ViewVC logotype

Contents of /code/trunk/ChangeLog

Parent Directory Parent Directory | Revision Log Revision Log

Revision 480 - (show annotations)
Sun Jan 3 17:29:59 2010 UTC (5 years, 10 months ago) by ph10
File size: 178695 byte(s)
Error occurred while calculating annotation data.
Update stack documentation to be clearer about pcre_dfa_exec().
1 ChangeLog for PCRE
2 ------------------
4 Version 8.01 11-Dec-09
5 ----------------------
7 1. If a pattern contained a conditional subpattern with only one branch (in
8 particular, this includes all (DEFINE) patterns), a call to pcre_study()
9 computed the wrong minimum data length (which is of course zero for such
10 subpatterns).
12 2. For patterns such as (?i)a(?-i)b|c where an option setting at the start of
13 the pattern is reset in the first branch, pcre_compile() failed with
14 "internal error: code overflow at offset...". This happened only when
15 the reset was to the original external option setting. (An optimization
16 abstracts leading options settings into an external setting, which was the
17 cause of this.)
19 3. A pattern such as ^(?!a(*SKIP)b) where a negative assertion contained one
20 of the verbs SKIP, PRUNE, or COMMIT, did not work correctly. When the
21 assertion pattern did not match (meaning that the assertion was true), it
22 was incorrectly treated as false if the SKIP had been reached during the
23 matching. This also applied to assertions used as conditions.
25 4. If an item that is not supported by pcre_dfa_exec() was encountered in an
26 assertion subpattern, including such a pattern used as a condition,
27 unpredictable results occurred, instead of the error return
30 5. The C++ GlobalReplace function was not working like Perl for the special
31 situation when an empty string is matched. It now does the fancy magic
32 stuff that is necessary.
34 6. In pcre_internal.h, obsolete includes to setjmp.h and stdarg.h have been
35 removed. (These were left over from very, very early versions of PCRE.)
37 7. Some cosmetic changes to the code to make life easier when compiling it
38 as part of something else:
40 (a) Change DEBUG to PCRE_DEBUG.
42 (b) In pcre_compile(), rename the member of the "branch_chain" structure
43 called "current" as "current_branch", to prevent a collision with the
44 Linux macro when compiled as a kernel module.
46 (c) In pcre_study(), rename the function set_bit() as set_table_bit(), to
47 prevent a collision with the Linux macro when compiled as a kernel
48 module.
50 8. In pcre_compile() there are some checks for integer overflows that used to
51 cast potentially large values to (double). This has been changed to that
52 when building, a check for int64_t is made, and if it is found, it is used
53 instead, thus avoiding the use of floating point arithmetic. (There is no
54 other use of FP in PCRE.) If int64_t is not found, the fallback is to
55 double.
57 9. Added two casts to avoid signed/unsigned warnings from VS Studio Express
58 2005 (difference between two addresses compared to an unsigned value).
60 10. Change the standard AC_CHECK_LIB test for libbz2 in configure.ac to a
61 custom one, because of the following reported problem in Windows:
63 - libbz2 uses the Pascal calling convention (WINAPI) for the functions
64 under Win32.
65 - The standard autoconf AC_CHECK_LIB fails to include "bzlib.h",
66 therefore missing the function definition.
67 - The compiler thus generates a "C" signature for the test function.
68 - The linker fails to find the "C" function.
69 - PCRE fails to configure if asked to do so against libbz2.
71 11. When running libtoolize from libtool-2.2.6b as part of autogen.sh, these
72 messages were output:
74 Consider adding `AC_CONFIG_MACRO_DIR([m4])' to configure.ac and
75 rerunning libtoolize, to keep the correct libtool macros in-tree.
76 Consider adding `-I m4' to ACLOCAL_AMFLAGS in Makefile.am.
78 I have done both of these things.
80 12. Although pcre_dfa_exec() does not use nearly as much stack as pcre_exec()
81 most of the time, it *can* run out if it is given a pattern that contains a
82 runaway infinite recursion. I updated the discussion in the pcrestack man
83 page.
87 Version 8.00 19-Oct-09
88 ----------------------
90 1. The table for translating pcre_compile() error codes into POSIX error codes
91 was out-of-date, and there was no check on the pcre_compile() error code
92 being within the table. This could lead to an OK return being given in
93 error.
95 2. Changed the call to open a subject file in pcregrep from fopen(pathname,
96 "r") to fopen(pathname, "rb"), which fixed a problem with some of the tests
97 in a Windows environment.
99 3. The pcregrep --count option prints the count for each file even when it is
100 zero, as does GNU grep. However, pcregrep was also printing all files when
101 --files-with-matches was added. Now, when both options are given, it prints
102 counts only for those files that have at least one match. (GNU grep just
103 prints the file name in this circumstance, but including the count seems
104 more useful - otherwise, why use --count?) Also ensured that the
105 combination -clh just lists non-zero counts, with no names.
107 4. The long form of the pcregrep -F option was incorrectly implemented as
108 --fixed_strings instead of --fixed-strings. This is an incompatible change,
109 but it seems right to fix it, and I didn't think it was worth preserving
110 the old behaviour.
112 5. The command line items --regex=pattern and --regexp=pattern were not
113 recognized by pcregrep, which required --regex pattern or --regexp pattern
114 (with a space rather than an '='). The man page documented the '=' forms,
115 which are compatible with GNU grep; these now work.
117 6. No libpcreposix.pc file was created for pkg-config; there was just
118 libpcre.pc and libpcrecpp.pc. The omission has been rectified.
120 7. Added #ifndef SUPPORT_UCP into the pcre_ucd.c module, to reduce its size
121 when UCP support is not needed, by modifying the Python script that
122 generates it from Unicode data files. This should not matter if the module
123 is correctly used as a library, but I received one complaint about 50K of
124 unwanted data. My guess is that the person linked everything into his
125 program rather than using a library. Anyway, it does no harm.
127 8. A pattern such as /\x{123}{2,2}+/8 was incorrectly compiled; the trigger
128 was a minimum greater than 1 for a wide character in a possessive
129 repetition. The same bug could also affect patterns like /(\x{ff}{0,2})*/8
130 which had an unlimited repeat of a nested, fixed maximum repeat of a wide
131 character. Chaos in the form of incorrect output or a compiling loop could
132 result.
134 9. The restrictions on what a pattern can contain when partial matching is
135 requested for pcre_exec() have been removed. All patterns can now be
136 partially matched by this function. In addition, if there are at least two
137 slots in the offset vector, the offset of the earliest inspected character
138 for the match and the offset of the end of the subject are set in them when
139 PCRE_ERROR_PARTIAL is returned.
141 10. Partial matching has been split into two forms: PCRE_PARTIAL_SOFT, which is
142 synonymous with PCRE_PARTIAL, for backwards compatibility, and
143 PCRE_PARTIAL_HARD, which causes a partial match to supersede a full match,
144 and may be more useful for multi-segment matching.
146 11. Partial matching with pcre_exec() is now more intuitive. A partial match
147 used to be given if ever the end of the subject was reached; now it is
148 given only if matching could not proceed because another character was
149 needed. This makes a difference in some odd cases such as Z(*FAIL) with the
150 string "Z", which now yields "no match" instead of "partial match". In the
151 case of pcre_dfa_exec(), "no match" is given if every matching path for the
152 final character ended with (*FAIL).
154 12. Restarting a match using pcre_dfa_exec() after a partial match did not work
155 if the pattern had a "must contain" character that was already found in the
156 earlier partial match, unless partial matching was again requested. For
157 example, with the pattern /dog.(body)?/, the "must contain" character is
158 "g". If the first part-match was for the string "dog", restarting with
159 "sbody" failed. This bug has been fixed.
161 13. The string returned by pcre_dfa_exec() after a partial match has been
162 changed so that it starts at the first inspected character rather than the
163 first character of the match. This makes a difference only if the pattern
164 starts with a lookbehind assertion or \b or \B (\K is not supported by
165 pcre_dfa_exec()). It's an incompatible change, but it makes the two
166 matching functions compatible, and I think it's the right thing to do.
168 14. Added a pcredemo man page, created automatically from the pcredemo.c file,
169 so that the demonstration program is easily available in environments where
170 PCRE has not been installed from source.
172 15. Arranged to add -DPCRE_STATIC to cflags in libpcre.pc, libpcreposix.cp,
173 libpcrecpp.pc and pcre-config when PCRE is not compiled as a shared
174 library.
176 16. Added REG_UNGREEDY to the pcreposix interface, at the request of a user.
177 It maps to PCRE_UNGREEDY. It is not, of course, POSIX-compatible, but it
178 is not the first non-POSIX option to be added. Clearly some people find
179 these options useful.
181 17. If a caller to the POSIX matching function regexec() passes a non-zero
182 value for nmatch with a NULL value for pmatch, the value of
183 nmatch is forced to zero.
185 18. RunGrepTest did not have a test for the availability of the -u option of
186 the diff command, as RunTest does. It now checks in the same way as
187 RunTest, and also checks for the -b option.
189 19. If an odd number of negated classes containing just a single character
190 interposed, within parentheses, between a forward reference to a named
191 subpattern and the definition of the subpattern, compilation crashed with
192 an internal error, complaining that it could not find the referenced
193 subpattern. An example of a crashing pattern is /(?&A)(([^m])(?<A>))/.
194 [The bug was that it was starting one character too far in when skipping
195 over the character class, thus treating the ] as data rather than
196 terminating the class. This meant it could skip too much.]
198 20. Added PCRE_NOTEMPTY_ATSTART in order to be able to correctly implement the
199 /g option in pcretest when the pattern contains \K, which makes it possible
200 to have an empty string match not at the start, even when the pattern is
201 anchored. Updated pcretest and pcredemo to use this option.
203 21. If the maximum number of capturing subpatterns in a recursion was greater
204 than the maximum at the outer level, the higher number was returned, but
205 with unset values at the outer level. The correct (outer level) value is
206 now given.
208 22. If (*ACCEPT) appeared inside capturing parentheses, previous releases of
209 PCRE did not set those parentheses (unlike Perl). I have now found a way to
210 make it do so. The string so far is captured, making this feature
211 compatible with Perl.
213 23. The tests have been re-organized, adding tests 11 and 12, to make it
214 possible to check the Perl 5.10 features against Perl 5.10.
216 24. Perl 5.10 allows subroutine calls in lookbehinds, as long as the subroutine
217 pattern matches a fixed length string. PCRE did not allow this; now it
218 does. Neither allows recursion.
220 25. I finally figured out how to implement a request to provide the minimum
221 length of subject string that was needed in order to match a given pattern.
222 (It was back references and recursion that I had previously got hung up
223 on.) This code has now been added to pcre_study(); it finds a lower bound
224 to the length of subject needed. It is not necessarily the greatest lower
225 bound, but using it to avoid searching strings that are too short does give
226 some useful speed-ups. The value is available to calling programs via
227 pcre_fullinfo().
229 26. While implementing 25, I discovered to my embarrassment that pcretest had
230 not been passing the result of pcre_study() to pcre_dfa_exec(), so the
231 study optimizations had never been tested with that matching function.
232 Oops. What is worse, even when it was passed study data, there was a bug in
233 pcre_dfa_exec() that meant it never actually used it. Double oops. There
234 were also very few tests of studied patterns with pcre_dfa_exec().
236 27. If (?| is used to create subpatterns with duplicate numbers, they are now
237 allowed to have the same name, even if PCRE_DUPNAMES is not set. However,
238 on the other side of the coin, they are no longer allowed to have different
239 names, because these cannot be distinguished in PCRE, and this has caused
240 confusion. (This is a difference from Perl.)
242 28. When duplicate subpattern names are present (necessarily with different
243 numbers, as required by 27 above), and a test is made by name in a
244 conditional pattern, either for a subpattern having been matched, or for
245 recursion in such a pattern, all the associated numbered subpatterns are
246 tested, and the overall condition is true if the condition is true for any
247 one of them. This is the way Perl works, and is also more like the way
248 testing by number works.
251 Version 7.9 11-Apr-09
252 ---------------------
254 1. When building with support for bzlib/zlib (pcregrep) and/or readline
255 (pcretest), all targets were linked against these libraries. This included
256 libpcre, libpcreposix, and libpcrecpp, even though they do not use these
257 libraries. This caused unwanted dependencies to be created. This problem
258 has been fixed, and now only pcregrep is linked with bzlib/zlib and only
259 pcretest is linked with readline.
261 2. The "typedef int BOOL" in pcre_internal.h that was included inside the
262 "#ifndef FALSE" condition by an earlier change (probably 7.8/18) has been
263 moved outside it again, because FALSE and TRUE are already defined in AIX,
264 but BOOL is not.
266 3. The pcre_config() function was treating the PCRE_MATCH_LIMIT and
267 PCRE_MATCH_LIMIT_RECURSION values as ints, when they should be long ints.
269 4. The pcregrep documentation said spaces were inserted as well as colons (or
270 hyphens) following file names and line numbers when outputting matching
271 lines. This is not true; no spaces are inserted. I have also clarified the
272 wording for the --colour (or --color) option.
274 5. In pcregrep, when --colour was used with -o, the list of matching strings
275 was not coloured; this is different to GNU grep, so I have changed it to be
276 the same.
278 6. When --colo(u)r was used in pcregrep, only the first matching substring in
279 each matching line was coloured. Now it goes on to look for further matches
280 of any of the test patterns, which is the same behaviour as GNU grep.
282 7. A pattern that could match an empty string could cause pcregrep to loop; it
283 doesn't make sense to accept an empty string match in pcregrep, so I have
284 locked it out (using PCRE's PCRE_NOTEMPTY option). By experiment, this
285 seems to be how GNU grep behaves.
287 8. The pattern (?(?=.*b)b|^) was incorrectly compiled as "match must be at
288 start or after a newline", because the conditional assertion was not being
289 correctly handled. The rule now is that both the assertion and what follows
290 in the first alternative must satisfy the test.
292 9. If auto-callout was enabled in a pattern with a conditional group whose
293 condition was an assertion, PCRE could crash during matching, both with
294 pcre_exec() and pcre_dfa_exec().
296 10. The PCRE_DOLLAR_ENDONLY option was not working when pcre_dfa_exec() was
297 used for matching.
299 11. Unicode property support in character classes was not working for
300 characters (bytes) greater than 127 when not in UTF-8 mode.
302 12. Added the -M command line option to pcretest.
304 14. Added the non-standard REG_NOTEMPTY option to the POSIX interface.
306 15. Added the PCRE_NO_START_OPTIMIZE match-time option.
308 16. Added comments and documentation about mis-use of no_arg in the C++
309 wrapper.
311 17. Implemented support for UTF-8 encoding in EBCDIC environments, a patch
312 from Martin Jerabek that uses macro names for all relevant character and
313 string constants.
315 18. Added to pcre_internal.h two configuration checks: (a) If both EBCDIC and
316 SUPPORT_UTF8 are set, give an error; (b) If SUPPORT_UCP is set without
317 SUPPORT_UTF8, define SUPPORT_UTF8. The "configure" script handles both of
318 these, but not everybody uses configure.
320 19. A conditional group that had only one branch was not being correctly
321 recognized as an item that could match an empty string. This meant that an
322 enclosing group might also not be so recognized, causing infinite looping
323 (and probably a segfault) for patterns such as ^"((?(?=[a])[^"])|b)*"$
324 with the subject "ab", where knowledge that the repeated group can match
325 nothing is needed in order to break the loop.
327 20. If a pattern that was compiled with callouts was matched using pcre_dfa_
328 exec(), but without supplying a callout function, matching went wrong.
330 21. If PCRE_ERROR_MATCHLIMIT occurred during a recursion, there was a memory
331 leak if the size of the offset vector was greater than 30. When the vector
332 is smaller, the saved offsets during recursion go onto a local stack
333 vector, but for larger vectors malloc() is used. It was failing to free
334 when the recursion yielded PCRE_ERROR_MATCH_LIMIT (or any other "abnormal"
335 error, in fact).
337 22. There was a missing #ifdef SUPPORT_UTF8 round one of the variables in the
338 heapframe that is used only when UTF-8 support is enabled. This caused no
339 problem, but was untidy.
341 23. Steven Van Ingelgem's patch to CMakeLists.txt to change the name
342 CMAKE_BINARY_DIR to PROJECT_BINARY_DIR so that it works when PCRE is
343 included within another project.
345 24. Steven Van Ingelgem's patches to add more options to the CMake support,
346 slightly modified by me:
348 (a) PCRE_BUILD_TESTS can be set OFF not to build the tests, including
349 not building pcregrep.
351 (b) PCRE_BUILD_PCREGREP can be see OFF not to build pcregrep, but only
352 if PCRE_BUILD_TESTS is also set OFF, because the tests use pcregrep.
354 25. Forward references, both numeric and by name, in patterns that made use of
355 duplicate group numbers, could behave incorrectly or give incorrect errors,
356 because when scanning forward to find the reference group, PCRE was not
357 taking into account the duplicate group numbers. A pattern such as
358 ^X(?3)(a)(?|(b)|(q))(Y) is an example.
360 26. Changed a few more instances of "const unsigned char *" to USPTR, making
361 the feature of a custom pointer more persuasive (as requested by a user).
363 27. Wrapped the definitions of fileno and isatty for Windows, which appear in
364 pcretest.c, inside #ifndefs, because it seems they are sometimes already
365 pre-defined.
367 28. Added support for (*UTF8) at the start of a pattern.
369 29. Arrange for flags added by the "release type" setting in CMake to be shown
370 in the configuration summary.
373 Version 7.8 05-Sep-08
374 ---------------------
376 1. Replaced UCP searching code with optimized version as implemented for Ad
377 Muncher (http://www.admuncher.com/) by Peter Kankowski. This uses a two-
378 stage table and inline lookup instead of a function, giving speed ups of 2
379 to 5 times on some simple patterns that I tested. Permission was given to
380 distribute the MultiStage2.py script that generates the tables (it's not in
381 the tarball, but is in the Subversion repository).
383 2. Updated the Unicode datatables to Unicode 5.1.0. This adds yet more
384 scripts.
386 3. Change 12 for 7.7 introduced a bug in pcre_study() when a pattern contained
387 a group with a zero qualifier. The result of the study could be incorrect,
388 or the function might crash, depending on the pattern.
390 4. Caseless matching was not working for non-ASCII characters in back
391 references. For example, /(\x{de})\1/8i was not matching \x{de}\x{fe}.
392 It now works when Unicode Property Support is available.
394 5. In pcretest, an escape such as \x{de} in the data was always generating
395 a UTF-8 string, even in non-UTF-8 mode. Now it generates a single byte in
396 non-UTF-8 mode. If the value is greater than 255, it gives a warning about
397 truncation.
399 6. Minor bugfix in pcrecpp.cc (change "" == ... to NULL == ...).
401 7. Added two (int) casts to pcregrep when printing the difference of two
402 pointers, in case they are 64-bit values.
404 8. Added comments about Mac OS X stack usage to the pcrestack man page and to
405 test 2 if it fails.
407 9. Added PCRE_CALL_CONVENTION just before the names of all exported functions,
408 and a #define of that name to empty if it is not externally set. This is to
409 allow users of MSVC to set it if necessary.
411 10. The PCRE_EXP_DEFN macro which precedes exported functions was missing from
412 the convenience functions in the pcre_get.c source file.
414 11. An option change at the start of a pattern that had top-level alternatives
415 could cause overwriting and/or a crash. This command provoked a crash in
416 some environments:
418 printf "/(?i)[\xc3\xa9\xc3\xbd]|[\xc3\xa9\xc3\xbdA]/8\n" | pcretest
420 This potential security problem was recorded as CVE-2008-2371.
422 12. For a pattern where the match had to start at the beginning or immediately
423 after a newline (e.g /.*anything/ without the DOTALL flag), pcre_exec() and
424 pcre_dfa_exec() could read past the end of the passed subject if there was
425 no match. To help with detecting such bugs (e.g. with valgrind), I modified
426 pcretest so that it places the subject at the end of its malloc-ed buffer.
428 13. The change to pcretest in 12 above threw up a couple more cases when pcre_
429 exec() might read past the end of the data buffer in UTF-8 mode.
431 14. A similar bug to 7.3/2 existed when the PCRE_FIRSTLINE option was set and
432 the data contained the byte 0x85 as part of a UTF-8 character within its
433 first line. This applied both to normal and DFA matching.
435 15. Lazy qualifiers were not working in some cases in UTF-8 mode. For example,
436 /^[^d]*?$/8 failed to match "abc".
438 16. Added a missing copyright notice to pcrecpp_internal.h.
440 17. Make it more clear in the documentation that values returned from
441 pcre_exec() in ovector are byte offsets, not character counts.
443 18. Tidied a few places to stop certain compilers from issuing warnings.
445 19. Updated the Virtual Pascal + BCC files to compile the latest v7.7, as
446 supplied by Stefan Weber. I made a further small update for 7.8 because
447 there is a change of source arrangements: the pcre_searchfuncs.c module is
448 replaced by pcre_ucd.c.
451 Version 7.7 07-May-08
452 ---------------------
454 1. Applied Craig's patch to sort out a long long problem: "If we can't convert
455 a string to a long long, pretend we don't even have a long long." This is
456 done by checking for the strtoq, strtoll, and _strtoi64 functions.
458 2. Applied Craig's patch to pcrecpp.cc to restore ABI compatibility with
459 pre-7.6 versions, which defined a global no_arg variable instead of putting
460 it in the RE class. (See also #8 below.)
462 3. Remove a line of dead code, identified by coverity and reported by Nuno
463 Lopes.
465 4. Fixed two related pcregrep bugs involving -r with --include or --exclude:
467 (1) The include/exclude patterns were being applied to the whole pathnames
468 of files, instead of just to the final components.
470 (2) If there was more than one level of directory, the subdirectories were
471 skipped unless they satisfied the include/exclude conditions. This is
472 inconsistent with GNU grep (and could even be seen as contrary to the
473 pcregrep specification - which I improved to make it absolutely clear).
474 The action now is always to scan all levels of directory, and just
475 apply the include/exclude patterns to regular files.
477 5. Added the --include_dir and --exclude_dir patterns to pcregrep, and used
478 --exclude_dir in the tests to avoid scanning .svn directories.
480 6. Applied Craig's patch to the QuoteMeta function so that it escapes the
481 NUL character as backslash + 0 rather than backslash + NUL, because PCRE
482 doesn't support NULs in patterns.
484 7. Added some missing "const"s to declarations of static tables in
485 pcre_compile.c and pcre_dfa_exec.c.
487 8. Applied Craig's patch to pcrecpp.cc to fix a problem in OS X that was
488 caused by fix #2 above. (Subsequently also a second patch to fix the
489 first patch. And a third patch - this was a messy problem.)
491 9. Applied Craig's patch to remove the use of push_back().
493 10. Applied Alan Lehotsky's patch to add REG_STARTEND support to the POSIX
494 matching function regexec().
496 11. Added support for the Oniguruma syntax \g<name>, \g<n>, \g'name', \g'n',
497 which, however, unlike Perl's \g{...}, are subroutine calls, not back
498 references. PCRE supports relative numbers with this syntax (I don't think
499 Oniguruma does).
501 12. Previously, a group with a zero repeat such as (...){0} was completely
502 omitted from the compiled regex. However, this means that if the group
503 was called as a subroutine from elsewhere in the pattern, things went wrong
504 (an internal error was given). Such groups are now left in the compiled
505 pattern, with a new opcode that causes them to be skipped at execution
506 time.
508 13. Added the PCRE_JAVASCRIPT_COMPAT option. This makes the following changes
509 to the way PCRE behaves:
511 (a) A lone ] character is dis-allowed (Perl treats it as data).
513 (b) A back reference to an unmatched subpattern matches an empty string
514 (Perl fails the current match path).
516 (c) A data ] in a character class must be notated as \] because if the
517 first data character in a class is ], it defines an empty class. (In
518 Perl it is not possible to have an empty class.) The empty class []
519 never matches; it forces failure and is equivalent to (*FAIL) or (?!).
520 The negative empty class [^] matches any one character, independently
521 of the DOTALL setting.
523 14. A pattern such as /(?2)[]a()b](abc)/ which had a forward reference to a
524 non-existent subpattern following a character class starting with ']' and
525 containing () gave an internal compiling error instead of "reference to
526 non-existent subpattern". Fortunately, when the pattern did exist, the
527 compiled code was correct. (When scanning forwards to check for the
528 existencd of the subpattern, it was treating the data ']' as terminating
529 the class, so got the count wrong. When actually compiling, the reference
530 was subsequently set up correctly.)
532 15. The "always fail" assertion (?!) is optimzed to (*FAIL) by pcre_compile;
533 it was being rejected as not supported by pcre_dfa_exec(), even though
534 other assertions are supported. I have made pcre_dfa_exec() support
535 (*FAIL).
537 16. The implementation of 13c above involved the invention of a new opcode,
538 OP_ALLANY, which is like OP_ANY but doesn't check the /s flag. Since /s
539 cannot be changed at match time, I realized I could make a small
540 improvement to matching performance by compiling OP_ALLANY instead of
541 OP_ANY for "." when DOTALL was set, and then removing the runtime tests
542 on the OP_ANY path.
544 17. Compiling pcretest on Windows with readline support failed without the
545 following two fixes: (1) Make the unistd.h include conditional on
546 HAVE_UNISTD_H; (2) #define isatty and fileno as _isatty and _fileno.
548 18. Changed CMakeLists.txt and cmake/FindReadline.cmake to arrange for the
549 ncurses library to be included for pcretest when ReadLine support is
550 requested, but also to allow for it to be overridden. This patch came from
551 Daniel Bergström.
553 19. There was a typo in the file ucpinternal.h where f0_rangeflag was defined
554 as 0x00f00000 instead of 0x00800000. Luckily, this would not have caused
555 any errors with the current Unicode tables. Thanks to Peter Kankowski for
556 spotting this.
559 Version 7.6 28-Jan-08
560 ---------------------
562 1. A character class containing a very large number of characters with
563 codepoints greater than 255 (in UTF-8 mode, of course) caused a buffer
564 overflow.
566 2. Patch to cut out the "long long" test in pcrecpp_unittest when
567 HAVE_LONG_LONG is not defined.
569 3. Applied Christian Ehrlicher's patch to update the CMake build files to
570 bring them up to date and include new features. This patch includes:
572 - Fixed PH's badly added libz and libbz2 support.
573 - Fixed a problem with static linking.
574 - Added pcredemo. [But later removed - see 7 below.]
575 - Fixed dftables problem and added an option.
576 - Added a number of HAVE_XXX tests, including HAVE_WINDOWS_H and
578 - Added readline support for pcretest.
579 - Added an listing of the option settings after cmake has run.
581 4. A user submitted a patch to Makefile that makes it easy to create
582 "pcre.dll" under mingw when using Configure/Make. I added stuff to
583 Makefile.am that cause it to include this special target, without
584 affecting anything else. Note that the same mingw target plus all
585 the other distribution libraries and programs are now supported
586 when configuring with CMake (see 6 below) instead of with
587 Configure/Make.
589 5. Applied Craig's patch that moves no_arg into the RE class in the C++ code.
590 This is an attempt to solve the reported problem "pcrecpp::no_arg is not
591 exported in the Windows port". It has not yet been confirmed that the patch
592 solves the problem, but it does no harm.
594 6. Applied Sheri's patch to CMakeLists.txt to add NON_STANDARD_LIB_PREFIX and
595 NON_STANDARD_LIB_SUFFIX for dll names built with mingw when configured
596 with CMake, and also correct the comment about stack recursion.
598 7. Remove the automatic building of pcredemo from the ./configure system and
599 from CMakeLists.txt. The whole idea of pcredemo.c is that it is an example
600 of a program that users should build themselves after PCRE is installed, so
601 building it automatically is not really right. What is more, it gave
602 trouble in some build environments.
604 8. Further tidies to CMakeLists.txt from Sheri and Christian.
607 Version 7.5 10-Jan-08
608 ---------------------
610 1. Applied a patch from Craig: "This patch makes it possible to 'ignore'
611 values in parens when parsing an RE using the C++ wrapper."
613 2. Negative specials like \S did not work in character classes in UTF-8 mode.
614 Characters greater than 255 were excluded from the class instead of being
615 included.
617 3. The same bug as (2) above applied to negated POSIX classes such as
618 [:^space:].
620 4. PCRECPP_STATIC was referenced in pcrecpp_internal.h, but nowhere was it
621 defined or documented. It seems to have been a typo for PCRE_STATIC, so
622 I have changed it.
624 5. The construct (?&) was not diagnosed as a syntax error (it referenced the
625 first named subpattern) and a construct such as (?&a) would reference the
626 first named subpattern whose name started with "a" (in other words, the
627 length check was missing). Both these problems are fixed. "Subpattern name
628 expected" is now given for (?&) (a zero-length name), and this patch also
629 makes it give the same error for \k'' (previously it complained that that
630 was a reference to a non-existent subpattern).
632 6. The erroneous patterns (?+-a) and (?-+a) give different error messages;
633 this is right because (?- can be followed by option settings as well as by
634 digits. I have, however, made the messages clearer.
636 7. Patterns such as (?(1)a|b) (a pattern that contains fewer subpatterns
637 than the number used in the conditional) now cause a compile-time error.
638 This is actually not compatible with Perl, which accepts such patterns, but
639 treats the conditional as always being FALSE (as PCRE used to), but it
640 seems to me that giving a diagnostic is better.
642 8. Change "alphameric" to the more common word "alphanumeric" in comments
643 and messages.
645 9. Fix two occurrences of "backslash" in comments that should have been
646 "backspace".
648 10. Remove two redundant lines of code that can never be obeyed (their function
649 was moved elsewhere).
651 11. The program that makes PCRE's Unicode character property table had a bug
652 which caused it to generate incorrect table entries for sequences of
653 characters that have the same character type, but are in different scripts.
654 It amalgamated them into a single range, with the script of the first of
655 them. In other words, some characters were in the wrong script. There were
656 thirteen such cases, affecting characters in the following ranges:
658 U+002b0 - U+002c1
659 U+0060c - U+0060d
660 U+0061e - U+00612
661 U+0064b - U+0065e
662 U+0074d - U+0076d
663 U+01800 - U+01805
664 U+01d00 - U+01d77
665 U+01d9b - U+01dbf
666 U+0200b - U+0200f
667 U+030fc - U+030fe
668 U+03260 - U+0327f
669 U+0fb46 - U+0fbb1
670 U+10450 - U+1049d
672 12. The -o option (show only the matching part of a line) for pcregrep was not
673 compatible with GNU grep in that, if there was more than one match in a
674 line, it showed only the first of them. It now behaves in the same way as
675 GNU grep.
677 13. If the -o and -v options were combined for pcregrep, it printed a blank
678 line for every non-matching line. GNU grep prints nothing, and pcregrep now
679 does the same. The return code can be used to tell if there were any
680 non-matching lines.
682 14. Added --file-offsets and --line-offsets to pcregrep.
684 15. The pattern (?=something)(?R) was not being diagnosed as a potentially
685 infinitely looping recursion. The bug was that positive lookaheads were not
686 being skipped when checking for a possible empty match (negative lookaheads
687 and both kinds of lookbehind were skipped).
689 16. Fixed two typos in the Windows-only code in pcregrep.c, and moved the
690 inclusion of <windows.h> to before rather than after the definition of
691 INVALID_FILE_ATTRIBUTES (patch from David Byron).
693 17. Specifying a possessive quantifier with a specific limit for a Unicode
694 character property caused pcre_compile() to compile bad code, which led at
695 runtime to PCRE_ERROR_INTERNAL (-14). Examples of patterns that caused this
696 are: /\p{Zl}{2,3}+/8 and /\p{Cc}{2}+/8. It was the possessive "+" that
697 caused the error; without that there was no problem.
699 18. Added --enable-pcregrep-libz and --enable-pcregrep-libbz2.
701 19. Added --enable-pcretest-libreadline.
703 20. In pcrecpp.cc, the variable 'count' was incremented twice in
704 RE::GlobalReplace(). As a result, the number of replacements returned was
705 double what it should be. I removed one of the increments, but Craig sent a
706 later patch that removed the other one (the right fix) and added unit tests
707 that check the return values (which was not done before).
709 21. Several CMake things:
711 (1) Arranged that, when cmake is used on Unix, the libraries end up with
712 the names libpcre and libpcreposix, not just pcre and pcreposix.
714 (2) The above change means that pcretest and pcregrep are now correctly
715 linked with the newly-built libraries, not previously installed ones.
719 22. In UTF-8 mode, with newline set to "any", a pattern such as .*a.*=.b.*
720 crashed when matching a string such as a\x{2029}b (note that \x{2029} is a
721 UTF-8 newline character). The key issue is that the pattern starts .*;
722 this means that the match must be either at the beginning, or after a
723 newline. The bug was in the code for advancing after a failed match and
724 checking that the new position followed a newline. It was not taking
725 account of UTF-8 characters correctly.
727 23. PCRE was behaving differently from Perl in the way it recognized POSIX
728 character classes. PCRE was not treating the sequence [:...:] as a
729 character class unless the ... were all letters. Perl, however, seems to
730 allow any characters between [: and :], though of course it rejects as
731 unknown any "names" that contain non-letters, because all the known class
732 names consist only of letters. Thus, Perl gives an error for [[:1234:]],
733 for example, whereas PCRE did not - it did not recognize a POSIX character
734 class. This seemed a bit dangerous, so the code has been changed to be
735 closer to Perl. The behaviour is not identical to Perl, because PCRE will
736 diagnose an unknown class for, for example, [[:l\ower:]] where Perl will
737 treat it as [[:lower:]]. However, PCRE does now give "unknown" errors where
738 Perl does, and where it didn't before.
740 24. Rewrite so as to remove the single use of %n from pcregrep because in some
741 Windows environments %n is disabled by default.
744 Version 7.4 21-Sep-07
745 ---------------------
747 1. Change 7.3/28 was implemented for classes by looking at the bitmap. This
748 means that a class such as [\s] counted as "explicit reference to CR or
749 LF". That isn't really right - the whole point of the change was to try to
750 help when there was an actual mention of one of the two characters. So now
751 the change happens only if \r or \n (or a literal CR or LF) character is
752 encountered.
754 2. The 32-bit options word was also used for 6 internal flags, but the numbers
755 of both had grown to the point where there were only 3 bits left.
756 Fortunately, there was spare space in the data structure, and so I have
757 moved the internal flags into a new 16-bit field to free up more option
758 bits.
760 3. The appearance of (?J) at the start of a pattern set the DUPNAMES option,
761 but did not set the internal JCHANGED flag - either of these is enough to
762 control the way the "get" function works - but the PCRE_INFO_JCHANGED
763 facility is supposed to tell if (?J) was ever used, so now (?J) at the
764 start sets both bits.
766 4. Added options (at build time, compile time, exec time) to change \R from
767 matching any Unicode line ending sequence to just matching CR, LF, or CRLF.
769 5. doc/pcresyntax.html was missing from the distribution.
771 6. Put back the definition of PCRE_ERROR_NULLWSLIMIT, for backward
772 compatibility, even though it is no longer used.
774 7. Added macro for snprintf to pcrecpp_unittest.cc and also for strtoll and
775 strtoull to pcrecpp.cc to select the available functions in WIN32 when the
776 windows.h file is present (where different names are used). [This was
777 reversed later after testing - see 16 below.]
779 8. Changed all #include <config.h> to #include "config.h". There were also
780 some further <pcre.h> cases that I changed to "pcre.h".
782 9. When pcregrep was used with the --colour option, it missed the line ending
783 sequence off the lines that it output.
785 10. It was pointed out to me that arrays of string pointers cause lots of
786 relocations when a shared library is dynamically loaded. A technique of
787 using a single long string with a table of offsets can drastically reduce
788 these. I have refactored PCRE in four places to do this. The result is
789 dramatic:
791 Originally: 290
792 After changing UCP table: 187
793 After changing error message table: 43
794 After changing table of "verbs" 36
795 After changing table of Posix names 22
797 Thanks to the folks working on Gregex for glib for this insight.
799 11. --disable-stack-for-recursion caused compiling to fail unless -enable-
800 unicode-properties was also set.
802 12. Updated the tests so that they work when \R is defaulted to ANYCRLF.
804 13. Added checks for ANY and ANYCRLF to pcrecpp.cc where it previously
805 checked only for CRLF.
807 14. Added casts to pcretest.c to avoid compiler warnings.
809 15. Added Craig's patch to various pcrecpp modules to avoid compiler warnings.
811 16. Added Craig's patch to remove the WINDOWS_H tests, that were not working,
812 and instead check for _strtoi64 explicitly, and avoid the use of snprintf()
813 entirely. This removes changes made in 7 above.
815 17. The CMake files have been updated, and there is now more information about
816 building with CMake in the NON-UNIX-USE document.
819 Version 7.3 28-Aug-07
820 ---------------------
822 1. In the rejigging of the build system that eventually resulted in 7.1, the
823 line "#include <pcre.h>" was included in pcre_internal.h. The use of angle
824 brackets there is not right, since it causes compilers to look for an
825 installed pcre.h, not the version that is in the source that is being
826 compiled (which of course may be different). I have changed it back to:
828 #include "pcre.h"
830 I have a vague recollection that the change was concerned with compiling in
831 different directories, but in the new build system, that is taken care of
832 by the VPATH setting the Makefile.
834 2. The pattern .*$ when run in not-DOTALL UTF-8 mode with newline=any failed
835 when the subject happened to end in the byte 0x85 (e.g. if the last
836 character was \x{1ec5}). *Character* 0x85 is one of the "any" newline
837 characters but of course it shouldn't be taken as a newline when it is part
838 of another character. The bug was that, for an unlimited repeat of . in
839 not-DOTALL UTF-8 mode, PCRE was advancing by bytes rather than by
840 characters when looking for a newline.
842 3. A small performance improvement in the DOTALL UTF-8 mode .* case.
844 4. Debugging: adjusted the names of opcodes for different kinds of parentheses
845 in debug output.
847 5. Arrange to use "%I64d" instead of "%lld" and "%I64u" instead of "%llu" for
848 long printing in the pcrecpp unittest when running under MinGW.
850 6. ESC_K was left out of the EBCDIC table.
852 7. Change 7.0/38 introduced a new limit on the number of nested non-capturing
853 parentheses; I made it 1000, which seemed large enough. Unfortunately, the
854 limit also applies to "virtual nesting" when a pattern is recursive, and in
855 this case 1000 isn't so big. I have been able to remove this limit at the
856 expense of backing off one optimization in certain circumstances. Normally,
857 when pcre_exec() would call its internal match() function recursively and
858 immediately return the result unconditionally, it uses a "tail recursion"
859 feature to save stack. However, when a subpattern that can match an empty
860 string has an unlimited repetition quantifier, it no longer makes this
861 optimization. That gives it a stack frame in which to save the data for
862 checking that an empty string has been matched. Previously this was taken
863 from the 1000-entry workspace that had been reserved. So now there is no
864 explicit limit, but more stack is used.
866 8. Applied Daniel's patches to solve problems with the import/export magic
867 syntax that is required for Windows, and which was going wrong for the
868 pcreposix and pcrecpp parts of the library. These were overlooked when this
869 problem was solved for the main library.
871 9. There were some crude static tests to avoid integer overflow when computing
872 the size of patterns that contain repeated groups with explicit upper
873 limits. As the maximum quantifier is 65535, the maximum group length was
874 set at 30,000 so that the product of these two numbers did not overflow a
875 32-bit integer. However, it turns out that people want to use groups that
876 are longer than 30,000 bytes (though not repeat them that many times).
877 Change 7.0/17 (the refactoring of the way the pattern size is computed) has
878 made it possible to implement the integer overflow checks in a much more
879 dynamic way, which I have now done. The artificial limitation on group
880 length has been removed - we now have only the limit on the total length of
881 the compiled pattern, which depends on the LINK_SIZE setting.
883 10. Fixed a bug in the documentation for get/copy named substring when
884 duplicate names are permitted. If none of the named substrings are set, the
885 functions return PCRE_ERROR_NOSUBSTRING (7); the doc said they returned an
886 empty string.
888 11. Because Perl interprets \Q...\E at a high level, and ignores orphan \E
889 instances, patterns such as [\Q\E] or [\E] or even [^\E] cause an error,
890 because the ] is interpreted as the first data character and the
891 terminating ] is not found. PCRE has been made compatible with Perl in this
892 regard. Previously, it interpreted [\Q\E] as an empty class, and [\E] could
893 cause memory overwriting.
895 10. Like Perl, PCRE automatically breaks an unlimited repeat after an empty
896 string has been matched (to stop an infinite loop). It was not recognizing
897 a conditional subpattern that could match an empty string if that
898 subpattern was within another subpattern. For example, it looped when
899 trying to match (((?(1)X|))*) but it was OK with ((?(1)X|)*) where the
900 condition was not nested. This bug has been fixed.
902 12. A pattern like \X?\d or \P{L}?\d in non-UTF-8 mode could cause a backtrack
903 past the start of the subject in the presence of bytes with the top bit
904 set, for example "\x8aBCD".
906 13. Added Perl 5.10 experimental backtracking controls (*FAIL), (*F), (*PRUNE),
907 (*SKIP), (*THEN), (*COMMIT), and (*ACCEPT).
909 14. Optimized (?!) to (*FAIL).
911 15. Updated the test for a valid UTF-8 string to conform to the later RFC 3629.
912 This restricts code points to be within the range 0 to 0x10FFFF, excluding
913 the "low surrogate" sequence 0xD800 to 0xDFFF. Previously, PCRE allowed the
914 full range 0 to 0x7FFFFFFF, as defined by RFC 2279. Internally, it still
915 does: it's just the validity check that is more restrictive.
917 16. Inserted checks for integer overflows during escape sequence (backslash)
918 processing, and also fixed erroneous offset values for syntax errors during
919 backslash processing.
921 17. Fixed another case of looking too far back in non-UTF-8 mode (cf 12 above)
922 for patterns like [\PPP\x8a]{1,}\x80 with the subject "A\x80".
924 18. An unterminated class in a pattern like (?1)\c[ with a "forward reference"
925 caused an overrun.
927 19. A pattern like (?:[\PPa*]*){8,} which had an "extended class" (one with
928 something other than just ASCII characters) inside a group that had an
929 unlimited repeat caused a loop at compile time (while checking to see
930 whether the group could match an empty string).
932 20. Debugging a pattern containing \p or \P could cause a crash. For example,
933 [\P{Any}] did so. (Error in the code for printing property names.)
935 21. An orphan \E inside a character class could cause a crash.
937 22. A repeated capturing bracket such as (A)? could cause a wild memory
938 reference during compilation.
940 23. There are several functions in pcre_compile() that scan along a compiled
941 expression for various reasons (e.g. to see if it's fixed length for look
942 behind). There were bugs in these functions when a repeated \p or \P was
943 present in the pattern. These operators have additional parameters compared
944 with \d, etc, and these were not being taken into account when moving along
945 the compiled data. Specifically:
947 (a) A item such as \p{Yi}{3} in a lookbehind was not treated as fixed
948 length.
950 (b) An item such as \pL+ within a repeated group could cause crashes or
951 loops.
953 (c) A pattern such as \p{Yi}+(\P{Yi}+)(?1) could give an incorrect
954 "reference to non-existent subpattern" error.
956 (d) A pattern like (\P{Yi}{2}\277)? could loop at compile time.
958 24. A repeated \S or \W in UTF-8 mode could give wrong answers when multibyte
959 characters were involved (for example /\S{2}/8g with "A\x{a3}BC").
961 25. Using pcregrep in multiline, inverted mode (-Mv) caused it to loop.
963 26. Patterns such as [\P{Yi}A] which include \p or \P and just one other
964 character were causing crashes (broken optimization).
966 27. Patterns such as (\P{Yi}*\277)* (group with possible zero repeat containing
967 \p or \P) caused a compile-time loop.
969 28. More problems have arisen in unanchored patterns when CRLF is a valid line
970 break. For example, the unstudied pattern [\r\n]A does not match the string
971 "\r\nA" because change 7.0/46 below moves the current point on by two
972 characters after failing to match at the start. However, the pattern \nA
973 *does* match, because it doesn't start till \n, and if [\r\n]A is studied,
974 the same is true. There doesn't seem any very clean way out of this, but
975 what I have chosen to do makes the common cases work: PCRE now takes note
976 of whether there can be an explicit match for \r or \n anywhere in the
977 pattern, and if so, 7.0/46 no longer applies. As part of this change,
978 there's a new PCRE_INFO_HASCRORLF option for finding out whether a compiled
979 pattern has explicit CR or LF references.
981 29. Added (*CR) etc for changing newline setting at start of pattern.
984 Version 7.2 19-Jun-07
985 ---------------------
987 1. If the fr_FR locale cannot be found for test 3, try the "french" locale,
988 which is apparently normally available under Windows.
990 2. Re-jig the pcregrep tests with different newline settings in an attempt
991 to make them independent of the local environment's newline setting.
993 3. Add code to configure.ac to remove -g from the CFLAGS default settings.
995 4. Some of the "internals" tests were previously cut out when the link size
996 was not 2, because the output contained actual offsets. The recent new
997 "Z" feature of pcretest means that these can be cut out, making the tests
998 usable with all link sizes.
1000 5. Implemented Stan Switzer's goto replacement for longjmp() when not using
1001 stack recursion. This gives a massive performance boost under BSD, but just
1002 a small improvement under Linux. However, it saves one field in the frame
1003 in all cases.
1005 6. Added more features from the forthcoming Perl 5.10:
1007 (a) (?-n) (where n is a string of digits) is a relative subroutine or
1008 recursion call. It refers to the nth most recently opened parentheses.
1010 (b) (?+n) is also a relative subroutine call; it refers to the nth next
1011 to be opened parentheses.
1013 (c) Conditions that refer to capturing parentheses can be specified
1014 relatively, for example, (?(-2)... or (?(+3)...
1016 (d) \K resets the start of the current match so that everything before
1017 is not part of it.
1019 (e) \k{name} is synonymous with \k<name> and \k'name' (.NET compatible).
1021 (f) \g{name} is another synonym - part of Perl 5.10's unification of
1022 reference syntax.
1024 (g) (?| introduces a group in which the numbering of parentheses in each
1025 alternative starts with the same number.
1027 (h) \h, \H, \v, and \V match horizontal and vertical whitespace.
1029 7. Added two new calls to pcre_fullinfo(): PCRE_INFO_OKPARTIAL and
1032 8. A pattern such as (.*(.)?)* caused pcre_exec() to fail by either not
1033 terminating or by crashing. Diagnosed by Viktor Griph; it was in the code
1034 for detecting groups that can match an empty string.
1036 9. A pattern with a very large number of alternatives (more than several
1037 hundred) was running out of internal workspace during the pre-compile
1038 phase, where pcre_compile() figures out how much memory will be needed. A
1039 bit of new cunning has reduced the workspace needed for groups with
1040 alternatives. The 1000-alternative test pattern now uses 12 bytes of
1041 workspace instead of running out of the 4096 that are available.
1043 10. Inserted some missing (unsigned int) casts to get rid of compiler warnings.
1045 11. Applied patch from Google to remove an optimization that didn't quite work.
1046 The report of the bug said:
1048 pcrecpp::RE("a*").FullMatch("aaa") matches, while
1049 pcrecpp::RE("a*?").FullMatch("aaa") does not, and
1050 pcrecpp::RE("a*?\\z").FullMatch("aaa") does again.
1052 12. If \p or \P was used in non-UTF-8 mode on a character greater than 127
1053 it matched the wrong number of bytes.
1056 Version 7.1 24-Apr-07
1057 ---------------------
1059 1. Applied Bob Rossi and Daniel G's patches to convert the build system to one
1060 that is more "standard", making use of automake and other Autotools. There
1061 is some re-arrangement of the files and adjustment of comments consequent
1062 on this.
1064 2. Part of the patch fixed a problem with the pcregrep tests. The test of -r
1065 for recursive directory scanning broke on some systems because the files
1066 are not scanned in any specific order and on different systems the order
1067 was different. A call to "sort" has been inserted into RunGrepTest for the
1068 approprate test as a short-term fix. In the longer term there may be an
1069 alternative.
1071 3. I had an email from Eric Raymond about problems translating some of PCRE's
1072 man pages to HTML (despite the fact that I distribute HTML pages, some
1073 people do their own conversions for various reasons). The problems
1074 concerned the use of low-level troff macros .br and .in. I have therefore
1075 removed all such uses from the man pages (some were redundant, some could
1076 be replaced by .nf/.fi pairs). The 132html script that I use to generate
1077 HTML has been updated to handle .nf/.fi and to complain if it encounters
1078 .br or .in.
1080 4. Updated comments in configure.ac that get placed in config.h.in and also
1081 arranged for config.h to be included in the distribution, with the name
1082 config.h.generic, for the benefit of those who have to compile without
1083 Autotools (compare pcre.h, which is now distributed as pcre.h.generic).
1085 5. Updated the support (such as it is) for Virtual Pascal, thanks to Stefan
1086 Weber: (1) pcre_internal.h was missing some function renames; (2) updated
1087 makevp.bat for the current PCRE, using the additional files
1088 makevp_c.txt, makevp_l.txt, and pcregexp.pas.
1090 6. A Windows user reported a minor discrepancy with test 2, which turned out
1091 to be caused by a trailing space on an input line that had got lost in his
1092 copy. The trailing space was an accident, so I've just removed it.
1094 7. Add -Wl,-R... flags in pcre-config.in for *BSD* systems, as I'm told
1095 that is needed.
1097 8. Mark ucp_table (in ucptable.h) and ucp_gentype (in pcre_ucp_searchfuncs.c)
1098 as "const" (a) because they are and (b) because it helps the PHP
1099 maintainers who have recently made a script to detect big data structures
1100 in the php code that should be moved to the .rodata section. I remembered
1101 to update Builducptable as well, so it won't revert if ucptable.h is ever
1102 re-created.
1104 9. Added some extra #ifdef SUPPORT_UTF8 conditionals into pcretest.c,
1105 pcre_printint.src, pcre_compile.c, pcre_study.c, and pcre_tables.c, in
1106 order to be able to cut out the UTF-8 tables in the latter when UTF-8
1107 support is not required. This saves 1.5-2K of code, which is important in
1108 some applications.
1110 Later: more #ifdefs are needed in pcre_ord2utf8.c and pcre_valid_utf8.c
1111 so as not to refer to the tables, even though these functions will never be
1112 called when UTF-8 support is disabled. Otherwise there are problems with a
1113 shared library.
1115 10. Fixed two bugs in the emulated memmove() function in pcre_internal.h:
1117 (a) It was defining its arguments as char * instead of void *.
1119 (b) It was assuming that all moves were upwards in memory; this was true
1120 a long time ago when I wrote it, but is no longer the case.
1122 The emulated memove() is provided for those environments that have neither
1123 memmove() nor bcopy(). I didn't think anyone used it these days, but that
1124 is clearly not the case, as these two bugs were recently reported.
1126 11. The script PrepareRelease is now distributed: it calls 132html, CleanTxt,
1127 and Detrail to create the HTML documentation, the .txt form of the man
1128 pages, and it removes trailing spaces from listed files. It also creates
1129 pcre.h.generic and config.h.generic from pcre.h and config.h. In the latter
1130 case, it wraps all the #defines with #ifndefs. This script should be run
1131 before "make dist".
1133 12. Fixed two fairly obscure bugs concerned with quantified caseless matching
1134 with Unicode property support.
1136 (a) For a maximizing quantifier, if the two different cases of the
1137 character were of different lengths in their UTF-8 codings (there are
1138 some cases like this - I found 11), and the matching function had to
1139 back up over a mixture of the two cases, it incorrectly assumed they
1140 were both the same length.
1142 (b) When PCRE was configured to use the heap rather than the stack for
1143 recursion during matching, it was not correctly preserving the data for
1144 the other case of a UTF-8 character when checking ahead for a match
1145 while processing a minimizing repeat. If the check also involved
1146 matching a wide character, but failed, corruption could cause an
1147 erroneous result when trying to check for a repeat of the original
1148 character.
1150 13. Some tidying changes to the testing mechanism:
1152 (a) The RunTest script now detects the internal link size and whether there
1153 is UTF-8 and UCP support by running ./pcretest -C instead of relying on
1154 values substituted by "configure". (The RunGrepTest script already did
1155 this for UTF-8.) The configure.ac script no longer substitutes the
1156 relevant variables.
1158 (b) The debugging options /B and /D in pcretest show the compiled bytecode
1159 with length and offset values. This means that the output is different
1160 for different internal link sizes. Test 2 is skipped for link sizes
1161 other than 2 because of this, bypassing the problem. Unfortunately,
1162 there was also a test in test 3 (the locale tests) that used /B and
1163 failed for link sizes other than 2. Rather than cut the whole test out,
1164 I have added a new /Z option to pcretest that replaces the length and
1165 offset values with spaces. This is now used to make test 3 independent
1166 of link size. (Test 2 will be tidied up later.)
1168 14. If erroroffset was passed as NULL to pcre_compile, it provoked a
1169 segmentation fault instead of returning the appropriate error message.
1171 15. In multiline mode when the newline sequence was set to "any", the pattern
1172 ^$ would give a match between the \r and \n of a subject such as "A\r\nB".
1173 This doesn't seem right; it now treats the CRLF combination as the line
1174 ending, and so does not match in that case. It's only a pattern such as ^$
1175 that would hit this one: something like ^ABC$ would have failed after \r
1176 and then tried again after \r\n.
1178 16. Changed the comparison command for RunGrepTest from "diff -u" to "diff -ub"
1179 in an attempt to make files that differ only in their line terminators
1180 compare equal. This works on Linux.
1182 17. Under certain error circumstances pcregrep might try to free random memory
1183 as it exited. This is now fixed, thanks to valgrind.
1185 19. In pcretest, if the pattern /(?m)^$/g<any> was matched against the string
1186 "abc\r\n\r\n", it found an unwanted second match after the second \r. This
1187 was because its rules for how to advance for /g after matching an empty
1188 string at the end of a line did not allow for this case. They now check for
1189 it specially.
1191 20. pcretest is supposed to handle patterns and data of any length, by
1192 extending its buffers when necessary. It was getting this wrong when the
1193 buffer for a data line had to be extended.
1195 21. Added PCRE_NEWLINE_ANYCRLF which is like ANY, but matches only CR, LF, or
1196 CRLF as a newline sequence.
1198 22. Code for handling Unicode properties in pcre_dfa_exec() wasn't being cut
1199 out by #ifdef SUPPORT_UCP. This did no harm, as it could never be used, but
1200 I have nevertheless tidied it up.
1202 23. Added some casts to kill warnings from HP-UX ia64 compiler.
1204 24. Added a man page for pcre-config.
1207 Version 7.0 19-Dec-06
1208 ---------------------
1210 1. Fixed a signed/unsigned compiler warning in pcre_compile.c, shown up by
1211 moving to gcc 4.1.1.
1213 2. The -S option for pcretest uses setrlimit(); I had omitted to #include
1214 sys/time.h, which is documented as needed for this function. It doesn't
1215 seem to matter on Linux, but it showed up on some releases of OS X.
1217 3. It seems that there are systems where bytes whose values are greater than
1218 127 match isprint() in the "C" locale. The "C" locale should be the
1219 default when a C program starts up. In most systems, only ASCII printing
1220 characters match isprint(). This difference caused the output from pcretest
1221 to vary, making some of the tests fail. I have changed pcretest so that:
1223 (a) When it is outputting text in the compiled version of a pattern, bytes
1224 other than 32-126 are always shown as hex escapes.
1226 (b) When it is outputting text that is a matched part of a subject string,
1227 it does the same, unless a different locale has been set for the match
1228 (using the /L modifier). In this case, it uses isprint() to decide.
1230 4. Fixed a major bug that caused incorrect computation of the amount of memory
1231 required for a compiled pattern when options that changed within the
1232 pattern affected the logic of the preliminary scan that determines the
1233 length. The relevant options are -x, and -i in UTF-8 mode. The result was
1234 that the computed length was too small. The symptoms of this bug were
1235 either the PCRE error "internal error: code overflow" from pcre_compile(),
1236 or a glibc crash with a message such as "pcretest: free(): invalid next
1237 size (fast)". Examples of patterns that provoked this bug (shown in
1238 pcretest format) are:
1240 /(?-x: )/x
1241 /(?x)(?-x: \s*#\s*)/
1242 /((?i)[\x{c0}])/8
1243 /(?i:[\x{c0}])/8
1245 HOWEVER: Change 17 below makes this fix obsolete as the memory computation
1246 is now done differently.
1248 5. Applied patches from Google to: (a) add a QuoteMeta function to the C++
1249 wrapper classes; (b) implement a new function in the C++ scanner that is
1250 more efficient than the old way of doing things because it avoids levels of
1251 recursion in the regex matching; (c) add a paragraph to the documentation
1252 for the FullMatch() function.
1254 6. The escape sequence \n was being treated as whatever was defined as
1255 "newline". Not only was this contrary to the documentation, which states
1256 that \n is character 10 (hex 0A), but it also went horribly wrong when
1257 "newline" was defined as CRLF. This has been fixed.
1259 7. In pcre_dfa_exec.c the value of an unsigned integer (the variable called c)
1260 was being set to -1 for the "end of line" case (supposedly a value that no
1261 character can have). Though this value is never used (the check for end of
1262 line is "zero bytes in current character"), it caused compiler complaints.
1263 I've changed it to 0xffffffff.
1265 8. In pcre_version.c, the version string was being built by a sequence of
1266 C macros that, in the event of PCRE_PRERELEASE being defined as an empty
1267 string (as it is for production releases) called a macro with an empty
1268 argument. The C standard says the result of this is undefined. The gcc
1269 compiler treats it as an empty string (which was what was wanted) but it is
1270 reported that Visual C gives an error. The source has been hacked around to
1271 avoid this problem.
1273 9. On the advice of a Windows user, included <io.h> and <fcntl.h> in Windows
1274 builds of pcretest, and changed the call to _setmode() to use _O_BINARY
1275 instead of 0x8000. Made all the #ifdefs test both _WIN32 and WIN32 (not all
1276 of them did).
1278 10. Originally, pcretest opened its input and output without "b"; then I was
1279 told that "b" was needed in some environments, so it was added for release
1280 5.0 to both the input and output. (It makes no difference on Unix-like
1281 systems.) Later I was told that it is wrong for the input on Windows. I've
1282 now abstracted the modes into two macros, to make it easier to fiddle with
1283 them, and removed "b" from the input mode under Windows.
1285 11. Added pkgconfig support for the C++ wrapper library, libpcrecpp.
1287 12. Added -help and --help to pcretest as an official way of being reminded
1288 of the options.
1290 13. Removed some redundant semicolons after macro calls in pcrecpparg.h.in
1291 and pcrecpp.cc because they annoy compilers at high warning levels.
1293 14. A bit of tidying/refactoring in pcre_exec.c in the main bumpalong loop.
1295 15. Fixed an occurrence of == in configure.ac that should have been = (shell
1296 scripts are not C programs :-) and which was not noticed because it works
1297 on Linux.
1299 16. pcretest is supposed to handle any length of pattern and data line (as one
1300 line or as a continued sequence of lines) by extending its input buffer if
1301 necessary. This feature was broken for very long pattern lines, leading to
1302 a string of junk being passed to pcre_compile() if the pattern was longer
1303 than about 50K.
1305 17. I have done a major re-factoring of the way pcre_compile() computes the
1306 amount of memory needed for a compiled pattern. Previously, there was code
1307 that made a preliminary scan of the pattern in order to do this. That was
1308 OK when PCRE was new, but as the facilities have expanded, it has become
1309 harder and harder to keep it in step with the real compile phase, and there
1310 have been a number of bugs (see for example, 4 above). I have now found a
1311 cunning way of running the real compile function in a "fake" mode that
1312 enables it to compute how much memory it would need, while actually only
1313 ever using a few hundred bytes of working memory and without too many
1314 tests of the mode. This should make future maintenance and development
1315 easier. A side effect of this work is that the limit of 200 on the nesting
1316 depth of parentheses has been removed (though this was never a serious
1317 limitation, I suspect). However, there is a downside: pcre_compile() now
1318 runs more slowly than before (30% or more, depending on the pattern). I
1319 hope this isn't a big issue. There is no effect on runtime performance.
1321 18. Fixed a minor bug in pcretest: if a pattern line was not terminated by a
1322 newline (only possible for the last line of a file) and it was a
1323 pattern that set a locale (followed by /Lsomething), pcretest crashed.
1325 19. Added additional timing features to pcretest. (1) The -tm option now times
1326 matching only, not compiling. (2) Both -t and -tm can be followed, as a
1327 separate command line item, by a number that specifies the number of
1328 repeats to use when timing. The default is 50000; this gives better
1329 precision, but takes uncomfortably long for very large patterns.
1331 20. Extended pcre_study() to be more clever in cases where a branch of a
1332 subpattern has no definite first character. For example, (a*|b*)[cd] would
1333 previously give no result from pcre_study(). Now it recognizes that the
1334 first character must be a, b, c, or d.
1336 21. There was an incorrect error "recursive call could loop indefinitely" if
1337 a subpattern (or the entire pattern) that was being tested for matching an
1338 empty string contained only one non-empty item after a nested subpattern.
1339 For example, the pattern (?>\x{100}*)\d(?R) provoked this error
1340 incorrectly, because the \d was being skipped in the check.
1342 22. The pcretest program now has a new pattern option /B and a command line
1343 option -b, which is equivalent to adding /B to every pattern. This causes
1344 it to show the compiled bytecode, without the additional information that
1345 -d shows. The effect of -d is now the same as -b with -i (and similarly, /D
1346 is the same as /B/I).
1348 23. A new optimization is now able automatically to treat some sequences such
1349 as a*b as a*+b. More specifically, if something simple (such as a character
1350 or a simple class like \d) has an unlimited quantifier, and is followed by
1351 something that cannot possibly match the quantified thing, the quantifier
1352 is automatically "possessified".
1354 24. A recursive reference to a subpattern whose number was greater than 39
1355 went wrong under certain circumstances in UTF-8 mode. This bug could also
1356 have affected the operation of pcre_study().
1358 25. Realized that a little bit of performance could be had by replacing
1359 (c & 0xc0) == 0xc0 with c >= 0xc0 when processing UTF-8 characters.
1361 26. Timing data from pcretest is now shown to 4 decimal places instead of 3.
1363 27. Possessive quantifiers such as a++ were previously implemented by turning
1364 them into atomic groups such as ($>a+). Now they have their own opcodes,
1365 which improves performance. This includes the automatically created ones
1366 from 23 above.
1368 28. A pattern such as (?=(\w+))\1: which simulates an atomic group using a
1369 lookahead was broken if it was not anchored. PCRE was mistakenly expecting
1370 the first matched character to be a colon. This applied both to named and
1371 numbered groups.
1373 29. The ucpinternal.h header file was missing its idempotency #ifdef.
1375 30. I was sent a "project" file called libpcre.a.dev which I understand makes
1376 building PCRE on Windows easier, so I have included it in the distribution.
1378 31. There is now a check in pcretest against a ridiculously large number being
1379 returned by pcre_exec() or pcre_dfa_exec(). If this happens in a /g or /G
1380 loop, the loop is abandoned.
1382 32. Forward references to subpatterns in conditions such as (?(2)...) where
1383 subpattern 2 is defined later cause pcre_compile() to search forwards in
1384 the pattern for the relevant set of parentheses. This search went wrong
1385 when there were unescaped parentheses in a character class, parentheses
1386 escaped with \Q...\E, or parentheses in a #-comment in /x mode.
1388 33. "Subroutine" calls and backreferences were previously restricted to
1389 referencing subpatterns earlier in the regex. This restriction has now
1390 been removed.
1392 34. Added a number of extra features that are going to be in Perl 5.10. On the
1393 whole, these are just syntactic alternatives for features that PCRE had
1394 previously implemented using the Python syntax or my own invention. The
1395 other formats are all retained for compatibility.
1397 (a) Named groups can now be defined as (?<name>...) or (?'name'...) as well
1398 as (?P<name>...). The new forms, as well as being in Perl 5.10, are
1399 also .NET compatible.
1401 (b) A recursion or subroutine call to a named group can now be defined as
1402 (?&name) as well as (?P>name).
1404 (c) A backreference to a named group can now be defined as \k<name> or
1405 \k'name' as well as (?P=name). The new forms, as well as being in Perl
1406 5.10, are also .NET compatible.
1408 (d) A conditional reference to a named group can now use the syntax
1409 (?(<name>) or (?('name') as well as (?(name).
1411 (e) A "conditional group" of the form (?(DEFINE)...) can be used to define
1412 groups (named and numbered) that are never evaluated inline, but can be
1413 called as "subroutines" from elsewhere. In effect, the DEFINE condition
1414 is always false. There may be only one alternative in such a group.
1416 (f) A test for recursion can be given as (?(R1).. or (?(R&name)... as well
1417 as the simple (?(R). The condition is true only if the most recent
1418 recursion is that of the given number or name. It does not search out
1419 through the entire recursion stack.
1421 (g) The escape \gN or \g{N} has been added, where N is a positive or
1422 negative number, specifying an absolute or relative reference.
1424 35. Tidied to get rid of some further signed/unsigned compiler warnings and
1425 some "unreachable code" warnings.
1427 36. Updated the Unicode property tables to Unicode version 5.0.0. Amongst other
1428 things, this adds five new scripts.
1430 37. Perl ignores orphaned \E escapes completely. PCRE now does the same.
1431 There were also incompatibilities regarding the handling of \Q..\E inside
1432 character classes, for example with patterns like [\Qa\E-\Qz\E] where the
1433 hyphen was adjacent to \Q or \E. I hope I've cleared all this up now.
1435 38. Like Perl, PCRE detects when an indefinitely repeated parenthesized group
1436 matches an empty string, and forcibly breaks the loop. There were bugs in
1437 this code in non-simple cases. For a pattern such as ^(a()*)* matched
1438 against aaaa the result was just "a" rather than "aaaa", for example. Two
1439 separate and independent bugs (that affected different cases) have been
1440 fixed.
1442 39. Refactored the code to abolish the use of different opcodes for small
1443 capturing bracket numbers. This is a tidy that I avoided doing when I
1444 removed the limit on the number of capturing brackets for 3.5 back in 2001.
1445 The new approach is not only tidier, it makes it possible to reduce the
1446 memory needed to fix the previous bug (38).
1448 40. Implemented PCRE_NEWLINE_ANY to recognize any of the Unicode newline
1449 sequences (http://unicode.org/unicode/reports/tr18/) as "newline" when
1450 processing dot, circumflex, or dollar metacharacters, or #-comments in /x
1451 mode.
1453 41. Add \R to match any Unicode newline sequence, as suggested in the Unicode
1454 report.
1456 42. Applied patch, originally from Ari Pollak, modified by Google, to allow
1457 copy construction and assignment in the C++ wrapper.
1459 43. Updated pcregrep to support "--newline=any". In the process, I fixed a
1460 couple of bugs that could have given wrong results in the "--newline=crlf"
1461 case.
1463 44. Added a number of casts and did some reorganization of signed/unsigned int
1464 variables following suggestions from Dair Grant. Also renamed the variable
1465 "this" as "item" because it is a C++ keyword.
1467 45. Arranged for dftables to add
1469 #include "pcre_internal.h"
1471 to pcre_chartables.c because without it, gcc 4.x may remove the array
1472 definition from the final binary if PCRE is built into a static library and
1473 dead code stripping is activated.
1475 46. For an unanchored pattern, if a match attempt fails at the start of a
1476 newline sequence, and the newline setting is CRLF or ANY, and the next two
1477 characters are CRLF, advance by two characters instead of one.
1480 Version 6.7 04-Jul-06
1481 ---------------------
1483 1. In order to handle tests when input lines are enormously long, pcretest has
1484 been re-factored so that it automatically extends its buffers when
1485 necessary. The code is crude, but this _is_ just a test program. The
1486 default size has been increased from 32K to 50K.
1488 2. The code in pcre_study() was using the value of the re argument before
1489 testing it for NULL. (Of course, in any sensible call of the function, it
1490 won't be NULL.)
1492 3. The memmove() emulation function in pcre_internal.h, which is used on
1493 systems that lack both memmove() and bcopy() - that is, hardly ever -
1494 was missing a "static" storage class specifier.
1496 4. When UTF-8 mode was not set, PCRE looped when compiling certain patterns
1497 containing an extended class (one that cannot be represented by a bitmap
1498 because it contains high-valued characters or Unicode property items, e.g.
1499 [\pZ]). Almost always one would set UTF-8 mode when processing such a
1500 pattern, but PCRE should not loop if you do not (it no longer does).
1501 [Detail: two cases were found: (a) a repeated subpattern containing an
1502 extended class; (b) a recursive reference to a subpattern that followed a
1503 previous extended class. It wasn't skipping over the extended class
1504 correctly when UTF-8 mode was not set.]
1506 5. A negated single-character class was not being recognized as fixed-length
1507 in lookbehind assertions such as (?<=[^f]), leading to an incorrect
1508 compile error "lookbehind assertion is not fixed length".
1510 6. The RunPerlTest auxiliary script was showing an unexpected difference
1511 between PCRE and Perl for UTF-8 tests. It turns out that it is hard to
1512 write a Perl script that can interpret lines of an input file either as
1513 byte characters or as UTF-8, which is what "perltest" was being required to
1514 do for the non-UTF-8 and UTF-8 tests, respectively. Essentially what you
1515 can't do is switch easily at run time between having the "use utf8;" pragma
1516 or not. In the end, I fudged it by using the RunPerlTest script to insert
1517 "use utf8;" explicitly for the UTF-8 tests.
1519 7. In multiline (/m) mode, PCRE was matching ^ after a terminating newline at
1520 the end of the subject string, contrary to the documentation and to what
1521 Perl does. This was true of both matching functions. Now it matches only at
1522 the start of the subject and immediately after *internal* newlines.
1524 8. A call of pcre_fullinfo() from pcretest to get the option bits was passing
1525 a pointer to an int instead of a pointer to an unsigned long int. This
1526 caused problems on 64-bit systems.
1528 9. Applied a patch from the folks at Google to pcrecpp.cc, to fix "another
1529 instance of the 'standard' template library not being so standard".
1531 10. There was no check on the number of named subpatterns nor the maximum
1532 length of a subpattern name. The product of these values is used to compute
1533 the size of the memory block for a compiled pattern. By supplying a very
1534 long subpattern name and a large number of named subpatterns, the size
1535 computation could be caused to overflow. This is now prevented by limiting
1536 the length of names to 32 characters, and the number of named subpatterns
1537 to 10,000.
1539 11. Subpatterns that are repeated with specific counts have to be replicated in
1540 the compiled pattern. The size of memory for this was computed from the
1541 length of the subpattern and the repeat count. The latter is limited to
1542 65535, but there was no limit on the former, meaning that integer overflow
1543 could in principle occur. The compiled length of a repeated subpattern is
1544 now limited to 30,000 bytes in order to prevent this.
1546 12. Added the optional facility to have named substrings with the same name.
1548 13. Added the ability to use a named substring as a condition, using the
1549 Python syntax: (?(name)yes|no). This overloads (?(R)... and names that
1550 are numbers (not recommended). Forward references are permitted.
1552 14. Added forward references in named backreferences (if you see what I mean).
1554 15. In UTF-8 mode, with the PCRE_DOTALL option set, a quantified dot in the
1555 pattern could run off the end of the subject. For example, the pattern
1556 "(?s)(.{1,5})"8 did this with the subject "ab".
1558 16. If PCRE_DOTALL or PCRE_MULTILINE were set, pcre_dfa_exec() behaved as if
1559 PCRE_CASELESS was set when matching characters that were quantified with ?
1560 or *.
1562 17. A character class other than a single negated character that had a minimum
1563 but no maximum quantifier - for example [ab]{6,} - was not handled
1564 correctly by pce_dfa_exec(). It would match only one character.
1566 18. A valid (though odd) pattern that looked like a POSIX character
1567 class but used an invalid character after [ (for example [[,abc,]]) caused
1568 pcre_compile() to give the error "Failed: internal error: code overflow" or
1569 in some cases to crash with a glibc free() error. This could even happen if
1570 the pattern terminated after [[ but there just happened to be a sequence of
1571 letters, a binary zero, and a closing ] in the memory that followed.
1573 19. Perl's treatment of octal escapes in the range \400 to \777 has changed
1574 over the years. Originally (before any Unicode support), just the bottom 8
1575 bits were taken. Thus, for example, \500 really meant \100. Nowadays the
1576 output from "man perlunicode" includes this:
1578 The regular expression compiler produces polymorphic opcodes. That
1579 is, the pattern adapts to the data and automatically switches to
1580 the Unicode character scheme when presented with Unicode data--or
1581 instead uses a traditional byte scheme when presented with byte
1582 data.
1584 Sadly, a wide octal escape does not cause a switch, and in a string with
1585 no other multibyte characters, these octal escapes are treated as before.
1586 Thus, in Perl, the pattern /\500/ actually matches \100 but the pattern
1587 /\500|\x{1ff}/ matches \500 or \777 because the whole thing is treated as a
1588 Unicode string.
1590 I have not perpetrated such confusion in PCRE. Up till now, it took just
1591 the bottom 8 bits, as in old Perl. I have now made octal escapes with
1592 values greater than \377 illegal in non-UTF-8 mode. In UTF-8 mode they
1593 translate to the appropriate multibyte character.
1595 29. Applied some refactoring to reduce the number of warnings from Microsoft
1596 and Borland compilers. This has included removing the fudge introduced
1597 seven years ago for the OS/2 compiler (see 2.02/2 below) because it caused
1598 a warning about an unused variable.
1600 21. PCRE has not included VT (character 0x0b) in the set of whitespace
1601 characters since release 4.0, because Perl (from release 5.004) does not.
1602 [Or at least, is documented not to: some releases seem to be in conflict
1603 with the documentation.] However, when a pattern was studied with
1604 pcre_study() and all its branches started with \s, PCRE still included VT
1605 as a possible starting character. Of course, this did no harm; it just
1606 caused an unnecessary match attempt.
1608 22. Removed a now-redundant internal flag bit that recorded the fact that case
1609 dependency changed within the pattern. This was once needed for "required
1610 byte" processing, but is no longer used. This recovers a now-scarce options
1611 bit. Also moved the least significant internal flag bit to the most-
1612 significant bit of the word, which was not previously used (hangover from
1613 the days when it was an int rather than a uint) to free up another bit for
1614 the future.
1616 23. Added support for CRLF line endings as well as CR and LF. As well as the
1617 default being selectable at build time, it can now be changed at runtime
1618 via the PCRE_NEWLINE_xxx flags. There are now options for pcregrep to
1619 specify that it is scanning data with non-default line endings.
1621 24. Changed the definition of CXXLINK to make it agree with the definition of
1622 LINK in the Makefile, by replacing LDFLAGS to CXXFLAGS.
1624 25. Applied Ian Taylor's patches to avoid using another stack frame for tail
1625 recursions. This makes a big different to stack usage for some patterns.
1627 26. If a subpattern containing a named recursion or subroutine reference such
1628 as (?P>B) was quantified, for example (xxx(?P>B)){3}, the calculation of
1629 the space required for the compiled pattern went wrong and gave too small a
1630 value. Depending on the environment, this could lead to "Failed: internal
1631 error: code overflow at offset 49" or "glibc detected double free or
1632 corruption" errors.
1634 27. Applied patches from Google (a) to support the new newline modes and (b) to
1635 advance over multibyte UTF-8 characters in GlobalReplace.
1637 28. Change free() to pcre_free() in pcredemo.c. Apparently this makes a
1638 difference for some implementation of PCRE in some Windows version.
1640 29. Added some extra testing facilities to pcretest:
1642 \q<number> in a data line sets the "match limit" value
1643 \Q<number> in a data line sets the "match recursion limt" value
1644 -S <number> sets the stack size, where <number> is in megabytes
1646 The -S option isn't available for Windows.
1649 Version 6.6 06-Feb-06
1650 ---------------------
1652 1. Change 16(a) for 6.5 broke things, because PCRE_DATA_SCOPE was not defined
1653 in pcreposix.h. I have copied the definition from pcre.h.
1655 2. Change 25 for 6.5 broke compilation in a build directory out-of-tree
1656 because pcre.h is no longer a built file.
1658 3. Added Jeff Friedl's additional debugging patches to pcregrep. These are
1659 not normally included in the compiled code.
1662 Version 6.5 01-Feb-06
1663 ---------------------
1665 1. When using the partial match feature with pcre_dfa_exec(), it was not
1666 anchoring the second and subsequent partial matches at the new starting
1667 point. This could lead to incorrect results. For example, with the pattern
1668 /1234/, partially matching against "123" and then "a4" gave a match.
1670 2. Changes to pcregrep:
1672 (a) All non-match returns from pcre_exec() were being treated as failures
1673 to match the line. Now, unless the error is PCRE_ERROR_NOMATCH, an
1674 error message is output. Some extra information is given for the
1676 probably the only errors that are likely to be caused by users (by
1677 specifying a regex that has nested indefinite repeats, for instance).
1678 If there are more than 20 of these errors, pcregrep is abandoned.
1680 (b) A binary zero was treated as data while matching, but terminated the
1681 output line if it was written out. This has been fixed: binary zeroes
1682 are now no different to any other data bytes.
1684 (c) Whichever of the LC_ALL or LC_CTYPE environment variables is set is
1685 used to set a locale for matching. The --locale=xxxx long option has
1686 been added (no short equivalent) to specify a locale explicitly on the
1687 pcregrep command, overriding the environment variables.
1689 (d) When -B was used with -n, some line numbers in the output were one less
1690 than they should have been.
1692 (e) Added the -o (--only-matching) option.
1694 (f) If -A or -C was used with -c (count only), some lines of context were
1695 accidentally printed for the final match.
1697 (g) Added the -H (--with-filename) option.
1699 (h) The combination of options -rh failed to suppress file names for files
1700 that were found from directory arguments.
1702 (i) Added the -D (--devices) and -d (--directories) options.
1704 (j) Added the -F (--fixed-strings) option.
1706 (k) Allow "-" to be used as a file name for -f as well as for a data file.
1708 (l) Added the --colo(u)r option.
1710 (m) Added Jeffrey Friedl's -S testing option, but within #ifdefs so that it
1711 is not present by default.
1713 3. A nasty bug was discovered in the handling of recursive patterns, that is,
1714 items such as (?R) or (?1), when the recursion could match a number of
1715 alternatives. If it matched one of the alternatives, but subsequently,
1716 outside the recursion, there was a failure, the code tried to back up into
1717 the recursion. However, because of the way PCRE is implemented, this is not
1718 possible, and the result was an incorrect result from the match.
1720 In order to prevent this happening, the specification of recursion has
1721 been changed so that all such subpatterns are automatically treated as
1722 atomic groups. Thus, for example, (?R) is treated as if it were (?>(?R)).
1724 4. I had overlooked the fact that, in some locales, there are characters for
1725 which isalpha() is true but neither isupper() nor islower() are true. In
1726 the fr_FR locale, for instance, the \xAA and \xBA characters (ordmasculine
1727 and ordfeminine) are like this. This affected the treatment of \w and \W
1728 when they appeared in character classes, but not when they appeared outside
1729 a character class. The bit map for "word" characters is now created
1730 separately from the results of isalnum() instead of just taking it from the
1731 upper, lower, and digit maps. (Plus the underscore character, of course.)
1733 5. The above bug also affected the handling of POSIX character classes such as
1734 [[:alpha:]] and [[:alnum:]]. These do not have their own bit maps in PCRE's
1735 permanent tables. Instead, the bit maps for such a class were previously
1736 created as the appropriate unions of the upper, lower, and digit bitmaps.
1737 Now they are created by subtraction from the [[:word:]] class, which has
1738 its own bitmap.
1740 6. The [[:blank:]] character class matches horizontal, but not vertical space.
1741 It is created by subtracting the vertical space characters (\x09, \x0a,
1742 \x0b, \x0c) from the [[:space:]] bitmap. Previously, however, the
1743 subtraction was done in the overall bitmap for a character class, meaning
1744 that a class such as [\x0c[:blank:]] was incorrect because \x0c would not
1745 be recognized. This bug has been fixed.
1747 7. Patches from the folks at Google:
1749 (a) pcrecpp.cc: "to handle a corner case that may or may not happen in
1750 real life, but is still worth protecting against".
1752 (b) pcrecpp.cc: "corrects a bug when negative radixes are used with
1753 regular expressions".
1755 (c) pcre_scanner.cc: avoid use of std::count() because not all systems
1756 have it.
1758 (d) Split off pcrecpparg.h from pcrecpp.h and had the former built by
1759 "configure" and the latter not, in order to fix a problem somebody had
1760 with compiling the Arg class on HP-UX.
1762 (e) Improve the error-handling of the C++ wrapper a little bit.
1764 (f) New tests for checking recursion limiting.
1766 8. The pcre_memmove() function, which is used only if the environment does not
1767 have a standard memmove() function (and is therefore rarely compiled),
1768 contained two bugs: (a) use of int instead of size_t, and (b) it was not
1769 returning a result (though PCRE never actually uses the result).
1771 9. In the POSIX regexec() interface, if nmatch is specified as a ridiculously
1772 large number - greater than INT_MAX/(3*sizeof(int)) - REG_ESPACE is
1773 returned instead of calling malloc() with an overflowing number that would
1774 most likely cause subsequent chaos.
1776 10. The debugging option of pcretest was not showing the NO_AUTO_CAPTURE flag.
1778 11. The POSIX flag REG_NOSUB is now supported. When a pattern that was compiled
1779 with this option is matched, the nmatch and pmatch options of regexec() are
1780 ignored.
1782 12. Added REG_UTF8 to the POSIX interface. This is not defined by POSIX, but is
1783 provided in case anyone wants to the the POSIX interface with UTF-8
1784 strings.
1786 13. Added CXXLDFLAGS to the Makefile parameters to provide settings only on the
1787 C++ linking (needed for some HP-UX environments).
1789 14. Avoid compiler warnings in get_ucpname() when compiled without UCP support
1790 (unused parameter) and in the pcre_printint() function (omitted "default"
1791 switch label when the default is to do nothing).
1793 15. Added some code to make it possible, when PCRE is compiled as a C++
1794 library, to replace subject pointers for pcre_exec() with a smart pointer
1795 class, thus making it possible to process discontinuous strings.
1797 16. The two macros PCRE_EXPORT and PCRE_DATA_SCOPE are confusing, and perform
1798 much the same function. They were added by different people who were trying
1799 to make PCRE easy to compile on non-Unix systems. It has been suggested
1800 that PCRE_EXPORT be abolished now that there is more automatic apparatus
1801 for compiling on Windows systems. I have therefore replaced it with
1802 PCRE_DATA_SCOPE. This is set automatically for Windows; if not set it
1803 defaults to "extern" for C or "extern C" for C++, which works fine on
1804 Unix-like systems. It is now possible to override the value of PCRE_DATA_
1805 SCOPE with something explicit in config.h. In addition:
1807 (a) pcreposix.h still had just "extern" instead of either of these macros;
1808 I have replaced it with PCRE_DATA_SCOPE.
1810 (b) Functions such as _pcre_xclass(), which are internal to the library,
1811 but external in the C sense, all had PCRE_EXPORT in their definitions.
1812 This is apparently wrong for the Windows case, so I have removed it.
1813 (It makes no difference on Unix-like systems.)
1815 17. Added a new limit, MATCH_LIMIT_RECURSION, which limits the depth of nesting
1816 of recursive calls to match(). This is different to MATCH_LIMIT because
1817 that limits the total number of calls to match(), not all of which increase
1818 the depth of recursion. Limiting the recursion depth limits the amount of
1819 stack (or heap if NO_RECURSE is set) that is used. The default can be set
1820 when PCRE is compiled, and changed at run time. A patch from Google adds
1821 this functionality to the C++ interface.
1823 18. Changes to the handling of Unicode character properties:
1825 (a) Updated the table to Unicode 4.1.0.
1827 (b) Recognize characters that are not in the table as "Cn" (undefined).
1829 (c) I revised the way the table is implemented to a much improved format
1830 which includes recognition of ranges. It now supports the ranges that
1831 are defined in UnicodeData.txt, and it also amalgamates other
1832 characters into ranges. This has reduced the number of entries in the
1833 table from around 16,000 to around 3,000, thus reducing its size
1834 considerably. I realized I did not need to use a tree structure after
1835 all - a binary chop search is just as efficient. Having reduced the
1836 number of entries, I extended their size from 6 bytes to 8 bytes to
1837 allow for more data.
1839 (d) Added support for Unicode script names via properties such as \p{Han}.
1841 19. In UTF-8 mode, a backslash followed by a non-Ascii character was not
1842 matching that character.
1844 20. When matching a repeated Unicode property with a minimum greater than zero,
1845 (for example \pL{2,}), PCRE could look past the end of the subject if it
1846 reached it while seeking the minimum number of characters. This could
1847 happen only if some of the characters were more than one byte long, because
1848 there is a check for at least the minimum number of bytes.
1850 21. Refactored the implementation of \p and \P so as to be more general, to
1851 allow for more different types of property in future. This has changed the
1852 compiled form incompatibly. Anybody with saved compiled patterns that use
1853 \p or \P will have to recompile them.
1855 22. Added "Any" and "L&" to the supported property types.
1857 23. Recognize \x{...} as a code point specifier, even when not in UTF-8 mode,
1858 but give a compile time error if the value is greater than 0xff.
1860 24. The man pages for pcrepartial, pcreprecompile, and pcre_compile2 were
1861 accidentally not being installed or uninstalled.
1863 25. The pcre.h file was built from pcre.h.in, but the only changes that were
1864 made were to insert the current release number. This seemed silly, because
1865 it made things harder for people building PCRE on systems that don't run
1866 "configure". I have turned pcre.h into a distributed file, no longer built
1867 by "configure", with the version identification directly included. There is
1868 no longer a pcre.h.in file.
1870 However, this change necessitated a change to the pcre-config script as
1871 well. It is built from pcre-config.in, and one of the substitutions was the
1872 release number. I have updated configure.ac so that ./configure now finds
1873 the release number by grepping pcre.h.
1875 26. Added the ability to run the tests under valgrind.
1878 Version 6.4 05-Sep-05
1879 ---------------------
1881 1. Change 6.0/10/(l) to pcregrep introduced a bug that caused separator lines
1882 "--" to be printed when multiple files were scanned, even when none of the
1883 -A, -B, or -C options were used. This is not compatible with Gnu grep, so I
1884 consider it to be a bug, and have restored the previous behaviour.
1886 2. A couple of code tidies to get rid of compiler warnings.
1888 3. The pcretest program used to cheat by referring to symbols in the library
1889 whose names begin with _pcre_. These are internal symbols that are not
1890 really supposed to be visible externally, and in some environments it is
1891 possible to suppress them. The cheating is now confined to including
1892 certain files from the library's source, which is a bit cleaner.
1894 4. Renamed pcre.in as pcre.h.in to go with pcrecpp.h.in; it also makes the
1895 file's purpose clearer.
1897 5. Reorganized pcre_ucp_findchar().
1900 Version 6.3 15-Aug-05
1901 ---------------------
1903 1. The file libpcre.pc.in did not have general read permission in the tarball.
1905 2. There were some problems when building without C++ support:
1907 (a) If C++ support was not built, "make install" and "make test" still
1908 tried to test it.
1910 (b) There were problems when the value of CXX was explicitly set. Some
1911 changes have been made to try to fix these, and ...
1913 (c) --disable-cpp can now be used to explicitly disable C++ support.
1915 (d) The use of @CPP_OBJ@ directly caused a blank line preceded by a
1916 backslash in a target when C++ was disabled. This confuses some
1917 versions of "make", apparently. Using an intermediate variable solves
1918 this. (Same for CPP_LOBJ.)
1920 3. $(LINK_FOR_BUILD) now includes $(CFLAGS_FOR_BUILD) and $(LINK)
1921 (non-Windows) now includes $(CFLAGS) because these flags are sometimes
1922 necessary on certain architectures.
1924 4. Added a setting of -export-symbols-regex to the link command to remove
1925 those symbols that are exported in the C sense, but actually are local
1926 within the library, and not documented. Their names all begin with
1927 "_pcre_". This is not a perfect job, because (a) we have to except some
1928 symbols that pcretest ("illegally") uses, and (b) the facility isn't always
1929 available (and never for static libraries). I have made a note to try to
1930 find a way round (a) in the future.
1933 Version 6.2 01-Aug-05
1934 ---------------------
1936 1. There was no test for integer overflow of quantifier values. A construction
1937 such as {1111111111111111} would give undefined results. What is worse, if
1938 a minimum quantifier for a parenthesized subpattern overflowed and became
1939 negative, the calculation of the memory size went wrong. This could have
1940 led to memory overwriting.
1942 2. Building PCRE using VPATH was broken. Hopefully it is now fixed.
1944 3. Added "b" to the 2nd argument of fopen() in dftables.c, for non-Unix-like
1945 operating environments where this matters.
1947 4. Applied Giuseppe Maxia's patch to add additional features for controlling
1948 PCRE options from within the C++ wrapper.
1950 5. Named capturing subpatterns were not being correctly counted when a pattern
1951 was compiled. This caused two problems: (a) If there were more than 100
1952 such subpatterns, the calculation of the memory needed for the whole
1953 compiled pattern went wrong, leading to an overflow error. (b) Numerical
1954 back references of the form \12, where the number was greater than 9, were
1955 not recognized as back references, even though there were sufficient
1956 previous subpatterns.
1958 6. Two minor patches to pcrecpp.cc in order to allow it to compile on older
1959 versions of gcc, e.g. 2.95.4.
1962 Version 6.1 21-Jun-05
1963 ---------------------
1965 1. There was one reference to the variable "posix" in pcretest.c that was not
1966 surrounded by "#if !defined NOPOSIX".
1968 2. Make it possible to compile pcretest without DFA support, UTF8 support, or
1969 the cross-check on the old pcre_info() function, for the benefit of the
1970 cut-down version of PCRE that is currently imported into Exim.
1972 3. A (silly) pattern starting with (?i)(?-i) caused an internal space
1973 allocation error. I've done the easy fix, which wastes 2 bytes for sensible
1974 patterns that start (?i) but I don't think that matters. The use of (?i) is
1975 just an example; this all applies to the other options as well.
1977 4. Since libtool seems to echo the compile commands it is issuing, the output
1978 from "make" can be reduced a bit by putting "@" in front of each libtool
1979 compile command.
1981 5. Patch from the folks at Google for configure.in to be a bit more thorough
1982 in checking for a suitable C++ installation before trying to compile the
1983 C++ stuff. This should fix a reported problem when a compiler was present,
1984 but no suitable headers.
1986 6. The man pages all had just "PCRE" as their title. I have changed them to
1987 be the relevant file name. I have also arranged that these names are
1988 retained in the file doc/pcre.txt, which is a concatenation in text format
1989 of all the man pages except the little individual ones for each function.
1991 7. The NON-UNIX-USE file had not been updated for the different set of source
1992 files that come with release 6. I also added a few comments about the C++
1993 wrapper.
1996 Version 6.0 07-Jun-05
1997 ---------------------
1999 1. Some minor internal re-organization to help with my DFA experiments.
2001 2. Some missing #ifdef SUPPORT_UCP conditionals in pcretest and printint that
2002 didn't matter for the library itself when fully configured, but did matter
2003 when compiling without UCP support, or within Exim, where the ucp files are
2004 not imported.
2006 3. Refactoring of the library code to split up the various functions into
2007 different source modules. The addition of the new DFA matching code (see
2008 below) to a single monolithic source would have made it really too
2009 unwieldy, quite apart from causing all the code to be include in a
2010 statically linked application, when only some functions are used. This is
2011 relevant even without the DFA addition now that patterns can be compiled in
2012 one application and matched in another.
2014 The downside of splitting up is that there have to be some external
2015 functions and data tables that are used internally in different modules of
2016 the library but which are not part of the API. These have all had their
2017 names changed to start with "_pcre_" so that they are unlikely to clash
2018 with other external names.
2020 4. Added an alternate matching function, pcre_dfa_exec(), which matches using
2021 a different (DFA) algorithm. Although it is slower than the original
2022 function, it does have some advantages for certain types of matching
2023 problem.
2025 5. Upgrades to pcretest in order to test the features of pcre_dfa_exec(),
2026 including restarting after a partial match.
2028 6. A patch for pcregrep that defines INVALID_FILE_ATTRIBUTES if it is not
2029 defined when compiling for Windows was sent to me. I have put it into the
2030 code, though I have no means of testing or verifying it.
2032 7. Added the pcre_refcount() auxiliary function.
2034 8. Added the PCRE_FIRSTLINE option. This constrains an unanchored pattern to
2035 match before or at the first newline in the subject string. In pcretest,
2036 the /f option on a pattern can be used to set this.
2038 9. A repeated \w when used in UTF-8 mode with characters greater than 256
2039 would behave wrongly. This has been present in PCRE since release 4.0.
2041 10. A number of changes to the pcregrep command:
2043 (a) Refactored how -x works; insert ^(...)$ instead of setting
2044 PCRE_ANCHORED and checking the length, in preparation for adding
2045 something similar for -w.
2047 (b) Added the -w (match as a word) option.
2049 (c) Refactored the way lines are read and buffered so as to have more
2050 than one at a time available.
2052 (d) Implemented a pcregrep test script.
2054 (e) Added the -M (multiline match) option. This allows patterns to match
2055 over several lines of the subject. The buffering ensures that at least
2056 8K, or the rest of the document (whichever is the shorter) is available
2057 for matching (and similarly the previous 8K for lookbehind assertions).
2059 (f) Changed the --help output so that it now says
2061 -w, --word-regex(p)
2063 instead of two lines, one with "regex" and the other with "regexp"
2064 because that confused at least one person since the short forms are the
2065 same. (This required a bit of code, as the output is generated
2066 automatically from a table. It wasn't just a text change.)
2068 (g) -- can be used to terminate pcregrep options if the next thing isn't an
2069 option but starts with a hyphen. Could be a pattern or a path name
2070 starting with a hyphen, for instance.
2072 (h) "-" can be given as a file name to represent stdin.
2074 (i) When file names are being printed, "(standard input)" is used for
2075 the standard input, for compatibility with GNU grep. Previously
2076 "<stdin>" was used.
2078 (j) The option --label=xxx can be used to supply a name to be used for
2079 stdin when file names are being printed. There is no short form.
2081 (k) Re-factored the options decoding logic because we are going to add
2082 two more options that take data. Such options can now be given in four
2083 different ways, e.g. "-fname", "-f name", "--file=name", "--file name".
2085 (l) Added the -A, -B, and -C options for requesting that lines of context
2086 around matches be printed.
2088 (m) Added the -L option to print the names of files that do not contain
2089 any matching lines, that is, the complement of -l.
2091 (n) The return code is 2 if any file cannot be opened, but pcregrep does
2092 continue to scan other files.
2094 (o) The -s option was incorrectly implemented. For compatibility with other
2095 greps, it now suppresses the error message for a non-existent or non-
2096 accessible file (but not the return code). There is a new option called
2097 -q that suppresses the output of matching lines, which was what -s was
2098 previously doing.
2100 (p) Added --include and --exclude options to specify files for inclusion
2101 and exclusion when recursing.
2103 11. The Makefile was not using the Autoconf-supported LDFLAGS macro properly.
2104 Hopefully, it now does.
2106 12. Missing cast in pcre_study().
2108 13. Added an "uninstall" target to the makefile.
2110 14. Replaced "extern" in the function prototypes in Makefile.in with
2111 "PCRE_DATA_SCOPE", which defaults to 'extern' or 'extern "C"' in the Unix
2112 world, but is set differently for Windows.
2114 15. Added a second compiling function called pcre_compile2(). The only
2115 difference is that it has an extra argument, which is a pointer to an
2116 integer error code. When there is a compile-time failure, this is set
2117 non-zero, in addition to the error test pointer being set to point to an
2118 error message. The new argument may be NULL if no error number is required
2119 (but then you may as well call pcre_compile(), which is now just a
2120 wrapper). This facility is provided because some applications need a
2121 numeric error indication, but it has also enabled me to tidy up the way
2122 compile-time errors are handled in the POSIX wrapper.
2124 16. Added VPATH=.libs to the makefile; this should help when building with one
2125 prefix path and installing with another. (Or so I'm told by someone who
2126 knows more about this stuff than I do.)
2128 17. Added a new option, REG_DOTALL, to the POSIX function regcomp(). This
2129 passes PCRE_DOTALL to the pcre_compile() function, making the "." character
2130 match everything, including newlines. This is not POSIX-compatible, but
2131 somebody wanted the feature. From pcretest it can be activated by using
2132 both the P and the s flags.
2134 18. AC_PROG_LIBTOOL appeared twice in Makefile.in. Removed one.
2136 19. libpcre.pc was being incorrectly installed as executable.
2138 20. A couple of places in pcretest check for end-of-line by looking for '\n';
2139 it now also looks for '\r' so that it will work unmodified on Windows.
2141 21. Added Google's contributed C++ wrapper to the distribution.
2143 22. Added some untidy missing memory free() calls in pcretest, to keep
2144 Electric Fence happy when testing.
2148 Version 5.0 13-Sep-04
2149 ---------------------
2151 1. Internal change: literal characters are no longer packed up into items
2152 containing multiple characters in a single byte-string. Each character
2153 is now matched using a separate opcode. However, there may be more than one
2154 byte in the character in UTF-8 mode.
2156 2. The pcre_callout_block structure has two new fields: pattern_position and
2157 next_item_length. These contain the offset in the pattern to the next match
2158 item, and its length, respectively.
2160 3. The PCRE_AUTO_CALLOUT option for pcre_compile() requests the automatic
2161 insertion of callouts before each pattern item. Added the /C option to
2162 pcretest to make use of this.
2164 4. On the advice of a Windows user, the lines
2166 #if defined(_WIN32) || defined(WIN32)
2167 _setmode( _fileno( stdout ), 0x8000 );
2168 #endif /* defined(_WIN32) || defined(WIN32) */
2170 have been added to the source of pcretest. This apparently does useful
2171 magic in relation to line terminators.
2173 5. Changed "r" and "w" in the calls to fopen() in pcretest to "rb" and "wb"
2174 for the benefit of those environments where the "b" makes a difference.
2176 6. The icc compiler has the same options as gcc, but "configure" doesn't seem
2177 to know about it. I have put a hack into configure.in that adds in code
2178 to set GCC=yes if CC=icc. This seems to end up at a point in the
2179 generated configure script that is early enough to affect the setting of
2180 compiler options, which is what is needed, but I have no means of testing
2181 whether it really works. (The user who reported this had patched the
2182 generated configure script, which of course I cannot do.)
2184 LATER: After change 22 below (new libtool files), the configure script
2185 seems to know about icc (and also ecc). Therefore, I have commented out
2186 this hack in configure.in.
2188 7. Added support for pkg-config (2 patches were sent in).
2190 8. Negated POSIX character classes that used a combination of internal tables
2191 were completely broken. These were [[:^alpha:]], [[:^alnum:]], and
2192 [[:^ascii]]. Typically, they would match almost any characters. The other
2193 POSIX classes were not broken in this way.
2195 9. Matching the pattern "\b.*?" against "ab cd", starting at offset 1, failed
2196 to find the match, as PCRE was deluded into thinking that the match had to
2197 start at the start point or following a newline. The same bug applied to
2198 patterns with negative forward assertions or any backward assertions
2199 preceding ".*" at the start, unless the pattern required a fixed first
2200 character. This was a failing pattern: "(?!.bcd).*". The bug is now fixed.
2202 10. In UTF-8 mode, when moving forwards in the subject after a failed match
2203 starting at the last subject character, bytes beyond the end of the subject
2204 string were read.
2206 11. Renamed the variable "class" as "classbits" to make life easier for C++
2207 users. (Previously there was a macro definition, but it apparently wasn't
2208 enough.)
2210 12. Added the new field "tables" to the extra data so that tables can be passed
2211 in at exec time, or the internal tables can be re-selected. This allows
2212 a compiled regex to be saved and re-used at a later time by a different
2213 program that might have everything at different addresses.
2215 13. Modified the pcre-config script so that, when run on Solaris, it shows a
2216 -R library as well as a -L library.
2218 14. The debugging options of pcretest (-d on the command line or D on a
2219 pattern) showed incorrect output for anything following an extended class
2220 that contained multibyte characters and which was followed by a quantifier.
2222 15. Added optional support for general category Unicode character properties
2223 via the \p, \P, and \X escapes. Unicode property support implies UTF-8
2224 support. It adds about 90K to the size of the library. The meanings of the
2225 inbuilt class escapes such as \d and \s have NOT been changed.
2227 16. Updated pcredemo.c to include calls to free() to release the memory for the
2228 compiled pattern.
2230 17. The generated file chartables.c was being created in the source directory
2231 instead of in the building directory. This caused the build to fail if the
2232 source directory was different from the building directory, and was
2233 read-only.
2235 18. Added some sample Win commands from Mark Tetrode into the NON-UNIX-USE
2236 file. No doubt somebody will tell me if they don't make sense... Also added
2237 Dan Mooney's comments about building on OpenVMS.
2239 19. Added support for partial matching via the PCRE_PARTIAL option for
2240 pcre_exec() and the \P data escape in pcretest.
2242 20. Extended pcretest with 3 new pattern features:
2244 (i) A pattern option of the form ">rest-of-line" causes pcretest to
2245 write the compiled pattern to the file whose name is "rest-of-line".
2246 This is a straight binary dump of the data, with the saved pointer to
2247 the character tables forced to be NULL. The study data, if any, is
2248 written too. After writing, pcretest reads a new pattern.
2250 (ii) If, instead of a pattern, "<rest-of-line" is given, pcretest reads a
2251 compiled pattern from the given file. There must not be any
2252 occurrences of "<" in the file name (pretty unlikely); if there are,
2253 pcretest will instead treat the initial "<" as a pattern delimiter.
2254 After reading in the pattern, pcretest goes on to read data lines as
2255 usual.
2257 (iii) The F pattern option causes pcretest to flip the bytes in the 32-bit
2258 and 16-bit fields in a compiled pattern, to simulate a pattern that
2259 was compiled on a host of opposite endianness.
2261 21. The pcre-exec() function can now cope with patterns that were compiled on
2262 hosts of opposite endianness, with this restriction:
2264 As for any compiled expression that is saved and used later, the tables
2265 pointer field cannot be preserved; the extra_data field in the arguments
2266 to pcre_exec() should be used to pass in a tables address if a value
2267 other than the default internal tables were used at compile time.
2269 22. Calling pcre_exec() with a negative value of the "ovecsize" parameter is
2270 now diagnosed as an error. Previously, most of the time, a negative number
2271 would have been treated as zero, but if in addition "ovector" was passed as
2272 NULL, a crash could occur.
2274 23. Updated the files ltmain.sh, config.sub, config.guess, and aclocal.m4 with
2275 new versions from the libtool 1.5 distribution (the last one is a copy of
2276 a file called libtool.m4). This seems to have fixed the need to patch
2277 "configure" to support Darwin 1.3 (which I used to do). However, I still
2278 had to patch ltmain.sh to ensure that ${SED} is set (it isn't on my
2279 workstation).
2281 24. Changed the PCRE licence to be the more standard "BSD" licence.
2284 Version 4.5 01-Dec-03
2285 ---------------------
2287 1. There has been some re-arrangement of the code for the match() function so
2288 that it can be compiled in a version that does not call itself recursively.
2289 Instead, it keeps those local variables that need separate instances for
2290 each "recursion" in a frame on the heap, and gets/frees frames whenever it
2291 needs to "recurse". Keeping track of where control must go is done by means
2292 of setjmp/longjmp. The whole thing is implemented by a set of macros that
2293 hide most of the details from the main code, and operates only if
2294 NO_RECURSE is defined while compiling pcre.c. If PCRE is built using the
2295 "configure" mechanism, "--disable-stack-for-recursion" turns on this way of
2296 operating.
2298 To make it easier for callers to provide specially tailored get/free
2299 functions for this usage, two new functions, pcre_stack_malloc, and
2300 pcre_stack_free, are used. They are always called in strict stacking order,
2301 and the size of block requested is always the same.
2303 The PCRE_CONFIG_STACKRECURSE info parameter can be used to find out whether
2304 PCRE has been compiled to use the stack or the heap for recursion. The
2305 -C option of pcretest uses this to show which version is compiled.
2307 A new data escape \S, is added to pcretest; it causes the amounts of store
2308 obtained and freed by both kinds of malloc/free at match time to be added
2309 to the output.
2311 2. Changed the locale test to use "fr_FR" instead of "fr" because that's
2312 what's available on my current Linux desktop machine.
2314 3. When matching a UTF-8 string, the test for a valid string at the start has
2315 been extended. If start_offset is not zero, PCRE now checks that it points
2316 to a byte that is the start of a UTF-8 character. If not, it returns
2317 PCRE_ERROR_BADUTF8_OFFSET (-11). Note: the whole string is still checked;
2318 this is necessary because there may be backward assertions in the pattern.
2319 When matching the same subject several times, it may save resources to use
2320 PCRE_NO_UTF8_CHECK on all but the first call if the string is long.
2322 4. The code for checking the validity of UTF-8 strings has been tightened so
2323 that it rejects (a) strings containing 0xfe or 0xff bytes and (b) strings
2324 containing "overlong sequences".
2326 5. Fixed a bug (appearing twice) that I could not find any way of exploiting!
2327 I had written "if ((digitab[*p++] && chtab_digit) == 0)" where the "&&"
2328 should have been "&", but it just so happened that all the cases this let
2329 through by mistake were picked up later in the function.
2331 6. I had used a variable called "isblank" - this is a C99 function, causing
2332 some compilers to warn. To avoid this, I renamed it (as "blankclass").
2334 7. Cosmetic: (a) only output another newline at the end of pcretest if it is
2335 prompting; (b) run "./pcretest /dev/null" at the start of the test script
2336 so the version is shown; (c) stop "make test" echoing "./RunTest".
2338 8. Added patches from David Burgess to enable PCRE to run on EBCDIC systems.
2340 9. The prototype for memmove() for systems that don't have it was using
2341 size_t, but the inclusion of the header that defines size_t was later. I've
2342 moved the #includes for the C headers earlier to avoid this.
2344 10. Added some adjustments to the code to make it easier to compiler on certain
2345 special systems:
2347 (a) Some "const" qualifiers were missing.
2348 (b) Added the macro EXPORT before all exported functions; by default this
2349 is defined to be empty.
2350 (c) Changed the dftables auxiliary program (that builds chartables.c) so
2351 that it reads its output file name as an argument instead of writing
2352 to the standard output and assuming this can be redirected.
2354 11. In UTF-8 mode, if a recursive reference (e.g. (?1)) followed a character
2355 class containing characters with values greater than 255, PCRE compilation
2356 went into a loop.
2358 12. A recursive reference to a subpattern that was within another subpattern
2359 that had a minimum quantifier of zero caused PCRE to crash. For example,
2360 (x(y(?2))z)? provoked this bug with a subject that got as far as the
2361 recursion. If the recursively-called subpattern itself had a zero repeat,
2362 that was OK.
2364 13. In pcretest, the buffer for reading a data line was set at 30K, but the
2365 buffer into which it was copied (for escape processing) was still set at
2366 1024, so long lines caused crashes.
2368 14. A pattern such as /[ab]{1,3}+/ failed to compile, giving the error
2369 "internal error: code overflow...". This applied to any character class
2370 that was followed by a possessive quantifier.
2372 15. Modified the Makefile to add libpcre.la as a prerequisite for
2373 libpcreposix.la because I was told this is needed for a parallel build to
2374 work.
2376 16. If a pattern that contained .* following optional items at the start was
2377 studied, the wrong optimizing data was generated, leading to matching
2378 errors. For example, studying /[ab]*.*c/ concluded, erroneously, that any
2379 matching string must start with a or b or c. The correct conclusion for
2380 this pattern is that a match can start with any character.
2383 Version 4.4 13-Aug-03
2384 ---------------------
2386 1. In UTF-8 mode, a character class containing characters with values between
2387 127 and 255 was not handled correctly if the compiled pattern was studied.
2388 In fixing this, I have also improved the studying algorithm for such
2389 classes (slightly).
2391 2. Three internal functions had redundant arguments passed to them. Removal
2392 might give a very teeny performance improvement.
2394 3. Documentation bug: the value of the capture_top field in a callout is *one
2395 more than* the number of the hightest numbered captured substring.
2397 4. The Makefile linked pcretest and pcregrep with -lpcre, which could result
2398 in incorrectly linking with a previously installed version. They now link
2399 explicitly with libpcre.la.
2401 5. configure.in no longer needs to recognize Cygwin specially.
2403 6. A problem in pcre.in for Windows platforms is fixed.
2405 7. If a pattern was successfully studied, and the -d (or /D) flag was given to
2406 pcretest, it used to include the size of the study block as part of its
2407 output. Unfortunately, the structure contains a field that has a different
2408 size on different hardware architectures. This meant that the tests that
2409 showed this size failed. As the block is currently always of a fixed size,
2410 this information isn't actually particularly useful in pcretest output, so
2411 I have just removed it.
2413 8. Three pre-processor statements accidentally did not start in column 1.
2414 Sadly, there are *still* compilers around that complain, even though
2415 standard C has not required this for well over a decade. Sigh.
2417 9. In pcretest, the code for checking callouts passed small integers in the
2418 callout_data field, which is a void * field. However, some picky compilers
2419 complained about the casts involved for this on 64-bit systems. Now
2420 pcretest passes the address of the small integer instead, which should get
2421 rid of the warnings.
2423 10. By default, when in UTF-8 mode, PCRE now checks for valid UTF-8 strings at
2424 both compile and run time, and gives an error if an invalid UTF-8 sequence
2425 is found. There is a option for disabling this check in cases where the
2426 string is known to be correct and/or the maximum performance is wanted.
2428 11. In response to a bug report, I changed one line in Makefile.in from
2430 -Wl,--out-implib,.libs/lib@WIN_PREFIX@pcreposix.dll.a \
2431 to
2432 -Wl,--out-implib,.libs/@WIN_PREFIX@libpcreposix.dll.a \
2434 to look similar to other lines, but I have no way of telling whether this
2435 is the right thing to do, as I do not use Windows. No doubt I'll get told
2436 if it's wrong...
2439 Version 4.3 21-May-03
2440 ---------------------
2442 1. Two instances of @WIN_PREFIX@ omitted from the Windows targets in the
2443 Makefile.
2445 2. Some refactoring to improve the quality of the code:
2447 (i) The utf8_table... variables are now declared "const".
2449 (ii) The code for \cx, which used the "case flipping" table to upper case
2450 lower case letters, now just substracts 32. This is ASCII-specific,
2451 but the whole concept of \cx is ASCII-specific, so it seems
2452 reasonable.
2454 (iii) PCRE was using its character types table to recognize decimal and
2455 hexadecimal digits in the pattern. This is silly, because it handles
2456 only 0-9, a-f, and A-F, but the character types table is locale-
2457 specific, which means strange things might happen. A private
2458 table is now used for this - though it costs 256 bytes, a table is
2459 much faster than multiple explicit tests. Of course, the standard
2460 character types table is still used for matching digits in subject
2461 strings against \d.
2463 (iv) Strictly, the identifier ESC_t is reserved by POSIX (all identifiers
2464 ending in _t are). So I've renamed it as ESC_tee.
2466 3. The first argument for regexec() in the POSIX wrapper should have been
2467 defined as "const".
2469 4. Changed pcretest to use malloc() for its buffers so that they can be
2470 Electric Fenced for debugging.
2472 5. There were several places in the code where, in UTF-8 mode, PCRE would try
2473 to read one or more bytes before the start of the subject string. Often this
2474 had no effect on PCRE's behaviour, but in some circumstances it could
2475 provoke a segmentation fault.
2477 6. A lookbehind at the start of a pattern in UTF-8 mode could also cause PCRE
2478 to try to read one or more bytes before the start of the subject string.
2480 7. A lookbehind in a pattern matched in non-UTF-8 mode on a PCRE compiled with
2481 UTF-8 support could misbehave in various ways if the subject string
2482 contained bytes with the 0x80 bit set and the 0x40 bit unset in a lookbehind
2483 area. (PCRE was not checking for the UTF-8 mode flag, and trying to move
2484 back over UTF-8 characters.)
2487 Version 4.2 14-Apr-03
2488 ---------------------
2490 1. Typo "#if SUPPORT_UTF8" instead of "#ifdef SUPPORT_UTF8" fixed.
2492 2. Changes to the building process, supplied by Ronald Landheer-Cieslak
2493 [ON_WINDOWS]: new variable, "#" on non-Windows platforms
2494 [NOT_ON_WINDOWS]: new variable, "#" on Windows platforms
2495 [WIN_PREFIX]: new variable, "cyg" for Cygwin
2496 * Makefile.in: use autoconf substitution for OBJEXT, EXEEXT, BUILD_OBJEXT
2498 Note: automatic setting of the BUILD variables is not yet working
2499 set CPPFLAGS and BUILD_CPPFLAGS (but don't use yet) - should be used at
2500 compile-time but not at link-time
2501 [LINK]: use for linking executables only
2502 make different versions for Windows and non-Windows
2503 [LINKLIB]: new variable, copy of UNIX-style LINK, used for linking
2504 libraries
2505 [LINK_FOR_BUILD]: new variable
2506 [OBJEXT]: use throughout
2507 [EXEEXT]: use throughout
2508 <winshared>: new target
2509 <wininstall>: new target
2510 <dftables.o>: use native compiler
2511 <dftables>: use native linker
2512 <install>: handle Windows platform correctly
2513 <clean>: ditto
2514 <check>: ditto
2515 copy DLL to top builddir before testing
2517 As part of these changes, -no-undefined was removed again. This was reported
2518 to give trouble on HP-UX 11.0, so getting rid of it seems like a good idea
2519 in any case.
2521 3. Some tidies to get rid of compiler warnings:
2523 . In the match_data structure, match_limit was an unsigned long int, whereas
2524 match_call_count was an int. I've made them both unsigned long ints.
2526 . In pcretest the fact that a const uschar * doesn't automatically cast to
2527 a void * provoked a warning.
2529 . Turning on some more compiler warnings threw up some "shadow" variables
2530 and a few more missing casts.
2532 4. If PCRE was complied with UTF-8 support, but called without the PCRE_UTF8
2533 option, a class that contained a single character with a value between 128
2534 and 255 (e.g. /[\xFF]/) caused PCRE to crash.
2536 5. If PCRE was compiled with UTF-8 support, but called without the PCRE_UTF8
2537 option, a class that contained several characters, but with at least one
2538 whose value was between 128 and 255 caused PCRE to crash.
2541 Version 4.1 12-Mar-03
2542 ---------------------
2544 1. Compiling with gcc -pedantic found a couple of places where casts were
2545 needed, and a string in dftables.c that was longer than standard compilers are
2546 required to support.
2548 2. Compiling with Sun's compiler found a few more places where the code could
2549 be tidied up in order to avoid warnings.
2551 3. The variables for cross-compiling were called HOST_CC and HOST_CFLAGS; the
2552 first of these names is deprecated in the latest Autoconf in favour of the name
2553 CC_FOR_BUILD, because "host" is typically used to mean the system on which the
2554 compiled code will be run. I can't find a reference for HOST_CFLAGS, but by
2555 analogy I have changed it to CFLAGS_FOR_BUILD.
2557 4. Added -no-undefined to the linking command in the Makefile, because this is
2558 apparently helpful for Windows. To make it work, also added "-L. -lpcre" to the
2559 linking step for the pcreposix library.
2561 5. PCRE was failing to diagnose the case of two named groups with the same
2562 name.
2564 6. A problem with one of PCRE's optimizations was discovered. PCRE remembers a
2565 literal character that is needed in the subject for a match, and scans along to
2566 ensure that it is present before embarking on the full matching process. This
2567 saves time in cases of nested unlimited repeats that are never going to match.
2568 Problem: the scan can take a lot of time if the subject is very long (e.g.
2569 megabytes), thus penalizing straightforward matches. It is now done only if the
2570 amount of subject to be scanned is less than 1000 bytes.
2572 7. A lesser problem with the same optimization is that it was recording the
2573 first character of an anchored pattern as "needed", thus provoking a search
2574 right along the subject, even when the first match of the pattern was going to
2575 fail. The "needed" character is now not set for anchored patterns, unless it
2576 follows something in the pattern that is of non-fixed length. Thus, it still
2577 fulfils its original purpose of finding quick non-matches in cases of nested
2578 unlimited repeats, but isn't used for simple anchored patterns such as /^abc/.
2581 Version 4.0 17-Feb-03
2582 ---------------------
2584 1. If a comment in an extended regex that started immediately after a meta-item
2585 extended to the end of string, PCRE compiled incorrect data. This could lead to
2586 all kinds of weird effects. Example: /#/ was bad; /()#/ was bad; /a#/ was not.
2588 2. Moved to autoconf 2.53 and libtool 1.4.2.
2590 3. Perl 5.8 no longer needs "use utf8" for doing UTF-8 things. Consequently,
2591 the special perltest8 script is no longer needed - all the tests can be run
2592 from a single perltest script.
2594 4. From 5.004, Perl has not included the VT character (0x0b) in the set defined
2595 by \s. It has now been removed in PCRE. This means it isn't recognized as
2596 whitespace in /x regexes too, which is the same as Perl. Note that the POSIX
2597 class [:space:] *does* include VT, thereby creating a mess.
2599 5. Added the class [:blank:] (a GNU extension from Perl 5.8) to match only
2600 space and tab.
2602 6. Perl 5.005 was a long time ago. It's time to amalgamate the tests that use
2603 its new features into the main test script, reducing the number of scripts.
2605 7. Perl 5.8 has changed the meaning of patterns like /a(?i)b/. Earlier versions
2606 were backward compatible, and made the (?i) apply to the whole pattern, as if
2607 /i were given. Now it behaves more logically, and applies the option setting
2608 only to what follows. PCRE has been changed to follow suit. However, if it
2609 finds options settings right at the start of the pattern, it extracts them into
2610 the global options, as before. Thus, they show up in the info data.
2612 8. Added support for the \Q...\E escape sequence. Characters in between are
2613 treated as literals. This is slightly different from Perl in that $ and @ are
2614 also handled as literals inside the quotes. In Perl, they will cause variable
2615 interpolation. Note the following examples:
2617 Pattern PCRE matches Perl matches
2619 \Qabc$xyz\E abc$xyz abc followed by the contents of $xyz
2620 \Qabc\$xyz\E abc\$xyz abc\$xyz
2621 \Qabc\E\$\Qxyz\E abc$xyz abc$xyz
2623 For compatibility with Perl, \Q...\E sequences are recognized inside character
2624 classes as well as outside them.
2626 9. Re-organized 3 code statements in pcretest to avoid "overflow in
2627 floating-point constant arithmetic" warnings from a Microsoft compiler. Added a
2628 (size_t) cast to one statement in pcretest and one in pcreposix to avoid
2629 signed/unsigned warnings.
2631 10. SunOS4 doesn't have strtoul(). This was used only for unpicking the -o
2632 option for pcretest, so I've replaced it by a simple function that does just
2633 that job.
2635 11. pcregrep was ending with code 0 instead of 2 for the commands "pcregrep" or
2636 "pcregrep -".
2638 12. Added "possessive quantifiers" ?+, *+, ++, and {,}+ which come from Sun's
2639 Java package. This provides some syntactic sugar for simple cases of what my
2640 documentation calls "once-only subpatterns". A pattern such as x*+ is the same
2641 as (?>x*). In other words, if what is inside (?>...) is just a single repeated
2642 item, you can use this simplified notation. Note that only makes sense with
2643 greedy quantifiers. Consequently, the use of the possessive quantifier forces
2644 greediness, whatever the setting of the PCRE_UNGREEDY option.
2646 13. A change of greediness default within a pattern was not taking effect at
2647 the current level for patterns like /(b+(?U)a+)/. It did apply to parenthesized
2648 subpatterns that followed. Patterns like /b+(?U)a+/ worked because the option
2649 was abstracted outside.
2651 14. PCRE now supports the \G assertion. It is true when the current matching
2652 position is at the start point of the match. This differs from \A when the
2653 starting offset is non-zero. Used with the /g option of pcretest (or similar
2654 code), it works in the same way as it does for Perl's /g option. If all
2655 alternatives of a regex begin with \G, the expression is anchored to the start
2656 match position, and the "anchored" flag is set in the compiled expression.
2658 15. Some bugs concerning the handling of certain option changes within patterns
2659 have been fixed. These applied to options other than (?ims). For example,
2660 "a(?x: b c )d" did not match "XabcdY" but did match "Xa b c dY". It should have
2661 been the other way round. Some of this was related to change 7 above.
2663 16. PCRE now gives errors for /[.x.]/ and /[=x=]/ as unsupported POSIX
2664 features, as Perl does. Previously, PCRE gave the warnings only for /[[.x.]]/
2665 and /[[=x=]]/. PCRE now also gives an error for /[:name:]/ because it supports
2666 POSIX classes only within a class (e.g. /[[:alpha:]]/).
2668 17. Added support for Perl's \C escape. This matches one byte, even in UTF8
2669 mode. Unlike ".", it always matches newline, whatever the setting of
2670 PCRE_DOTALL. However, PCRE does not permit \C to appear in lookbehind
2671 assertions. Perl allows it, but it doesn't (in general) work because it can't
2672 calculate the length of the lookbehind. At least, that's the case for Perl
2673 5.8.0 - I've been told they are going to document that it doesn't work in
2674 future.
2676 18. Added an error diagnosis for escapes that PCRE does not support: these are
2677 \L, \l, \N, \P, \p, \U, \u, and \X.
2679 19. Although correctly diagnosing a missing ']' in a character class, PCRE was
2680 reading past the end of the pattern in cases such as /[abcd/.
2682 20. PCRE was getting more memory than necessary for patterns with classes that
2683 contained both POSIX named classes and other characters, e.g. /[[:space:]abc/.
2685 21. Added some code, conditional on #ifdef VPCOMPAT, to make life easier for
2686 compiling PCRE for use with Virtual Pascal.
2688 22. Small fix to the Makefile to make it work properly if the build is done
2689 outside the source tree.
2691 23. Added a new extension: a condition to go with recursion. If a conditional
2692 subpattern starts with (?(R) the "true" branch is used if recursion has
2693 happened, whereas the "false" branch is used only at the top level.
2695 24. When there was a very long string of literal characters (over 255 bytes
2696 without UTF support, over 250 bytes with UTF support), the computation of how
2697 much memory was required could be incorrect, leading to segfaults or other
2698 strange effects.
2700 25. PCRE was incorrectly assuming anchoring (either to start of subject or to
2701 start of line for a non-DOTALL pattern) when a pattern started with (.*) and
2702 there was a subsequent back reference to those brackets. This meant that, for
2703 example, /(.*)\d+\1/ failed to match "abc123bc". Unfortunately, it isn't
2704 possible to check for precisely this case. All we can do is abandon the
2705 optimization if .* occurs inside capturing brackets when there are any back
2706 references whatsoever. (See below for a better fix that came later.)
2708 26. The handling of the optimization for finding the first character of a
2709 non-anchored pattern, and for finding a character that is required later in the
2710 match were failing in some cases. This didn't break the matching; it just
2711 failed to optimize when it could. The way this is done has been re-implemented.
2713 27. Fixed typo in error message for invalid (?R item (it said "(?p").
2715 28. Added a new feature that provides some of the functionality that Perl
2716 provides with (?{...}). The facility is termed a "callout". The way it is done
2717 in PCRE is for the caller to provide an optional function, by setting
2718 pcre_callout to its entry point. Like pcre_malloc and pcre_free, this is a
2719 global variable. By default it is unset, which disables all calling out. To get
2720 the function called, the regex must include (?C) at appropriate points. This
2721 is, in fact, equivalent to (?C0), and any number <= 255 may be given with (?C).
2722 This provides a means of identifying different callout points. When PCRE
2723 reaches such a point in the regex, if pcre_callout has been set, the external
2724 function is called. It is provided with data in a structure called
2725 pcre_callout_block, which is defined in pcre.h. If the function returns 0,
2726 matching continues; if it returns a non-zero value, the match at the current
2727 point fails. However, backtracking will occur if possible. [This was changed
2728 later and other features added - see item 49 below.]
2730 29. pcretest is upgraded to test the callout functionality. It provides a
2731 callout function that displays information. By default, it shows the start of
2732 the match and the current position in the text. There are some new data escapes
2733 to vary what happens:
2735 \C+ in addition, show current contents of captured substrings
2736 \C- do not supply a callout function
2737 \C!n return 1 when callout number n is reached
2738 \C!n!m return 1 when callout number n is reached for the mth time
2740 30. If pcregrep was called with the -l option and just a single file name, it
2741 output "<stdin>" if a match was found, instead of the file name.
2743 31. Improve the efficiency of the POSIX API to PCRE. If the number of capturing
2744 slots is less than POSIX_MALLOC_THRESHOLD, use a block on the stack to pass to
2745 pcre_exec(). This saves a malloc/free per call. The default value of
2746 POSIX_MALLOC_THRESHOLD is 10; it can be changed by --with-posix-malloc-threshold
2747 when configuring.
2749 32. The default maximum size of a compiled pattern is 64K. There have been a
2750 few cases of people hitting this limit. The code now uses macros to handle the
2751 storing of links as offsets within the compiled pattern. It defaults to 2-byte
2752 links, but this can be changed to 3 or 4 bytes by --with-link-size when
2753 configuring. Tests 2 and 5 work only with 2-byte links because they output
2754 debugging information about compiled patterns.
2756 33. Internal code re-arrangements:
2758 (a) Moved the debugging function for printing out a compiled regex into
2759 its own source file (printint.c) and used #include to pull it into
2760 pcretest.c and, when DEBUG is defined, into pcre.c, instead of having two
2761 separate copies.
2763 (b) Defined the list of op-code names for debugging as a macro in
2764 internal.h so that it is next to the definition of the opcodes.
2766 (c) Defined a table of op-code lengths for simpler skipping along compiled
2767 code. This is again a macro in internal.h so that it is next to the
2768 definition of the opcodes.
2770 34. Added support for recursive calls to individual subpatterns, along the
2771 lines of Robin Houston's patch (but implemented somewhat differently).
2773 35. Further mods to the Makefile to help Win32. Also, added code to pcregrep to
2774 allow it to read and process whole directories in Win32. This code was
2775 contributed by Lionel Fourquaux; it has not been tested by me.
2777 36. Added support for named subpatterns. The Python syntax (?P<name>...) is
2778 used to name a group. Names consist of alphanumerics and underscores, and must
2779 be unique. Back references use the syntax (?P=name) and recursive calls use
2780 (?P>name) which is a PCRE extension to the Python extension. Groups still have
2781 numbers. The function pcre_fullinfo() can be used after compilation to extract
2782 a name/number map. There are three relevant calls:
2784 PCRE_INFO_NAMEENTRYSIZE yields the size of each entry in the map
2785 PCRE_INFO_NAMECOUNT yields the number of entries
2786 PCRE_INFO_NAMETABLE yields a pointer to the map.
2788 The map is a vector of fixed-size entries. The size of each entry depends on
2789 the length of the longest name used. The first two bytes of each entry are the
2790 group number, most significant byte first. There follows the corresponding
2791 name, zero terminated. The names are in alphabetical order.
2793 37. Make the maximum literal string in the compiled code 250 for the non-UTF-8
2794 case instead of 255. Making it the same both with and without UTF-8 support
2795 means that the same test output works with both.
2797 38. There was a case of malloc(0) in the POSIX testing code in pcretest. Avoid
2798 calling malloc() with a zero argument.
2800 39. Change 25 above had to resort to a heavy-handed test for the .* anchoring
2801 optimization. I've improved things by keeping a bitmap of backreferences with
2802 numbers 1-31 so that if .* occurs inside capturing brackets that are not in
2803 fact referenced, the optimization can be applied. It is unlikely that a
2804 relevant occurrence of .* (i.e. one which might indicate anchoring or forcing
2805 the match to follow \n) will appear inside brackets with a number greater than
2806 31, but if it does, any back reference > 31 suppresses the optimization.
2808 40. Added a new compile-time option PCRE_NO_AUTO_CAPTURE. This has the effect
2809 of disabling numbered capturing parentheses. Any opening parenthesis that is
2810 not followed by ? behaves as if it were followed by ?: but named parentheses
2811 can still be used for capturing (and they will acquire numbers in the usual
2812 way).
2814 41. Redesigned the return codes from the match() function into yes/no/error so
2815 that errors can be passed back from deep inside the nested calls. A malloc
2816 failure while inside a recursive subpattern call now causes the
2817 PCRE_ERROR_NOMEMORY return instead of quietly going wrong.
2819 42. It is now possible to set a limit on the number of times the match()
2820 function is called in a call to pcre_exec(). This facility makes it possible to
2821 limit the amount of recursion and backtracking, though not in a directly
2822 obvious way, because the match() function is used in a number of different
2823 circumstances. The count starts from zero for each position in the subject
2824 string (for non-anchored patterns). The default limit is, for compatibility, a
2825 large number, namely 10 000 000. You can change this in two ways:
2827 (a) When configuring PCRE before making, you can use --with-match-limit=n
2828 to set a default value for the compiled library.
2830 (b) For each call to pcre_exec(), you can pass a pcre_extra block in which
2831 a different value is set. See 45 below.
2833 If the limit is exceeded, pcre_exec() returns PCRE_ERROR_MATCHLIMIT.
2835 43. Added a new function pcre_config(int, void *) to enable run-time extraction
2836 of things that can be changed at compile time. The first argument specifies
2837 what is wanted and the second points to where the information is to be placed.
2838 The current list of available information is:
2842 The output is an integer that is set to one if UTF-8 support is available;
2843 otherwise it is set to zero.
2847 The output is an integer that it set to the value of the code that is used for
2848 newline. It is either LF (10) or CR (13).
2852 The output is an integer that contains the number of bytes used for internal
2853 linkage in compiled expressions. The value is 2, 3, or 4. See item 32 above.
2857 The output is an integer that contains the threshold above which the POSIX
2858 interface uses malloc() for output vectors. See item 31 above.
2862 The output is an unsigned integer that contains the default limit of the number
2863 of match() calls in a pcre_exec() execution. See 42 above.
2865 44. pcretest has been upgraded by the addition of the -C option. This causes it
2866 to extract all the available output from the new pcre_config() function, and to
2867 output it. The program then exits immediately.
2869 45. A need has arisen to pass over additional data with calls to pcre_exec() in
2870 order to support additional features. One way would have been to define
2871 pcre_exec2() (for example) with extra arguments, but this would not have been
2872 extensible, and would also have required all calls to the original function to
2873 be mapped to the new one. Instead, I have chosen to extend the mechanism that
2874 is used for passing in "extra" data from pcre_study().
2876 The pcre_extra structure is now exposed and defined in pcre.h. It currently
2877 contains the following fields:
2879 flags a bitmap indicating which of the following fields are set
2880 study_data opaque data from pcre_study()
2881 match_limit a way of specifying a limit on match() calls for a specific
2882 call to pcre_exec()
2883 callout_data data for callouts (see 49 below)
2885 The flag bits are also defined in pcre.h, and are
2891 The pcre_study() function now returns one of these new pcre_extra blocks, with
2892 the actual study data pointed to by the study_data field, and the
2893 PCRE_EXTRA_STUDY_DATA flag set. This can be passed directly to pcre_exec() as
2894 before. That is, this change is entirely upwards-compatible and requires no
2895 change to existing code.
2897 If you want to pass in additional data to pcre_exec(), you can either place it
2898 in a pcre_extra block provided by pcre_study(), or create your own pcre_extra
2899 block.
2901 46. pcretest has been extended to test the PCRE_EXTRA_MATCH_LIMIT feature. If a
2902 data string contains the escape sequence \M, pcretest calls pcre_exec() several
2903 times with different match limits, until it finds the minimum value needed for
2904 pcre_exec() to complete. The value is then output. This can be instructive; for
2905 most simple matches the number is quite small, but for pathological cases it
2906 gets very large very quickly.
2908 47. There's a new option for pcre_fullinfo() called PCRE_INFO_STUDYSIZE. It
2909 returns the size of the data block pointed to by the study_data field in a
2910 pcre_extra block, that is, the value that was passed as the argument to
2911 pcre_malloc() when PCRE was getting memory in which to place the information
2912 created by pcre_study(). The fourth argument should point to a size_t variable.
2913 pcretest has been extended so that this information is shown after a successful
2914 pcre_study() call when information about the compiled regex is being displayed.
2916 48. Cosmetic change to Makefile: there's no need to have / after $(DESTDIR)
2917 because what follows is always an absolute path. (Later: it turns out that this
2918 is more than cosmetic for MinGW, because it doesn't like empty path
2919 components.)
2921 49. Some changes have been made to the callout feature (see 28 above):
2923 (i) A callout function now has three choices for what it returns:
2925 0 => success, carry on matching
2926 > 0 => failure at this point, but backtrack if possible
2927 < 0 => serious error, return this value from pcre_exec()
2929 Negative values should normally be chosen from the set of PCRE_ERROR_xxx
2930 values. In particular, returning PCRE_ERROR_NOMATCH forces a standard
2931 "match failed" error. The error number PCRE_ERROR_CALLOUT is reserved for
2932 use by callout functions. It will never be used by PCRE itself.
2934 (ii) The pcre_extra structure (see 45 above) has a void * field called
2935 callout_data, with corresponding flag bit PCRE_EXTRA_CALLOUT_DATA. The
2936 pcre_callout_block structure has a field of the same name. The contents of
2937 the field passed in the pcre_extra structure are passed to the callout
2938 function in the corresponding field in the callout block. This makes it
2939 easier to use the same callout-containing regex from multiple threads. For
2940 testing, the pcretest program has a new data escape
2942 \C*n pass the number n (may be negative) as callout_data
2944 If the callout function in pcretest receives a non-zero value as
2945 callout_data, it returns that value.
2947 50. Makefile wasn't handling CFLAGS properly when compiling dftables. Also,
2948 there were some redundant $(CFLAGS) in commands that are now specified as
2949 $(LINK), which already includes $(CFLAGS).
2951 51. Extensions to UTF-8 support are listed below. These all apply when (a) PCRE
2952 has been compiled with UTF-8 support *and* pcre_compile() has been compiled
2953 with the PCRE_UTF8 flag. Patterns that are compiled without that flag assume
2954 one-byte characters throughout. Note that case-insensitive matching applies
2955 only to characters whose values are less than 256. PCRE doesn't support the
2956 notion of cases for higher-valued characters.
2958 (i) A character class whose characters are all within 0-255 is handled as
2959 a bit map, and the map is inverted for negative classes. Previously, a
2960 character > 255 always failed to match such a class; however it should
2961 match if the class was a negative one (e.g. [^ab]). This has been fixed.
2963 (ii) A negated character class with a single character < 255 is coded as
2964 "not this character" (OP_NOT). This wasn't working properly when the test
2965 character was multibyte, either singly or repeated.
2967 (iii) Repeats of multibyte characters are now handled correctly in UTF-8
2968 mode, for example: \x{100}{2,3}.
2970 (iv) The character escapes \b, \B, \d, \D, \s, \S, \w, and \W (either
2971 singly or repeated) now correctly test multibyte characters. However,
2972 PCRE doesn't recognize any characters with values greater than 255 as
2973 digits, spaces, or word characters. Such characters always match \D, \S,
2974 and \W, and never match \d, \s, or \w.
2976 (v) Classes may now contain characters and character ranges with values
2977 greater than 255. For example: [ab\x{100}-\x{400}].
2979 (vi) pcregrep now has a --utf-8 option (synonym -u) which makes it call
2980 PCRE in UTF-8 mode.
2982 52. The info request value PCRE_INFO_FIRSTCHAR has been renamed
2983 PCRE_INFO_FIRSTBYTE because it is a byte value. However, the old name is
2984 retained for backwards compatibility. (Note that LASTLITERAL is also a byte
2985 value.)
2987 53. The single man page has become too large. I have therefore split it up into
2988 a number of separate man pages. These also give rise to individual HTML pages;
2989 these are now put in a separate directory, and there is an index.html page that
2990 lists them all. Some hyperlinking between the pages has been installed.
2992 54. Added convenience functions for handling named capturing parentheses.
2994 55. Unknown escapes inside character classes (e.g. [\M]) and escapes that
2995 aren't interpreted therein (e.g. [\C]) are literals in Perl. This is now also
2996 true in PCRE, except when the PCRE_EXTENDED option is set, in which case they
2997 are faulted.
2999 56. Introduced HOST_CC and HOST_CFLAGS which can be set in the environment when
3000 calling configure. These values are used when compiling the dftables.c program
3001 which is run to generate the source of the default character tables. They
3002 default to the values of CC and CFLAGS. If you are cross-compiling PCRE,
3003 you will need to set these values.
3005 57. Updated the building process for Windows DLL, as provided by Fred Cox.
3008 Version 3.9 02-Jan-02
3009 ---------------------
3011 1. A bit of extraneous text had somehow crept into the pcregrep documentation.
3013 2. If --disable-static was given, the building process failed when trying to
3014 build pcretest and pcregrep. (For some reason it was using libtool to compile
3015 them, which is not right, as they aren't part of the library.)
3018 Version 3.8 18-Dec-01
3019 ---------------------
3021 1. The experimental UTF-8 code was completely screwed up. It was packing the
3022 bytes in the wrong order. How dumb can you get?
3025 Version 3.7 29-Oct-01
3026 ---------------------
3028 1. In updating pcretest to check change 1 of version 3.6, I screwed up.
3029 This caused pcretest, when used on the test data, to segfault. Unfortunately,
3030 this didn't happen under Solaris 8, where I normally test things.
3032 2. The Makefile had to be changed to make it work on BSD systems, where 'make'
3033 doesn't seem to recognize that ./xxx and xxx are the same file. (This entry
3034 isn't in ChangeLog distributed with 3.7 because I forgot when I hastily made
3035 this fix an hour or so after the initial 3.7 release.)
3038 Version 3.6 23-Oct-01
3039 ---------------------
3041 1. Crashed with /(sens|respons)e and \1ibility/ and "sense and sensibility" if
3042 offsets passed as NULL with zero offset count.
3044 2. The config.guess and config.sub files had not been updated when I moved to
3045 the latest autoconf.
3048 Version 3.5 15-Aug-01
3049 ---------------------
3051 1. Added some missing #if !defined NOPOSIX conditionals in pcretest.c that
3052 had been forgotten.
3054 2. By using declared but undefined structures, we can avoid using "void"
3055 definitions in pcre.h while keeping the internal definitions of the structures
3056 private.
3058 3. The distribution is now built using autoconf 2.50 and libtool 1.4. From a
3059 user point of view, this means that both static and shared libraries are built
3060 by default, but this can be individually controlled. More of the work of
3061 handling this static/shared cases is now inside libtool instead of PCRE's make
3062 file.
3064 4. The pcretest utility is now installed along with pcregrep because it is
3065 useful for users (to test regexs) and by doing this, it automatically gets
3066 relinked by libtool. The documentation has been turned into a man page, so
3067 there are now .1, .txt, and .html versions in /doc.
3069 5. Upgrades to pcregrep:
3070 (i) Added long-form option names like gnu grep.
3071 (ii) Added --help to list all options with an explanatory phrase.
3072 (iii) Added -r, --recursive to recurse into sub-directories.
3073 (iv) Added -f, --file to read patterns from a file.
3075 6. pcre_exec() was referring to its "code" argument before testing that
3076 argument for NULL (and giving an error if it was NULL).
3078 7. Upgraded Makefile.in to allow for compiling in a different directory from
3079 the source directory.
3081 8. Tiny buglet in pcretest: when pcre_fullinfo() was called to retrieve the
3082 options bits, the pointer it was passed was to an int instead of to an unsigned
3083 long int. This mattered only on 64-bit systems.
3085 9. Fixed typo (3.4/1) in pcre.h again. Sigh. I had changed pcre.h (which is
3086 generated) instead of pcre.in, which it its source. Also made the same change
3087 in several of the .c files.
3089 10. A new release of gcc defines printf() as a macro, which broke pcretest
3090 because it had an ifdef in the middle of a string argument for printf(). Fixed
3091 by using separate calls to printf().
3093 11. Added --enable-newline-is-cr and --enable-newline-is-lf to the configure
3094 script, to force use of CR or LF instead of \n in the source. On non-Unix
3095 systems, the value can be set in config.h.
3097 12. The limit of 200 on non-capturing parentheses is a _nesting_ limit, not an
3098 absolute limit. Changed the text of the error message to make this clear, and
3099 likewise updated the man page.
3101 13. The limit of 99 on the number of capturing subpatterns has been removed.
3102 The new limit is 65535, which I hope will not be a "real" limit.
3105 Version 3.4 22-Aug-00
3106 ---------------------
3108 1. Fixed typo in pcre.h: unsigned const char * changed to const unsigned char *.
3110 2. Diagnose condition (?(0) as an error instead of crashing on matching.
3113 Version 3.3 01-Aug-00
3114 ---------------------
3116 1. If an octal character was given, but the value was greater than \377, it
3117 was not getting masked to the least significant bits, as documented. This could
3118 lead to crashes in some systems.
3120 2. Perl 5.6 (if not earlier versions) accepts classes like [a-\d] and treats
3121 the hyphen as a literal. PCRE used to give an error; it now behaves like Perl.
3123 3. Added the functions pcre_free_substring() and pcre_free_substring_list().
3124 These just pass their arguments on to (pcre_free)(), but they are provided
3125 because some uses of PCRE bind it to non-C systems that can call its functions,
3126 but cannot call free() or pcre_free() directly.
3128 4. Add "make test" as a synonym for "make check". Corrected some comments in
3129 the Makefile.
3131 5. Add $(DESTDIR)/ in front of all the paths in the "install" target in the
3132 Makefile.
3134 6. Changed the name of pgrep to pcregrep, because Solaris has introduced a
3135 command called pgrep for grepping around the active processes.
3137 7. Added the beginnings of support for UTF-8 character strings.
3139 8. Arranged for the Makefile to pass over the settings of CC, CFLAGS, and
3140 RANLIB to ./ltconfig so that they are used by libtool. I think these are all
3141 the relevant ones. (AR is not passed because ./ltconfig does its own figuring
3142 out for the ar command.)
3145 Version 3.2 12-May-00
3146 ---------------------
3148 This is purely a bug fixing release.
3150 1. If the pattern /((Z)+|A)*/ was matched agained ZABCDEFG it matched Z instead
3151 of ZA. This was just one example of several cases that could provoke this bug,
3152 which was introduced by change 9 of version 2.00. The code for breaking
3153 infinite loops after an iteration that matches an empty string was't working
3154 correctly.
3156 2. The pcretest program was not imitating Perl correctly for the pattern /a*/g
3157 when matched against abbab (for example). After matching an empty string, it
3158 wasn't forcing anchoring when setting PCRE_NOTEMPTY for the next attempt; this
3159 caused it to match further down the string than it should.
3161 3. The code contained an inclusion of sys/types.h. It isn't clear why this
3162 was there because it doesn't seem to be needed, and it causes trouble on some
3163 systems, as it is not a Standard C header. It has been removed.
3165 4. Made 4 silly changes to the source to avoid stupid compiler warnings that
3166 were reported on the Macintosh. The changes were from
3168 while ((c = *(++ptr)) != 0 && c != '\n');
3169 to
3170 while ((c = *(++ptr)) != 0 && c != '\n') ;
3172 Totally extraordinary, but if that's what it takes...
3174 5. PCRE is being used in one environment where neither memmove() nor bcopy() is
3175 available. Added HAVE_BCOPY and an autoconf test for it; if neither
3176 HAVE_MEMMOVE nor HAVE_BCOPY is set, use a built-in emulation function which
3177 assumes the way PCRE uses memmove() (always moving upwards).
3179 6. PCRE is being used in one environment where strchr() is not available. There
3180 was only one use in pcre.c, and writing it out to avoid strchr() probably gives
3181 faster code anyway.
3184 Version 3.1 09-Feb-00
3185 ---------------------
3187 The only change in this release is the fixing of some bugs in Makefile.in for
3188 the "install" target:
3190 (1) It was failing to install pcreposix.h.
3192 (2) It was overwriting the pcre.3 man page with the pcreposix.3 man page.
3195 Version 3.0 01-Feb-00
3196 ---------------------
3198 1. Add support for the /+ modifier to perltest (to output $` like it does in
3199 pcretest).
3201 2. Add support for the /g modifier to perltest.
3203 3. Fix pcretest so that it behaves even more like Perl for /g when the pattern
3204 matches null strings.
3206 4. Fix perltest so that it doesn't do unwanted things when fed an empty
3207 pattern. Perl treats empty patterns specially - it reuses the most recent
3208 pattern, which is not what we want. Replace // by /(?#)/ in order to avoid this
3209 effect.
3211 5. The POSIX interface was broken in that it was just handing over the POSIX
3212 captured string vector to pcre_exec(), but (since release 2.00) PCRE has
3213 required a bigger vector, with some working space on the end. This means that
3214 the POSIX wrapper now has to get and free some memory, and copy the results.
3216 6. Added some simple autoconf support, placing the test data and the
3217 documentation in separate directories, re-organizing some of the
3218 information files, and making it build pcre-config (a GNU standard). Also added
3219 libtool support for building PCRE as a shared library, which is now the
3220 default.
3222 7. Got rid of the leading zero in the definition of PCRE_MINOR because 08 and
3223 09 are not valid octal constants. Single digits will be used for minor values
3224 less than 10.
3226 8. Defined REG_EXTENDED and REG_NOSUB as zero in the POSIX header, so that
3227 existing programs that set these in the POSIX interface can use PCRE without
3228 modification.
3230 9. Added a new function, pcre_fullinfo() with an extensible interface. It can
3231 return all that pcre_info() returns, plus additional data. The pcre_info()
3232 function is retained for compatibility, but is considered to be obsolete.
3234 10. Added experimental recursion feature (?R) to handle one common case that
3235 Perl 5.6 will be able to do with (?p{...}).
3237 11. Added support for POSIX character classes like [:alpha:], which Perl is
3238 adopting.
3241 Version 2.08 31-Aug-99
3242 ----------------------
3244 1. When startoffset was not zero and the pattern began with ".*", PCRE was not
3245 trying to match at the startoffset position, but instead was moving forward to
3246 the next newline as if a previous match had failed.
3248 2. pcretest was not making use of PCRE_NOTEMPTY when repeating for /g and /G,
3249 and could get into a loop if a null string was matched other than at the start
3250 of the subject.
3252 3. Added definitions of PCRE_MAJOR and PCRE_MINOR to pcre.h so the version can
3253 be distinguished at compile time, and for completeness also added PCRE_DATE.
3255 5. Added Paul Sokolovsky's minor changes to make it easy to compile a Win32 DLL
3256 in GnuWin32 environments.
3259 Version 2.07 29-Jul-99
3260 ----------------------
3262 1. The documentation is now supplied in plain text form and HTML as well as in
3263 the form of man page sources.
3265 2. C++ compilers don't like assigning (void *) values to other pointer types.
3266 In particular this affects malloc(). Although there is no problem in Standard
3267 C, I've put in casts to keep C++ compilers happy.
3269 3. Typo on pcretest.c; a cast of (unsigned char *) in the POSIX regexec() call
3270 should be (const char *).
3272 4. If NOPOSIX is defined, pcretest.c compiles without POSIX support. This may
3273 be useful for non-Unix systems who don't want to bother with the POSIX stuff.
3274 However, I haven't made this a standard facility. The documentation doesn't
3275 mention it, and the Makefile doesn't support it.
3277 5. The Makefile now contains an "install" target, with editable destinations at
3278 the top of the file. The pcretest program is not installed.
3280 6. pgrep -V now gives the PCRE version number and date.
3282 7. Fixed bug: a zero repetition after a literal string (e.g. /abcde{0}/) was
3283 causing the entire string to be ignored, instead of just the last character.
3285 8. If a pattern like /"([^\\"]+|\\.)*"/ is applied in the normal way to a
3286 non-matching string, it can take a very, very long time, even for strings of
3287 quite modest length, because of the nested recursion. PCRE now does better in
3288 some of these cases. It does this by remembering the last required literal
3289 character in the pattern, and pre-searching the subject to ensure it is present
3290 before running the real match. In other words, it applies a heuristic to detect
3291 some types of certain failure quickly, and in the above example, if presented
3292 with a string that has no trailing " it gives "no match" very quickly.
3294 9. A new runtime option PCRE_NOTEMPTY causes null string matches to be ignored;
3295 other alternatives are tried instead.
3298 Version 2.06 09-Jun-99
3299 ----------------------
3301 1. Change pcretest's output for amount of store used to show just the code
3302 space, because the remainder (the data block) varies in size between 32-bit and
3303 64-bit systems.
3305 2. Added an extra argument to pcre_exec() to supply an offset in the subject to
3306 start matching at. This allows lookbehinds to work when searching for multiple
3307 occurrences in a string.
3309 3. Added additional options to pcretest for testing multiple occurrences:
3311 /+ outputs the rest of the string that follows a match
3312 /g loops for multiple occurrences, using the new startoffset argument
3313 /G loops for multiple occurrences by passing an incremented pointer
3315 4. PCRE wasn't doing the "first character" optimization for patterns starting
3316 with \b or \B, though it was doing it for other lookbehind assertions. That is,
3317 it wasn't noticing that a match for a pattern such as /\bxyz/ has to start with
3318 the letter 'x'. On long subject strings, this gives a significant speed-up.
3321 Version 2.05 21-Apr-99
3322 ----------------------
3324 1. Changed the type of magic_number from int to long int so that it works
3325 properly on 16-bit systems.
3327 2. Fixed a bug which caused patterns starting with .* not to work correctly
3328 when the subject string contained newline characters. PCRE was assuming
3329 anchoring for such patterns in all cases, which is not correct because .* will
3330 not pass a newline unless PCRE_DOTALL is set. It now assumes anchoring only if
3331 DOTALL is set at top level; otherwise it knows that patterns starting with .*
3332 must be retried after every newline in the subject.
3335 Version 2.04 18-Feb-99
3336 ----------------------
3338 1. For parenthesized subpatterns with repeats whose minimum was zero, the
3339 computation of the store needed to hold the pattern was incorrect (too large).
3340 If such patterns were nested a few deep, this could multiply and become a real
3341 problem.
3343 2. Added /M option to pcretest to show the memory requirement of a specific
3344 pattern. Made -m a synonym of -s (which does this globally) for compatibility.
3346 3. Subpatterns of the form (regex){n,m} (i.e. limited maximum) were being
3347 compiled in such a way that the backtracking after subsequent failure was
3348 pessimal. Something like (a){0,3} was compiled as (a)?(a)?(a)? instead of
3349 ((a)((a)(a)?)?)? with disastrous performance if the maximum was of any size.
3352 Version 2.03 02-Feb-99
3353 ----------------------
3355 1. Fixed typo and small mistake in man page.
3357 2. Added 4th condition (GPL supersedes if conflict) and created separate
3358 LICENCE file containing the conditions.
3360 3. Updated pcretest so that patterns such as /abc\/def/ work like they do in
3361 Perl, that is the internal \ allows the delimiter to be included in the
3362 pattern. Locked out the use of \ as a delimiter. If \ immediately follows
3363 the final delimiter, add \ to the end of the pattern (to test the error).
3365 4. Added the convenience functions for extracting substrings after a successful
3366 match. Updated pcretest to make it able to test these functions.
3369 Version 2.02 14-Jan-99
3370 ----------------------
3372 1. Initialized the working variables associated with each extraction so that
3373 their saving and restoring doesn't refer to uninitialized store.
3375 2. Put dummy code into study.c in order to trick the optimizer of the IBM C
3376 compiler for OS/2 into generating correct code. Apparently IBM isn't going to
3377 fix the problem.
3379 3. Pcretest: the timing code wasn't using LOOPREPEAT for timing execution
3380 calls, and wasn't printing the correct value for compiling calls. Increased the
3381 default value of LOOPREPEAT, and the number of significant figures in the
3382 times.
3384 4. Changed "/bin/rm" in the Makefile to "-rm" so it works on Windows NT.
3386 5. Renamed "deftables" as "dftables" to get it down to 8 characters, to avoid
3387 a building problem on Windows NT with a FAT file system.
3390 Version 2.01 21-Oct-98
3391 ----------------------
3393 1. Changed the API for pcre_compile() to allow for the provision of a pointer
3394 to character tables built by pcre_maketables() in the current locale. If NULL
3395 is passed, the default tables are used.
3398 Version 2.00 24-Sep-98
3399 ----------------------
3401 1. Since the (>?) facility is in Perl 5.005, don't require PCRE_EXTRA to enable
3402 it any more.
3404 2. Allow quantification of (?>) groups, and make it work correctly.
3406 3. The first character computation wasn't working for (?>) groups.
3408 4. Correct the implementation of \Z (it is permitted to match on the \n at the
3409 end of the subject) and add 5.005's \z, which really does match only at the
3410 very end of the subject.
3412 5. Remove the \X "cut" facility; Perl doesn't have it, and (?> is neater.
3414 6. Remove the ability to specify CASELESS, MULTILINE, DOTALL, and
3415 DOLLAR_END_ONLY at runtime, to make it possible to implement the Perl 5.005
3416 localized options. All options to pcre_study() were also removed.
3418 7. Add other new features from 5.005:
3420 $(?<= positive lookbehind
3421 $(?<! negative lookbehind
3422 (?imsx-imsx) added the unsetting capability
3423 such a setting is global if at outer level; local otherwise
3424 (?imsx-imsx:) non-capturing groups with option setting
3425 (?(cond)re|re) conditional pattern matching
3427 A backreference to itself in a repeated group matches the previous
3428 captured string.
3430 8. General tidying up of studying (both automatic and via "study")
3431 consequential on the addition of new assertions.
3433 9. As in 5.005, unlimited repeated groups that could match an empty substring
3434 are no longer faulted at compile time. Instead, the loop is forcibly broken at
3435 runtime if any iteration does actually match an empty substring.
3437 10. Include the RunTest script in the distribution.
3439 11. Added tests from the Perl 5.005_02 distribution. This showed up a few
3440 discrepancies, some of which were old and were also with respect to 5.004. They
3441 have now been fixed.
3444 Version 1.09 28-Apr-98
3445 ----------------------
3447 1. A negated single character class followed by a quantifier with a minimum
3448 value of one (e.g. [^x]{1,6} ) was not compiled correctly. This could lead to
3449 program crashes, or just wrong answers. This did not apply to negated classes
3450 containing more than one character, or to minima other than one.
3453 Version 1.08 27-Mar-98
3454 ----------------------
3456 1. Add PCRE_UNGREEDY to invert the greediness of quantifiers.
3458 2. Add (?U) and (?X) to set PCRE_UNGREEDY and PCRE_EXTRA respectively. The
3459 latter must appear before anything that relies on it in the pattern.
3462 Version 1.07 16-Feb-98
3463 ----------------------
3465 1. A pattern such as /((a)*)*/ was not being diagnosed as in error (unlimited
3466 repeat of a potentially empty string).
3469 Version 1.06 23-Jan-98
3470 ----------------------
3472 1. Added Markus Oberhumer's little patches for C++.
3474 2. Literal strings longer than 255 characters were broken.
3477 Version 1.05 23-Dec-97
3478 ----------------------
3480 1. Negated character classes containing more than one character were failing if
3481 PCRE_CASELESS was set at run time.
3484 Version 1.04 19-Dec-97
3485 ----------------------
3487 1. Corrected the man page, where some "const" qualifiers had been omitted.
3489 2. Made debugging output print "{0,xxx}" instead of just "{,xxx}" to agree with
3490 input syntax.
3492 3. Fixed memory leak which occurred when a regex with back references was
3493 matched with an offsets vector that wasn't big enough. The temporary memory
3494 that is used in this case wasn't being freed if the match failed.
3496 4. Tidied pcretest to ensure it frees memory that it gets.
3498 5. Temporary memory was being obtained in the case where the passed offsets
3499 vector was exactly big enough.
3501 6. Corrected definition of offsetof() from change 5 below.
3503 7. I had screwed up change 6 below and broken the rules for the use of
3504 setjmp(). Now fixed.
3507 Version 1.03 18-Dec-97
3508 ----------------------
3510 1. A erroneous regex with a missing opening parenthesis was correctly
3511 diagnosed, but PCRE attempted to access brastack[-1], which could cause crashes
3512 on some systems.
3514 2. Replaced offsetof(real_pcre, code) by offsetof(real_pcre, code[0]) because
3515 it was reported that one broken compiler failed on the former because "code" is
3516 also an independent variable.
3518 3. The erroneous regex a[]b caused an array overrun reference.
3520 4. A regex ending with a one-character negative class (e.g. /[^k]$/) did not
3521 fail on data ending with that character. (It was going on too far, and checking
3522 the next character, typically a binary zero.) This was specific to the
3523 optimized code for single-character negative classes.
3525 5. Added a contributed patch from the TIN world which does the following:
3527 + Add an undef for memmove, in case the the system defines a macro for it.
3529 + Add a definition of offsetof(), in case there isn't one. (I don't know
3530 the reason behind this - offsetof() is part of the ANSI standard - but
3531 it does no harm).
3533 + Reduce the ifdef's in pcre.c using macro DPRINTF, thereby eliminating
3534 most of the places where whitespace preceded '#'. I have given up and
3535 allowed the remaining 2 cases to be at the margin.
3537 + Rename some variables in pcre to eliminate shadowing. This seems very
3538 pedantic, but does no harm, of course.
3540 6. Moved the call to setjmp() into its own function, to get rid of warnings
3541 from gcc -Wall, and avoided calling it at all unless PCRE_EXTRA is used.
3543 7. Constructs such as \d{8,} were compiling into the equivalent of
3544 \d{8}\d{0,65527} instead of \d{8}\d* which didn't make much difference to the
3545 outcome, but in this particular case used more store than had been allocated,
3546 which caused the bug to be discovered because it threw up an internal error.
3548 8. The debugging code in both pcre and pcretest for outputting the compiled
3549 form of a regex was going wrong in the case of back references followed by
3550 curly-bracketed repeats.
3553 Version 1.02 12-Dec-97
3554 ----------------------
3556 1. Typos in pcre.3 and comments in the source fixed.
3558 2. Applied a contributed patch to get rid of places where it used to remove
3559 'const' from variables, and fixed some signed/unsigned and uninitialized
3560 variable warnings.
3562 3. Added the "runtest" target to Makefile.
3564 4. Set default compiler flag to -O2 rather than just -O.
3567 Version 1.01 19-Nov-97
3568 ----------------------
3570 1. PCRE was failing to diagnose unlimited repeat of empty string for patterns
3571 like /([ab]*)*/, that is, for classes with more than one character in them.
3573 2. Likewise, it wasn't diagnosing patterns with "once-only" subpatterns, such
3574 as /((?>a*))*/ (a PCRE_EXTRA facility).
3577 Version 1.00 18-Nov-97
3578 ----------------------
3580 1. Added compile-time macros to support systems such as SunOS4 which don't have
3581 memmove() or strerror() but have other things that can be used instead.
3583 2. Arranged that "make clean" removes the executables.
3586 Version 0.99 27-Oct-97
3587 ----------------------
3589 1. Fixed bug in code for optimizing classes with only one character. It was
3590 initializing a 32-byte map regardless, which could cause it to run off the end
3591 of the memory it had got.
3593 2. Added, conditional on PCRE_EXTRA, the proposed (?>REGEX) construction.
3596 Version 0.98 22-Oct-97
3597 ----------------------
3599 1. Fixed bug in code for handling temporary memory usage when there are more
3600 back references than supplied space in the ovector. This could cause segfaults.
3603 Version 0.97 21-Oct-97
3604 ----------------------
3606 1. Added the \X "cut" facility, conditional on PCRE_EXTRA.
3608 2. Optimized negated single characters not to use a bit map.
3610 3. Brought error texts together as macro definitions; clarified some of them;
3611 fixed one that was wrong - it said "range out of order" when it meant "invalid
3612 escape sequence".
3614 4. Changed some char * arguments to const char *.
3616 5. Added PCRE_NOTBOL and PCRE_NOTEOL (from POSIX).
3618 6. Added the POSIX-style API wrapper in pcreposix.a and testing facilities in
3619 pcretest.
3622 Version 0.96 16-Oct-97
3623 ----------------------
3625 1. Added a simple "pgrep" utility to the distribution.
3627 2. Fixed an incompatibility with Perl: "{" is now treated as a normal character
3628 unless it appears in one of the precise forms "{ddd}", "{ddd,}", or "{ddd,ddd}"
3629 where "ddd" means "one or more decimal digits".
3631 3. Fixed serious bug. If a pattern had a back reference, but the call to
3632 pcre_exec() didn't supply a large enough ovector to record the related
3633 identifying subpattern, the match always failed. PCRE now remembers the number
3634 of the largest back reference, and gets some temporary memory in which to save
3635 the offsets during matching if necessary, in order to ensure that
3636 backreferences always work.
3638 4. Increased the compatibility with Perl in a number of ways:
3640 (a) . no longer matches \n by default; an option PCRE_DOTALL is provided
3641 to request this handling. The option can be set at compile or exec time.
3643 (b) $ matches before a terminating newline by default; an option
3644 PCRE_DOLLAR_ENDONLY is provided to override this (but not in multiline
3645 mode). The option can be set at compile or exec time.
3647 (c) The handling of \ followed by a digit other than 0 is now supposed to be
3648 the same as Perl's. If the decimal number it represents is less than 10
3649 or there aren't that many previous left capturing parentheses, an octal
3650 escape is read. Inside a character class, it's always an octal escape,
3651 even if it is a single digit.
3653 (d) An escaped but undefined alphabetic character is taken as a literal,
3654 unless PCRE_EXTRA is set. Currently this just reserves the remaining
3655 escapes.
3657 (e) {0} is now permitted. (The previous item is removed from the compiled
3658 pattern).
3660 5. Changed all the names of code files so that the basic parts are no longer
3661 than 10 characters, and abolished the teeny "globals.c" file.
3663 6. Changed the handling of character classes; they are now done with a 32-byte
3664 bit map always.
3666 7. Added the -d and /D options to pcretest to make it possible to look at the
3667 internals of compilation without having to recompile pcre.
3670 Version 0.95 23-Sep-97
3671 ----------------------
3673 1. Fixed bug in pre-pass concerning escaped "normal" characters such as \x5c or
3674 \x20 at the start of a run of normal characters. These were being treated as
3675 real characters, instead of the source characters being re-checked.
3678 Version 0.94 18-Sep-97
3679 ----------------------
3681 1. The functions are now thread-safe, with the caveat that the global variables
3682 containing pointers to malloc() and free() or alternative functions are the
3683 same for all threads.
3685 2. Get pcre_study() to generate a bitmap of initial characters for non-
3686 anchored patterns when this is possible, and use it if passed to pcre_exec().
3689 Version 0.93 15-Sep-97
3690 ----------------------
3692 1. /(b)|(:+)/ was computing an incorrect first character.
3694 2. Add pcre_study() to the API and the passing of pcre_extra to pcre_exec(),
3695 but not actually doing anything yet.
3697 3. Treat "-" characters in classes that cannot be part of ranges as literals,
3698 as Perl does (e.g. [-az] or [az-]).
3700 4. Set the anchored flag if a branch starts with .* or .*? because that tests
3701 all possible positions.
3703 5. Split up into different modules to avoid including unneeded functions in a
3704 compiled binary. However, compile and exec are still in one module. The "study"
3705 function is split off.
3707 6. The character tables are now in a separate module whose source is generated
3708 by an auxiliary program - but can then be edited by hand if required. There are
3709 now no calls to isalnum(), isspace(), isdigit(), isxdigit(), tolower() or
3710 toupper() in the code.
3712 7. Turn the malloc/free funtions variables into pcre_malloc and pcre_free and
3713 make them global. Abolish the function for setting them, as the caller can now
3714 set them directly.
3717 Version 0.92 11-Sep-97
3718 ----------------------
3720 1. A repeat with a fixed maximum and a minimum of 1 for an ordinary character
3721 (e.g. /a{1,3}/) was broken (I mis-optimized it).
3723 2. Caseless matching was not working in character classes if the characters in
3724 the pattern were in upper case.
3726 3. Make ranges like [W-c] work in the same way as Perl for caseless matching.
3728 4. Make PCRE_ANCHORED public and accept as a compile option.
3730 5. Add an options word to pcre_exec() and accept PCRE_ANCHORED and
3731 PCRE_CASELESS at run time. Add escapes \A and \I to pcretest to cause it to
3732 pass them.
3734 6. Give an error if bad option bits passed at compile or run time.
3736 7. Add PCRE_MULTILINE at compile and exec time, and (?m) as well. Add \M to
3737 pcretest to cause it to pass that flag.
3739 8. Add pcre_info(), to get the number of identifying subpatterns, the stored
3740 options, and the first character, if set.
3742 9. Recognize C+ or C{n,m} where n >= 1 as providing a fixed starting character.
3745 Version 0.91 10-Sep-97
3746 ----------------------
3748 1. PCRE was failing to diagnose unlimited repeats of subpatterns that could
3749 match the empty string as in /(a*)*/. It was looping and ultimately crashing.
3751 2. PCRE was looping on encountering an indefinitely repeated back reference to
3752 a subpattern that had matched an empty string, e.g. /(a|)\1*/. It now does what
3753 Perl does - treats the match as successful.
3755 ****


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

  ViewVC Help
Powered by ViewVC 1.1.5