 usable with all link sizes.
1075 5. Implemented Stan Switzer's goto replacement for longjmp() when not using
1076 stack recursion. This gives a massive performance boost under BSD, but just
1077 a small improvement under Linux. However, it saves one field in the frame
1078 in all cases.
1080 6. Added more features from the forthcoming Perl 5.10:
1082 (a) (?-n) (where n is a string of digits) is a relative subroutine or
1083 recursion call. It refers to the nth most recently opened parentheses.
1085 (b) (?+n) is also a relative subroutine call; it refers to the nth next
1086 to be opened parentheses.
1088 (c) Conditions that refer to capturing parentheses can be specified
1089 relatively, for example, (?(-2)... or (?(+3)...
1091 (d) \K resets the start of the current match so that everything before
1092 is not part of it.
1094 (e) \k{name} is synonymous with \k<name> and \k'name' (.NET compatible).
1096 (f) \g{name} is another synonym - part of Perl 5.10's unification of
1097 reference syntax.
1099 (g) (?| introduces a group in which the numbering of parentheses in each
1100 alternative starts with the same number.
1102 (h) \h, \H, \v, and \V match horizontal and vertical whitespace.
1104 7. Added two new calls to pcre_fullinfo(): PCRE_INFO_OKPARTIAL and
1107 8. A pattern such as (.*(.)?)* caused pcre_exec() to fail by either not
1108 terminating or by crashing. Diagnosed by Viktor Griph; it was in the code
1109 for detecting groups that can match an empty string.
1111 9. A pattern with a very large number of alternatives (more than several
1112 hundred) was running out of internal workspace during the pre-compile
1113 phase, where pcre_compile() figures out how much memory will be needed. A
1114 bit of new cunning has reduced the workspace needed for groups with
1115 alternatives. The 1000-alternative test pattern now uses 12 bytes of
1116 workspace instead of running out of the 4096 that are available.
1118 10. Inserted some missing (unsigned int) casts to get rid of compiler warnings.
1120 11. Applied patch from Google to remove an optimization that didn't quite work.
1121 The report of the bug said:
1123 pcrecpp::RE("a*").FullMatch("aaa") matches, while
1124 pcrecpp::RE("a*?").FullMatch("aaa") does not, and
1125 pcrecpp::RE("a*?\\z").FullMatch("aaa") does again.
1127 12. If \p or \P was used in non-UTF-8 mode on a character greater than 127
1128 it matched the wrong number of bytes.
1131 Version 7.1 24-Apr-07
1132 ---------------------
1134 1. Applied Bob Rossi and Daniel G's patches to convert the build system to one
1135 that is more "standard", making use of automake and other Autotools. There
1136 is some re-arrangement of the files and adjustment of comments consequent
1137 on this.
1139 2. Part of the patch fixed a problem with the pcregrep tests. The test of -r
1140 for recursive directory scanning broke on some systems because the files
1141 are not scanned in any specific order and on different systems the order
1142 was different. A call to "sort" has been inserted into RunGrepTest for the
1143 approprate test as a short-term fix. In the longer term there may be an
1144 alternative.
1146 3. I had an email from Eric Raymond about problems translating some of PCRE's
1147 man pages to HTML (despite the fact that I distribute HTML pages, some
1148 people do their own conversions for various reasons). The problems
1149 concerned the use of low-level troff macros .br and .in. I have therefore
1150 removed all such uses from the man pages (some were redundant, some could
1151 be replaced by .nf/.fi pairs). The 132html script that I use to generate
1152 HTML has been updated to handle .nf/.fi and to complain if it encounters
1153 .br or .in.
1155 4. Updated comments in configure.ac that get placed in config.h.in and also
1156 arranged for config.h to be included in the distribution, with the name
1157 config.h.generic, for the benefit of those who have to compile without
1158 Autotools (compare pcre.h, which is now distributed as pcre.h.generic).
1160 5. Updated the support (such as it is) for Virtual Pascal, thanks to Stefan
1161 Weber: (1) pcre_internal.h was missing some function renames; (2) updated
1162 makevp.bat for the current PCRE, using the additional files
1163 makevp_c.txt, makevp_l.txt, and pcregexp.pas.
1165 6. A Windows user reported a minor discrepancy with test 2, which turned out
1166 to be caused by a trailing space on an input line that had got lost in his
1167 copy. The trailing space was an accident, so I've just removed it.
1169 7. Add -Wl,-R... flags in pcre-config.in for *BSD* systems, as I'm told
1170 that is needed.
1172 8. Mark ucp_table (in ucptable.h) and ucp_gentype (in pcre_ucp_searchfuncs.c)
1173 as "const" (a) because they are and (b) because it helps the PHP
1174 maintainers who have recently made a script to detect big data structures
1175 in the php code that should be moved to the .rodata section. I remembered
1176 to update Builducptable as well, so it won't revert if ucptable.h is ever
1177 re-created.
1179 9. Added some extra #ifdef SUPPORT_UTF8 conditionals into pcretest.c,
1180 pcre_printint.src, pcre_compile.c, pcre_study.c, and pcre_tables.c, in
1181 order to be able to cut out the UTF-8 tables in the latter when UTF-8
1182 support is not required. This saves 1.5-2K of code, which is important in
1183 some applications.
1185 Later: more #ifdefs are needed in pcre_ord2utf8.c and pcre_valid_utf8.c
1186 so as not to refer to the tables, even though these functions will never be
1187 called when UTF-8 support is disabled. Otherwise there are problems with a
1188 shared library.
1190 10. Fixed two bugs in the emulated memmove() function in pcre_internal.h:
1192 (a) It was defining its arguments as char * instead of void *.
1194 (b) It was assuming that all moves were upwards in memory; this was true
1195 a long time ago when I wrote it, but is no longer the case.
1197 The emulated memove() is provided for those environments that have neither
1198 memmove() nor bcopy(). I didn't think anyone used it these days, but that
1199 is clearly not the case, as these two bugs were recently reported.
1201 11. The script PrepareRelease is now distributed: it calls 132html, CleanTxt,
1202 and Detrail to create the HTML documentation, the .txt form of the man
1203 pages, and it removes trailing spaces from listed files. It also creates
1204 pcre.h.generic and config.h.generic from pcre.h and config.h. In the latter
1205 case, it wraps all the #defines with #ifndefs. This script should be run
1206 before "make dist".
1208 12. Fixed two fairly obscure bugs concerned with quantified caseless matching
1209 with Unicode property support.
1211 (a) For a maximizing quantifier, if the two different cases of the
1212 character were of different lengths in their UTF-8 codings (there are
1213 some cases like this - I found 11), and the matching function had to
1214 back up over a mixture of the two cases, it incorrectly assumed they
1215 were both the same length.
1217 (b) When PCRE was configured to use the heap rather than the stack for
1218 recursion during matching, it was not correctly preserving the data for
1219 the other case of a UTF-8 character when checking ahead for a match
1220 while processing a minimizing repeat. If the check also involved
1221 matching a wide character, but failed, corruption could cause an
1222 erroneous result when trying to check for a repeat of the original
1223 character.
1225 13. Some tidying changes to the testing mechanism:
1227 (a) The RunTest script now detects the internal link size and whether there
1228 is UTF-8 and UCP support by running ./pcretest -C instead of relying on
1229 values substituted by "configure". (The RunGrepTest script already did
1230 this for UTF-8.) The configure.ac script no longer substitutes the
1231 relevant variables.
1233 (b) The debugging options /B and /D in pcretest show the compiled bytecode
1234 with length and offset values. This means that the output is different
1235 for different internal link sizes. Test 2 is skipped for link sizes
1236 other than 2 because of this, bypassing the problem. Unfortunately,
1237 there was also a test in test 3 (the locale tests) that used /B and
1238 failed for link sizes other than 2. Rather than cut the whole test out,
1239 I have added a new /Z option to pcretest that replaces the length and
1240 offset values with spaces. This is now used to make test 3 independent
1241 of link size. (Test 2 will be tidied up later.)
1243 14. If erroroffset was passed as NULL to pcre_compile, it provoked a
1244 segmentation fault instead of returning the appropriate error message.
1246 15. In multiline mode when the newline sequence was set to "any", the pattern
1247 ^$ would give a match between the \r and \n of a subject such as "A\r\nB".
1248 This doesn't seem right; it now treats the CRLF combination as the line
1249 ending, and so does not match in that case. It's only a pattern such as ^$
1250 that would hit this one: something like ^ABC$ would have failed after \r
1251 and then tried again after \r\n.
1253 16. Changed the comparison command for RunGrepTest from "diff -u" to "diff -ub"
1254 in an attempt to make files that differ only in their line terminators
1255 compare equal. This works on Linux.
1257 17. Under certain error circumstances pcregrep might try to free random memory
1258 as it exited. This is now fixed, thanks to valgrind.
1260 19. In pcretest, if the pattern /(?m)^$/g<any> was matched against the string
1261 "abc\r\n\r\n", it found an unwanted second match after the second \r. This
1262 was because its rules for how to advance for /g after matching an empty
1263 string at the end of a line did not allow for this case. They now check for
1264 it specially.
1266 20. pcretest is supposed to handle patterns and data of any length, by
1267 extending its buffers when necessary. It was getting this wrong when the
1268 buffer for a data line had to be extended.
1270 21. Added PCRE_NEWLINE_ANYCRLF which is like ANY, but matches only CR, LF, or
1271 CRLF as a newline sequence.
1273 22. Code for handling Unicode properties in pcre_dfa_exec() wasn't being cut
1274 out by #ifdef SUPPORT_UCP. This did no harm, as it could never be used, but
1275 I have nevertheless tidied it up.
1277 23. Added some casts to kill warnings from HP-UX ia64 compiler.
1279 24. Added a man page for pcre-config.
1282 Version 7.0 19-Dec-06
1283 ---------------------
1285 1. Fixed a signed/unsigned compiler warning in pcre_compile.c, shown up by
1286 moving to gcc 4.1.1.
1288 2. The -S option for pcretest uses setrlimit(); I had omitted to #include
1289 sys/time.h, which is documented as needed for this function. It doesn't
1290 seem to matter on Linux, but it showed up on some releases of OS X.
1292 3. It seems that there are systems where bytes whose values are greater than
1293 127 match isprint() in the "C" locale. The "C" locale should be the
1294 default when a C program starts up. In most systems, only ASCII printing
1295 characters match isprint(). This difference caused the output from pcretest
1296 to vary, making some of the tests fail. I have changed pcretest so that:
1298 (a) When it is outputting text in the compiled version of a pattern, bytes
1299 other than 32-126 are always shown as hex escapes.
1301 (b) When it is outputting text that is a matched part of a subject string,
1302 it does the same, unless a different locale has been set for the match
1303 (using the /L modifier). In this case, it uses isprint() to decide.
1305 4. Fixed a major bug that caused incorrect computation of the amount of memory
1306 required for a compiled pattern when options that changed within the
1307 pattern affected the logic of the preliminary scan that determines the
1308 length. The relevant options are -x, and -i in UTF-8 mode. The result was
1309 that the computed length was too small. The symptoms of this bug were
1310 either the PCRE error "internal error: code overflow" from pcre_compile(),
1311 or a glibc crash with a message such as "pcretest: free(): invalid next
1312 size (fast)". Examples of patterns that provoked this bug (shown in
1313 pcretest format) are:
1315 /(?-x: )/x
1316 /(?x)(?-x: \s*#\s*)/
1317 /((?i)[\x{c0}])/8
1318 /(?i:[\x{c0}])/8
1320 HOWEVER: Change 17 below makes this fix obsolete as the memory computation
1321 is now done differently.
1323 5. Applied patches from Google to: (a) add a QuoteMeta function to the C++
1324 wrapper classes; (b) implement a new function in the C++ scanner that is
1325 more efficient than the old way of doing things because it avoids levels of
1326 recursion in the regex matching; (c) add a paragraph to the documentation
1327 for the FullMatch() function.
1329 6. The escape sequence \n was being treated as whatever was defined as
1330 "newline". Not only was this contrary to the documentation, which states
1331 that \n is character 10 (hex 0A), but it also went horribly wrong when
1332 "newline" was defined as CRLF. This has been fixed.
1334 7. In pcre_dfa_exec.c the value of an unsigned integer (the variable called c)
1335 was being set to -1 for the "end of line" case (supposedly a value that no
1336 character can have). Though this value is never used (the check for end of
1337 line is "zero bytes in current character"), it caused compiler complaints.
1338 I've changed it to 0xffffffff.
1340 8. In pcre_version.c, the version string was being built by a sequence of
1341 C macros that, in the event of PCRE_PRERELEASE being defined as an empty
1342 string (as it is for production releases) called a macro with an empty
1343 argument. The C standard says the result of this is undefined. The gcc
1344 compiler treats it as an empty string (which was what was wanted) but it is
1345 reported that Visual C gives an error. The source has been hacked around to
1346 avoid this problem.
1348 9. On the advice of a Windows user, included <io.h> and <fcntl.h> in Windows
1349 builds of pcretest, and changed the call to _setmode() to use _O_BINARY
1350 instead of 0x8000. Made all the #ifdefs test both _WIN32 and WIN32 (not all
1351 of them did).
1353 10. Originally, pcretest opened its input and output without "b"; then I was
1354 told that "b" was needed in some environments, so it was added for release
1355 5.0 to both the input and output. (It makes no difference on Unix-like
1356 systems.) Later I was told that it is wrong for the input on Windows. I've
1357 now abstracted the modes into two macros, to make it easier to fiddle with
1358 them, and removed "b" from the input mode under Windows.
1360 11. Added pkgconfig support for the C++ wrapper library, libpcrecpp.
1362 12. Added -help and --help to pcretest as an official way of being reminded
1363 of the options.
1365 13. Removed some redundant semicolons after macro calls in pcrecpparg.h.in
1366 and pcrecpp.cc because they annoy compilers at high warning levels.
1368 14. A bit of tidying/refactoring in pcre_exec.c in the main bumpalong loop.
1370 15. Fixed an occurrence of == in configure.ac that should have been = (shell
1371 scripts are not C programs :-) and which was not noticed because it works
1372 on Linux.
1374 16. pcretest is supposed to handle any length of pattern and data line (as one
1375 line or as a continued sequence of lines) by extending its input buffer if
1376 necessary. This feature was broken for very long pattern lines, leading to
1377 a string of junk being passed to pcre_compile() if the pattern was longer
1378 than about 50K.
1380 17. I have done a major re-factoring of the way pcre_compile() computes the
1381 amount of memory needed for a compiled pattern. Previously, there was code
1382 that made a preliminary scan of the pattern in order to do this. That was
1383 OK when PCRE was new, but as the facilities have expanded, it has become
1384 harder and harder to keep it in step with the real compile phase, and there
1385 have been a number of bugs (see for example, 4 above). I have now found a
1386 cunning way of running the real compile function in a "fake" mode that
1387 enables it to compute how much memory it would need, while actually only
1388 ever using a few hundred bytes of working memory and without too many
1389 tests of the mode. This should make future maintenance and development
1390 easier. A side effect of this work is that the limit of 200 on the nesting
1391 depth of parentheses has been removed (though this was never a serious
1392 limitation, I suspect). However, there is a downside: pcre_compile() now
1393 runs more slowly than before (30% or more, depending on the pattern). I
1394 hope this isn't a big issue. There is no effect on runtime performance.
1396 18. Fixed a minor bug in pcretest: if a pattern line was not terminated by a
1397 newline (only possible for the last line of a file) and it was a
1398 pattern that set a locale (followed by /Lsomething), pcretest crashed.
1400 19. Added additional timing features to pcretest. (1) The -tm option now times
1401 matching only, not compiling. (2) Both -t and -tm can be followed, as a
1402 separate command line item, by a number that specifies the number of
1403 repeats to use when timing. The default is 50000; this gives better
1404 precision, but takes uncomfortably long for very large patterns.
1406 20. Extended pcre_study() to be more clever in cases where a branch of a
1407 subpattern has no definite first character. For example, (a*|b*)[cd] would
1408 previously give no result from pcre_study(). Now it recognizes that the
1409 first character must be a, b, c, or d.
1411 21. There was an incorrect error "recursive call could loop indefinitely" if
1412 a subpattern (or the entire pattern) that was being tested for matching an
1413 empty string contained only one non-empty item after a nested subpattern.
1414 For example, the pattern (?>\x{100}*)\d(?R) provoked this error
1415 incorrectly, because the \d was being skipped in the check.
1417 22. The pcretest program now has a new pattern option /B and a command line
1418 option -b, which is equivalent to adding /B to every pattern. This causes
1419 it to show the compiled bytecode, without the additional information that
1420 -d shows. The effect of -d is now the same as -b with -i (and similarly, /D
1421 is the same as /B/I).
1423 23. A new optimization is now able automatically to treat some sequences such
1424 as a*b as a*+b. More specifically, if something simple (such as a character
1425 or a simple class like \d) has an unlimited quantifier, and is followed by
1426 something that cannot possibly match the quantified thing, the quantifier
1427 is automatically "possessified".
1429 24. A recursive reference to a subpattern whose number was greater than 39
1430 went wrong under certain circumstances in UTF-8 mode. This bug could also
1431 have affected the operation of pcre_study().
1433 25. Realized that a little bit of performance could be had by replacing
1434 (c & 0xc0) == 0xc0 with c >= 0xc0 when processing UTF-8 characters.
1436 26. Timing data from pcretest is now shown to 4 decimal places instead of 3.
1438 27. Possessive quantifiers such as a++ were previously implemented by turning
1439 them into atomic groups such as ($>a+). Now they have their own opcodes,
1440 which improves performance. This includes the automatically created ones
1441 from 23 above.
1443 28. A pattern such as (?=(\w+))\1: which simulates an atomic group using a
1444 lookahead was broken if it was not anchored. PCRE was mistakenly expecting
1445 the first matched character to be a colon. This applied both to named and
1446 numbered groups.
1448 29. The ucpinternal.h header file was missing its idempotency #ifdef.
1450 30. I was sent a "project" file called libpcre.a.dev which I understand makes
1451 building PCRE on Windows easier, so I have included it in the distribution.
1453 31. There is now a check in pcretest against a ridiculously large number being
1454 returned by pcre_exec() or pcre_dfa_exec(). If this happens in a /g or /G
1455 loop, the loop is abandoned.
1457 32. Forward references to subpatterns in conditions such as (?(2)...) where
1458 subpattern 2 is defined later cause pcre_compile() to search forwards in
1459 the pattern for the relevant set of parentheses. This search went wrong
1460 when there were unescaped parentheses in a character class, parentheses
1461 escaped with \Q...\E, or parentheses in a #-comment in /x mode.
1463 33. "Subroutine" calls and backreferences were previously restricted to
1464 referencing subpatterns earlier in the regex. This restriction has now
1465 been removed.
1467 34. Added a number of extra features that are going to be in Perl 5.10. On the
1468 whole, these are just syntactic alternatives for features that PCRE had
1469 previously implemented using the Python syntax or my own invention. The
1470 other formats are all retained for compatibility.
1472 (a) Named groups can now be defined as (?<name>...) or (?'name'...) as well
1473 as (?P<name>...). The new forms, as well as being in Perl 5.10, are
1474 also .NET compatible.
1476 (b) A recursion or subroutine call to a named group can now be defined as
1477 (?&name) as well as (?P>name).
1479 (c) A backreference to a named group can now be defined as \k<name> or
1480 \k'name' as well as (?P=name). The new forms, as well as being in Perl
1481 5.10, are also .NET compatible.
1483 (d) A conditional reference to a named group can now use the syntax
1484 (?(<name>) or (?('name') as well as (?(name).
1486 (e) A "conditional group" of the form (?(DEFINE)...) can be used to define
1487 groups (named and numbered) that are never evaluated inline, but can be
1488 called as "subroutines" from elsewhere. In effect, the DEFINE condition
1489 is always false. There may be only one alternative in such a group.
1491 (f) A test for recursion can be given as (?(R1).. or (?(R&name)... as well
1492 as the simple (?(R). The condition is true only if the most recent
1493 recursion is that of the given number or name. It does not search out
1494 through the entire recursion stack.
1496 (g) The escape \gN or \g{N} has been added, where N is a positive or
1497 negative number, specifying an absolute or relative reference.
1499 35. Tidied to get rid of some further signed/unsigned compiler warnings and
1500 some "unreachable code" warnings.
1502 36. Updated the Unicode property tables to Unicode version 5.0.0. Amongst other
1503 things, this adds five new scripts.
1505 37. Perl ignores orphaned \E escapes completely. PCRE now does the same.
1506 There were also incompatibilities regarding the handling of \Q..\E inside
1507 character classes, for example with patterns like [\Qa\E-\Qz\E] where the
1508 hyphen was adjacent to \Q or \E. I hope I've cleared all this up now.
1510 38. Like Perl, PCRE detects when an indefinitely repeated parenthesized group
1511 matches an empty string, and forcibly breaks the loop. There were bugs in
1512 this code in non-simple cases. For a pattern such as ^(a()*)* matched
1513 against aaaa the result was just "a" rather than "aaaa", for example. Two
1514 separate and independent bugs (that affected different cases) have been
1515 fixed.
1517 39. Refactored the code to abolish the use of different opcodes for small
1518 capturing bracket numbers. This is a tidy that I avoided doing when I
1519 removed the limit on the number of capturing brackets for 3.5 back in 2001.
1520 The new approach is not only tidier, it makes it possible to reduce the
1521 memory needed to fix the previous bug (38).
1523 40. Implemented PCRE_NEWLINE_ANY to recognize any of the Unicode newline
1524 sequences (http://unicode.org/unicode/reports/tr18/) as "newline" when
1525 processing dot, circumflex, or dollar metacharacters, or #-comments in /x
1526 mode.
1528 41. Add \R to match any Unicode newline sequence, as suggested in the Unicode
1529 report.
1531 42. Applied patch, originally from Ari Pollak, modified by Google, to allow
1532 copy construction and assignment in the C++ wrapper.
1534 43. Updated pcregrep to support "--newline=any". In the process, I fixed a
1535 couple of bugs that could have given wrong results in the "--newline=crlf"
1536 case.
1538 44. Added a number of casts and did some reorganization of signed/unsigned int
1539 variables following suggestions from Dair Grant. Also renamed the variable
1540 "this" as "item" because it is a C++ keyword.
1542 45. Arranged for dftables to add
1544 #include "pcre_internal.h"
1546 to pcre_chartables.c because without it, gcc 4.x may remove the array
1547 definition from the final binary if PCRE is built into a static library and
1548 dead code stripping is activated.
1550 46. For an unanchored pattern, if a match attempt fails at the start of a
1551 newline sequence, and the newline setting is CRLF or ANY, and the next two
1552 characters are CRLF, advance by two characters instead of one.
1555 Version 6.7 04-Jul-06
1556 ---------------------
1558 1. In order to handle tests when input lines are enormously long, pcretest has
1559 been re-factored so that it automatically extends its buffers when
1560 necessary. The code is crude, but this _is_ just a test program. The
1561 default size has been increased from 32K to 50K.
1563 2. The code in pcre_study() was using the value of the re argument before
1564 testing it for NULL. (Of course, in any sensible call of the function, it
1565 won't be NULL.)
1567 3. The memmove() emulation function in pcre_internal.h, which is used on
1568 systems that lack both memmove() and bcopy() - that is, hardly ever -
1569 was missing a "static" storage class specifier.
1571 4. When UTF-8 mode was not set, PCRE looped when compiling certain patterns
1572 containing an extended class (one that cannot be represented by a bitmap
1573 because it contains high-valued characters or Unicode property items, e.g.
1574 [\pZ]). Almost always one would set UTF-8 mode when processing such a
1575 pattern, but PCRE should not loop if you do not (it no longer does).
1576 [Detail: two cases were found: (a) a repeated subpattern containing an
1577 extended class; (b) a recursive reference to a subpattern that followed a
1578 previous extended class. It wasn't skipping over the extended class
1579 correctly when UTF-8 mode was not set.]
1581 5. A negated single-character class was not being recognized as fixed-length
1582 in lookbehind assertions such as (?<=[^f]), leading to an incorrect
1583 compile error "lookbehind assertion is not fixed length".
1585 6. The RunPerlTest auxiliary script was showing an unexpected difference
1586 between PCRE and Perl for UTF-8 tests. It turns out that it is hard to
1587 write a Perl script that can interpret lines of an input file either as
1588 byte characters or as UTF-8, which is what "perltest" was being required to
1589 do for the non-UTF-8 and UTF-8 tests, respectively. Essentially what you
1590 can't do is switch easily at run time between having the "use utf8;" pragma
1591 or not. In the end, I fudged it by using the RunPerlTest script to insert
1592 "use utf8;" explicitly for the UTF-8 tests.
1594 7. In multiline (/m) mode, PCRE was matching ^ after a terminating newline at
1595 the end of the subject string, contrary to the documentation and to what
1596 Perl does. This was true of both matching functions. Now it matches only at
1597 the start of the subject and immediately after *internal* newlines.
1599 8. A call of pcre_fullinfo() from pcretest to get the option bits was passing
1600 a pointer to an int instead of a pointer to an unsigned long int. This
1601 caused problems on 64-bit systems.
1603 9. Applied a patch from the folks at Google to pcrecpp.cc, to fix "another
1604 instance of the 'standard' template library not being so standard".
1606 10. There was no check on the number of named subpatterns nor the maximum
1607 length of a subpattern name. The product of these values is used to compute
1608 the size of the memory block for a compiled pattern. By supplying a very
1609 long subpattern name and a large number of named subpatterns, the size
1610 computation could be caused to overflow. This is now prevented by limiting
1611 the length of names to 32 characters, and the number of named subpatterns
1612 to 10,000.
1614 11. Subpatterns that are repeated with specific counts have to be replicated in
1615 the compiled pattern. The size of memory for this was computed from the
1616 length of the subpattern and the repeat count. The latter is limited to
1617 65535, but there was no limit on the former, meaning that integer overflow
1618 could in principle occur. The compiled length of a repeated subpattern is
1619 now limited to 30,000 bytes in order to prevent this.
1621 12. Added the optional facility to have named substrings with the same name.
1623 13. Added the ability to use a named substring as a condition, using the
1624 Python syntax: (?(name)yes|no). This overloads (?(R)... and names that
1625 are numbers (not recommended). Forward references are permitted.
1627 14. Added forward references in named backreferences (if you see what I mean).
1629 15. In UTF-8 mode, with the PCRE_DOTALL option set, a quantified dot in the
1630 pattern could run off the end of the subject. For example, the pattern
1631 "(?s)(.{1,5})"8 did this with the subject "ab".
1633 16. If PCRE_DOTALL or PCRE_MULTILINE were set, pcre_dfa_exec() behaved as if
1634 PCRE_CASELESS was set when matching characters that were quantified with ?
1635 or *.
1637 17. A character class other than a single negated character that had a minimum
1638 but no maximum quantifier - for example [ab]{6,} - was not handled
1639 correctly by pce_dfa_exec(). It would match only one character.
1641 18. A valid (though odd) pattern that looked like a POSIX character
1642 class but used an invalid character after [ (for example [[,abc,]]) caused
1643 pcre_compile() to give the error "Failed: internal error: code overflow" or
1644 in some cases to crash with a glibc free() error. This could even happen if
1645 the pattern terminated after [[ but there just happened to be a sequence of
1646 letters, a binary zero, and a closing ] in the memory that followed.
1648 19. Perl's treatment of octal escapes in the range \400 to \777 has changed
1649 over the years. Originally (before any Unicode support), just the bottom 8
1650 bits were taken. Thus, for example, \500 really meant \100. Nowadays the
1651 output from "man perlunicode" includes this:
1653 The regular expression compiler produces polymorphic opcodes. That
1654 is, the pattern adapts to the data and automatically switches to
1655 the Unicode character scheme when presented with Unicode data--or
1656 instead uses a traditional byte scheme when presented with byte
1657 data.
1659 Sadly, a wide octal escape does not cause a switch, and in a string with
1660 no other multibyte characters, these octal escapes are treated as before.
1661 Thus, in Perl, the pattern /\500/ actually matches \100 but the pattern
1662 /\500|\x{1ff}/ matches \500 or \777 because the whole thing is treated as a
1663 Unicode string.
1665 I have not perpetrated such confusion in PCRE. Up till now, it took just
1666 the bottom 8 bits, as in old Perl. I have now made octal escapes with
1667 values greater than \377 illegal in non-UTF-8 mode. In UTF-8 mode they
1668 translate to the appropriate multibyte character.
1670 29. Applied some refactoring to reduce the number of warnings from Microsoft
1671 and Borland compilers. This has included removing the fudge introduced
1672 seven years ago for the OS/2 compiler (see 2.02/2 below) because it caused
1673 a warning about an unused variable.
1675 21. PCRE has not included VT (character 0x0b) in the set of whitespace
1676 characters since release 4.0, because Perl (from release 5.004) does not.
1677 [Or at least, is documented not to: some releases seem to be in conflict
1678 with the documentation.] However, when a pattern was studied with
1679 pcre_study() and all its branches started with \s, PCRE still included VT
1680 as a possible starting character. Of course, this did no harm; it just
1681 caused an unnecessary match attempt.
1683 22. Removed a now-redundant internal flag bit that recorded the fact that case
1684 dependency changed within the pattern. This was once needed for "required
1685 byte" processing, but is no longer used. This recovers a now-scarce options
1686 bit. Also moved the least significant internal flag bit to the most-
1687 significant bit of the word, which was not previously used (hangover from
1688 the days when it was an int rather than a uint) to free up another bit for
1689 the future.
1691 23. Added support for CRLF line endings as well as CR and LF. As well as the
1692 default being selectable at build time, it can now be changed at runtime
1693 via the PCRE_NEWLINE_xxx flags. There are now options for pcregrep to
1694 specify that it is scanning data with non-default line endings.
1696 24. Changed the definition of CXXLINK to make it agree with the definition of
1697 LINK in the Makefile, by replacing LDFLAGS to CXXFLAGS.
1699 25. Applied Ian Taylor's patches to avoid using another stack frame for tail
1700 recursions. This makes a big different to stack usage for some patterns.
1702 26. If a subpattern containing a named recursion or subroutine reference such
1703 as (?P>B) was quantified, for example (xxx(?P>B)){3}, the calculation of
1704 the space required for the compiled pattern went wrong and gave too small a
1705 value. Depending on the environment, this could lead to "Failed: internal
1706 error: code overflow at offset 49" or "glibc detected double free or
1707 corruption" errors.
1709 27. Applied patches from Google (a) to support the new newline modes and (b) to
1710 advance over multibyte UTF-8 characters in GlobalReplace.
1712 28. Change free() to pcre_free() in pcredemo.c. Apparently this makes a
1713 difference for some implementation of PCRE in some Windows version.
1715 29. Added some extra testing facilities to pcretest:
1717 \q<number> in a data line sets the "match limit" value
1718 \Q<number> in a data line sets the "match recursion limt" value
1719 -S <number> sets the stack size, where <number> is in megabytes
1721 The -S option isn't available for Windows.
1724 Version 6.6 06-Feb-06
1725 ---------------------
1727 1. Change 16(a) for 6.5 broke things, because PCRE_DATA_SCOPE was not defined
1728 in pcreposix.h. I have copied the definition from pcre.h.
1730 2. Change 25 for 6.5 broke compilation in a build directory out-of-tree
1731 because pcre.h is no longer a built file.
1733 3. Added Jeff Friedl's additional debugging patches to pcregrep. These are
1734 not normally included in the compiled code.
1737 Version 6.5 01-Feb-06
1738 ---------------------
1740 1. When using the partial match feature with pcre_dfa_exec(), it was not
1741 anchoring the second and subsequent partial matches at the new starting
1742 point. This could lead to incorrect results. For example, with the pattern
1743 /1234/, partially matching against "123" and then "a4" gave a match.
1745 2. Changes to pcregrep:
1747 (a) All non-match returns from pcre_exec() were being treated as failures
1748 to match the line. Now, unless the error is PCRE_ERROR_NOMATCH, an
1749 error message is output. Some extra information is given for the
1751 probably the only errors that are likely to be caused by users (by
1752 specifying a regex that has nested indefinite repeats, for instance).
1753 If there are more than 20 of these errors, pcregrep is abandoned.
1755 (b) A binary zero was treated as data while matching, but terminated the
1756 output line if it was written out. This has been fixed: binary zeroes
1757 are now no different to any other data bytes.
1759 (c) Whichever of the LC_ALL or LC_CTYPE environment variables is set is
1760 used to set a locale for matching. The --locale=xxxx long option has
1761 been added (no short equivalent) to specify a locale explicitly on the
1762 pcregrep command, overriding the environment variables.
1764 (d) When -B was used with -n, some line numbers in the output were one less
1765 than they should have been.
1767 (e) Added the -o (--only-matching) option.
1769 (f) If -A or -C was used with -c (count only), some lines of context were
1770 accidentally printed for the final match.
1772 (g) Added the -H (--with-filename) option.
1774 (h) The combination of options -rh failed to suppress file names for files
1775 that were found from directory arguments.
1777 (i) Added the -D (--devices) and -d (--directories) options.
1779 (j) Added the -F (--fixed-strings) option.
1781 (k) Allow "-" to be used as a file name for -f as well as for a data file.
1783 (l) Added the --colo(u)r option.
1785 (m) Added Jeffrey Friedl's -S testing option, but within #ifdefs so that it
1786 is not present by default.
1788 3. A nasty bug was discovered in the handling of recursive patterns, that is,
1789 items such as (?R) or (?1), when the recursion could match a number of
1790 alternatives. If it matched one of the alternatives, but subsequently,
1791 outside the recursion, there was a failure, the code tried to back up into
1792 the recursion. However, because of the way PCRE is implemented, this is not
1793 possible, and the result was an incorrect result from the match.
1795 In order to prevent this happening, the specification of recursion has
1796 been changed so that all such subpatterns are automatically treated as
1797 atomic groups. Thus, for example, (?R) is treated as if it were (?>(?R)).
1799 4. I had overlooked the fact that, in some locales, there are characters for
1800 which isalpha() is true but neither isupper() nor islower() are true. In
1801 the fr_FR locale, for instance, the \xAA and \xBA characters (ordmasculine
1802 and ordfeminine) are like this. This affected the treatment of \w and \W
1803 when they appeared in character classes, but not when they appeared outside
1804 a character class. The bit map for "word" characters is now created
1805 separately from the results of isalnum() instead of just taking it from the
1806 upper, lower, and digit maps. (Plus the underscore character, of course.)
1808 5. The above bug also affected the handling of POSIX character classes such as
1809 [[:alpha:]] and [[:alnum:]]. These do not have their own bit maps in PCRE's
1810 permanent tables. Instead, the bit maps for such a class were previously
1811 created as the appropriate unions of the upper, lower, and digit bitmaps.
1812 Now they are created by subtraction from the [[:word:]] class, which has
1813 its own bitmap.
1815 6. The [[:blank:]] character class matches horizontal, but not vertical space.
1816 It is created by subtracting the vertical space characters (\x09, \x0a,
1817 \x0b, \x0c) from the [[:space:]] bitmap. Previously, however, the
1818 subtraction was done in the overall bitmap for a character class, meaning
1819 that a class such as [\x0c[:blank:]] was incorrect because \x0c would not
1820 be recognized. This bug has been fixed.
1822 7. Patches from the folks at Google:
1824 (a) pcrecpp.cc: "to handle a corner case that may or may not happen in
1825 real life, but is still worth protecting against".
1827 (b) pcrecpp.cc: "corrects a bug when negative radixes are used with
1828 regular expressions".
1830 (c) pcre_scanner.cc: avoid use of std::count() because not all systems
1831 have it.
1833 (d) Split off pcrecpparg.h from pcrecpp.h and had the former built by
1834 "configure" and the latter not, in order to fix a problem somebody had
1835 with compiling the Arg class on HP-UX.
1837 (e) Improve the error-handling of the C++ wrapper a little bit.
1839 (f) New tests for checking recursion limiting.
1841 8. The pcre_memmove() function, which is used only if the environment does not
1842 have a standard memmove() function (and is therefore rarely compiled),
1843 contained two bugs: (a) use of int instead of size_t, and (b) it was not
1844 returning a result (though PCRE never actually uses the result).
1846 9. In the POSIX regexec() interface, if nmatch is specified as a ridiculously
1847 large number - greater than INT_MAX/(3*sizeof(int)) - REG_ESPACE is
1848 returned instead of calling malloc() with an overflowing number that would
1849 most likely cause subsequent chaos.
1851 10. The debugging option of pcretest was not showing the NO_AUTO_CAPTURE flag.
1853 11. The POSIX flag REG_NOSUB is now supported. When a pattern that was compiled
1854 with this option is matched, the nmatch and pmatch options of regexec() are
1855 ignored.
1857 12. Added REG_UTF8 to the POSIX interface. This is not defined by POSIX, but is
1858 provided in case anyone wants to the the POSIX interface with UTF-8
1859 strings.
1861 13. Added CXXLDFLAGS to the Makefile parameters to provide settings only on the
1862 C++ linking (needed for some HP-UX environments).
1864 14. Avoid compiler warnings in get_ucpname() when compiled without UCP support
1865 (unused parameter) and in the pcre_printint() function (omitted "default"
1866 switch label when the default is to do nothing).
1868 15. Added some code to make it possible, when PCRE is compiled as a C++
1869 library, to replace subject pointers for pcre_exec() with a smart pointer
1870 class, thus making it possible to process discontinuous strings.
1872 16. The two macros PCRE_EXPORT and PCRE_DATA_SCOPE are confusing, and perform
1873 much the same function. They were added by different people who were trying
1874 to make PCRE easy to compile on non-Unix systems. It has been suggested
1875 that PCRE_EXPORT be abolished now that there is more automatic apparatus
1876 for compiling on Windows systems. I have therefore replaced it with
1877 PCRE_DATA_SCOPE. This is set automatically for Windows; if not set it
1878 defaults to "extern" for C or "extern C" for C++, which works fine on
1879 Unix-like systems. It is now possible to override the value of PCRE_DATA_
1880 SCOPE with something explicit in config.h. In addition:
1882 (a) pcreposix.h still had just "extern" instead of either of these macros;
1883 I have replaced it with PCRE_DATA_SCOPE.
1885 (b) Functions such as _pcre_xclass(), which are internal to the library,
1886 but external in the C sense, all had PCRE_EXPORT in their definitions.
1887 This is apparently wrong for the Windows case, so I have removed it.
1888 (It makes no difference on Unix-like systems.)
1890 17. Added a new limit, MATCH_LIMIT_RECURSION, which limits the depth of nesting
1891 of recursive calls to match(). This is different to MATCH_LIMIT because
1892 that limits the total number of calls to match(), not all of which increase
1893 the depth of recursion. Limiting the recursion depth limits the amount of
1894 stack (or heap if NO_RECURSE is set) that is used. The default can be set
1895 when PCRE is compiled, and changed at run time. A patch from Google adds
1896 this functionality to the C++ interface.
1898 18. Changes to the handling of Unicode character properties:
1900 (a) Updated the table to Unicode 4.1.0.
1902 (b) Recognize characters that are not in the table as "Cn" (undefined).
1904 (c) I revised the way the table is implemented to a much improved format
1905 which includes recognition of ranges. It now supports the ranges that
1906 are defined in UnicodeData.txt, and it also amalgamates other
1907 characters into ranges. This has reduced the number of entries in the
1908 table from around 16,000 to around 3,000, thus reducing its size
1909 considerably. I realized I did not need to use a tree structure after
1910 all - a binary chop search is just as efficient. Having reduced the
1911 number of entries, I extended their size from 6 bytes to 8 bytes to
1912 allow for more data.
1914 (d) Added support for Unicode script names via properties such as \p{Han}.
1916 19. In UTF-8 mode, a backslash followed by a non-Ascii character was not
1917 matching that character.
1919 20. When matching a repeated Unicode property with a minimum greater than zero,
1920 (for example \pL{2,}), PCRE could look past the end of the subject if it
1921 reached it while seeking the minimum number of characters. This could
1922 happen only if some of the characters were more than one byte long, because
1923 there is a check for at least the minimum number of bytes.
1925 21. Refactored the implementation of \p and \P so as to be more general, to
1926 allow for more different types of property in future. This has changed the
1927 compiled form incompatibly. Anybody with saved compiled patterns that use
1928 \p or \P will have to recompile them.
1930 22. Added "Any" and "L&" to the supported property types.
1932 23. Recognize \x{...} as a code point specifier, even when not in UTF-8 mode,
1933 but give a compile time error if the value is greater than 0xff.
1935 24. The man pages for pcrepartial, pcreprecompile, and pcre_compile2 were
1936 accidentally not being installed or uninstalled.
1938 25. The pcre.h file was built from pcre.h.in, but the only changes that were
1939 made were to insert the current release number. This seemed silly, because
1940 it made things harder for people building PCRE on systems that don't run
1941 "configure". I have turned pcre.h into a distributed file, no longer built
1942 by "configure", with the version identification directly included. There is
1943 no longer a pcre.h.in file.
1945 However, this change necessitated a change to the pcre-config script as
1946 well. It is built from pcre-config.in, and one of the substitutions was the
1947 release number. I have updated configure.ac so that ./configure now finds
1948 the release number by grepping pcre.h.
1950 26. Added the ability to run the tests under valgrind.
1953 Version 6.4 05-Sep-05
1954 ---------------------
1956 1. Change 6.0/10/(l) to pcregrep introduced a bug that caused separator lines
1957 "--" to be printed when multiple files were scanned, even when none of the
1958 -A, -B, or -C options were used. This is not compatible with Gnu grep, so I
1959 consider it to be a bug, and have restored the previous behaviour.
1961 2. A couple of code tidies to get rid of compiler warnings.
1963 3. The pcretest program used to cheat by referring to symbols in the library
1964 whose names begin with _pcre_. These are internal symbols that are not
1965 really supposed to be visible externally, and in some environments it is
1966 possible to suppress them. The cheating is now confined to including
1967 certain files from the library's source, which is a bit cleaner.
1969 4. Renamed pcre.in as pcre.h.in to go with pcrecpp.h.in; it also makes the
1970 file's purpose clearer.
1972 5. Reorganized pcre_ucp_findchar().
1975 Version 6.3 15-Aug-05
1976 ---------------------
1978 1. The file libpcre.pc.in did not have general read permission in the tarball.
1980 2. There were some problems when building without C++ support:
1982 (a) If C++ support was not built, "make install" and "make test" still
1983 tried to test it.
1985 (b) There were problems when the value of CXX was explicitly set. Some
1986 changes have been made to try to fix these, and ...
1988 (c) --disable-cpp can now be used to explicitly disable C++ support.
1990 (d) The use of @CPP_OBJ@ directly caused a blank line preceded by a
1991 backslash in a target when C++ was disabled. This confuses some
1992 versions of "make", apparently. Using an intermediate variable solves
1993 this. (Same for CPP_LOBJ.)
1995 3. $(LINK_FOR_BUILD) now includes $(CFLAGS_FOR_BUILD) and $(LINK)
1996 (non-Windows) now includes $(CFLAGS) because these flags are sometimes
1997 necessary on certain architectures.
1999 4. Added a setting of -export-symbols-regex to the link command to remove
2000 those symbols that are exported in the C sense, but actually are local
2001 within the library, and not documented. Their names all begin with
2002 "_pcre_". This is not a perfect job, because (a) we have to except some
2003 symbols that pcretest ("illegally") uses, and (b) the facility isn't always
2004 available (and never for static libraries). I have made a note to try to
2005 find a way round (a) in the future.
2008 Version 6.2 01-Aug-05
2009 ---------------------
2011 1. There was no test for integer overflow of quantifier values. A construction
2012 such as {1111111111111111} would give undefined results. What is worse, if
2013 a minimum quantifier for a parenthesized subpattern overflowed and became
2014 negative, the calculation of the memory size went wrong. This could have
2015 led to memory overwriting.
2017 2. Building PCRE using VPATH was broken. Hopefully it is now fixed.
2019 3. Added "b" to the 2nd argument of fopen() in dftables.c, for non-Unix-like
2020 operating environments where this matters.
2022 4. Applied Giuseppe Maxia's patch to add additional features for controlling
2023 PCRE options from within the C++ wrapper.
2025 5. Named capturing subpatterns were not being correctly counted when a pattern
2026 was compiled. This caused two problems: (a) If there were more than 100
2027 such subpatterns, the calculation of the memory needed for the whole
2028 compiled pattern went wrong, leading to an overflow error. (b) Numerical
2029 back references of the form \12, where the number was greater than 9, were
2030 not recognized as back references, even though there were sufficient
2031 previous subpatterns.
2033 6. Two minor patches to pcrecpp.cc in order to allow it to compile on older
2034 versions of gcc, e.g. 2.95.4.
2037 Version 6.1 21-Jun-05
2038 ---------------------
2040 1. There was one reference to the variable "posix" in pcretest.c that was not
2041 surrounded by "#if !defined NOPOSIX".
2043 2. Make it possible to compile pcretest without DFA support, UTF8 support, or
2044 the cross-check on the old pcre_info() function, for the benefit of the
2045 cut-down version of PCRE that is currently imported into Exim.
2047 3. A (silly) pattern starting with (?i)(?-i) caused an internal space
2048 allocation error. I've done the easy fix, which wastes 2 bytes for sensible
2049 patterns that start (?i) but I don't think that matters. The use of (?i) is
2050 just an example; this all applies to the other options as well.
2052 4. Since libtool seems to echo the compile commands it is issuing, the output
2053 from "make" can be reduced a bit by putting "@" in front of each libtool
2054 compile command.
2056 5. Patch from the folks at Google for configure.in to be a bit more thorough
2057 in checking for a suitable C++ installation before trying to compile the
2058 C++ stuff. This should fix a reported problem when a compiler was present,
2059 but no suitable headers.
2061 6. The man pages all had just "PCRE" as their title. I have changed them to
2062 be the relevant file name. I have also arranged that these names are
2063 retained in the file doc/pcre.txt, which is a concatenation in text format
2064 of all the man pages except the little individual ones for each function.
2066 7. The NON-UNIX-USE file had not been updated for the different set of source
2067 files that come with release 6. I also added a few comments about the C++
2068 wrapper.
2071 Version 6.0 07-Jun-05
2072 ---------------------
2074 1. Some minor internal re-organization to help with my DFA experiments.
2076 2. Some missing #ifdef SUPPORT_UCP conditionals in pcretest and printint that
2077 didn't matter for the library itself when fully configured, but did matter
2078 when compiling without UCP support, or within Exim, where the ucp files are
2079 not imported.
2081 3. Refactoring of the library code to split up the various functions into
2082 different source modules. The addition of the new DFA matching code (see
2083 below) to a single monolithic source would have made it really too
2084 unwieldy, quite apart from causing all the code to be include in a
2085 statically linked application, when only some functions are used. This is
2086 relevant even without the DFA addition now that patterns can be compiled in
2087 one application and matched in another.
2089 The downside of splitting up is that there have to be some external
2090 functions and data tables that are used internally in different modules of
2091 the library but which are not part of the API. These have all had their
2092 names changed to start with "_pcre_" so that they are unlikely to clash
2093 with other external names.
2095 4. Added an alternate matching function, pcre_dfa_exec(), which matches using
2096 a different (DFA) algorithm. Although it is slower than the original
2097 function, it does have some advantages for certain types of matching
2098 problem.
2100 5. Upgrades to pcretest in order to test the features of pcre_dfa_exec(),
2101 including restarting after a partial match.
2103 6. A patch for pcregrep that defines INVALID_FILE_ATTRIBUTES if it is not
2104 defined when compiling for Windows was sent to me. I have put it into the
2105 code, though I have no means of testing or verifying it.
2107 7. Added the pcre_refcount() auxiliary function.
2109 8. Added the PCRE_FIRSTLINE option. This constrains an unanchored pattern to
2110 match before or at the first newline in the subject string. In pcretest,
2111 the /f option on a pattern can be used to set this.
2113 9. A repeated \w when used in UTF-8 mode with characters greater than 256
2114 would behave wrongly. This has been present in PCRE since release 4.0.
2116 10. A number of changes to the pcregrep command:
2118 (a) Refactored how -x works; insert ^(...)$ instead of setting
2119 PCRE_ANCHORED and checking the length, in preparation for adding
2120 something similar for -w.
2122 (b) Added the -w (match as a word) option.
2124 (c) Refactored the way lines are read and buffered so as to have more
2125 than one at a time available.
2127 (d) Implemented a pcregrep test script.
2129 (e) Added the -M (multiline match) option. This allows patterns to match
2130 over several lines of the subject. The buffering ensures that at least
2131 8K, or the rest of the document (whichever is the shorter) is available
2132 for matching (and similarly the previous 8K for lookbehind assertions).
2134 (f) Changed the --help output so that it now says
2136 -w, --word-regex(p)
2138 instead of two lines, one with "regex" and the other with "regexp"
2139 because that confused at least one person since the short forms are the
2140 same. (This required a bit of code, as the output is generated
2141 automatically from a table. It wasn't just a text change.)
2143 (g) -- can be used to terminate pcregrep options if the next thing isn't an
2144 option but starts with a hyphen. Could be a pattern or a path name
2145 starting with a hyphen, for instance.
2147 (h) "-" can be given as a file name to represent stdin.
2149 (i) When file names are being printed, "(standard input)" is used for
2150 the standard input, for compatibility with GNU grep. Previously
2151 "<stdin>" was used.
2153 (j) The option --label=xxx can be used to supply a name to be used for
2154 stdin when file names are being printed. There is no short form.
2156 (k) Re-factored the options decoding logic because we are going to add
2157 two more options that take data. Such options can now be given in four
2158 different ways, e.g. "-fname", "-f name", "--file=name", "--file name".
2160 (l) Added the -A, -B, and -C options for requesting that lines of context
2161 around matches be printed.
2163 (m) Added the -L option to print the names of files that do not contain
2164 any matching lines, that is, the complement of -l.
2166 (n) The return code is 2 if any file cannot be opened, but pcregrep does
2167 continue to scan other files.
2169 (o) The -s option was incorrectly implemented. For compatibility with other
2170 greps, it now suppresses the error message for a non-existent or non-
2171 accessible file (but not the return code). There is a new option called
2172 -q that suppresses the output of matching lines, which was what -s was
2173 previously doing.
2175 (p) Added --include and --exclude options to specify files for inclusion
2176 and exclusion when recursing.
2178 11. The Makefile was not using the Autoconf-supported LDFLAGS macro properly.
2179 Hopefully, it now does.
2181 12. Missing cast in pcre_study().
2183 13. Added an "uninstall" target to the makefile.
2185 14. Replaced "extern" in the function prototypes in Makefile.in with
2186 "PCRE_DATA_SCOPE", which defaults to 'extern' or 'extern "C"' in the Unix
2187 world, but is set differently for Windows.
2189 15. Added a second compiling function called pcre_compile2(). The only
2190 difference is that it has an extra argument, which is a pointer to an
2191 integer error code. When there is a compile-time failure, this is set
2192 non-zero, in addition to the error test pointer being set to point to an
2193 error message. The new argument may be NULL if no error number is required
2194 (but then you may as well call pcre_compile(), which is now just a
2195 wrapper). This facility is provided because some applications need a
2196 numeric error indication, but it has also enabled me to tidy up the way
2197 compile-time errors are handled in the POSIX wrapper.
2199 16. Added VPATH=.libs to the makefile; this should help when building with one
2200 prefix path and installing with another. (Or so I'm told by someone who
2201 knows more about this stuff than I do.)
2203 17. Added a new option, REG_DOTALL, to the POSIX function regcomp(). This
2204 passes PCRE_DOTALL to the pcre_compile() function, making the "." character
2205 match everything, including newlines. This is not POSIX-compatible, but
2206 somebody wanted the feature. From pcretest it can be activated by using
2207 both the P and the s flags.
2209 18. AC_PROG_LIBTOOL appeared twice in Makefile.in. Removed one.
2211 19. libpcre.pc was being incorrectly installed as executable.
2213 20. A couple of places in pcretest check for end-of-line by looking for '\n';
2214 it now also looks for '\r' so that it will work unmodified on Windows.
2216 21. Added Google's contributed C++ wrapper to the distribution.
2218 22. Added some untidy missing memory free() calls in pcretest, to keep
2219 Electric Fence happy when testing.
2223 Version 5.0 13-Sep-04
2224 ---------------------
2226 1. Internal change: literal characters are no longer packed up into items
2227 containing multiple characters in a single byte-string. Each character
2228 is now matched using a separate opcode. However, there may be more than one
2229 byte in the character in UTF-8 mode.
2231 2. The pcre_callout_block structure has two new fields: pattern_position and
2232 next_item_length. These contain the offset in the pattern to the next match
2233 item, and its length, respectively.
2235 3. The PCRE_AUTO_CALLOUT option for pcre_compile() requests the automatic
2236 insertion of callouts before each pattern item. Added the /C option to
2237 pcretest to make use of this.
2239 4. On the advice of a Windows user, the lines
2241 #if defined(_WIN32) || defined(WIN32)
2242 _setmode( _fileno( stdout ), 0x8000 );
2243 #endif /* defined(_WIN32) || defined(WIN32) */
2245 have been added to the source of pcretest. This apparently does useful
2246 magic in relation to line terminators.
2248 5. Changed "r" and "w" in the calls to fopen() in pcretest to "rb" and "wb"
2249 for the benefit of those environments where the "b" makes a difference.
2251 6. The icc compiler has the same options as gcc, but "configure" doesn't seem
2252 to know about it. I have put a hack into configure.in that adds in code
2253 to set GCC=yes if CC=icc. This seems to end up at a point in the
2254 generated configure script that is early enough to affect the setting of
2255 compiler options, which is what is needed, but I have no means of testing
2256 whether it really works. (The user who reported this had patched the
2257 generated configure script, which of course I cannot do.)
2259 LATER: After change 22 below (new libtool files), the configure script
2260 seems to know about icc (and also ecc). Therefore, I have commented out
2261 this hack in configure.in.
2263 7. Added support for pkg-config (2 patches were sent in).
2265 8. Negated POSIX character classes that used a combination of internal tables
2266 were completely broken. These were [[:^alpha:]], [[:^alnum:]], and
2267 [[:^ascii]]. Typically, they would match almost any characters. The other
2268 POSIX classes were not broken in this way.
2270 9. Matching the pattern "\b.*?" against "ab cd", starting at offset 1, failed
2271 to find the match, as PCRE was deluded into thinking that the match had to
2272 start at the start point or following a newline. The same bug applied to
2273 patterns with negative forward assertions or any backward assertions
2274 preceding ".*" at the start, unless the pattern required a fixed first
2275 character. This was a failing pattern: "(?!.bcd).*". The bug is now fixed.
2277 10. In UTF-8 mode, when moving forwards in the subject after a failed match
2278 starting at the last subject character, bytes beyond the end of the subject
2279 string were read.
2281 11. Renamed the variable "class" as "classbits" to make life easier for C++
2282 users. (Previously there was a macro definition, but it apparently wasn't
2283 enough.)
2285 12. Added the new field "tables" to the extra data so that tables can be passed
2286 in at exec time, or the internal tables can be re-selected. This allows
2287 a compiled regex to be saved and re-used at a later time by a different
2288 program that might have everything at different addresses.
2290 13. Modified the pcre-config script so that, when run on Solaris, it shows a
2291 -R library as well as a -L library.
2293 14. The debugging options of pcretest (-d on the command line or D on a
2294 pattern) showed incorrect output for anything following an extended class
2295 that contained multibyte characters and which was followed by a quantifier.
2297 15. Added optional support for general category Unicode character properties
2298 via the \p, \P, and \X escapes. Unicode property support implies UTF-8
2299 support. It adds about 90K to the size of the library. The meanings of the
2300 inbuilt class escapes such as \d and \s have NOT been changed.
2302 16. Updated pcredemo.c to include calls to free() to release the memory for the
2303 compiled pattern.
2305 17. The generated file chartables.c was being created in the source directory
2306 instead of in the building directory. This caused the build to fail if the
2307 source directory was different from the building directory, and was
2308 read-only.
2310 18. Added some sample Win commands from Mark Tetrode into the NON-UNIX-USE
2311 file. No doubt somebody will tell me if they don't make sense... Also added
2312 Dan Mooney's comments about building on OpenVMS.
2314 19. Added support for partial matching via the PCRE_PARTIAL option for
2315 pcre_exec() and the \P data escape in pcretest.
2317 20. Extended pcretest with 3 new pattern features:
2319 (i) A pattern option of the form ">rest-of-line" causes pcretest to
2320 write the compiled pattern to the file whose name is "rest-of-line".
2321 This is a straight binary dump of the data, with the saved pointer to
2322 the character tables forced to be NULL. The study data, if any, is
2323 written too. After writing, pcretest reads a new pattern.
2325 (ii) If, instead of a pattern, "<rest-of-line" is given, pcretest reads a
2326 compiled pattern from the given file. There must not be any
2327 occurrences of "<" in the file name (pretty unlikely); if there are,
2328 pcretest will instead treat the initial "<" as a pattern delimiter.
2329 After reading in the pattern, pcretest goes on to read data lines as
2330 usual.
2332 (iii) The F pattern option causes pcretest to flip the bytes in the 32-bit
2333 and 16-bit fields in a compiled pattern, to simulate a pattern that
2334 was compiled on a host of opposite endianness.
2336 21. The pcre-exec() function can now cope with patterns that were compiled on
2337 hosts of opposite endianness, with this restriction:
2339 As for any compiled expression that is saved and used later, the tables
2340 pointer field cannot be preserved; the extra_data field in the arguments
2341 to pcre_exec() should be used to pass in a tables address if a value
2342 other than the default internal tables were used at compile time.
2344 22. Calling pcre_exec() with a negative value of the "ovecsize" parameter is
2345 now diagnosed as an error. Previously, most of the time, a negative number
2346 would have been treated as zero, but if in addition "ovector" was passed as
2347 NULL, a crash could occur.
2349 23. Updated the files ltmain.sh, config.sub, config.guess, and aclocal.m4 with
2350 new versions from the libtool 1.5 distribution (the last one is a copy of
2351 a file called libtool.m4). This seems to have fixed the need to patch
2352 "configure" to support Darwin 1.3 (which I used to do). However, I still
2353 had to patch ltmain.sh to ensure that ${SED} is set (it isn't on my
2354 workstation).
2356 24. Changed the PCRE licence to be the more standard "BSD" licence.
2359 Version 4.5 01-Dec-03
2360 ---------------------
2362 1. There has been some re-arrangement of the code for the match() function so
2363 that it can be compiled in a version that does not call itself recursively.
2364 Instead, it keeps those local variables that need separate instances for
2365 each "recursion" in a frame on the heap, and gets/frees frames whenever it
2366 needs to "recurse". Keeping track of where control must go is done by means
2367 of setjmp/longjmp. The whole thing is implemented by a set of macros that
2368 hide most of the details from the main code, and operates only if
2369 NO_RECURSE is defined while compiling pcre.c. If PCRE is built using the
2370 "configure" mechanism, "--disable-stack-for-recursion" turns on this way of
2371 operating.
2373 To make it easier for callers to provide specially tailored get/free
2374 functions for this usage, two new functions, pcre_stack_malloc, and
2375 pcre_stack_free, are used. They are always called in strict stacking order,
2376 and the size of block requested is always the same.
2378 The PCRE_CONFIG_STACKRECURSE info parameter can be used to find out whether
2379 PCRE has been compiled to use the stack or the heap for recursion. The
2380 -C option of pcretest uses this to show which version is compiled.
2382 A new data escape \S, is added to pcretest; it causes the amounts of store
2383 obtained and freed by both kinds of malloc/free at match time to be added
2384 to the output.
2386 2. Changed the locale test to use "fr_FR" instead of "fr" because that's
2387 what's available on my current Linux desktop machine.
2389 3. When matching a UTF-8 string, the test for a valid string at the start has
2390 been extended. If start_offset is not zero, PCRE now checks that it points
2391 to a byte that is the start of a UTF-8 character. If not, it returns
2392 PCRE_ERROR_BADUTF8_OFFSET (-11). Note: the whole string is still checked;
2393 this is necessary because there may be backward assertions in the pattern.
2394 When matching the same subject several times, it may save resources to use
2395 PCRE_NO_UTF8_CHECK on all but the first call if the string is long.
2397 4. The code for checking the validity of UTF-8 strings has been tightened so
2398 that it rejects (a) strings containing 0xfe or 0xff bytes and (b) strings
2399 containing "overlong sequences".
2401 5. Fixed a bug (appearing twice) that I could not find any way of exploiting!
2402 I had written "if ((digitab[*p++] && chtab_digit) == 0)" where the "&&"
2403 should have been "&", but it just so happened that all the cases this let
2404 through by mistake were picked up later in the function.
2406 6. I had used a variable called "isblank" - this is a C99 function, causing
2407 some compilers to warn. To avoid this, I renamed it (as "blankclass").
2409 7. Cosmetic: (a) only output another newline at the end of pcretest if it is
2410 prompting; (b) run "./pcretest /dev/null" at the start of the test script
2411 so the version is shown; (c) stop "make test" echoing "./RunTest".
2413 8. Added patches from David Burgess to enable PCRE to run on EBCDIC systems.
2415 9. The prototype for memmove() for systems that don't have it was using
2416 size_t, but the inclusion of the header that defines size_t was later. I've
2417 moved the #includes for the C headers earlier to avoid this.
2419 10. Added some adjustments to the code to make it easier to compiler on certain
2420 special systems:
2422 (a) Some "const" qualifiers were missing.
2423 (b) Added the macro EXPORT before all exported functions; by default this
2424 is defined to be empty.
2425 (c) Changed the dftables auxiliary program (that builds chartables.c) so
2426 that it reads its output file name as an argument instead of writing
2427 to the standard output and assuming this can be redirected.
2429 11. In UTF-8 mode, if a recursive reference (e.g. (?1)) followed a character
2430 class containing characters with values greater than 255, PCRE compilation
2431 went into a loop.
2433 12. A recursive reference to a subpattern that was within another subpattern
2434 that had a minimum quantifier of zero caused PCRE to crash. For example,
2435 (x(y(?2))z)? provoked this bug with a subject that got as far as the
2436 recursion. If the recursively-called subpattern itself had a zero repeat,
2437 that was OK.
2439 13. In pcretest, the buffer for reading a data line was set at 30K, but the
2440 buffer into which it was copied (for escape processing) was still set at
2441 1024, so long lines caused crashes.
2443 14. A pattern such as /[ab]{1,3}+/ failed to compile, giving the error
2444 "internal error: code overflow...". This applied to any character class
2445 that was followed by a possessive quantifier.
2447 15. Modified the Makefile to add libpcre.la as a prerequisite for
2448 libpcreposix.la because I was told this is needed for a parallel build to
2449 work.
2451 16. If a pattern that contained .* following optional items at the start was
2452 studied, the wrong optimizing data was generated, leading to matching
2453 errors. For example, studying /[ab]*.*c/ concluded, erroneously, that any
2454 matching string must start with a or b or c. The correct conclusion for
2455 this pattern is that a match can start with any character.
2458 Version 4.4 13-Aug-03
2459 ---------------------
2461 1. In UTF-8 mode, a character class containing characters with values between
2462 127 and 255 was not handled correctly if the compiled pattern was studied.
2463 In fixing this, I have also improved the studying algorithm for such
2464 classes (slightly).
2466 2. Three internal functions had redundant arguments passed to them. Removal
2467 might give a very teeny performance improvement.
2469 3. Documentation bug: the value of the capture_top field in a callout is *one
2470 more than* the number of the hightest numbered captured substring.
2472 4. The Makefile linked pcretest and pcregrep with -lpcre, which could result
2473 in incorrectly linking with a previously installed version. They now link
2474 explicitly with libpcre.la.
2476 5. configure.in no longer needs to recognize Cygwin specially.
2478 6. A problem in pcre.in for Windows platforms is fixed.
2480 7. If a pattern was successfully studied, and the -d (or /D) flag was given to
2481 pcretest, it used to include the size of the study block as part of its
2482 output. Unfortunately, the structure contains a field that has a different
2483 size on different hardware architectures. This meant that the tests that
2484 showed this size failed. As the block is currently always of a fixed size,
2485 this information isn't actually particularly useful in pcretest output, so
2486 I have just removed it.
2488 8. Three pre-processor statements accidentally did not start in column 1.
2489 Sadly, there are *still* compilers around that complain, even though
2490 standard C has not required this for well over a decade. Sigh.
2492 9. In pcretest, the code for checking callouts passed small integers in the
2493 callout_data field, which is a void * field. However, some picky compilers
2494 complained about the casts involved for this on 64-bit systems. Now
2495 pcretest passes the address of the small integer instead, which should get
2496 rid of the warnings.
2498 10. By default, when in UTF-8 mode, PCRE now checks for valid UTF-8 strings at
2499 both compile and run time, and gives an error if an invalid UTF-8 sequence
2500 is found. There is a option for disabling this check in cases where the
2501 string is known to be correct and/or the maximum performance is wanted.
2503 11. In response to a bug report, I changed one line in Makefile.in from
2505 -Wl,--out-implib,.libs/lib@WIN_PREFIX@pcreposix.dll.a \
2506 to
2507 -Wl,--out-implib,.libs/@WIN_PREFIX@libpcreposix.dll.a \
2509 to look similar to other lines, but I have no way of telling whether this
2510 is the right thing to do, as I do not use Windows. No doubt I'll get told
2511 if it's wrong...
2514 Version 4.3 21-May-03
2515 ---------------------
2517 1. Two instances of @WIN_PREFIX@ omitted from the Windows targets in the
2518 Makefile.
2520 2. Some refactoring to improve the quality of the code:
2522 (i) The utf8_table... variables are now declared "const".
2524 (ii) The code for \cx, which used the "case flipping" table to upper case
2525 lower case letters, now just substracts 32. This is ASCII-specific,
2526 but the whole concept of \cx is ASCII-specific, so it seems
2527 reasonable.
2529 (iii) PCRE was using its character types table to recognize decimal and
2530 hexadecimal digits in the pattern. This is silly, because it handles
2531 only 0-9, a-f, and A-F, but the character types table is locale-
2532 specific, which means strange things might happen. A private
2533 table is now used for this - though it costs 256 bytes, a table is
2534 much faster than multiple explicit tests. Of course, the standard
2535 character types table is still used for matching digits in subject
2536 strings against \d.
2538 (iv) Strictly, the identifier ESC_t is reserved by POSIX (all identifiers
2539 ending in _t are). So I've renamed it as ESC_tee.
2541 3. The first argument for regexec() in the POSIX wrapper should have been
2542 defined as "const".
2544 4. Changed pcretest to use malloc() for its buffers so that they can be
2545 Electric Fenced for debugging.
2547 5. There were several places in the code where, in UTF-8 mode, PCRE would try
2548 to read one or more bytes before the start of the subject string. Often this
2549 had no effect on PCRE's behaviour, but in some circumstances it could
2550 provoke a segmentation fault.
2552 6. A lookbehind at the start of a pattern in UTF-8 mode could also cause PCRE
2553 to try to read one or more bytes before the start of the subject string.
2555 7. A lookbehind in a pattern matched in non-UTF-8 mode on a PCRE compiled with
2556 UTF-8 support could misbehave in various ways if the subject string
2557 contained bytes with the 0x80 bit set and the 0x40 bit unset in a lookbehind
2558 area. (PCRE was not checking for the UTF-8 mode flag, and trying to move
2559 back over UTF-8 characters.)
2562 Version 4.2 14-Apr-03
2563 ---------------------
2565 1. Typo "#if SUPPORT_UTF8" instead of "#ifdef SUPPORT_UTF8" fixed.
2567 2. Changes to the building process, supplied by Ronald Landheer-Cieslak
2568 [ON_WINDOWS]: new variable, "#" on non-Windows platforms
2569 [NOT_ON_WINDOWS]: new variable, "#" on Windows platforms
2570 [WIN_PREFIX]: new variable, "cyg" for Cygwin
2571 * Makefile.in: use autoconf substitution for OBJEXT, EXEEXT, BUILD_OBJEXT
2573 Note: automatic setting of the BUILD variables is not yet working
2574 set CPPFLAGS and BUILD_CPPFLAGS (but don't use yet) - should be used at
2575 compile-time but not at link-time
2576 [LINK]: use for linking executables only
2577 make different versions for Windows and non-Windows
2578 [LINKLIB]: new variable, copy of UNIX-style LINK, used for linking
2579 libraries
2580 [LINK_FOR_BUILD]: new variable
2581 [OBJEXT]: use throughout
2582 [EXEEXT]: use throughout
2583 <winshared>: new target
2584 <wininstall>: new target
2585 <dftables.o>: use native compiler
2586 <dftables>: use native linker
2587 <install>: handle Windows platform correctly
2588 <clean>: ditto
2589 <check>: ditto
2590 copy DLL to top builddir before testing
2592 As part of these changes, -no-undefined was removed again. This was reported
2593 to give trouble on HP-UX 11.0, so getting rid of it seems like a good idea
2594 in any case.
2596 3. Some tidies to get rid of compiler warnings:
2598 . In the match_data structure, match_limit was an unsigned long int, whereas
2599 match_call_count was an int. I've made them both unsigned long ints.
2601 . In pcretest the fact that a const uschar * doesn't automatically cast to
2602 a void * provoked a warning.
2604 . Turning on some more compiler warnings threw up some "shadow" variables
2605 and a few more missing casts.
2607 4. If PCRE was complied with UTF-8 support, but called without the PCRE_UTF8
2608 option, a class that contained a single character with a value between 128
2609 and 255 (e.g. /[\xFF]/) caused PCRE to crash.
2611 5. If PCRE was compiled with UTF-8 support, but called without the PCRE_UTF8
2612 option, a class that contained several characters, but with at least one
2613 whose value was between 128 and 255 caused PCRE to crash.
2616 Version 4.1 12-Mar-03
2617 ---------------------
2619 1. Compiling with gcc -pedantic found a couple of places where casts were
2620 needed, and a string in dftables.c that was longer than standard compilers are
2621 required to support.
2623 2. Compiling with Sun's compiler found a few more places where the code could
2624 be tidied up in order to avoid warnings.
2626 3. The variables for cross-compiling were called HOST_CC and HOST_CFLAGS; the
2627 first of these names is deprecated in the latest Autoconf in favour of the name
2628 CC_FOR_BUILD, because "host" is typically used to mean the system on which the
2629 compiled code will be run. I can't find a reference for HOST_CFLAGS, but by
2630 analogy I have changed it to CFLAGS_FOR_BUILD.
2632 4. Added -no-undefined to the linking command in the Makefile, because this is
2633 apparently helpful for Windows. To make it work, also added "-L. -lpcre" to the
2634 linking step for the pcreposix library.
2636 5. PCRE was failing to diagnose the case of two named groups with the same
2637 name.
2639 6. A problem with one of PCRE's optimizations was discovered. PCRE remembers a
2640 literal character that is needed in the subject for a match, and scans along to
2641 ensure that it is present before embarking on the full matching process. This
2642 saves time in cases of nested unlimited repeats that are never going to match.
2643 Problem: the scan can take a lot of time if the subject is very long (e.g.
2644 megabytes), thus penalizing straightforward matches. It is now done only if the
2645 amount of subject to be scanned is less than 1000 bytes.
2647 7. A lesser problem with the same optimization is that it was recording the
2648 first character of an anchored pattern as "needed", thus provoking a search
2649 right along the subject, even when the first match of the pattern was going to
2650 fail. The "needed" character is now not set for anchored patterns, unless it
2651 follows something in the pattern that is of non-fixed length. Thus, it still
2652 fulfils its original purpose of finding quick non-matches in cases of nested
2653 unlimited repeats, but isn't used for simple anchored patterns such as /^abc/.
2656 Version 4.0 17-Feb-03
2657 ---------------------
2659 1. If a comment in an extended regex that started immediately after a meta-item
2660 extended to the end of string, PCRE compiled incorrect data. This could lead to
2661 all kinds of weird effects. Example: /#/ was bad; /()#/ was bad; /a#/ was not.
2663 2. Moved to autoconf 2.53 and libtool 1.4.2.
2665 3. Perl 5.8 no longer needs "use utf8" for doing UTF-8 things. Consequently,
2666 the special perltest8 script is no longer needed - all the tests can be run
2667 from a single perltest script.
2669 4. From 5.004, Perl has not included the VT character (0x0b) in the set defined
2670 by \s. It has now been removed in PCRE. This means it isn't recognized as
2671 whitespace in /x regexes too, which is the same as Perl. Note that the POSIX
2672 class [:space:] *does* include VT, thereby creating a mess.
2674 5. Added the class [:blank:] (a GNU extension from Perl 5.8) to match only
2675 space and tab.
2677 6. Perl 5.005 was a long time ago. It's time to amalgamate the tests that use
2678 its new features into the main test script, reducing the number of scripts.
2680 7. Perl 5.8 has changed the meaning of patterns like /a(?i)b/. Earlier versions
2681 were backward compatible, and made the (?i) apply to the whole pattern, as if
2682 /i were given. Now it behaves more logically, and applies the option setting
2683 only to what follows. PCRE has been changed to follow suit. However, if it
2684 finds options settings right at the start of the pattern, it extracts them into
2685 the global options, as before. Thus, they show up in the info data.
2687 8. Added support for the \Q...\E escape sequence. Characters in between are
2688 treated as literals. This is slightly different from Perl in that $ and @ are
2689 also handled as literals inside the quotes. In Perl, they will cause variable
2690 interpolation. Note the following examples:
2692 Pattern PCRE matches Perl matches
2694 \Qabc$xyz\E abc$xyz abc followed by the contents of $xyz
2695 \Qabc\$xyz\E abc\$xyz abc\$xyz
2696 \Qabc\E\$\Qxyz\E abc$xyz abc$xyz
2698 For compatibility with Perl, \Q...\E sequences are recognized inside character
2699 classes as well as outside them.
2701 9. Re-organized 3 code statements in pcretest to avoid "overflow in
2702 floating-point constant arithmetic" warnings from a Microsoft compiler. Added a
2703 (size_t) cast to one statement in pcretest and one in pcreposix to avoid
2704 signed/unsigned warnings.
2706 10. SunOS4 doesn't have strtoul(). This was used only for unpicking the -o
2707 option for pcretest, so I've replaced it by a simple function that does just
2708 that job.
2710 11. pcregrep was ending with code 0 instead of 2 for the commands "pcregrep" or
2711 "pcregrep -".
2713 12. Added "possessive quantifiers" ?+, *+, ++, and {,}+ which come from Sun's
2714 Java package. This provides some syntactic sugar for simple cases of what my
2715 documentation calls "once-only subpatterns". A pattern such as x*+ is the same
2716 as (?>x*). In other words, if what is inside (?>...) is just a single repeated
2717 item, you can use this simplified notation. Note that only makes sense with
2718 greedy quantifiers. Consequently, the use of the possessive quantifier forces
2719 greediness, whatever the setting of the PCRE_UNGREEDY option.
2721 13. A change of greediness default within a pattern was not taking effect at
2722 the current level for patterns like /(b+(?U)a+)/. It did apply to parenthesized
2723 subpatterns that followed. Patterns like /b+(?U)a+/ worked because the option
2724 was abstracted outside.
2726 14. PCRE now supports the \G assertion. It is true when the current matching
2727 position is at the start point of the match. This differs from \A when the
2728 starting offset is non-zero. Used with the /g option of pcretest (or similar
2729 code), it works in the same way as it does for Perl's /g option. If all
2730 alternatives of a regex begin with \G, the expression is anchored to the start
2731 match position, and the "anchored" flag is set in the compiled expression.
2733 15. Some bugs concerning the handling of certain option changes within patterns
2734 have been fixed. These applied to options other than (?ims). For example,
2735 "a(?x: b c )d" did not match "XabcdY" but did match "Xa b c dY". It should have
2736 been the other way round. Some of this was related to change 7 above.
2738 16. PCRE now gives errors for /[.x.]/ and /[=x=]/ as unsupported POSIX
2739 features, as Perl does. Previously, PCRE gave the warnings only for /[[.x.]]/
2740 and /[[=x=]]/. PCRE now also gives an error for /[:name:]/ because it supports
2741 POSIX classes only within a class (e.g. /[[:alpha:]]/).
2743 17. Added support for Perl's \C escape. This matches one byte, even in UTF8
2744 mode. Unlike ".", it always matches newline, whatever the setting of
2745 PCRE_DOTALL. However, PCRE does not permit \C to appear in lookbehind
2746 assertions. Perl allows it, but it doesn't (in general) work because it can't
2747 calculate the length of the lookbehind. At least, that's the case for Perl
2748 5.8.0 - I've been told they are going to document that it doesn't work in
2749 future.
2751 18. Added an error diagnosis for escapes that PCRE does not support: these are
2752 \L, \l, \N, \P, \p, \U, \u, and \X.
2754 19. Although correctly diagnosing a missing ']' in a character class, PCRE was
2755 reading past the end of the pattern in cases such as /[abcd/.
2757 20. PCRE was getting more memory than necessary for patterns with classes that
2758 contained both POSIX named classes and other characters, e.g. /[[:space:]abc/.
2760 21. Added some code, conditional on #ifdef VPCOMPAT, to make life easier for
2761 compiling PCRE for use with Virtual Pascal.
2763 22. Small fix to the Makefile to make it work properly if the build is done
2764 outside the source tree.
2766 23. Added a new extension: a condition to go with recursion. If a conditional
2767 subpattern starts with (?(R) the "true" branch is used if recursion has
2768 happened, whereas the "false" branch is used only at the top level.
2770 24. When there was a very long string of literal characters (over 255 bytes
2771 without UTF support, over 250 bytes with UTF support), the computation of how
2772 much memory was required could be incorrect, leading to segfaults or other
2773 strange effects.
2775 25. PCRE was incorrectly assuming anchoring (either to start of subject or to
2776 start of line for a non-DOTALL pattern) when a pattern started with (.*) and
2777 there was a subsequent back reference to those brackets. This meant that, for
2778 example, /(.*)\d+\1/ failed to match "abc123bc". Unfortunately, it isn't
2779 possible to check for precisely this case. All we can do is abandon the
2780 optimization if .* occurs inside capturing brackets when there are any back
2781 references whatsoever. (See below for a better fix that came later.)
2783 26. The handling of the optimization for finding the first character of a
2784 non-anchored pattern, and for finding a character that is required later in the
2785 match were failing in some cases. This didn't break the matching; it just
2786 failed to optimize when it could. The way this is done has been re-implemented.
2788 27. Fixed typo in error message for invalid (?R item (it said "(?p").
2790 28. Added a new feature that provides some of the functionality that Perl
2791 provides with (?{...}). The facility is termed a "callout". The way it is done
2792 in PCRE is for the caller to provide an optional function, by setting
2793 pcre_callout to its entry point. Like pcre_malloc and pcre_free, this is a
2794 global variable. By default it is unset, which disables all calling out. To get
2795 the function called, the regex must include (?C) at appropriate points. This
2796 is, in fact, equivalent to (?C0), and any number <= 255 may be given with (?C).
2797 This provides a means of identifying different callout points. When PCRE
2798 reaches such a point in the regex, if pcre_callout has been set, the external
2799 function is called. It is provided with data in a structure called
2800 pcre_callout_block, which is defined in pcre.h. If the function returns 0,
2801 matching continues; if it returns a non-zero value, the match at the current
2802 point fails. However, backtracking will occur if possible. [This was changed
2803 later and other features added - see item 49 below.]
2805 29. pcretest is upgraded to test the callout functionality. It provides a
2806 callout function that displays information. By default, it shows the start of
2807 the match and the current position in the text. There are some new data escapes
2808 to vary what happens:
2810 \C+ in addition, show current contents of captured substrings
2811 \C- do not supply a callout function
2812 \C!n return 1 when callout number n is reached
2813 \C!n!m return 1 when callout number n is reached for the mth time
2815 30. If pcregrep was called with the -l option and just a single file name, it
2816 output "<stdin>" if a match was found, instead of the file name.
2818 31. Improve the efficiency of the POSIX API to PCRE. If the number of capturing
2819 slots is less than POSIX_MALLOC_THRESHOLD, use a block on the stack to pass to
2820 pcre_exec(). This saves a malloc/free per call. The default value of
2821 POSIX_MALLOC_THRESHOLD is 10; it can be changed by --with-posix-malloc-threshold
2822 when configuring.
2824 32. The default maximum size of a compiled pattern is 64K. There have been a
2825 few cases of people hitting this limit. The code now uses macros to handle the
2826 storing of links as offsets within the compiled pattern. It defaults to 2-byte
2827 links, but this can be changed to 3 or 4 bytes by --with-link-size when
2828 configuring. Tests 2 and 5 work only with 2-byte links because they output
2829 debugging information about compiled patterns.
2831 33. Internal code re-arrangements:
2833 (a) Moved the debugging function for printing out a compiled regex into
2834 its own source file (printint.c) and used #include to pull it into
2835 pcretest.c and, when DEBUG is defined, into pcre.c, instead of having two
2836 separate copies.
2838 (b) Defined the list of op-code names for debugging as a macro in
2839 internal.h so that it is next to the definition of the opcodes.
2841 (c) Defined a table of op-code lengths for simpler skipping along compiled
2842 code. This is again a macro in internal.h so that it is next to the
2843 definition of the opcodes.
2845 34. Added support for recursive calls to individual subpatterns, along the
2846 lines of Robin Houston's patch (but implemented somewhat differently).
2848 35. Further mods to the Makefile to help Win32. Also, added code to pcregrep to
2849 allow it to read and process whole directories in Win32. This code was
2850 contributed by Lionel Fourquaux; it has not been tested by me.
2852 36. Added support for named subpatterns. The Python syntax (?P<name>...) is
2853 used to name a group. Names consist of alphanumerics and underscores, and must
2854 be unique. Back references use the syntax (?P=name) and recursive calls use
2855 (?P>name) which is a PCRE extension to the Python extension. Groups still have
2856 numbers. The function pcre_fullinfo() can be used after compilation to extract
2857 a name/number map. There are three relevant calls:
2859 PCRE_INFO_NAMEENTRYSIZE yields the size of each entry in the map
2860 PCRE_INFO_NAMECOUNT yields the number of entries
2861 PCRE_INFO_NAMETABLE yields a pointer to the map.
2863 The map is a vector of fixed-size entries. The size of each entry depends on
2864 the length of the longest name used. The first two bytes of each entry are the
2865 group number, most significant byte first. There follows the corresponding
2866 name, zero terminated. The names are in alphabetical order.
2868 37. Make the maximum literal string in the compiled code 250 for the non-UTF-8
2869 case instead of 255. Making it the same both with and without UTF-8 support
2870 means that the same test output works with both.
2872 38. There was a case of malloc(0) in the POSIX testing code in pcretest. Avoid
2873 calling malloc() with a zero argument.
2875 39. Change 25 above had to resort to a heavy-handed test for the .* anchoring
2876 optimization. I've improved things by keeping a bitmap of backreferences with
2877 numbers 1-31 so that if .* occurs inside capturing brackets that are not in
2878 fact referenced, the optimization can be applied. It is unlikely that a
2879 relevant occurrence of .* (i.e. one which might indicate anchoring or forcing
2880 the match to follow \n) will appear inside brackets with a number greater than
2881 31, but if it does, any back reference > 31 suppresses the optimization.
2883 40. Added a new compile-time option PCRE_NO_AUTO_CAPTURE. This has the effect
2884 of disabling numbered capturing parentheses. Any opening parenthesis that is
2885 not followed by ? behaves as if it were followed by ?: but named parentheses
2886 can still be used for capturing (and they will acquire numbers in the usual
2887 way).
2889 41. Redesigned the return codes from the match() function into yes/no/error so
2890 that errors can be passed back from deep inside the nested calls. A malloc
2891 failure while inside a recursive subpattern call now causes the
2892 PCRE_ERROR_NOMEMORY return instead of quietly going wrong.
2894 42. It is now possible to set a limit on the number of times the match()
2895 function is called in a call to pcre_exec(). This facility makes it possible to
2896 limit the amount of recursion and backtracking, though not in a directly
2897 obvious way, because the match() function is used in a number of different
2898 circumstances. The count starts from zero for each position in the subject
2899 string (for non-anchored patterns). The default limit is, for compatibility, a
2900 large number, namely 10 000 000. You can change this in two ways:
2902 (a) When configuring PCRE before making, you can use --with-match-limit=n
2903 to set a default value for the compiled library.
2905 (b) For each call to pcre_exec(), you can pass a pcre_extra block in which
2906 a different value is set. See 45 below.
2908 If the limit is exceeded, pcre_exec() returns PCRE_ERROR_MATCHLIMIT.
2910 43. Added a new function pcre_config(int, void *) to enable run-time extraction
2911 of things that can be changed at compile time. The first argument specifies
2912 what is wanted and the second points to where the information is to be placed.
2913 The current list of available information is:
2917 The output is an integer that is set to one if UTF-8 support is available;
2918 otherwise it is set to zero.
2922 The output is an integer that it set to the value of the code that is used for
2923 newline. It is either LF (10) or CR (13).
2927 The output is an integer that contains the number of bytes used for internal
2928 linkage in compiled expressions. The value is 2, 3, or 4. See item 32 above.
2932 The output is an integer that contains the threshold above which the POSIX
2933 interface uses malloc() for output vectors. See item 31 above.
2937 The output is an unsigned integer that contains the default limit of the number
2938 of match() calls in a pcre_exec() execution. See 42 above.
2940 44. pcretest has been upgraded by the addition of the -C option. This causes it
2941 to extract all the available output from the new pcre_config() function, and to
2942 output it. The program then exits immediately.
2944 45. A need has arisen to pass over additional data with calls to pcre_exec() in
2945 order to support additional features. One way would have been to define
2946 pcre_exec2() (for example) with extra arguments, but this would not have been
2947 extensible, and would also have required all calls to the original function to
2948 be mapped to the new one. Instead, I have chosen to extend the mechanism that
2949 is used for passing in "extra" data from pcre_study().
2951 The pcre_extra structure is now exposed and defined in pcre.h. It currently
2952 contains the following fields:
2954 flags a bitmap indicating which of the following fields are set
2955 study_data opaque data from pcre_study()
2956 match_limit a way of specifying a limit on match() calls for a specific
2957 call to pcre_exec()
2958 callout_data data for callouts (see 49 below)
2960 The flag bits are also defined in pcre.h, and are
2966 The pcre_study() function now returns one of these new pcre_extra blocks, with
2967 the actual study data pointed to by the study_data field, and the
2968 PCRE_EXTRA_STUDY_DATA flag set. This can be passed directly to pcre_exec() as
2969 before. That is, this change is entirely upwards-compatible and requires no
2970 change to existing code.
2972 If you want to pass in additional data to pcre_exec(), you can either place it
2973 in a pcre_extra block provided by pcre_study(), or create your own pcre_extra
2974 block.
2976 46. pcretest has been extended to test the PCRE_EXTRA_MATCH_LIMIT feature. If a
2977 data string contains the escape sequence \M, pcretest calls pcre_exec() several
2978 times with different match limits, until it finds the minimum value needed for
2979 pcre_exec() to complete. The value is then output. This can be instructive; for
2980 most simple matches the number is quite small, but for pathological cases it
2981 gets very large very quickly.
2983 47. There's a new option for pcre_fullinfo() called PCRE_INFO_STUDYSIZE. It
2984 returns the size of the data block pointed to by the study_data field in a
2985 pcre_extra block, that is, the value that was passed as the argument to
2986 pcre_malloc() when PCRE was getting memory in which to place the information
2987 created by pcre_study(). The fourth argument should point to a size_t variable.
2988 pcretest has been extended so that this information is shown after a successful
2989 pcre_study() call when information about the compiled regex is being displayed.
2991 48. Cosmetic change to Makefile: there's no need to have / after $(DESTDIR)
2992 because what follows is always an absolute path. (Later: it turns out that this
2993 is more than cosmetic for MinGW, because it doesn't like empty path
2994 components.)
2996 49. Some changes have been made to the callout feature (see 28 above):
2998 (i) A callout function now has three choices for what it returns:
3000 0 => success, carry on matching
3001 > 0 => failure at this point, but backtrack if possible
3002 < 0 => serious error, return this value from pcre_exec()
3004 Negative values should normally be chosen from the set of PCRE_ERROR_xxx
3005 values. In particular, returning PCRE_ERROR_NOMATCH forces a standard
3006 "match failed" error. The error number PCRE_ERROR_CALLOUT is reserved for
3007 use by callout functions. It will never be used by PCRE itself.
3009 (ii) The pcre_extra structure (see 45 above) has a void * field called
3010 callout_data, with corresponding flag bit PCRE_EXTRA_CALLOUT_DATA. The
3011 pcre_callout_block structure has a field of the same name. The contents of
3012 the field passed in the pcre_extra structure are passed to the callout
3013 function in the corresponding field in the callout block. This makes it
3014 easier to use the same callout-containing regex from multiple threads. For
3015 testing, the pcretest program has a new data escape
3017 \C*n pass the number n (may be negative) as callout_data
3019 If the callout function in pcretest receives a non-zero value as
3020 callout_data, it returns that value.
3022 50. Makefile wasn't handling CFLAGS properly when compiling dftables. Also,
3023 there were some redundant $(CFLAGS) in commands that are now specified as
3024 $(LINK), which already includes $(CFLAGS).
3026 51. Extensions to UTF-8 support are listed below. These all apply when (a) PCRE
3027 has been compiled with UTF-8 support *and* pcre_compile() has been compiled
3028 with the PCRE_UTF8 flag. Patterns that are compiled without that flag assume
3029 one-byte characters throughout. Note that case-insensitive matching applies
3030 only to characters whose values are less than 256. PCRE doesn't support the
3031 notion of cases for higher-valued characters.
3033 (i) A character class whose characters are all within 0-255 is handled as
3034 a bit map, and the map is inverted for negative classes. Previously, a
3035 character > 255 always failed to match such a class; however it should
3036 match if the class was a negative one (e.g. [^ab]). This has been fixed.
3038 (ii) A negated character class with a single character < 255 is coded as
3039 "not this character" (OP_NOT). This wasn't working properly when the test
3040 character was multibyte, either singly or repeated.
3042 (iii) Repeats of multibyte characters are now handled correctly in UTF-8
3043 mode, for example: \x{100}{2,3}.
3045 (iv) The character escapes \b, \B, \d, \D, \s, \S, \w, and \W (either
3046 singly or repeated) now correctly test multibyte characters. However,
3047 PCRE doesn't recognize any characters with values greater than 255 as
3048 digits, spaces, or word characters. Such characters always match \D, \S,
3049 and \W, and never match \d, \s, or \w.
3051 (v) Classes may now contain characters and character ranges with values
3052 greater than 255. For example: [ab\x{100}-\x{400}].
3054 (vi) pcregrep now has a --utf-8 option (synonym -u) which makes it call
3055 PCRE in UTF-8 mode.
3057 52. The info request value PCRE_INFO_FIRSTCHAR has been renamed
3058 PCRE_INFO_FIRSTBYTE because it is a byte value. However, the old name is
3059 retained for backwards compatibility. (Note that LASTLITERAL is also a byte
3060 value.)
3062 53. The single man page has become too large. I have therefore split it up into
3063 a number of separate man pages. These also give rise to individual HTML pages;
3064 these are now put in a separate directory, and there is an index.html page that
3065 lists them all. Some hyperlinking between the pages has been installed.
3067 54. Added convenience functions for handling named capturing parentheses.
3069 55. Unknown escapes inside character classes (e.g. [\M]) and escapes that
3070 aren't interpreted therein (e.g. [\C]) are literals in Perl. This is now also
3071 true in PCRE, except when the PCRE_EXTENDED option is set, in which case they
3072 are faulted.
3074 56. Introduced HOST_CC and HOST_CFLAGS which can be set in the environment when
3075 calling configure. These values are used when compiling the dftables.c program
3076 which is run to generate the source of the default character tables. They
3077 default to the values of CC and CFLAGS. If you are cross-compiling PCRE,
3078 you will need to set these values.
3080 57. Updated the building process for Windows DLL, as provided by Fred Cox.
3083 Version 3.9 02-Jan-02
3084 ---------------------
3086 1. A bit of extraneous text had somehow crept into the pcregrep documentation.
3088 2. If --disable-static was given, the building process failed when trying to
3089 build pcretest and pcregrep. (For some reason it was using libtool to compile
3090 them, which is not right, as they aren't part of the library.)
3093 Version 3.8 18-Dec-01
3094 ---------------------
3096 1. The experimental UTF-8 code was completely screwed up. It was packing the
3097 bytes in the wrong order. How dumb can you get?
3100 Version 3.7 29-Oct-01
3101 ---------------------
3103 1. In updating pcretest to check change 1 of version 3.6, I screwed up.
3104 This caused pcretest, when used on the test data, to segfault. Unfortunately,
3105 this didn't happen under Solaris 8, where I normally test things.
3107 2. The Makefile had to be changed to make it work on BSD systems, where 'make'
3108 doesn't seem to recognize that ./xxx and xxx are the same file. (This entry
3109 isn't in ChangeLog distributed with 3.7 because I forgot when I hastily made
3110 this fix an hour or so after the initial 3.7 release.)
3113 Version 3.6 23-Oct-01
3114 ---------------------
3116 1. Crashed with /(sens|respons)e and \1ibility/ and "sense and sensibility" if
3117 offsets passed as NULL with zero offset count.
3119 2. The config.guess and config.sub files had not been updated when I moved to
3120 the latest autoconf.
3123 Version 3.5 15-Aug-01
3124 ---------------------
3126 1. Added some missing #if !defined NOPOSIX conditionals in pcretest.c that
3127 had been forgotten.
3129 2. By using declared but undefined structures, we can avoid using "void"
3130 definitions in pcre.h while keeping the internal definitions of the structures
3131 private.
3133 3. The distribution is now built using autoconf 2.50 and libtool 1.4. From a
3134 user point of view, this means that both static and shared libraries are built
3135 by default, but this can be individually controlled. More of the work of
3136 handling this static/shared cases is now inside libtool instead of PCRE's make
3137 file.
3139 4. The pcretest utility is now installed along with pcregrep because it is
3140 useful for users (to test regexs) and by doing this, it automatically gets
3141 relinked by libtool. The documentation has been turned into a man page, so
3142 there are now .1, .txt, and .html versions in /doc.
3144 5. Upgrades to pcregrep:
3145 (i) Added long-form option names like gnu grep.
3146 (ii) Added --help to list all options with an explanatory phrase.
3147 (iii) Added -r, --recursive to recurse into sub-directories.
3148 (iv) Added -f, --file to read patterns from a file.
3150 6. pcre_exec() was referring to its "code" argument before testing that
3151 argument for NULL (and giving an error if it was NULL).
3153 7. Upgraded Makefile.in to allow for compiling in a different directory from
3154 the source directory.
3156 8. Tiny buglet in pcretest: when pcre_fullinfo() was called to retrieve the
3157 options bits, the pointer it was passed was to an int instead of to an unsigned
3158 long int. This mattered only on 64-bit systems.
3160 9. Fixed typo (3.4/1) in pcre.h again. Sigh. I had changed pcre.h (which is
3161 generated) instead of pcre.in, which it its source. Also made the same change
3162 in several of the .c files.
3164 10. A new release of gcc defines printf() as a macro, which broke pcretest
3165 because it had an ifdef in the middle of a string argument for printf(). Fixed
3166 by using separate calls to printf().
3168 11. Added --enable-newline-is-cr and --enable-newline-is-lf to the configure
3169 script, to force use of CR or LF instead of \n in the source. On non-Unix
3170 systems, the value can be set in config.h.
3172 12. The limit of 200 on non-capturing parentheses is a _nesting_ limit, not an
3173 absolute limit. Changed the text of the error message to make this clear, and
3174 likewise updated the man page.
3176 13. The limit of 99 on the number of capturing subpatterns has been removed.
3177 The new limit is 65535, which I hope will not be a "real" limit.
3180 Version 3.4 22-Aug-00
3181 ---------------------
3183 1. Fixed typo in pcre.h: unsigned const char * changed to const unsigned char *.
3185 2. Diagnose condition (?(0) as an error instead of crashing on matching.
3188 Version 3.3 01-Aug-00
3189 ---------------------
3191 1. If an octal character was given, but the value was greater than \377, it
3192 was not getting masked to the least significant bits, as documented. This could
3193 lead to crashes in some systems.
3195 2. Perl 5.6 (if not earlier versions) accepts classes like [a-\d] and treats
3196 the hyphen as a literal. PCRE used to give an error; it now behaves like Perl.
3198 3. Added the functions pcre_free_substring() and pcre_free_substring_list().
3199 These just pass their arguments on to (pcre_free)(), but they are provided
3200 because some uses of PCRE bind it to non-C systems that can call its functions,
3201 but cannot call free() or pcre_free() directly.
3203 4. Add "make test" as a synonym for "make check". Corrected some comments in
3204 the Makefile.
3206 5. Add $(DESTDIR)/ in front of all the paths in the "install" target in the
3207 Makefile.
3209 6. Changed the name of pgrep to pcregrep, because Solaris has introduced a
3210 command called pgrep for grepping around the active processes.
3212 7. Added the beginnings of support for UTF-8 character strings.
3214 8. Arranged for the Makefile to pass over the settings of CC, CFLAGS, and
3215 RANLIB to ./ltconfig so that they are used by libtool. I think these are all
3216 the relevant ones. (AR is not passed because ./ltconfig does its own figuring
3217 out for the ar command.)
3220 Version 3.2 12-May-00
3221 ---------------------
3223 This is purely a bug fixing release.
3225 1. If the pattern /((Z)+|A)*/ was matched agained ZABCDEFG it matched Z instead
3226 of ZA. This was just one example of several cases that could provoke this bug,
3227 which was introduced by change 9 of version 2.00. The code for breaking
3228 infinite loops after an iteration that matches an empty string was't working
3229 correctly.
3231 2. The pcretest program was not imitating Perl correctly for the pattern /a*/g
3232 when matched against abbab (for example). After matching an empty string, it
3233 wasn't forcing anchoring when setting PCRE_NOTEMPTY for the next attempt; this
3234 caused it to match further down the string than it should.
3236 3. The code contained an inclusion of sys/types.h. It isn't clear why this
3237 was there because it doesn't seem to be needed, and it causes trouble on some
3238 systems, as it is not a Standard C header. It has been removed.
3240 4. Made 4 silly changes to the source to avoid stupid compiler warnings that
3241 were reported on the Macintosh. The changes were from
3243 while ((c = *(++ptr)) != 0 && c != '\n');
3244 to
3245 while ((c = *(++ptr)) != 0 && c != '\n') ;
3247 Totally extraordinary, but if that's what it takes...
3249 5. PCRE is being used in one environment where neither memmove() nor bcopy() is
3250 available. Added HAVE_BCOPY and an autoconf test for it; if neither
3251 HAVE_MEMMOVE nor HAVE_BCOPY is set, use a built-in emulation function which
3252 assumes the way PCRE uses memmove() (always moving upwards).
3254 6. PCRE is being used in one environment where strchr() is not available. There
3255 was only one use in pcre.c, and writing it out to avoid strchr() probably gives
3256 faster code anyway.
3259 Version 3.1 09-Feb-00
3260 ---------------------
3262 The only change in this release is the fixing of some bugs in Makefile.in for
3263 the "install" target:
3265 (1) It was failing to install pcreposix.h.
3267 (2) It was overwriting the pcre.3 man page with the pcreposix.3 man page.
3270 Version 3.0 01-Feb-00
3271 ---------------------
3273 1. Add support for the /+ modifier to perltest (to output $` like it does in
3274 pcretest).
3276 2. Add support for the /g modifier to perltest.
3278 3. Fix pcretest so that it behaves even more like Perl for /g when the pattern
3279 matches null strings.
3281 4. Fix perltest so that it doesn't do unwanted things when fed an empty
3282 pattern. Perl treats empty patterns specially - it reuses the most recent
3283 pattern, which is not what we want. Replace // by /(?#)/ in order to avoid this
3284 effect.
3286 5. The POSIX interface was broken in that it was just handing over the POSIX
3287 captured string vector to pcre_exec(), but (since release 2.00) PCRE has
3288 required a bigger vector, with some working space on the end. This means that
3289 the POSIX wrapper now has to get and free some memory, and copy the results.
3291 6. Added some simple autoconf support, placing the test data and the
3292 documentation in separate directories, re-organizing some of the
3293 information files, and making it build pcre-config (a GNU standard). Also added
3294 libtool support for building PCRE as a shared library, which is now the
3295 default.
3297 7. Got rid of the leading zero in the definition of PCRE_MINOR because 08 and
3298 09 are not valid octal constants. Single digits will be used for minor values
3299 less than 10.
3301 8. Defined REG_EXTENDED and REG_NOSUB as zero in the POSIX header, so that
3302 existing programs that set these in the POSIX interface can use PCRE without
3303 modification.
3305 9. Added a new function, pcre_fullinfo() with an extensible interface. It can
3306 return all that pcre_info() returns, plus additional data. The pcre_info()
3307 function is retained for compatibility, but is considered to be obsolete.
3309 10. Added experimental recursion feature (?R) to handle one common case that
3310 Perl 5.6 will be able to do with (?p{...}).
3312 11. Added support for POSIX character classes like [:alpha:], which Perl is
3313 adopting.
3316 Version 2.08 31-Aug-99
3317 ----------------------
3319 1. When startoffset was not zero and the pattern began with ".*", PCRE was not
3320 trying to match at the startoffset position, but instead was moving forward to
3321 the next newline as if a previous match had failed.
3323 2. pcretest was not making use of PCRE_NOTEMPTY when repeating for /g and /G,
3324 and could get into a loop if a null string was matched other than at the start
3325 of the subject.
3327 3. Added definitions of PCRE_MAJOR and PCRE_MINOR to pcre.h so the version can
3328 be distinguished at compile time, and for completeness also added PCRE_DATE.
3330 5. Added Paul Sokolovsky's minor changes to make it easy to compile a Win32 DLL
3331 in GnuWin32 environments.
3334 Version 2.07 29-Jul-99
3335 ----------------------
3337 1. The documentation is now supplied in plain text form and HTML as well as in
3338 the form of man page sources.
3340 2. C++ compilers don't like assigning (void *) values to other pointer types.
3341 In particular this affects malloc(). Although there is no problem in Standard
3342 C, I've put in casts to keep C++ compilers happy.
3344 3. Typo on pcretest.c; a cast of (unsigned char *) in the POSIX regexec() call
3345 should be (const char *).
3347 4. If NOPOSIX is defined, pcretest.c compiles without POSIX support. This may
3348 be useful for non-Unix systems who don't want to bother with the POSIX stuff.
3349 However, I haven't made this a standard facility. The documentation doesn't
3350 mention it, and the Makefile doesn't support it.
3352 5. The Makefile now contains an "install" target, with editable destinations at
3353 the top of the file. The pcretest program is not installed.
3355 6. pgrep -V now gives the PCRE version number and date.
3357 7. Fixed bug: a zero repetition after a literal string (e.g. /abcde{0}/) was
3358 causing the entire string to be ignored, instead of just the last character.
3360 8. If a pattern like /"([^\\"]+|\\.)*"/ is applied in the normal way to a
3361 non-matching string, it can take a very, very long time, even for strings of
3362 quite modest length, because of the nested recursion. PCRE now does better in
3363 some of these cases. It does this by remembering the last required literal
3364 character in the pattern, and pre-searching the subject to ensure it is present
3365 before running the real match. In other words, it applies a heuristic to detect
3366 some types of certain failure quickly, and in the above example, if presented
3367 with a string that has no trailing " it gives "no match" very quickly.
3369 9. A new runtime option PCRE_NOTEMPTY causes null string matches to be ignored;
3370 other alternatives are tried instead.
3373 Version 2.06 09-Jun-99
3374 ----------------------
3376 1. Change pcretest's output for amount of store used to show just the code
3377 space, because the remainder (the data block) varies in size between 32-bit and
3378 64-bit systems.
3380 2. Added an extra argument to pcre_exec() to supply an offset in the subject to
3381 start matching at. This allows lookbehinds to work when searching for multiple
3382 occurrences in a string.
3384 3. Added additional options to pcretest for testing multiple occurrences:
3386 /+ outputs the rest of the string that follows a match
3387 /g loops for multiple occurrences, using the new startoffset argument
3388 /G loops for multiple occurrences by passing an incremented pointer
3390 4. PCRE wasn't doing the "first character" optimization for patterns starting
3391 with \b or \B, though it was doing it for other lookbehind assertions. That is,
3392 it wasn't noticing that a match for a pattern such as /\bxyz/ has to start with
3393 the letter 'x'. On long subject strings, this gives a significant speed-up.
3396 Version 2.05 21-Apr-99
3397 ----------------------
3399 1. Changed the type of magic_number from int to long int so that it works
3400 properly on 16-bit systems.
3402 2. Fixed a bug which caused patterns starting with .* not to work correctly
3403 when the subject string contained newline characters. PCRE was assuming
3404 anchoring for such patterns in all cases, which is not correct because .* will
3405 not pass a newline unless PCRE_DOTALL is set. It now assumes anchoring only if
3406 DOTALL is set at top level; otherwise it knows that patterns starting with .*
3407 must be retried after every newline in the subject.
3410 Version 2.04 18-Feb-99
3411 ----------------------
3413 1. For parenthesized subpatterns with repeats whose minimum was zero, the
3414 computation of the store needed to hold the pattern was incorrect (too large).
3415 If such patterns were nested a few deep, this could multiply and become a real
3416 problem.
3418 2. Added /M option to pcretest to show the memory requirement of a specific
3419 pattern. Made -m a synonym of -s (which does this globally) for compatibility.
3421 3. Subpatterns of the form (regex){n,m} (i.e. limited maximum) were being
3422 compiled in such a way that the backtracking after subsequent failure was
3423 pessimal. Something like (a){0,3} was compiled as (a)?(a)?(a)? instead of
3424 ((a)((a)(a)?)?)? with disastrous performance if the maximum was of any size.
3427 Version 2.03 02-Feb-99
3428 ----------------------
3430 1. Fixed typo and small mistake in man page.
3432 2. Added 4th condition (GPL supersedes if conflict) and created separate
3433 LICENCE file containing the conditions.
3435 3. Updated pcretest so that patterns such as /abc\/def/ work like they do in
3436 Perl, that is the internal \ allows the delimiter to be included in the
3437 pattern. Locked out the use of \ as a delimiter. If \ immediately follows
3438 the final delimiter, add \ to the end of the pattern (to test the error).
3440 4. Added the convenience functions for extracting substrings after a successful
3441 match. Updated pcretest to make it able to test these functions.
3444 Version 2.02 14-Jan-99
3445 ----------------------
3447 1. Initialized the working variables associated with each extraction so that
3448 their saving and restoring doesn't refer to uninitialized store.
3450 2. Put dummy code into study.c in order to trick the optimizer of the IBM C
3451 compiler for OS/2 into generating correct code. Apparently IBM isn't going to
3452 fix the problem.
3454 3. Pcretest: the timing code wasn't using LOOPREPEAT for timing execution
3455 calls, and wasn't printing the correct value for compiling calls. Increased the
3456 default value of LOOPREPEAT, and the number of significant figures in the
3457 times.
3459 4. Changed "/bin/rm" in the Makefile to "-rm" so it works on Windows NT.
3461 5. Renamed "deftables" as "dftables" to get it down to 8 characters, to avoid
3462 a building problem on Windows NT with a FAT file system.
3465 Version 2.01 21-Oct-98
3466 ----------------------
3468 1. Changed the API for pcre_compile() to allow for the provision of a pointer
3469 to character tables built by pcre_maketables() in the current locale. If NULL
3470 is passed, the default tables are used.
3473 Version 2.00 24-Sep-98
3474 ----------------------
3476 1. Since the (>?) facility is in Perl 5.005, don't require PCRE_EXTRA to enable
3477 it any more.
3479 2. Allow quantification of (?>) groups, and make it work correctly.
3481 3. The first character computation wasn't working for (?>) groups.
3483 4. Correct the implementation of \Z (it is permitted to match on the \n at the
3484 end of the subject) and add 5.005's \z, which really does match only at the
3485 very end of the subject.
3487 5. Remove the \X "cut" facility; Perl doesn't have it, and (?> is neater.
3489 6. Remove the ability to specify CASELESS, MULTILINE, DOTALL, and
3490 DOLLAR_END_ONLY at runtime, to make it possible to implement the Perl 5.005
3491 localized options. All options to pcre_study() were also removed.
3493 7. Add other new features from 5.005:
3495 $(?<= positive lookbehind
3496 $(?<! negative lookbehind
3497 (?imsx-imsx) added the unsetting capability
3498 such a setting is global if at outer level; local otherwise
3499 (?imsx-imsx:) non-capturing groups with option setting
3500 (?(cond)re|re) conditional pattern matching
3502 A backreference to itself in a repeated group matches the previous
3503 captured string.
3505 8. General tidying up of studying (both automatic and via "study")
3506 consequential on the addition of new assertions.
3508 9. As in 5.005, unlimited repeated groups that could match an empty substring
3509 are no longer faulted at compile time. Instead, the loop is forcibly broken at
3510 runtime if any iteration does actually match an empty substring.
3512 10. Include the RunTest script in the distribution.
3514 11. Added tests from the Perl 5.005_02 distribution. This showed up a few
3515 discrepancies, some of which were old and were also with respect to 5.004. They
3516 have now been fixed.
3519 Version 1.09 28-Apr-98
3520 ----------------------
3522 1. A negated single character class followed by a quantifier with a minimum
3523 value of one (e.g. [^x]{1,6} ) was not compiled correctly. This could lead to
3524 program crashes, or just wrong answers. This did not apply to negated classes
3525 containing more than one character, or to minima other than one.
3528 Version 1.08 27-Mar-98
3529 ----------------------
3531 1. Add PCRE_UNGREEDY to invert the greediness of quantifiers.
3533 2. Add (?U) and (?X) to set PCRE_UNGREEDY and PCRE_EXTRA respectively. The
3534 latter must appear before anything that relies on it in the pattern.
3537 Version 1.07 16-Feb-98
3538 ----------------------
3540 1. A pattern such as /((a)*)*/ was not being diagnosed as in error (unlimited
3541 repeat of a potentially empty string).
3544 Version 1.06 23-Jan-98
3545 ----------------------
3547 1. Added Markus Oberhumer's little patches for C++.
3549 2. Literal strings longer than 255 characters were broken.
3552 Version 1.05 23-Dec-97
3553 ----------------------
3555 1. Negated character classes containing more than one character were failing if
3556 PCRE_CASELESS was set at run time.
3559 Version 1.04 19-Dec-97
3560 ----------------------
3562 1. Corrected the man page, where some "const" qualifiers had been omitted.
3564 2. Made debugging output print "{0,xxx}" instead of just "{,xxx}" to agree with
3565 input syntax.
3567 3. Fixed memory leak which occurred when a regex with back references was
3568 matched with an offsets vector that wasn't big enough. The temporary memory
3569 that is used in this case wasn't being freed if the match failed.
3571 4. Tidied pcretest to ensure it frees memory that it gets.
3573 5. Temporary memory was being obtained in the case where the passed offsets
3574 vector was exactly big enough.
3576 6. Corrected definition of offsetof() from change 5 below.
3578 7. I had screwed up change 6 below and broken the rules for the use of
3579 setjmp(). Now fixed.
3582 Version 1.03 18-Dec-97
3583 ----------------------
3585 1. A erroneous regex with a missing opening parenthesis was correctly
3586 diagnosed, but PCRE attempted to access brastack[-1], which could cause crashes
3587 on some systems.
3589 2. Replaced offsetof(real_pcre, code) by offsetof(real_pcre, code[0]) because
3590 it was reported that one broken compiler failed on the former because "code" is
3591 also an independent variable.
3593 3. The erroneous regex a[]b caused an array overrun reference.
3595 4. A regex ending with a one-character negative class (e.g. /[^k]$/) did not
3596 fail on data ending with that character. (It was going on too far, and checking
3597 the next character, typically a binary zero.) This was specific to the
3598 optimized code for single-character negative classes.
3600 5. Added a contributed patch from the TIN world which does the following:
3602 + Add an undef for memmove, in case the the system defines a macro for it.
3604 + Add a definition of offsetof(), in case there isn't one. (I don't know
3605 the reason behind this - offsetof() is part of the ANSI standard - but
3606 it does no harm).
3608 + Reduce the ifdef's in pcre.c using macro DPRINTF, thereby eliminating
3609 most of the places where whitespace preceded '#'. I have given up and
3610 allowed the remaining 2 cases to be at the margin.
3612 + Rename some variables in pcre to eliminate shadowing. This seems very
3613 pedantic, but does no harm, of course.
3615 6. Moved the call to setjmp() into its own function, to get rid of warnings
3616 from gcc -Wall, and avoided calling it at all unless PCRE_EXTRA is used.
3618 7. Constructs such as \d{8,} were compiling into the equivalent of
3619 \d{8}\d{0,65527} instead of \d{8}\d* which didn't make much difference to the
3620 outcome, but in this particular case used more store than had been allocated,
3621 which caused the bug to be discovered because it threw up an internal error.
3623 8. The debugging code in both pcre and pcretest for outputting the compiled
3624 form of a regex was going wrong in the case of back references followed by
3625 curly-bracketed repeats.
3628 Version 1.02 12-Dec-97
3629 ----------------------
3631 1. Typos in pcre.3 and comments in the source fixed.
3633 2. Applied a contributed patch to get rid of places where it used to remove
3634 'const' from variables, and fixed some signed/unsigned and uninitialized
3635 variable warnings.
3637 3. Added the "runtest" target to Makefile.
3639 4. Set default compiler flag to -O2 rather than just -O.
3642 Version 1.01 19-Nov-97
3643 ----------------------
3645 1. PCRE was failing to diagnose unlimited repeat of empty string for patterns
3646 like /([ab]*)*/, that is, for classes with more than one character in them.
3648 2. Likewise, it wasn't diagnosing patterns with "once-only" subpatterns, such
3649 as /((?>a*))*/ (a PCRE_EXTRA facility).
3652 Version 1.00 18-Nov-97
3653 ----------------------
3655 1. Added compile-time macros to support systems such as SunOS4 which don't have
3656 memmove() or strerror() but have other things that can be used instead.
3658 2. Arranged that "make clean" removes the executables.
3661 Version 0.99 27-Oct-97
3662 ----------------------
3664 1. Fixed bug in code for optimizing classes with only one character. It was
3665 initializing a 32-byte map regardless, which could cause it to run off the end
3666 of the memory it had got.
3668 2. Added, conditional on PCRE_EXTRA, the proposed (?>REGEX) construction.
3671 Version 0.98 22-Oct-97
3672 ----------------------
3674 1. Fixed bug in code for handling temporary memory usage when there are more
3675 back references than supplied space in the ovector. This could cause segfaults.
3678 Version 0.97 21-Oct-97
3679 ----------------------
3681 1. Added the \X "cut" facility, conditional on PCRE_EXTRA.
3683 2. Optimized negated single characters not to use a bit map.
3685 3. Brought error texts together as macro definitions; clarified some of them;
3686 fixed one that was wrong - it said "range out of order" when it meant "invalid
3687 escape sequence".
3689 4. Changed some char * arguments to const char *.
3691 5. Added PCRE_NOTBOL and PCRE_NOTEOL (from POSIX).
3693 6. Added the POSIX-style API wrapper in pcreposix.a and testing facilities in
3694 pcretest.
3697 Version 0.96 16-Oct-97
3698 ----------------------
3700 1. Added a simple "pgrep" utility to the distribution.
3702 2. Fixed an incompatibility with Perl: "{" is now treated as a normal character
3703 unless it appears in one of the precise forms "{ddd}", "{ddd,}", or "{ddd,ddd}"
3704 where "ddd" means "one or more decimal digits".
3706 3. Fixed serious bug. If a pattern had a back reference, but the call to
3707 pcre_exec() didn't supply a large enough ovector to record the related
3708 identifying subpattern, the match always failed. PCRE now remembers the number
3709 of the largest back reference, and gets some temporary memory in which to save
3710 the offsets during matching if necessary, in order to ensure that
3711 backreferences always work.
3713 4. Increased the compatibility with Perl in a number of ways:
3715 (a) . no longer matches \n by default; an option PCRE_DOTALL is provided
3716 to request this handling. The option can be set at compile or exec time.
3718 (b) $ matches before a terminating newline by default; an option
3719 PCRE_DOLLAR_ENDONLY is provided to override this (but not in multiline
3720 mode). The option can be set at compile or exec time.
3722 (c) The handling of \ followed by a digit other than 0 is now supposed to be
3723 the same as Perl's. If the decimal number it represents is less than 10
3724 or there aren't that many previous left capturing parentheses, an octal
3725 escape is read. Inside a character class, it's always an octal escape,
3726 even if it is a single digit.
3728 (d) An escaped but undefined alphabetic character is taken as a literal,
3729 unless PCRE_EXTRA is set. Currently this just reserves the remaining
3730 escapes.
3732 (e) {0} is now permitted. (The previous item is removed from the compiled
3733 pattern).
3735 5. Changed all the names of code files so that the basic parts are no longer
3736 than 10 characters, and abolished the teeny "globals.c" file.
3738 6. Changed the handling of character classes; they are now done with a 32-byte
3739 bit map always.
3741 7. Added the -d and /D options to pcretest to make it possible to look at the
3742 internals of compilation without having to recompile pcre.
3745 Version 0.95 23-Sep-97
3746 ----------------------
3748 1. Fixed bug in pre-pass concerning escaped "normal" characters such as \x5c or
3749 \x20 at the start of a run of normal characters. These were being treated as
3750 real characters, instead of the source characters being re-checked.
3753 Version 0.94 18-Sep-97
3754 ----------------------
3756 1. The functions are now thread-safe, with the caveat that the global variables
3757 containing pointers to malloc() and free() or alternative functions are the
3758 same for all threads.
3760 2. Get pcre_study() to generate a bitmap of initial characters for non-
3761 anchored patterns when this is possible, and use it if passed to pcre_exec().
3764 Version 0.93 15-Sep-97
3765 ----------------------
3767 1. /(b)|(:+)/ was computing an incorrect first character.
3769 2. Add pcre_study() to the API and the passing of pcre_extra to pcre_exec(),
3770 but not actually doing anything yet.
3772 3. Treat "-" characters in classes that cannot be part of ranges as literals,
3773 as Perl does (e.g. [-az] or [az-]).
3775 4. Set the anchored flag if a branch starts with .* or .*? because that tests
3776 all possible positions.
3778 5. Split up into different modules to avoid including unneeded functions in a
3779 compiled binary. However, compile and exec are still in one module. The "study"
3780 function is split off.
3782 6. The character tables are now in a separate module whose source is generated
3783 by an auxiliary program - but can then be edited by hand if required. There are
3784 now no calls to isalnum(), isspace(), isdigit(), isxdigit(), tolower() or
3785 toupper() in the code.
3787 7. Turn the malloc/free funtions variables into pcre_malloc and pcre_free and
3788 make them global. Abolish the function for setting them, as the caller can now
3789 set them directly.
3792 Version 0.92 11-Sep-97
3793 ----------------------
3795 1. A repeat with a fixed maximum and a minimum of 1 for an ordinary character
3796 (e.g. /a{1,3}/) was broken (I mis-optimized it).
3798 2. Caseless matching was not working in character classes if the characters in
3799 the pattern were in upper case.
3801 3. Make ranges like [W-c] work in the same way as Perl for caseless matching.
3803 4. Make PCRE_ANCHORED public and accept as a compile option.
3805 5. Add an options word to pcre_exec() and accept PCRE_ANCHORED and
3806 PCRE_CASELESS at run time. Add escapes \A and \I to pcretest to cause it to
3807 pass them.
3809 6. Give an error if bad option bits passed at compile or run time.
3811 7. Add PCRE_MULTILINE at compile and exec time, and (?m) as well. Add \M to
3812 pcretest to cause it to pass that flag.
3814 8. Add pcre_info(), to get the number of identifying subpatterns, the stored
3815 options, and the first character, if set.
3817 9. Recognize C+ or C{n,m} where n >= 1 as providing a fixed starting character.
3820 Version 0.91 10-Sep-97
3821 ----------------------
3823 1. PCRE was failing to diagnose unlimited repeats of subpatterns that could
3824 match the empty string as in /(a*)*/. It was looping and ultimately crashing.
3826 2. PCRE was looping on encountering an indefinitely repeated back reference to
3827 a subpattern that had matched an empty string, e.g. /(a|)\1*/. It now does what
3828 Perl does - treats the match as successful.
3830 ****


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

  ViewVC Help
Powered by ViewVC 1.1.5