/[pcre]/code/trunk/ChangeLog
ViewVC logotype

Contents of /code/trunk/ChangeLog

Parent Directory Parent Directory | Revision Log Revision Log


Revision 496 - (show annotations) (download)
Tue Mar 2 19:11:17 2010 UTC (4 years, 6 months ago) by ph10
File size: 182334 byte(s)
Fix DEFINE bug for forward reference with a possessive quantifier.

1 ChangeLog for PCRE
2 ------------------
3
4 Version 8.02 01-Mar-2010
5 ------------------------
6
7 1. The Unicode data tables have been updated to Unicode 5.2.0.
8
9 2. Added the option --libs-cpp to pcre-config, but only when C++ support is
10 configured.
11
12 3. Updated the licensing terms in the pcregexp.pas file, as agreed with the
13 original author of that file, following a query about its status.
14
15 4. On systems that do not have stdint.h (e.g. Solaris), check for and include
16 inttypes.h instead. This fixes a bug that was introduced by change 8.01/8.
17
18 5. A pattern such as (?&t)*+(?(DEFINE)(?<t>.)) which has a possessive
19 quantifier applied to a forward-referencing subroutine call, could compile
20 incorrect code or give the error "internal error: previously-checked
21 referenced subpattern not found".
22
23
24 Version 8.01 19-Jan-2010
25 ------------------------
26
27 1. If a pattern contained a conditional subpattern with only one branch (in
28 particular, this includes all (*DEFINE) patterns), a call to pcre_study()
29 computed the wrong minimum data length (which is of course zero for such
30 subpatterns). This could cause incorrect "no match" results.
31
32 2. For patterns such as (?i)a(?-i)b|c where an option setting at the start of
33 the pattern is reset in the first branch, pcre_compile() failed with
34 "internal error: code overflow at offset...". This happened only when
35 the reset was to the original external option setting. (An optimization
36 abstracts leading options settings into an external setting, which was the
37 cause of this.)
38
39 3. A pattern such as ^(?!a(*SKIP)b) where a negative assertion contained one
40 of the verbs SKIP, PRUNE, or COMMIT, did not work correctly. When the
41 assertion pattern did not match (meaning that the assertion was true), it
42 was incorrectly treated as false if the SKIP had been reached during the
43 matching. This also applied to assertions used as conditions.
44
45 4. If an item that is not supported by pcre_dfa_exec() was encountered in an
46 assertion subpattern, including such a pattern used as a condition,
47 unpredictable results occurred, instead of the error return
48 PCRE_ERROR_DFA_UITEM.
49
50 5. The C++ GlobalReplace function was not working like Perl for the special
51 situation when an empty string is matched. It now does the fancy magic
52 stuff that is necessary.
53
54 6. In pcre_internal.h, obsolete includes to setjmp.h and stdarg.h have been
55 removed. (These were left over from very, very early versions of PCRE.)
56
57 7. Some cosmetic changes to the code to make life easier when compiling it
58 as part of something else:
59
60 (a) Change DEBUG to PCRE_DEBUG.
61
62 (b) In pcre_compile(), rename the member of the "branch_chain" structure
63 called "current" as "current_branch", to prevent a collision with the
64 Linux macro when compiled as a kernel module.
65
66 (c) In pcre_study(), rename the function set_bit() as set_table_bit(), to
67 prevent a collision with the Linux macro when compiled as a kernel
68 module.
69
70 8. In pcre_compile() there are some checks for integer overflows that used to
71 cast potentially large values to (double). This has been changed to that
72 when building, a check for int64_t is made, and if it is found, it is used
73 instead, thus avoiding the use of floating point arithmetic. (There is no
74 other use of FP in PCRE.) If int64_t is not found, the fallback is to
75 double.
76
77 9. Added two casts to avoid signed/unsigned warnings from VS Studio Express
78 2005 (difference between two addresses compared to an unsigned value).
79
80 10. Change the standard AC_CHECK_LIB test for libbz2 in configure.ac to a
81 custom one, because of the following reported problem in Windows:
82
83 - libbz2 uses the Pascal calling convention (WINAPI) for the functions
84 under Win32.
85 - The standard autoconf AC_CHECK_LIB fails to include "bzlib.h",
86 therefore missing the function definition.
87 - The compiler thus generates a "C" signature for the test function.
88 - The linker fails to find the "C" function.
89 - PCRE fails to configure if asked to do so against libbz2.
90
91 11. When running libtoolize from libtool-2.2.6b as part of autogen.sh, these
92 messages were output:
93
94 Consider adding `AC_CONFIG_MACRO_DIR([m4])' to configure.ac and
95 rerunning libtoolize, to keep the correct libtool macros in-tree.
96 Consider adding `-I m4' to ACLOCAL_AMFLAGS in Makefile.am.
97
98 I have done both of these things.
99
100 12. Although pcre_dfa_exec() does not use nearly as much stack as pcre_exec()
101 most of the time, it *can* run out if it is given a pattern that contains a
102 runaway infinite recursion. I updated the discussion in the pcrestack man
103 page.
104
105 13. Now that we have gone to the x.xx style of version numbers, the minor
106 version may start with zero. Using 08 or 09 is a bad idea because users
107 might check the value of PCRE_MINOR in their code, and 08 or 09 may be
108 interpreted as invalid octal numbers. I've updated the previous comment in
109 configure.ac, and also added a check that gives an error if 08 or 09 are
110 used.
111
112 14. Change 8.00/11 was not quite complete: code had been accidentally omitted,
113 causing partial matching to fail when the end of the subject matched \W
114 in a UTF-8 pattern where \W was quantified with a minimum of 3.
115
116 15. There were some discrepancies between the declarations in pcre_internal.h
117 of _pcre_is_newline(), _pcre_was_newline(), and _pcre_valid_utf8() and
118 their definitions. The declarations used "const uschar *" and the
119 definitions used USPTR. Even though USPTR is normally defined as "const
120 unsigned char *" (and uschar is typedeffed as "unsigned char"), it was
121 reported that: "This difference in casting confuses some C++ compilers, for
122 example, SunCC recognizes above declarations as different functions and
123 generates broken code for hbpcre." I have changed the declarations to use
124 USPTR.
125
126 16. GNU libtool is named differently on some systems. The autogen.sh script now
127 tries several variants such as glibtoolize (MacOSX) and libtoolize1x
128 (FreeBSD).
129
130 17. Applied Craig's patch that fixes an HP aCC compile error in pcre 8.00
131 (strtoXX undefined when compiling pcrecpp.cc). The patch contains this
132 comment: "Figure out how to create a longlong from a string: strtoll and
133 equivalent. It's not enough to call AC_CHECK_FUNCS: hpux has a strtoll, for
134 instance, but it only takes 2 args instead of 3!"
135
136 18. A subtle bug concerned with back references has been fixed by a change of
137 specification, with a corresponding code fix. A pattern such as
138 ^(xa|=?\1a)+$ which contains a back reference inside the group to which it
139 refers, was giving matches when it shouldn't. For example, xa=xaaa would
140 match that pattern. Interestingly, Perl (at least up to 5.11.3) has the
141 same bug. Such groups have to be quantified to be useful, or contained
142 inside another quantified group. (If there's no repetition, the reference
143 can never match.) The problem arises because, having left the group and
144 moved on to the rest of the pattern, a later failure that backtracks into
145 the group uses the captured value from the final iteration of the group
146 rather than the correct earlier one. I have fixed this in PCRE by forcing
147 any group that contains a reference to itself to be an atomic group; that
148 is, there cannot be any backtracking into it once it has completed. This is
149 similar to recursive and subroutine calls.
150
151
152 Version 8.00 19-Oct-09
153 ----------------------
154
155 1. The table for translating pcre_compile() error codes into POSIX error codes
156 was out-of-date, and there was no check on the pcre_compile() error code
157 being within the table. This could lead to an OK return being given in
158 error.
159
160 2. Changed the call to open a subject file in pcregrep from fopen(pathname,
161 "r") to fopen(pathname, "rb"), which fixed a problem with some of the tests
162 in a Windows environment.
163
164 3. The pcregrep --count option prints the count for each file even when it is
165 zero, as does GNU grep. However, pcregrep was also printing all files when
166 --files-with-matches was added. Now, when both options are given, it prints
167 counts only for those files that have at least one match. (GNU grep just
168 prints the file name in this circumstance, but including the count seems
169 more useful - otherwise, why use --count?) Also ensured that the
170 combination -clh just lists non-zero counts, with no names.
171
172 4. The long form of the pcregrep -F option was incorrectly implemented as
173 --fixed_strings instead of --fixed-strings. This is an incompatible change,
174 but it seems right to fix it, and I didn't think it was worth preserving
175 the old behaviour.
176
177 5. The command line items --regex=pattern and --regexp=pattern were not
178 recognized by pcregrep, which required --regex pattern or --regexp pattern
179 (with a space rather than an '='). The man page documented the '=' forms,
180 which are compatible with GNU grep; these now work.
181
182 6. No libpcreposix.pc file was created for pkg-config; there was just
183 libpcre.pc and libpcrecpp.pc. The omission has been rectified.
184
185 7. Added #ifndef SUPPORT_UCP into the pcre_ucd.c module, to reduce its size
186 when UCP support is not needed, by modifying the Python script that
187 generates it from Unicode data files. This should not matter if the module
188 is correctly used as a library, but I received one complaint about 50K of
189 unwanted data. My guess is that the person linked everything into his
190 program rather than using a library. Anyway, it does no harm.
191
192 8. A pattern such as /\x{123}{2,2}+/8 was incorrectly compiled; the trigger
193 was a minimum greater than 1 for a wide character in a possessive
194 repetition. The same bug could also affect patterns like /(\x{ff}{0,2})*/8
195 which had an unlimited repeat of a nested, fixed maximum repeat of a wide
196 character. Chaos in the form of incorrect output or a compiling loop could
197 result.
198
199 9. The restrictions on what a pattern can contain when partial matching is
200 requested for pcre_exec() have been removed. All patterns can now be
201 partially matched by this function. In addition, if there are at least two
202 slots in the offset vector, the offset of the earliest inspected character
203 for the match and the offset of the end of the subject are set in them when
204 PCRE_ERROR_PARTIAL is returned.
205
206 10. Partial matching has been split into two forms: PCRE_PARTIAL_SOFT, which is
207 synonymous with PCRE_PARTIAL, for backwards compatibility, and
208 PCRE_PARTIAL_HARD, which causes a partial match to supersede a full match,
209 and may be more useful for multi-segment matching.
210
211 11. Partial matching with pcre_exec() is now more intuitive. A partial match
212 used to be given if ever the end of the subject was reached; now it is
213 given only if matching could not proceed because another character was
214 needed. This makes a difference in some odd cases such as Z(*FAIL) with the
215 string "Z", which now yields "no match" instead of "partial match". In the
216 case of pcre_dfa_exec(), "no match" is given if every matching path for the
217 final character ended with (*FAIL).
218
219 12. Restarting a match using pcre_dfa_exec() after a partial match did not work
220 if the pattern had a "must contain" character that was already found in the
221 earlier partial match, unless partial matching was again requested. For
222 example, with the pattern /dog.(body)?/, the "must contain" character is
223 "g". If the first part-match was for the string "dog", restarting with
224 "sbody" failed. This bug has been fixed.
225
226 13. The string returned by pcre_dfa_exec() after a partial match has been
227 changed so that it starts at the first inspected character rather than the
228 first character of the match. This makes a difference only if the pattern
229 starts with a lookbehind assertion or \b or \B (\K is not supported by
230 pcre_dfa_exec()). It's an incompatible change, but it makes the two
231 matching functions compatible, and I think it's the right thing to do.
232
233 14. Added a pcredemo man page, created automatically from the pcredemo.c file,
234 so that the demonstration program is easily available in environments where
235 PCRE has not been installed from source.
236
237 15. Arranged to add -DPCRE_STATIC to cflags in libpcre.pc, libpcreposix.cp,
238 libpcrecpp.pc and pcre-config when PCRE is not compiled as a shared
239 library.
240
241 16. Added REG_UNGREEDY to the pcreposix interface, at the request of a user.
242 It maps to PCRE_UNGREEDY. It is not, of course, POSIX-compatible, but it
243 is not the first non-POSIX option to be added. Clearly some people find
244 these options useful.
245
246 17. If a caller to the POSIX matching function regexec() passes a non-zero
247 value for nmatch with a NULL value for pmatch, the value of
248 nmatch is forced to zero.
249
250 18. RunGrepTest did not have a test for the availability of the -u option of
251 the diff command, as RunTest does. It now checks in the same way as
252 RunTest, and also checks for the -b option.
253
254 19. If an odd number of negated classes containing just a single character
255 interposed, within parentheses, between a forward reference to a named
256 subpattern and the definition of the subpattern, compilation crashed with
257 an internal error, complaining that it could not find the referenced
258 subpattern. An example of a crashing pattern is /(?&A)(([^m])(?<A>))/.
259 [The bug was that it was starting one character too far in when skipping
260 over the character class, thus treating the ] as data rather than
261 terminating the class. This meant it could skip too much.]
262
263 20. Added PCRE_NOTEMPTY_ATSTART in order to be able to correctly implement the
264 /g option in pcretest when the pattern contains \K, which makes it possible
265 to have an empty string match not at the start, even when the pattern is
266 anchored. Updated pcretest and pcredemo to use this option.
267
268 21. If the maximum number of capturing subpatterns in a recursion was greater
269 than the maximum at the outer level, the higher number was returned, but
270 with unset values at the outer level. The correct (outer level) value is
271 now given.
272
273 22. If (*ACCEPT) appeared inside capturing parentheses, previous releases of
274 PCRE did not set those parentheses (unlike Perl). I have now found a way to
275 make it do so. The string so far is captured, making this feature
276 compatible with Perl.
277
278 23. The tests have been re-organized, adding tests 11 and 12, to make it
279 possible to check the Perl 5.10 features against Perl 5.10.
280
281 24. Perl 5.10 allows subroutine calls in lookbehinds, as long as the subroutine
282 pattern matches a fixed length string. PCRE did not allow this; now it
283 does. Neither allows recursion.
284
285 25. I finally figured out how to implement a request to provide the minimum
286 length of subject string that was needed in order to match a given pattern.
287 (It was back references and recursion that I had previously got hung up
288 on.) This code has now been added to pcre_study(); it finds a lower bound
289 to the length of subject needed. It is not necessarily the greatest lower
290 bound, but using it to avoid searching strings that are too short does give
291 some useful speed-ups. The value is available to calling programs via
292 pcre_fullinfo().
293
294 26. While implementing 25, I discovered to my embarrassment that pcretest had
295 not been passing the result of pcre_study() to pcre_dfa_exec(), so the
296 study optimizations had never been tested with that matching function.
297 Oops. What is worse, even when it was passed study data, there was a bug in
298 pcre_dfa_exec() that meant it never actually used it. Double oops. There
299 were also very few tests of studied patterns with pcre_dfa_exec().
300
301 27. If (?| is used to create subpatterns with duplicate numbers, they are now
302 allowed to have the same name, even if PCRE_DUPNAMES is not set. However,
303 on the other side of the coin, they are no longer allowed to have different
304 names, because these cannot be distinguished in PCRE, and this has caused
305 confusion. (This is a difference from Perl.)
306
307 28. When duplicate subpattern names are present (necessarily with different
308 numbers, as required by 27 above), and a test is made by name in a
309 conditional pattern, either for a subpattern having been matched, or for
310 recursion in such a pattern, all the associated numbered subpatterns are
311 tested, and the overall condition is true if the condition is true for any
312 one of them. This is the way Perl works, and is also more like the way
313 testing by number works.
314
315
316 Version 7.9 11-Apr-09
317 ---------------------
318
319 1. When building with support for bzlib/zlib (pcregrep) and/or readline
320 (pcretest), all targets were linked against these libraries. This included
321 libpcre, libpcreposix, and libpcrecpp, even though they do not use these
322 libraries. This caused unwanted dependencies to be created. This problem
323 has been fixed, and now only pcregrep is linked with bzlib/zlib and only
324 pcretest is linked with readline.
325
326 2. The "typedef int BOOL" in pcre_internal.h that was included inside the
327 "#ifndef FALSE" condition by an earlier change (probably 7.8/18) has been
328 moved outside it again, because FALSE and TRUE are already defined in AIX,
329 but BOOL is not.
330
331 3. The pcre_config() function was treating the PCRE_MATCH_LIMIT and
332 PCRE_MATCH_LIMIT_RECURSION values as ints, when they should be long ints.
333
334 4. The pcregrep documentation said spaces were inserted as well as colons (or
335 hyphens) following file names and line numbers when outputting matching
336 lines. This is not true; no spaces are inserted. I have also clarified the
337 wording for the --colour (or --color) option.
338
339 5. In pcregrep, when --colour was used with -o, the list of matching strings
340 was not coloured; this is different to GNU grep, so I have changed it to be
341 the same.
342
343 6. When --colo(u)r was used in pcregrep, only the first matching substring in
344 each matching line was coloured. Now it goes on to look for further matches
345 of any of the test patterns, which is the same behaviour as GNU grep.
346
347 7. A pattern that could match an empty string could cause pcregrep to loop; it
348 doesn't make sense to accept an empty string match in pcregrep, so I have
349 locked it out (using PCRE's PCRE_NOTEMPTY option). By experiment, this
350 seems to be how GNU grep behaves.
351
352 8. The pattern (?(?=.*b)b|^) was incorrectly compiled as "match must be at
353 start or after a newline", because the conditional assertion was not being
354 correctly handled. The rule now is that both the assertion and what follows
355 in the first alternative must satisfy the test.
356
357 9. If auto-callout was enabled in a pattern with a conditional group whose
358 condition was an assertion, PCRE could crash during matching, both with
359 pcre_exec() and pcre_dfa_exec().
360
361 10. The PCRE_DOLLAR_ENDONLY option was not working when pcre_dfa_exec() was
362 used for matching.
363
364 11. Unicode property support in character classes was not working for
365 characters (bytes) greater than 127 when not in UTF-8 mode.
366
367 12. Added the -M command line option to pcretest.
368
369 14. Added the non-standard REG_NOTEMPTY option to the POSIX interface.
370
371 15. Added the PCRE_NO_START_OPTIMIZE match-time option.
372
373 16. Added comments and documentation about mis-use of no_arg in the C++
374 wrapper.
375
376 17. Implemented support for UTF-8 encoding in EBCDIC environments, a patch
377 from Martin Jerabek that uses macro names for all relevant character and
378 string constants.
379
380 18. Added to pcre_internal.h two configuration checks: (a) If both EBCDIC and
381 SUPPORT_UTF8 are set, give an error; (b) If SUPPORT_UCP is set without
382 SUPPORT_UTF8, define SUPPORT_UTF8. The "configure" script handles both of
383 these, but not everybody uses configure.
384
385 19. A conditional group that had only one branch was not being correctly
386 recognized as an item that could match an empty string. This meant that an
387 enclosing group might also not be so recognized, causing infinite looping
388 (and probably a segfault) for patterns such as ^"((?(?=[a])[^"])|b)*"$
389 with the subject "ab", where knowledge that the repeated group can match
390 nothing is needed in order to break the loop.
391
392 20. If a pattern that was compiled with callouts was matched using pcre_dfa_
393 exec(), but without supplying a callout function, matching went wrong.
394
395 21. If PCRE_ERROR_MATCHLIMIT occurred during a recursion, there was a memory
396 leak if the size of the offset vector was greater than 30. When the vector
397 is smaller, the saved offsets during recursion go onto a local stack
398 vector, but for larger vectors malloc() is used. It was failing to free
399 when the recursion yielded PCRE_ERROR_MATCH_LIMIT (or any other "abnormal"
400 error, in fact).
401
402 22. There was a missing #ifdef SUPPORT_UTF8 round one of the variables in the
403 heapframe that is used only when UTF-8 support is enabled. This caused no
404 problem, but was untidy.
405
406 23. Steven Van Ingelgem's patch to CMakeLists.txt to change the name
407 CMAKE_BINARY_DIR to PROJECT_BINARY_DIR so that it works when PCRE is
408 included within another project.
409
410 24. Steven Van Ingelgem's patches to add more options to the CMake support,
411 slightly modified by me:
412
413 (a) PCRE_BUILD_TESTS can be set OFF not to build the tests, including
414 not building pcregrep.
415
416 (b) PCRE_BUILD_PCREGREP can be see OFF not to build pcregrep, but only
417 if PCRE_BUILD_TESTS is also set OFF, because the tests use pcregrep.
418
419 25. Forward references, both numeric and by name, in patterns that made use of
420 duplicate group numbers, could behave incorrectly or give incorrect errors,
421 because when scanning forward to find the reference group, PCRE was not
422 taking into account the duplicate group numbers. A pattern such as
423 ^X(?3)(a)(?|(b)|(q))(Y) is an example.
424
425 26. Changed a few more instances of "const unsigned char *" to USPTR, making
426 the feature of a custom pointer more persuasive (as requested by a user).
427
428 27. Wrapped the definitions of fileno and isatty for Windows, which appear in
429 pcretest.c, inside #ifndefs, because it seems they are sometimes already
430 pre-defined.
431
432 28. Added support for (*UTF8) at the start of a pattern.
433
434 29. Arrange for flags added by the "release type" setting in CMake to be shown
435 in the configuration summary.
436
437
438 Version 7.8 05-Sep-08
439 ---------------------
440
441 1. Replaced UCP searching code with optimized version as implemented for Ad
442 Muncher (http://www.admuncher.com/) by Peter Kankowski. This uses a two-
443 stage table and inline lookup instead of a function, giving speed ups of 2
444 to 5 times on some simple patterns that I tested. Permission was given to
445 distribute the MultiStage2.py script that generates the tables (it's not in
446 the tarball, but is in the Subversion repository).
447
448 2. Updated the Unicode datatables to Unicode 5.1.0. This adds yet more
449 scripts.
450
451 3. Change 12 for 7.7 introduced a bug in pcre_study() when a pattern contained
452 a group with a zero qualifier. The result of the study could be incorrect,
453 or the function might crash, depending on the pattern.
454
455 4. Caseless matching was not working for non-ASCII characters in back
456 references. For example, /(\x{de})\1/8i was not matching \x{de}\x{fe}.
457 It now works when Unicode Property Support is available.
458
459 5. In pcretest, an escape such as \x{de} in the data was always generating
460 a UTF-8 string, even in non-UTF-8 mode. Now it generates a single byte in
461 non-UTF-8 mode. If the value is greater than 255, it gives a warning about
462 truncation.
463
464 6. Minor bugfix in pcrecpp.cc (change "" == ... to NULL == ...).
465
466 7. Added two (int) casts to pcregrep when printing the difference of two
467 pointers, in case they are 64-bit values.
468
469 8. Added comments about Mac OS X stack usage to the pcrestack man page and to
470 test 2 if it fails.
471
472 9. Added PCRE_CALL_CONVENTION just before the names of all exported functions,
473 and a #define of that name to empty if it is not externally set. This is to
474 allow users of MSVC to set it if necessary.
475
476 10. The PCRE_EXP_DEFN macro which precedes exported functions was missing from
477 the convenience functions in the pcre_get.c source file.
478
479 11. An option change at the start of a pattern that had top-level alternatives
480 could cause overwriting and/or a crash. This command provoked a crash in
481 some environments:
482
483 printf "/(?i)[\xc3\xa9\xc3\xbd]|[\xc3\xa9\xc3\xbdA]/8\n" | pcretest
484
485 This potential security problem was recorded as CVE-2008-2371.
486
487 12. For a pattern where the match had to start at the beginning or immediately
488 after a newline (e.g /.*anything/ without the DOTALL flag), pcre_exec() and
489 pcre_dfa_exec() could read past the end of the passed subject if there was
490 no match. To help with detecting such bugs (e.g. with valgrind), I modified
491 pcretest so that it places the subject at the end of its malloc-ed buffer.
492
493 13. The change to pcretest in 12 above threw up a couple more cases when pcre_
494 exec() might read past the end of the data buffer in UTF-8 mode.
495
496 14. A similar bug to 7.3/2 existed when the PCRE_FIRSTLINE option was set and
497 the data contained the byte 0x85 as part of a UTF-8 character within its
498 first line. This applied both to normal and DFA matching.
499
500 15. Lazy qualifiers were not working in some cases in UTF-8 mode. For example,
501 /^[^d]*?$/8 failed to match "abc".
502
503 16. Added a missing copyright notice to pcrecpp_internal.h.
504
505 17. Make it more clear in the documentation that values returned from
506 pcre_exec() in ovector are byte offsets, not character counts.
507
508 18. Tidied a few places to stop certain compilers from issuing warnings.
509
510 19. Updated the Virtual Pascal + BCC files to compile the latest v7.7, as
511 supplied by Stefan Weber. I made a further small update for 7.8 because
512 there is a change of source arrangements: the pcre_searchfuncs.c module is
513 replaced by pcre_ucd.c.
514
515
516 Version 7.7 07-May-08
517 ---------------------
518
519 1. Applied Craig's patch to sort out a long long problem: "If we can't convert
520 a string to a long long, pretend we don't even have a long long." This is
521 done by checking for the strtoq, strtoll, and _strtoi64 functions.
522
523 2. Applied Craig's patch to pcrecpp.cc to restore ABI compatibility with
524 pre-7.6 versions, which defined a global no_arg variable instead of putting
525 it in the RE class. (See also #8 below.)
526
527 3. Remove a line of dead code, identified by coverity and reported by Nuno
528 Lopes.
529
530 4. Fixed two related pcregrep bugs involving -r with --include or --exclude:
531
532 (1) The include/exclude patterns were being applied to the whole pathnames
533 of files, instead of just to the final components.
534
535 (2) If there was more than one level of directory, the subdirectories were
536 skipped unless they satisfied the include/exclude conditions. This is
537 inconsistent with GNU grep (and could even be seen as contrary to the
538 pcregrep specification - which I improved to make it absolutely clear).
539 The action now is always to scan all levels of directory, and just
540 apply the include/exclude patterns to regular files.
541
542 5. Added the --include_dir and --exclude_dir patterns to pcregrep, and used
543 --exclude_dir in the tests to avoid scanning .svn directories.
544
545 6. Applied Craig's patch to the QuoteMeta function so that it escapes the
546 NUL character as backslash + 0 rather than backslash + NUL, because PCRE
547 doesn't support NULs in patterns.
548
549 7. Added some missing "const"s to declarations of static tables in
550 pcre_compile.c and pcre_dfa_exec.c.
551
552 8. Applied Craig's patch to pcrecpp.cc to fix a problem in OS X that was
553 caused by fix #2 above. (Subsequently also a second patch to fix the
554 first patch. And a third patch - this was a messy problem.)
555
556 9. Applied Craig's patch to remove the use of push_back().
557
558 10. Applied Alan Lehotsky's patch to add REG_STARTEND support to the POSIX
559 matching function regexec().
560
561 11. Added support for the Oniguruma syntax \g<name>, \g<n>, \g'name', \g'n',
562 which, however, unlike Perl's \g{...}, are subroutine calls, not back
563 references. PCRE supports relative numbers with this syntax (I don't think
564 Oniguruma does).
565
566 12. Previously, a group with a zero repeat such as (...){0} was completely
567 omitted from the compiled regex. However, this means that if the group
568 was called as a subroutine from elsewhere in the pattern, things went wrong
569 (an internal error was given). Such groups are now left in the compiled
570 pattern, with a new opcode that causes them to be skipped at execution
571 time.
572
573 13. Added the PCRE_JAVASCRIPT_COMPAT option. This makes the following changes
574 to the way PCRE behaves:
575
576 (a) A lone ] character is dis-allowed (Perl treats it as data).
577
578 (b) A back reference to an unmatched subpattern matches an empty string
579 (Perl fails the current match path).
580
581 (c) A data ] in a character class must be notated as \] because if the
582 first data character in a class is ], it defines an empty class. (In
583 Perl it is not possible to have an empty class.) The empty class []
584 never matches; it forces failure and is equivalent to (*FAIL) or (?!).
585 The negative empty class [^] matches any one character, independently
586 of the DOTALL setting.
587
588 14. A pattern such as /(?2)[]a()b](abc)/ which had a forward reference to a
589 non-existent subpattern following a character class starting with ']' and
590 containing () gave an internal compiling error instead of "reference to
591 non-existent subpattern". Fortunately, when the pattern did exist, the
592 compiled code was correct. (When scanning forwards to check for the
593 existencd of the subpattern, it was treating the data ']' as terminating
594 the class, so got the count wrong. When actually compiling, the reference
595 was subsequently set up correctly.)
596
597 15. The "always fail" assertion (?!) is optimzed to (*FAIL) by pcre_compile;
598 it was being rejected as not supported by pcre_dfa_exec(), even though
599 other assertions are supported. I have made pcre_dfa_exec() support
600 (*FAIL).
601
602 16. The implementation of 13c above involved the invention of a new opcode,
603 OP_ALLANY, which is like OP_ANY but doesn't check the /s flag. Since /s
604 cannot be changed at match time, I realized I could make a small
605 improvement to matching performance by compiling OP_ALLANY instead of
606 OP_ANY for "." when DOTALL was set, and then removing the runtime tests
607 on the OP_ANY path.
608
609 17. Compiling pcretest on Windows with readline support failed without the
610 following two fixes: (1) Make the unistd.h include conditional on
611 HAVE_UNISTD_H; (2) #define isatty and fileno as _isatty and _fileno.
612
613 18. Changed CMakeLists.txt and cmake/FindReadline.cmake to arrange for the
614 ncurses library to be included for pcretest when ReadLine support is
615 requested, but also to allow for it to be overridden. This patch came from
616 Daniel Bergström.
617
618 19. There was a typo in the file ucpinternal.h where f0_rangeflag was defined
619 as 0x00f00000 instead of 0x00800000. Luckily, this would not have caused
620 any errors with the current Unicode tables. Thanks to Peter Kankowski for
621 spotting this.
622
623
624 Version 7.6 28-Jan-08
625 ---------------------
626
627 1. A character class containing a very large number of characters with
628 codepoints greater than 255 (in UTF-8 mode, of course) caused a buffer
629 overflow.
630
631 2. Patch to cut out the "long long" test in pcrecpp_unittest when
632 HAVE_LONG_LONG is not defined.
633
634 3. Applied Christian Ehrlicher's patch to update the CMake build files to
635 bring them up to date and include new features. This patch includes:
636
637 - Fixed PH's badly added libz and libbz2 support.
638 - Fixed a problem with static linking.
639 - Added pcredemo. [But later removed - see 7 below.]
640 - Fixed dftables problem and added an option.
641 - Added a number of HAVE_XXX tests, including HAVE_WINDOWS_H and
642 HAVE_LONG_LONG.
643 - Added readline support for pcretest.
644 - Added an listing of the option settings after cmake has run.
645
646 4. A user submitted a patch to Makefile that makes it easy to create
647 "pcre.dll" under mingw when using Configure/Make. I added stuff to
648 Makefile.am that cause it to include this special target, without
649 affecting anything else. Note that the same mingw target plus all
650 the other distribution libraries and programs are now supported
651 when configuring with CMake (see 6 below) instead of with
652 Configure/Make.
653
654 5. Applied Craig's patch that moves no_arg into the RE class in the C++ code.
655 This is an attempt to solve the reported problem "pcrecpp::no_arg is not
656 exported in the Windows port". It has not yet been confirmed that the patch
657 solves the problem, but it does no harm.
658
659 6. Applied Sheri's patch to CMakeLists.txt to add NON_STANDARD_LIB_PREFIX and
660 NON_STANDARD_LIB_SUFFIX for dll names built with mingw when configured
661 with CMake, and also correct the comment about stack recursion.
662
663 7. Remove the automatic building of pcredemo from the ./configure system and
664 from CMakeLists.txt. The whole idea of pcredemo.c is that it is an example
665 of a program that users should build themselves after PCRE is installed, so
666 building it automatically is not really right. What is more, it gave
667 trouble in some build environments.
668
669 8. Further tidies to CMakeLists.txt from Sheri and Christian.
670
671
672 Version 7.5 10-Jan-08
673 ---------------------
674
675 1. Applied a patch from Craig: "This patch makes it possible to 'ignore'
676 values in parens when parsing an RE using the C++ wrapper."
677
678 2. Negative specials like \S did not work in character classes in UTF-8 mode.
679 Characters greater than 255 were excluded from the class instead of being
680 included.
681
682 3. The same bug as (2) above applied to negated POSIX classes such as
683 [:^space:].
684
685 4. PCRECPP_STATIC was referenced in pcrecpp_internal.h, but nowhere was it
686 defined or documented. It seems to have been a typo for PCRE_STATIC, so
687 I have changed it.
688
689 5. The construct (?&) was not diagnosed as a syntax error (it referenced the
690 first named subpattern) and a construct such as (?&a) would reference the
691 first named subpattern whose name started with "a" (in other words, the
692 length check was missing). Both these problems are fixed. "Subpattern name
693 expected" is now given for (?&) (a zero-length name), and this patch also
694 makes it give the same error for \k'' (previously it complained that that
695 was a reference to a non-existent subpattern).
696
697 6. The erroneous patterns (?+-a) and (?-+a) give different error messages;
698 this is right because (?- can be followed by option settings as well as by
699 digits. I have, however, made the messages clearer.
700
701 7. Patterns such as (?(1)a|b) (a pattern that contains fewer subpatterns
702 than the number used in the conditional) now cause a compile-time error.
703 This is actually not compatible with Perl, which accepts such patterns, but
704 treats the conditional as always being FALSE (as PCRE used to), but it
705 seems to me that giving a diagnostic is better.
706
707 8. Change "alphameric" to the more common word "alphanumeric" in comments
708 and messages.
709
710 9. Fix two occurrences of "backslash" in comments that should have been
711 "backspace".
712
713 10. Remove two redundant lines of code that can never be obeyed (their function
714 was moved elsewhere).
715
716 11. The program that makes PCRE's Unicode character property table had a bug
717 which caused it to generate incorrect table entries for sequences of
718 characters that have the same character type, but are in different scripts.
719 It amalgamated them into a single range, with the script of the first of
720 them. In other words, some characters were in the wrong script. There were
721 thirteen such cases, affecting characters in the following ranges:
722
723 U+002b0 - U+002c1
724 U+0060c - U+0060d
725 U+0061e - U+00612
726 U+0064b - U+0065e
727 U+0074d - U+0076d
728 U+01800 - U+01805
729 U+01d00 - U+01d77
730 U+01d9b - U+01dbf
731 U+0200b - U+0200f
732 U+030fc - U+030fe
733 U+03260 - U+0327f
734 U+0fb46 - U+0fbb1
735 U+10450 - U+1049d
736
737 12. The -o option (show only the matching part of a line) for pcregrep was not
738 compatible with GNU grep in that, if there was more than one match in a
739 line, it showed only the first of them. It now behaves in the same way as
740 GNU grep.
741
742 13. If the -o and -v options were combined for pcregrep, it printed a blank
743 line for every non-matching line. GNU grep prints nothing, and pcregrep now
744 does the same. The return code can be used to tell if there were any
745 non-matching lines.
746
747 14. Added --file-offsets and --line-offsets to pcregrep.
748
749 15. The pattern (?=something)(?R) was not being diagnosed as a potentially
750 infinitely looping recursion. The bug was that positive lookaheads were not
751 being skipped when checking for a possible empty match (negative lookaheads
752 and both kinds of lookbehind were skipped).
753
754 16. Fixed two typos in the Windows-only code in pcregrep.c, and moved the
755 inclusion of <windows.h> to before rather than after the definition of
756 INVALID_FILE_ATTRIBUTES (patch from David Byron).
757
758 17. Specifying a possessive quantifier with a specific limit for a Unicode
759 character property caused pcre_compile() to compile bad code, which led at
760 runtime to PCRE_ERROR_INTERNAL (-14). Examples of patterns that caused this
761 are: /\p{Zl}{2,3}+/8 and /\p{Cc}{2}+/8. It was the possessive "+" that
762 caused the error; without that there was no problem.
763
764 18. Added --enable-pcregrep-libz and --enable-pcregrep-libbz2.
765
766 19. Added --enable-pcretest-libreadline.
767
768 20. In pcrecpp.cc, the variable 'count' was incremented twice in
769 RE::GlobalReplace(). As a result, the number of replacements returned was
770 double what it should be. I removed one of the increments, but Craig sent a
771 later patch that removed the other one (the right fix) and added unit tests
772 that check the return values (which was not done before).
773
774 21. Several CMake things:
775
776 (1) Arranged that, when cmake is used on Unix, the libraries end up with
777 the names libpcre and libpcreposix, not just pcre and pcreposix.
778
779 (2) The above change means that pcretest and pcregrep are now correctly
780 linked with the newly-built libraries, not previously installed ones.
781
782 (3) Added PCRE_SUPPORT_LIBREADLINE, PCRE_SUPPORT_LIBZ, PCRE_SUPPORT_LIBBZ2.
783
784 22. In UTF-8 mode, with newline set to "any", a pattern such as .*a.*=.b.*
785 crashed when matching a string such as a\x{2029}b (note that \x{2029} is a
786 UTF-8 newline character). The key issue is that the pattern starts .*;
787 this means that the match must be either at the beginning, or after a
788 newline. The bug was in the code for advancing after a failed match and
789 checking that the new position followed a newline. It was not taking
790 account of UTF-8 characters correctly.
791
792 23. PCRE was behaving differently from Perl in the way it recognized POSIX
793 character classes. PCRE was not treating the sequence [:...:] as a
794 character class unless the ... were all letters. Perl, however, seems to
795 allow any characters between [: and :], though of course it rejects as
796 unknown any "names" that contain non-letters, because all the known class
797 names consist only of letters. Thus, Perl gives an error for [[:1234:]],
798 for example, whereas PCRE did not - it did not recognize a POSIX character
799 class. This seemed a bit dangerous, so the code has been changed to be
800 closer to Perl. The behaviour is not identical to Perl, because PCRE will
801 diagnose an unknown class for, for example, [[:l\ower:]] where Perl will
802 treat it as [[:lower:]]. However, PCRE does now give "unknown" errors where
803 Perl does, and where it didn't before.
804
805 24. Rewrite so as to remove the single use of %n from pcregrep because in some
806 Windows environments %n is disabled by default.
807
808
809 Version 7.4 21-Sep-07
810 ---------------------
811
812 1. Change 7.3/28 was implemented for classes by looking at the bitmap. This
813 means that a class such as [\s] counted as "explicit reference to CR or
814 LF". That isn't really right - the whole point of the change was to try to
815 help when there was an actual mention of one of the two characters. So now
816 the change happens only if \r or \n (or a literal CR or LF) character is
817 encountered.
818
819 2. The 32-bit options word was also used for 6 internal flags, but the numbers
820 of both had grown to the point where there were only 3 bits left.
821 Fortunately, there was spare space in the data structure, and so I have
822 moved the internal flags into a new 16-bit field to free up more option
823 bits.
824
825 3. The appearance of (?J) at the start of a pattern set the DUPNAMES option,
826 but did not set the internal JCHANGED flag - either of these is enough to
827 control the way the "get" function works - but the PCRE_INFO_JCHANGED
828 facility is supposed to tell if (?J) was ever used, so now (?J) at the
829 start sets both bits.
830
831 4. Added options (at build time, compile time, exec time) to change \R from
832 matching any Unicode line ending sequence to just matching CR, LF, or CRLF.
833
834 5. doc/pcresyntax.html was missing from the distribution.
835
836 6. Put back the definition of PCRE_ERROR_NULLWSLIMIT, for backward
837 compatibility, even though it is no longer used.
838
839 7. Added macro for snprintf to pcrecpp_unittest.cc and also for strtoll and
840 strtoull to pcrecpp.cc to select the available functions in WIN32 when the
841 windows.h file is present (where different names are used). [This was
842 reversed later after testing - see 16 below.]
843
844 8. Changed all #include <config.h> to #include "config.h". There were also
845 some further <pcre.h> cases that I changed to "pcre.h".
846
847 9. When pcregrep was used with the --colour option, it missed the line ending
848 sequence off the lines that it output.
849
850 10. It was pointed out to me that arrays of string pointers cause lots of
851 relocations when a shared library is dynamically loaded. A technique of
852 using a single long string with a table of offsets can drastically reduce
853 these. I have refactored PCRE in four places to do this. The result is
854 dramatic:
855
856 Originally: 290
857 After changing UCP table: 187
858 After changing error message table: 43
859 After changing table of "verbs" 36
860 After changing table of Posix names 22
861
862 Thanks to the folks working on Gregex for glib for this insight.
863
864 11. --disable-stack-for-recursion caused compiling to fail unless -enable-
865 unicode-properties was also set.
866
867 12. Updated the tests so that they work when \R is defaulted to ANYCRLF.
868
869 13. Added checks for ANY and ANYCRLF to pcrecpp.cc where it previously
870 checked only for CRLF.
871
872 14. Added casts to pcretest.c to avoid compiler warnings.
873
874 15. Added Craig's patch to various pcrecpp modules to avoid compiler warnings.
875
876 16. Added Craig's patch to remove the WINDOWS_H tests, that were not working,
877 and instead check for _strtoi64 explicitly, and avoid the use of snprintf()
878 entirely. This removes changes made in 7 above.
879
880 17. The CMake files have been updated, and there is now more information about
881 building with CMake in the NON-UNIX-USE document.
882
883
884 Version 7.3 28-Aug-07
885 ---------------------
886
887 1. In the rejigging of the build system that eventually resulted in 7.1, the
888 line "#include <pcre.h>" was included in pcre_internal.h. The use of angle
889 brackets there is not right, since it causes compilers to look for an
890 installed pcre.h, not the version that is in the source that is being
891 compiled (which of course may be different). I have changed it back to:
892
893 #include "pcre.h"
894
895 I have a vague recollection that the change was concerned with compiling in
896 different directories, but in the new build system, that is taken care of
897 by the VPATH setting the Makefile.
898
899 2. The pattern .*$ when run in not-DOTALL UTF-8 mode with newline=any failed
900 when the subject happened to end in the byte 0x85 (e.g. if the last
901 character was \x{1ec5}). *Character* 0x85 is one of the "any" newline
902 characters but of course it shouldn't be taken as a newline when it is part
903 of another character. The bug was that, for an unlimited repeat of . in
904 not-DOTALL UTF-8 mode, PCRE was advancing by bytes rather than by
905 characters when looking for a newline.
906
907 3. A small performance improvement in the DOTALL UTF-8 mode .* case.
908
909 4. Debugging: adjusted the names of opcodes for different kinds of parentheses
910 in debug output.
911
912 5. Arrange to use "%I64d" instead of "%lld" and "%I64u" instead of "%llu" for
913 long printing in the pcrecpp unittest when running under MinGW.
914
915 6. ESC_K was left out of the EBCDIC table.
916
917 7. Change 7.0/38 introduced a new limit on the number of nested non-capturing
918 parentheses; I made it 1000, which seemed large enough. Unfortunately, the
919 limit also applies to "virtual nesting" when a pattern is recursive, and in
920 this case 1000 isn't so big. I have been able to remove this limit at the
921 expense of backing off one optimization in certain circumstances. Normally,
922 when pcre_exec() would call its internal match() function recursively and
923 immediately return the result unconditionally, it uses a "tail recursion"
924 feature to save stack. However, when a subpattern that can match an empty
925 string has an unlimited repetition quantifier, it no longer makes this
926 optimization. That gives it a stack frame in which to save the data for
927 checking that an empty string has been matched. Previously this was taken
928 from the 1000-entry workspace that had been reserved. So now there is no
929 explicit limit, but more stack is used.
930
931 8. Applied Daniel's patches to solve problems with the import/export magic
932 syntax that is required for Windows, and which was going wrong for the
933 pcreposix and pcrecpp parts of the library. These were overlooked when this
934 problem was solved for the main library.
935
936 9. There were some crude static tests to avoid integer overflow when computing
937 the size of patterns that contain repeated groups with explicit upper
938 limits. As the maximum quantifier is 65535, the maximum group length was
939 set at 30,000 so that the product of these two numbers did not overflow a
940 32-bit integer. However, it turns out that people want to use groups that
941 are longer than 30,000 bytes (though not repeat them that many times).
942 Change 7.0/17 (the refactoring of the way the pattern size is computed) has
943 made it possible to implement the integer overflow checks in a much more
944 dynamic way, which I have now done. The artificial limitation on group
945 length has been removed - we now have only the limit on the total length of
946 the compiled pattern, which depends on the LINK_SIZE setting.
947
948 10. Fixed a bug in the documentation for get/copy named substring when
949 duplicate names are permitted. If none of the named substrings are set, the
950 functions return PCRE_ERROR_NOSUBSTRING (7); the doc said they returned an
951 empty string.
952
953 11. Because Perl interprets \Q...\E at a high level, and ignores orphan \E
954 instances, patterns such as [\Q\E] or [\E] or even [^\E] cause an error,
955 because the ] is interpreted as the first data character and the
956 terminating ] is not found. PCRE has been made compatible with Perl in this
957 regard. Previously, it interpreted [\Q\E] as an empty class, and [\E] could
958 cause memory overwriting.
959
960 10. Like Perl, PCRE automatically breaks an unlimited repeat after an empty
961 string has been matched (to stop an infinite loop). It was not recognizing
962 a conditional subpattern that could match an empty string if that
963 subpattern was within another subpattern. For example, it looped when
964 trying to match (((?(1)X|))*) but it was OK with ((?(1)X|)*) where the
965 condition was not nested. This bug has been fixed.
966
967 12. A pattern like \X?\d or \P{L}?\d in non-UTF-8 mode could cause a backtrack
968 past the start of the subject in the presence of bytes with the top bit
969 set, for example "\x8aBCD".
970
971 13. Added Perl 5.10 experimental backtracking controls (*FAIL), (*F), (*PRUNE),
972 (*SKIP), (*THEN), (*COMMIT), and (*ACCEPT).
973
974 14. Optimized (?!) to (*FAIL).
975
976 15. Updated the test for a valid UTF-8 string to conform to the later RFC 3629.
977 This restricts code points to be within the range 0 to 0x10FFFF, excluding
978 the "low surrogate" sequence 0xD800 to 0xDFFF. Previously, PCRE allowed the
979 full range 0 to 0x7FFFFFFF, as defined by RFC 2279. Internally, it still
980 does: it's just the validity check that is more restrictive.
981
982 16. Inserted checks for integer overflows during escape sequence (backslash)
983 processing, and also fixed erroneous offset values for syntax errors during
984 backslash processing.
985
986 17. Fixed another case of looking too far back in non-UTF-8 mode (cf 12 above)
987 for patterns like [\PPP\x8a]{1,}\x80 with the subject "A\x80".
988
989 18. An unterminated class in a pattern like (?1)\c[ with a "forward reference"
990 caused an overrun.
991
992 19. A pattern like (?:[\PPa*]*){8,} which had an "extended class" (one with
993 something other than just ASCII characters) inside a group that had an
994 unlimited repeat caused a loop at compile time (while checking to see
995 whether the group could match an empty string).
996
997 20. Debugging a pattern containing \p or \P could cause a crash. For example,
998 [\P{Any}] did so. (Error in the code for printing property names.)
999
1000 21. An orphan \E inside a character class could cause a crash.
1001
1002 22. A repeated capturing bracket such as (A)? could cause a wild memory
1003 reference during compilation.
1004
1005 23. There are several functions in pcre_compile() that scan along a compiled
1006 expression for various reasons (e.g. to see if it's fixed length for look
1007 behind). There were bugs in these functions when a repeated \p or \P was
1008 present in the pattern. These operators have additional parameters compared
1009 with \d, etc, and these were not being taken into account when moving along
1010 the compiled data. Specifically:
1011
1012 (a) A item such as \p{Yi}{3} in a lookbehind was not treated as fixed
1013 length.
1014
1015 (b) An item such as \pL+ within a repeated group could cause crashes or
1016 loops.
1017
1018 (c) A pattern such as \p{Yi}+(\P{Yi}+)(?1) could give an incorrect
1019 "reference to non-existent subpattern" error.
1020
1021 (d) A pattern like (\P{Yi}{2}\277)? could loop at compile time.
1022
1023 24. A repeated \S or \W in UTF-8 mode could give wrong answers when multibyte
1024 characters were involved (for example /\S{2}/8g with "A\x{a3}BC").
1025
1026 25. Using pcregrep in multiline, inverted mode (-Mv) caused it to loop.
1027
1028 26. Patterns such as [\P{Yi}A] which include \p or \P and just one other
1029 character were causing crashes (broken optimization).
1030
1031 27. Patterns such as (\P{Yi}*\277)* (group with possible zero repeat containing
1032 \p or \P) caused a compile-time loop.
1033
1034 28. More problems have arisen in unanchored patterns when CRLF is a valid line
1035 break. For example, the unstudied pattern [\r\n]A does not match the string
1036 "\r\nA" because change 7.0/46 below moves the current point on by two
1037 characters after failing to match at the start. However, the pattern \nA
1038 *does* match, because it doesn't start till \n, and if [\r\n]A is studied,
1039 the same is true. There doesn't seem any very clean way out of this, but
1040 what I have chosen to do makes the common cases work: PCRE now takes note
1041 of whether there can be an explicit match for \r or \n anywhere in the
1042 pattern, and if so, 7.0/46 no longer applies. As part of this change,
1043 there's a new PCRE_INFO_HASCRORLF option for finding out whether a compiled
1044 pattern has explicit CR or LF references.
1045
1046 29. Added (*CR) etc for changing newline setting at start of pattern.
1047
1048
1049 Version 7.2 19-Jun-07
1050 ---------------------
1051
1052 1. If the fr_FR locale cannot be found for test 3, try the "french" locale,
1053 which is apparently normally available under Windows.
1054
1055 2. Re-jig the pcregrep tests with different newline settings in an attempt
1056 to make them independent of the local environment's newline setting.
1057
1058 3. Add code to configure.ac to remove -g from the CFLAGS default settings.
1059
1060 4. Some of the "internals" tests were previously cut out when the link size
1061 was not 2, because the output contained actual offsets. The recent new
1062 "Z" feature of pcretest means that these can be cut out, making the tests
1063 usable with all link sizes.
1064
1065 5. Implemented Stan Switzer's goto replacement for longjmp() when not using
1066 stack recursion. This gives a massive performance boost under BSD, but just
1067 a small improvement under Linux. However, it saves one field in the frame
1068 in all cases.
1069
1070 6. Added more features from the forthcoming Perl 5.10:
1071
1072 (a) (?-n) (where n is a string of digits) is a relative subroutine or
1073 recursion call. It refers to the nth most recently opened parentheses.
1074
1075 (b) (?+n) is also a relative subroutine call; it refers to the nth next
1076 to be opened parentheses.
1077
1078 (c) Conditions that refer to capturing parentheses can be specified
1079 relatively, for example, (?(-2)... or (?(+3)...
1080
1081 (d) \K resets the start of the current match so that everything before
1082 is not part of it.
1083
1084 (e) \k{name} is synonymous with \k<name> and \k'name' (.NET compatible).
1085
1086 (f) \g{name} is another synonym - part of Perl 5.10's unification of
1087 reference syntax.
1088
1089 (g) (?| introduces a group in which the numbering of parentheses in each
1090 alternative starts with the same number.
1091
1092 (h) \h, \H, \v, and \V match horizontal and vertical whitespace.
1093
1094 7. Added two new calls to pcre_fullinfo(): PCRE_INFO_OKPARTIAL and
1095 PCRE_INFO_JCHANGED.
1096
1097 8. A pattern such as (.*(.)?)* caused pcre_exec() to fail by either not
1098 terminating or by crashing. Diagnosed by Viktor Griph; it was in the code
1099 for detecting groups that can match an empty string.
1100
1101 9. A pattern with a very large number of alternatives (more than several
1102 hundred) was running out of internal workspace during the pre-compile
1103 phase, where pcre_compile() figures out how much memory will be needed. A
1104 bit of new cunning has reduced the workspace needed for groups with
1105 alternatives. The 1000-alternative test pattern now uses 12 bytes of
1106 workspace instead of running out of the 4096 that are available.
1107
1108 10. Inserted some missing (unsigned int) casts to get rid of compiler warnings.
1109
1110 11. Applied patch from Google to remove an optimization that didn't quite work.
1111 The report of the bug said:
1112
1113 pcrecpp::RE("a*").FullMatch("aaa") matches, while
1114 pcrecpp::RE("a*?").FullMatch("aaa") does not, and
1115 pcrecpp::RE("a*?\\z").FullMatch("aaa") does again.
1116
1117 12. If \p or \P was used in non-UTF-8 mode on a character greater than 127
1118 it matched the wrong number of bytes.
1119
1120
1121 Version 7.1 24-Apr-07
1122 ---------------------
1123
1124 1. Applied Bob Rossi and Daniel G's patches to convert the build system to one
1125 that is more "standard", making use of automake and other Autotools. There
1126 is some re-arrangement of the files and adjustment of comments consequent
1127 on this.
1128
1129 2. Part of the patch fixed a problem with the pcregrep tests. The test of -r
1130 for recursive directory scanning broke on some systems because the files
1131 are not scanned in any specific order and on different systems the order
1132 was different. A call to "sort" has been inserted into RunGrepTest for the
1133 approprate test as a short-term fix. In the longer term there may be an
1134 alternative.
1135
1136 3. I had an email from Eric Raymond about problems translating some of PCRE's
1137 man pages to HTML (despite the fact that I distribute HTML pages, some
1138 people do their own conversions for various reasons). The problems
1139 concerned the use of low-level troff macros .br and .in. I have therefore
1140 removed all such uses from the man pages (some were redundant, some could
1141 be replaced by .nf/.fi pairs). The 132html script that I use to generate
1142 HTML has been updated to handle .nf/.fi and to complain if it encounters
1143 .br or .in.
1144
1145 4. Updated comments in configure.ac that get placed in config.h.in and also
1146 arranged for config.h to be included in the distribution, with the name
1147 config.h.generic, for the benefit of those who have to compile without
1148 Autotools (compare pcre.h, which is now distributed as pcre.h.generic).
1149
1150 5. Updated the support (such as it is) for Virtual Pascal, thanks to Stefan
1151 Weber: (1) pcre_internal.h was missing some function renames; (2) updated
1152 makevp.bat for the current PCRE, using the additional files
1153 makevp_c.txt, makevp_l.txt, and pcregexp.pas.
1154
1155 6. A Windows user reported a minor discrepancy with test 2, which turned out
1156 to be caused by a trailing space on an input line that had got lost in his
1157 copy. The trailing space was an accident, so I've just removed it.
1158
1159 7. Add -Wl,-R... flags in pcre-config.in for *BSD* systems, as I'm told
1160 that is needed.
1161
1162 8. Mark ucp_table (in ucptable.h) and ucp_gentype (in pcre_ucp_searchfuncs.c)
1163 as "const" (a) because they are and (b) because it helps the PHP
1164 maintainers who have recently made a script to detect big data structures
1165 in the php code that should be moved to the .rodata section. I remembered
1166 to update Builducptable as well, so it won't revert if ucptable.h is ever
1167 re-created.
1168
1169 9. Added some extra #ifdef SUPPORT_UTF8 conditionals into pcretest.c,
1170 pcre_printint.src, pcre_compile.c, pcre_study.c, and pcre_tables.c, in
1171 order to be able to cut out the UTF-8 tables in the latter when UTF-8
1172 support is not required. This saves 1.5-2K of code, which is important in
1173 some applications.
1174
1175 Later: more #ifdefs are needed in pcre_ord2utf8.c and pcre_valid_utf8.c
1176 so as not to refer to the tables, even though these functions will never be
1177 called when UTF-8 support is disabled. Otherwise there are problems with a
1178 shared library.
1179
1180 10. Fixed two bugs in the emulated memmove() function in pcre_internal.h:
1181
1182 (a) It was defining its arguments as char * instead of void *.
1183
1184 (b) It was assuming that all moves were upwards in memory; this was true
1185 a long time ago when I wrote it, but is no longer the case.
1186
1187 The emulated memove() is provided for those environments that have neither
1188 memmove() nor bcopy(). I didn't think anyone used it these days, but that
1189 is clearly not the case, as these two bugs were recently reported.
1190
1191 11. The script PrepareRelease is now distributed: it calls 132html, CleanTxt,
1192 and Detrail to create the HTML documentation, the .txt form of the man
1193 pages, and it removes trailing spaces from listed files. It also creates
1194 pcre.h.generic and config.h.generic from pcre.h and config.h. In the latter
1195 case, it wraps all the #defines with #ifndefs. This script should be run
1196 before "make dist".
1197
1198 12. Fixed two fairly obscure bugs concerned with quantified caseless matching
1199 with Unicode property support.
1200
1201 (a) For a maximizing quantifier, if the two different cases of the
1202 character were of different lengths in their UTF-8 codings (there are
1203 some cases like this - I found 11), and the matching function had to
1204 back up over a mixture of the two cases, it incorrectly assumed they
1205 were both the same length.
1206
1207 (b) When PCRE was configured to use the heap rather than the stack for
1208 recursion during matching, it was not correctly preserving the data for
1209 the other case of a UTF-8 character when checking ahead for a match
1210 while processing a minimizing repeat. If the check also involved
1211 matching a wide character, but failed, corruption could cause an
1212 erroneous result when trying to check for a repeat of the original
1213 character.
1214
1215 13. Some tidying changes to the testing mechanism:
1216
1217 (a) The RunTest script now detects the internal link size and whether there
1218 is UTF-8 and UCP support by running ./pcretest -C instead of relying on
1219 values substituted by "configure". (The RunGrepTest script already did
1220 this for UTF-8.) The configure.ac script no longer substitutes the
1221 relevant variables.
1222
1223 (b) The debugging options /B and /D in pcretest show the compiled bytecode
1224 with length and offset values. This means that the output is different
1225 for different internal link sizes. Test 2 is skipped for link sizes
1226 other than 2 because of this, bypassing the problem. Unfortunately,
1227 there was also a test in test 3 (the locale tests) that used /B and
1228 failed for link sizes other than 2. Rather than cut the whole test out,
1229 I have added a new /Z option to pcretest that replaces the length and
1230 offset values with spaces. This is now used to make test 3 independent
1231 of link size. (Test 2 will be tidied up later.)
1232
1233 14. If erroroffset was passed as NULL to pcre_compile, it provoked a
1234 segmentation fault instead of returning the appropriate error message.
1235
1236 15. In multiline mode when the newline sequence was set to "any", the pattern
1237 ^$ would give a match between the \r and \n of a subject such as "A\r\nB".
1238 This doesn't seem right; it now treats the CRLF combination as the line
1239 ending, and so does not match in that case. It's only a pattern such as ^$
1240 that would hit this one: something like ^ABC$ would have failed after \r
1241 and then tried again after \r\n.
1242
1243 16. Changed the comparison command for RunGrepTest from "diff -u" to "diff -ub"
1244 in an attempt to make files that differ only in their line terminators
1245 compare equal. This works on Linux.
1246
1247 17. Under certain error circumstances pcregrep might try to free random memory
1248 as it exited. This is now fixed, thanks to valgrind.
1249
1250 19. In pcretest, if the pattern /(?m)^$/g<any> was matched against the string
1251 "abc\r\n\r\n", it found an unwanted second match after the second \r. This
1252 was because its rules for how to advance for /g after matching an empty
1253 string at the end of a line did not allow for this case. They now check for
1254 it specially.
1255
1256 20. pcretest is supposed to handle patterns and data of any length, by
1257 extending its buffers when necessary. It was getting this wrong when the
1258 buffer for a data line had to be extended.
1259
1260 21. Added PCRE_NEWLINE_ANYCRLF which is like ANY, but matches only CR, LF, or
1261 CRLF as a newline sequence.
1262
1263 22. Code for handling Unicode properties in pcre_dfa_exec() wasn't being cut
1264 out by #ifdef SUPPORT_UCP. This did no harm, as it could never be used, but
1265 I have nevertheless tidied it up.
1266
1267 23. Added some casts to kill warnings from HP-UX ia64 compiler.
1268
1269 24. Added a man page for pcre-config.
1270
1271
1272 Version 7.0 19-Dec-06
1273 ---------------------
1274
1275 1. Fixed a signed/unsigned compiler warning in pcre_compile.c, shown up by
1276 moving to gcc 4.1.1.
1277
1278 2. The -S option for pcretest uses setrlimit(); I had omitted to #include
1279 sys/time.h, which is documented as needed for this function. It doesn't
1280 seem to matter on Linux, but it showed up on some releases of OS X.
1281
1282 3. It seems that there are systems where bytes whose values are greater than
1283 127 match isprint() in the "C" locale. The "C" locale should be the
1284 default when a C program starts up. In most systems, only ASCII printing
1285 characters match isprint(). This difference caused the output from pcretest
1286 to vary, making some of the tests fail. I have changed pcretest so that:
1287
1288 (a) When it is outputting text in the compiled version of a pattern, bytes
1289 other than 32-126 are always shown as hex escapes.
1290
1291 (b) When it is outputting text that is a matched part of a subject string,
1292 it does the same, unless a different locale has been set for the match
1293 (using the /L modifier). In this case, it uses isprint() to decide.
1294
1295 4. Fixed a major bug that caused incorrect computation of the amount of memory
1296 required for a compiled pattern when options that changed within the
1297 pattern affected the logic of the preliminary scan that determines the
1298 length. The relevant options are -x, and -i in UTF-8 mode. The result was
1299 that the computed length was too small. The symptoms of this bug were
1300 either the PCRE error "internal error: code overflow" from pcre_compile(),
1301 or a glibc crash with a message such as "pcretest: free(): invalid next
1302 size (fast)". Examples of patterns that provoked this bug (shown in
1303 pcretest format) are:
1304
1305 /(?-x: )/x
1306 /(?x)(?-x: \s*#\s*)/
1307 /((?i)[\x{c0}])/8
1308 /(?i:[\x{c0}])/8
1309
1310 HOWEVER: Change 17 below makes this fix obsolete as the memory computation
1311 is now done differently.
1312
1313 5. Applied patches from Google to: (a) add a QuoteMeta function to the C++
1314 wrapper classes; (b) implement a new function in the C++ scanner that is
1315 more efficient than the old way of doing things because it avoids levels of
1316 recursion in the regex matching; (c) add a paragraph to the documentation
1317 for the FullMatch() function.
1318
1319 6. The escape sequence \n was being treated as whatever was defined as
1320 "newline". Not only was this contrary to the documentation, which states
1321 that \n is character 10 (hex 0A), but it also went horribly wrong when
1322 "newline" was defined as CRLF. This has been fixed.
1323
1324 7. In pcre_dfa_exec.c the value of an unsigned integer (the variable called c)
1325 was being set to -1 for the "end of line" case (supposedly a value that no
1326 character can have). Though this value is never used (the check for end of
1327 line is "zero bytes in current character"), it caused compiler complaints.
1328 I've changed it to 0xffffffff.
1329
1330 8. In pcre_version.c, the version string was being built by a sequence of
1331 C macros that, in the event of PCRE_PRERELEASE being defined as an empty
1332 string (as it is for production releases) called a macro with an empty
1333 argument. The C standard says the result of this is undefined. The gcc
1334 compiler treats it as an empty string (which was what was wanted) but it is
1335 reported that Visual C gives an error. The source has been hacked around to
1336 avoid this problem.
1337
1338 9. On the advice of a Windows user, included <io.h> and <fcntl.h> in Windows
1339 builds of pcretest, and changed the call to _setmode() to use _O_BINARY
1340 instead of 0x8000. Made all the #ifdefs test both _WIN32 and WIN32 (not all
1341 of them did).
1342
1343 10. Originally, pcretest opened its input and output without "b"; then I was
1344 told that "b" was needed in some environments, so it was added for release
1345 5.0 to both the input and output. (It makes no difference on Unix-like
1346 systems.) Later I was told that it is wrong for the input on Windows. I've
1347 now abstracted the modes into two macros, to make it easier to fiddle with
1348 them, and removed "b" from the input mode under Windows.
1349
1350 11. Added pkgconfig support for the C++ wrapper library, libpcrecpp.
1351
1352 12. Added -help and --help to pcretest as an official way of being reminded
1353 of the options.
1354
1355 13. Removed some redundant semicolons after macro calls in pcrecpparg.h.in
1356 and pcrecpp.cc because they annoy compilers at high warning levels.
1357
1358 14. A bit of tidying/refactoring in pcre_exec.c in the main bumpalong loop.
1359
1360 15. Fixed an occurrence of == in configure.ac that should have been = (shell
1361 scripts are not C programs :-) and which was not noticed because it works
1362 on Linux.
1363
1364 16. pcretest is supposed to handle any length of pattern and data line (as one
1365 line or as a continued sequence of lines) by extending its input buffer if
1366 necessary. This feature was broken for very long pattern lines, leading to
1367 a string of junk being passed to pcre_compile() if the pattern was longer
1368 than about 50K.
1369
1370 17. I have done a major re-factoring of the way pcre_compile() computes the
1371 amount of memory needed for a compiled pattern. Previously, there was code
1372 that made a preliminary scan of the pattern in order to do this. That was
1373 OK when PCRE was new, but as the facilities have expanded, it has become
1374 harder and harder to keep it in step with the real compile phase, and there
1375 have been a number of bugs (see for example, 4 above). I have now found a
1376 cunning way of running the real compile function in a "fake" mode that
1377 enables it to compute how much memory it would need, while actually only
1378 ever using a few hundred bytes of working memory and without too many
1379 tests of the mode. This should make future maintenance and development
1380 easier. A side effect of this work is that the limit of 200 on the nesting
1381 depth of parentheses has been removed (though this was never a serious
1382 limitation, I suspect). However, there is a downside: pcre_compile() now
1383 runs more slowly than before (30% or more, depending on the pattern). I
1384 hope this isn't a big issue. There is no effect on runtime performance.
1385
1386 18. Fixed a minor bug in pcretest: if a pattern line was not terminated by a
1387 newline (only possible for the last line of a file) and it was a
1388 pattern that set a locale (followed by /Lsomething), pcretest crashed.
1389
1390 19. Added additional timing features to pcretest. (1) The -tm option now times
1391 matching only, not compiling. (2) Both -t and -tm can be followed, as a
1392 separate command line item, by a number that specifies the number of
1393 repeats to use when timing. The default is 50000; this gives better
1394 precision, but takes uncomfortably long for very large patterns.
1395
1396 20. Extended pcre_study() to be more clever in cases where a branch of a
1397 subpattern has no definite first character. For example, (a*|b*)[cd] would
1398 previously give no result from pcre_study(). Now it recognizes that the
1399 first character must be a, b, c, or d.
1400
1401 21. There was an incorrect error "recursive call could loop indefinitely" if
1402 a subpattern (or the entire pattern) that was being tested for matching an
1403 empty string contained only one non-empty item after a nested subpattern.
1404 For example, the pattern (?>\x{100}*)\d(?R) provoked this error
1405 incorrectly, because the \d was being skipped in the check.
1406
1407 22. The pcretest program now has a new pattern option /B and a command line
1408 option -b, which is equivalent to adding /B to every pattern. This causes
1409 it to show the compiled bytecode, without the additional information that
1410 -d shows. The effect of -d is now the same as -b with -i (and similarly, /D
1411 is the same as /B/I).
1412
1413 23. A new optimization is now able automatically to treat some sequences such
1414 as a*b as a*+b. More specifically, if something simple (such as a character
1415 or a simple class like \d) has an unlimited quantifier, and is followed by
1416 something that cannot possibly match the quantified thing, the quantifier
1417 is automatically "possessified".
1418
1419 24. A recursive reference to a subpattern whose number was greater than 39
1420 went wrong under certain circumstances in UTF-8 mode. This bug could also
1421 have affected the operation of pcre_study().
1422
1423 25. Realized that a little bit of performance could be had by replacing
1424 (c & 0xc0) == 0xc0 with c >= 0xc0 when processing UTF-8 characters.
1425
1426 26. Timing data from pcretest is now shown to 4 decimal places instead of 3.
1427
1428 27. Possessive quantifiers such as a++ were previously implemented by turning
1429 them into atomic groups such as ($>a+). Now they have their own opcodes,
1430 which improves performance. This includes the automatically created ones
1431 from 23 above.
1432
1433 28. A pattern such as (?=(\w+))\1: which simulates an atomic group using a
1434 lookahead was broken if it was not anchored. PCRE was mistakenly expecting
1435 the first matched character to be a colon. This applied both to named and
1436 numbered groups.
1437
1438 29. The ucpinternal.h header file was missing its idempotency #ifdef.
1439
1440 30. I was sent a "project" file called libpcre.a.dev which I understand makes
1441 building PCRE on Windows easier, so I have included it in the distribution.
1442
1443 31. There is now a check in pcretest against a ridiculously large number being
1444 returned by pcre_exec() or pcre_dfa_exec(). If this happens in a /g or /G
1445 loop, the loop is abandoned.
1446
1447 32. Forward references to subpatterns in conditions such as (?(2)...) where
1448 subpattern 2 is defined later cause pcre_compile() to search forwards in
1449 the pattern for the relevant set of parentheses. This search went wrong
1450 when there were unescaped parentheses in a character class, parentheses
1451 escaped with \Q...\E, or parentheses in a #-comment in /x mode.
1452
1453 33. "Subroutine" calls and backreferences were previously restricted to
1454 referencing subpatterns earlier in the regex. This restriction has now
1455 been removed.
1456
1457 34. Added a number of extra features that are going to be in Perl 5.10. On the
1458 whole, these are just syntactic alternatives for features that PCRE had
1459 previously implemented using the Python syntax or my own invention. The
1460 other formats are all retained for compatibility.
1461
1462 (a) Named groups can now be defined as (?<name>...) or (?'name'...) as well
1463 as (?P<name>...). The new forms, as well as being in Perl 5.10, are
1464 also .NET compatible.
1465
1466 (b) A recursion or subroutine call to a named group can now be defined as
1467 (?&name) as well as (?P>name).
1468
1469 (c) A backreference to a named group can now be defined as \k<name> or
1470 \k'name' as well as (?P=name). The new forms, as well as being in Perl
1471 5.10, are also .NET compatible.
1472
1473 (d) A conditional reference to a named group can now use the syntax
1474 (?(<name>) or (?('name') as well as (?(name).
1475
1476 (e) A "conditional group" of the form (?(DEFINE)...) can be used to define
1477 groups (named and numbered) that are never evaluated inline, but can be
1478 called as "subroutines" from elsewhere. In effect, the DEFINE condition
1479 is always false. There may be only one alternative in such a group.
1480
1481 (f) A test for recursion can be given as (?(R1).. or (?(R&name)... as well
1482 as the simple (?(R). The condition is true only if the most recent
1483 recursion is that of the given number or name. It does not search out
1484 through the entire recursion stack.
1485
1486 (g) The escape \gN or \g{N} has been added, where N is a positive or
1487 negative number, specifying an absolute or relative reference.
1488
1489 35. Tidied to get rid of some further signed/unsigned compiler warnings and
1490 some "unreachable code" warnings.
1491
1492 36. Updated the Unicode property tables to Unicode version 5.0.0. Amongst other
1493 things, this adds five new scripts.
1494
1495 37. Perl ignores orphaned \E escapes completely. PCRE now does the same.
1496 There were also incompatibilities regarding the handling of \Q..\E inside
1497 character classes, for example with patterns like [\Qa\E-\Qz\E] where the
1498 hyphen was adjacent to \Q or \E. I hope I've cleared all this up now.
1499
1500 38. Like Perl, PCRE detects when an indefinitely repeated parenthesized group
1501 matches an empty string, and forcibly breaks the loop. There were bugs in
1502 this code in non-simple cases. For a pattern such as ^(a()*)* matched
1503 against aaaa the result was just "a" rather than "aaaa", for example. Two
1504 separate and independent bugs (that affected different cases) have been
1505 fixed.
1506
1507 39. Refactored the code to abolish the use of different opcodes for small
1508 capturing bracket numbers. This is a tidy that I avoided doing when I
1509 removed the limit on the number of capturing brackets for 3.5 back in 2001.
1510 The new approach is not only tidier, it makes it possible to reduce the
1511 memory needed to fix the previous bug (38).
1512
1513 40. Implemented PCRE_NEWLINE_ANY to recognize any of the Unicode newline
1514 sequences (http://unicode.org/unicode/reports/tr18/) as "newline" when
1515 processing dot, circumflex, or dollar metacharacters, or #-comments in /x
1516 mode.
1517
1518 41. Add \R to match any Unicode newline sequence, as suggested in the Unicode
1519 report.
1520
1521 42. Applied patch, originally from Ari Pollak, modified by Google, to allow
1522 copy construction and assignment in the C++ wrapper.
1523
1524 43. Updated pcregrep to support "--newline=any". In the process, I fixed a
1525 couple of bugs that could have given wrong results in the "--newline=crlf"
1526 case.
1527
1528 44. Added a number of casts and did some reorganization of signed/unsigned int
1529 variables following suggestions from Dair Grant. Also renamed the variable
1530 "this" as "item" because it is a C++ keyword.
1531
1532 45. Arranged for dftables to add
1533
1534 #include "pcre_internal.h"
1535
1536 to pcre_chartables.c because without it, gcc 4.x may remove the array
1537 definition from the final binary if PCRE is built into a static library and
1538 dead code stripping is activated.
1539
1540 46. For an unanchored pattern, if a match attempt fails at the start of a
1541 newline sequence, and the newline setting is CRLF or ANY, and the next two
1542 characters are CRLF, advance by two characters instead of one.
1543
1544
1545 Version 6.7 04-Jul-06
1546 ---------------------
1547
1548 1. In order to handle tests when input lines are enormously long, pcretest has
1549 been re-factored so that it automatically extends its buffers when
1550 necessary. The code is crude, but this _is_ just a test program. The
1551 default size has been increased from 32K to 50K.
1552
1553 2. The code in pcre_study() was using the value of the re argument before
1554 testing it for NULL. (Of course, in any sensible call of the function, it
1555 won't be NULL.)
1556
1557 3. The memmove() emulation function in pcre_internal.h, which is used on
1558 systems that lack both memmove() and bcopy() - that is, hardly ever -
1559 was missing a "static" storage class specifier.
1560
1561 4. When UTF-8 mode was not set, PCRE looped when compiling certain patterns
1562 containing an extended class (one that cannot be represented by a bitmap
1563 because it contains high-valued characters or Unicode property items, e.g.
1564 [\pZ]). Almost always one would set UTF-8 mode when processing such a
1565 pattern, but PCRE should not loop if you do not (it no longer does).
1566 [Detail: two cases were found: (a) a repeated subpattern containing an
1567 extended class; (b) a recursive reference to a subpattern that followed a
1568 previous extended class. It wasn't skipping over the extended class
1569 correctly when UTF-8 mode was not set.]
1570
1571 5. A negated single-character class was not being recognized as fixed-length
1572 in lookbehind assertions such as (?<=[^f]), leading to an incorrect
1573 compile error "lookbehind assertion is not fixed length".
1574
1575 6. The RunPerlTest auxiliary script was showing an unexpected difference
1576 between PCRE and Perl for UTF-8 tests. It turns out that it is hard to
1577 write a Perl script that can interpret lines of an input file either as
1578 byte characters or as UTF-8, which is what "perltest" was being required to
1579 do for the non-UTF-8 and UTF-8 tests, respectively. Essentially what you
1580 can't do is switch easily at run time between having the "use utf8;" pragma
1581 or not. In the end, I fudged it by using the RunPerlTest script to insert
1582 "use utf8;" explicitly for the UTF-8 tests.
1583
1584 7. In multiline (/m) mode, PCRE was matching ^ after a terminating newline at
1585 the end of the subject string, contrary to the documentation and to what
1586 Perl does. This was true of both matching functions. Now it matches only at
1587 the start of the subject and immediately after *internal* newlines.
1588
1589 8. A call of pcre_fullinfo() from pcretest to get the option bits was passing
1590 a pointer to an int instead of a pointer to an unsigned long int. This
1591 caused problems on 64-bit systems.
1592
1593 9. Applied a patch from the folks at Google to pcrecpp.cc, to fix "another
1594 instance of the 'standard' template library not being so standard".
1595
1596 10. There was no check on the number of named subpatterns nor the maximum
1597 length of a subpattern name. The product of these values is used to compute
1598 the size of the memory block for a compiled pattern. By supplying a very
1599 long subpattern name and a large number of named subpatterns, the size
1600 computation could be caused to overflow. This is now prevented by limiting
1601 the length of names to 32 characters, and the number of named subpatterns
1602 to 10,000.
1603
1604 11. Subpatterns that are repeated with specific counts have to be replicated in
1605 the compiled pattern. The size of memory for this was computed from the
1606 length of the subpattern and the repeat count. The latter is limited to
1607 65535, but there was no limit on the former, meaning that integer overflow
1608 could in principle occur. The compiled length of a repeated subpattern is
1609 now limited to 30,000 bytes in order to prevent this.
1610
1611 12. Added the optional facility to have named substrings with the same name.
1612
1613 13. Added the ability to use a named substring as a condition, using the
1614 Python syntax: (?(name)yes|no). This overloads (?(R)... and names that
1615 are numbers (not recommended). Forward references are permitted.
1616
1617 14. Added forward references in named backreferences (if you see what I mean).
1618
1619 15. In UTF-8 mode, with the PCRE_DOTALL option set, a quantified dot in the
1620 pattern could run off the end of the subject. For example, the pattern
1621 "(?s)(.{1,5})"8 did this with the subject "ab".
1622
1623 16. If PCRE_DOTALL or PCRE_MULTILINE were set, pcre_dfa_exec() behaved as if
1624 PCRE_CASELESS was set when matching characters that were quantified with ?
1625 or *.
1626
1627 17. A character class other than a single negated character that had a minimum
1628 but no maximum quantifier - for example [ab]{6,} - was not handled
1629 correctly by pce_dfa_exec(). It would match only one character.
1630
1631 18. A valid (though odd) pattern that looked like a POSIX character
1632 class but used an invalid character after [ (for example [[,abc,]]) caused
1633 pcre_compile() to give the error "Failed: internal error: code overflow" or
1634 in some cases to crash with a glibc free() error. This could even happen if
1635 the pattern terminated after [[ but there just happened to be a sequence of
1636 letters, a binary zero, and a closing ] in the memory that followed.
1637
1638 19. Perl's treatment of octal escapes in the range \400 to \777 has changed
1639 over the years. Originally (before any Unicode support), just the bottom 8
1640 bits were taken. Thus, for example, \500 really meant \100. Nowadays the
1641 output from "man perlunicode" includes this:
1642
1643 The regular expression compiler produces polymorphic opcodes. That
1644 is, the pattern adapts to the data and automatically switches to
1645 the Unicode character scheme when presented with Unicode data--or
1646 instead uses a traditional byte scheme when presented with byte
1647 data.
1648
1649 Sadly, a wide octal escape does not cause a switch, and in a string with
1650 no other multibyte characters, these octal escapes are treated as before.
1651 Thus, in Perl, the pattern /\500/ actually matches \100 but the pattern
1652 /\500|\x{1ff}/ matches \500 or \777 because the whole thing is treated as a
1653 Unicode string.
1654
1655 I have not perpetrated such confusion in PCRE. Up till now, it took just
1656 the bottom 8 bits, as in old Perl. I have now made octal escapes with
1657 values greater than \377 illegal in non-UTF-8 mode. In UTF-8 mode they
1658 translate to the appropriate multibyte character.
1659
1660 29. Applied some refactoring to reduce the number of warnings from Microsoft
1661 and Borland compilers. This has included removing the fudge introduced
1662 seven years ago for the OS/2 compiler (see 2.02/2 below) because it caused
1663 a warning about an unused variable.
1664
1665 21. PCRE has not included VT (character 0x0b) in the set of whitespace
1666 characters since release 4.0, because Perl (from release 5.004) does not.
1667 [Or at least, is documented not to: some releases seem to be in conflict
1668 with the documentation.] However, when a pattern was studied with
1669 pcre_study() and all its branches started with \s, PCRE still included VT
1670 as a possible starting character. Of course, this did no harm; it just
1671 caused an unnecessary match attempt.
1672
1673 22. Removed a now-redundant internal flag bit that recorded the fact that case
1674 dependency changed within the pattern. This was once needed for "required
1675 byte" processing, but is no longer used. This recovers a now-scarce options
1676 bit. Also moved the least significant internal flag bit to the most-
1677 significant bit of the word, which was not previously used (hangover from
1678 the days when it was an int rather than a uint) to free up another bit for
1679 the future.
1680
1681 23. Added support for CRLF line endings as well as CR and LF. As well as the
1682 default being selectable at build time, it can now be changed at runtime
1683 via the PCRE_NEWLINE_xxx flags. There are now options for pcregrep to
1684 specify that it is scanning data with non-default line endings.
1685
1686 24. Changed the definition of CXXLINK to make it agree with the definition of
1687 LINK in the Makefile, by replacing LDFLAGS to CXXFLAGS.
1688
1689 25. Applied Ian Taylor's patches to avoid using another stack frame for tail
1690 recursions. This makes a big different to stack usage for some patterns.
1691
1692 26. If a subpattern containing a named recursion or subroutine reference such
1693 as (?P>B) was quantified, for example (xxx(?P>B)){3}, the calculation of
1694 the space required for the compiled pattern went wrong and gave too small a
1695 value. Depending on the environment, this could lead to "Failed: internal
1696 error: code overflow at offset 49" or "glibc detected double free or
1697 corruption" errors.
1698
1699 27. Applied patches from Google (a) to support the new newline modes and (b) to
1700 advance over multibyte UTF-8 characters in GlobalReplace.
1701
1702 28. Change free() to pcre_free() in pcredemo.c. Apparently this makes a
1703 difference for some implementation of PCRE in some Windows version.
1704
1705 29. Added some extra testing facilities to pcretest:
1706
1707 \q<number> in a data line sets the "match limit" value
1708 \Q<number> in a data line sets the "match recursion limt" value
1709 -S <number> sets the stack size, where <number> is in megabytes
1710
1711 The -S option isn't available for Windows.
1712
1713
1714 Version 6.6 06-Feb-06
1715 ---------------------
1716
1717 1. Change 16(a) for 6.5 broke things, because PCRE_DATA_SCOPE was not defined
1718 in pcreposix.h. I have copied the definition from pcre.h.
1719
1720 2. Change 25 for 6.5 broke compilation in a build directory out-of-tree
1721 because pcre.h is no longer a built file.
1722
1723 3. Added Jeff Friedl's additional debugging patches to pcregrep. These are
1724 not normally included in the compiled code.
1725
1726
1727 Version 6.5 01-Feb-06
1728 ---------------------
1729
1730 1. When using the partial match feature with pcre_dfa_exec(), it was not
1731 anchoring the second and subsequent partial matches at the new starting
1732 point. This could lead to incorrect results. For example, with the pattern
1733 /1234/, partially matching against "123" and then "a4" gave a match.
1734
1735 2. Changes to pcregrep:
1736
1737 (a) All non-match returns from pcre_exec() were being treated as failures
1738 to match the line. Now, unless the error is PCRE_ERROR_NOMATCH, an
1739 error message is output. Some extra information is given for the
1740 PCRE_ERROR_MATCHLIMIT and PCRE_ERROR_RECURSIONLIMIT errors, which are
1741 probably the only errors that are likely to be caused by users (by
1742 specifying a regex that has nested indefinite repeats, for instance).
1743 If there are more than 20 of these errors, pcregrep is abandoned.
1744
1745 (b) A binary zero was treated as data while matching, but terminated the
1746 output line if it was written out. This has been fixed: binary zeroes
1747 are now no different to any other data bytes.
1748
1749 (c) Whichever of the LC_ALL or LC_CTYPE environment variables is set is
1750 used to set a locale for matching. The --locale=xxxx long option has
1751 been added (no short equivalent) to specify a locale explicitly on the
1752 pcregrep command, overriding the environment variables.
1753
1754 (d) When -B was used with -n, some line numbers in the output were one less
1755 than they should have been.
1756
1757 (e) Added the -o (--only-matching) option.
1758
1759 (f) If -A or -C was used with -c (count only), some lines of context were
1760 accidentally printed for the final match.
1761
1762 (g) Added the -H (--with-filename) option.
1763
1764 (h) The combination of options -rh failed to suppress file names for files
1765 that were found from directory arguments.
1766
1767 (i) Added the -D (--devices) and -d (--directories) options.
1768
1769 (j) Added the -F (--fixed-strings) option.
1770
1771 (k) Allow "-" to be used as a file name for -f as well as for a data file.
1772
1773 (l) Added the --colo(u)r option.
1774
1775 (m) Added Jeffrey Friedl's -S testing option, but within #ifdefs so that it
1776 is not present by default.
1777
1778 3. A nasty bug was discovered in the handling of recursive patterns, that is,
1779 items such as (?R) or (?1), when the recursion could match a number of
1780 alternatives. If it matched one of the alternatives, but subsequently,
1781 outside the recursion, there was a failure, the code tried to back up into
1782 the recursion. However, because of the way PCRE is implemented, this is not
1783 possible, and the result was an incorrect result from the match.
1784
1785 In order to prevent this happening, the specification of recursion has
1786 been changed so that all such subpatterns are automatically treated as
1787 atomic groups. Thus, for example, (?R) is treated as if it were (?>(?R)).
1788
1789 4. I had overlooked the fact that, in some locales, there are characters for
1790 which isalpha() is true but neither isupper() nor islower() are true. In
1791 the fr_FR locale, for instance, the \xAA and \xBA characters (ordmasculine
1792 and ordfeminine) are like this. This affected the treatment of \w and \W
1793 when they appeared in character classes, but not when they appeared outside
1794 a character class. The bit map for "word" characters is now created
1795 separately from the results of isalnum() instead of just taking it from the
1796 upper, lower, and digit maps. (Plus the underscore character, of course.)
1797
1798 5. The above bug also affected the handling of POSIX character classes such as
1799 [[:alpha:]] and [[:alnum:]]. These do not have their own bit maps in PCRE's
1800 permanent tables. Instead, the bit maps for such a class were previously
1801 created as the appropriate unions of the upper, lower, and digit bitmaps.
1802 Now they are created by subtraction from the [[:word:]] class, which has
1803 its own bitmap.
1804
1805 6. The [[:blank:]] character class matches horizontal, but not vertical space.
1806 It is created by subtracting the vertical space characters (\x09, \x0a,
1807 \x0b, \x0c) from the [[:space:]] bitmap. Previously, however, the
1808 subtraction was done in the overall bitmap for a character class, meaning
1809 that a class such as [\x0c[:blank:]] was incorrect because \x0c would not
1810 be recognized. This bug has been fixed.
1811
1812 7. Patches from the folks at Google:
1813
1814 (a) pcrecpp.cc: "to handle a corner case that may or may not happen in
1815 real life, but is still worth protecting against".
1816
1817 (b) pcrecpp.cc: "corrects a bug when negative radixes are used with
1818 regular expressions".
1819
1820 (c) pcre_scanner.cc: avoid use of std::count() because not all systems
1821 have it.
1822
1823 (d) Split off pcrecpparg.h from pcrecpp.h and had the former built by
1824 "configure" and the latter not, in order to fix a problem somebody had
1825 with compiling the Arg class on HP-UX.
1826
1827 (e) Improve the error-handling of the C++ wrapper a little bit.
1828
1829 (f) New tests for checking recursion limiting.
1830
1831 8. The pcre_memmove() function, which is used only if the environment does not
1832 have a standard memmove() function (and is therefore rarely compiled),
1833 contained two bugs: (a) use of int instead of size_t, and (b) it was not
1834 returning a result (though PCRE never actually uses the result).
1835
1836 9. In the POSIX regexec() interface, if nmatch is specified as a ridiculously
1837 large number - greater than INT_MAX/(3*sizeof(int)) - REG_ESPACE is
1838 returned instead of calling malloc() with an overflowing number that would
1839 most likely cause subsequent chaos.
1840
1841 10. The debugging option of pcretest was not showing the NO_AUTO_CAPTURE flag.
1842
1843 11. The POSIX flag REG_NOSUB is now supported. When a pattern that was compiled
1844 with this option is matched, the nmatch and pmatch options of regexec() are
1845 ignored.
1846
1847 12. Added REG_UTF8 to the POSIX interface. This is not defined by POSIX, but is
1848 provided in case anyone wants to the the POSIX interface with UTF-8
1849 strings.
1850
1851 13. Added CXXLDFLAGS to the Makefile parameters to provide settings only on the
1852 C++ linking (needed for some HP-UX environments).
1853
1854 14. Avoid compiler warnings in get_ucpname() when compiled without UCP support
1855 (unused parameter) and in the pcre_printint() function (omitted "default"
1856 switch label when the default is to do nothing).
1857
1858 15. Added some code to make it possible, when PCRE is compiled as a C++
1859 library, to replace subject pointers for pcre_exec() with a smart pointer
1860 class, thus making it possible to process discontinuous strings.
1861
1862 16. The two macros PCRE_EXPORT and PCRE_DATA_SCOPE are confusing, and perform
1863 much the same function. They were added by different people who were trying
1864 to make PCRE easy to compile on non-Unix systems. It has been suggested
1865 that PCRE_EXPORT be abolished now that there is more automatic apparatus
1866 for compiling on Windows systems. I have therefore replaced it with
1867 PCRE_DATA_SCOPE. This is set automatically for Windows; if not set it
1868 defaults to "extern" for C or "extern C" for C++, which works fine on
1869 Unix-like systems. It is now possible to override the value of PCRE_DATA_
1870 SCOPE with something explicit in config.h. In addition:
1871
1872 (a) pcreposix.h still had just "extern" instead of either of these macros;
1873 I have replaced it with PCRE_DATA_SCOPE.
1874
1875 (b) Functions such as _pcre_xclass(), which are internal to the library,
1876 but external in the C sense, all had PCRE_EXPORT in their definitions.
1877 This is apparently wrong for the Windows case, so I have removed it.
1878 (It makes no difference on Unix-like systems.)
1879
1880 17. Added a new limit, MATCH_LIMIT_RECURSION, which limits the depth of nesting
1881 of recursive calls to match(). This is different to MATCH_LIMIT because
1882 that limits the total number of calls to match(), not all of which increase
1883 the depth of recursion. Limiting the recursion depth limits the amount of
1884 stack (or heap if NO_RECURSE is set) that is used. The default can be set
1885 when PCRE is compiled, and changed at run time. A patch from Google adds
1886 this functionality to the C++ interface.
1887
1888 18. Changes to the handling of Unicode character properties:
1889
1890 (a) Updated the table to Unicode 4.1.0.
1891
1892 (b) Recognize characters that are not in the table as "Cn" (undefined).
1893
1894 (c) I revised the way the table is implemented to a much improved format
1895 which includes recognition of ranges. It now supports the ranges that
1896 are defined in UnicodeData.txt, and it also amalgamates other
1897 characters into ranges. This has reduced the number of entries in the
1898 table from around 16,000 to around 3,000, thus reducing its size
1899 considerably. I realized I did not need to use a tree structure after
1900 all - a binary chop search is just as efficient. Having reduced the
1901 number of entries, I extended their size from 6 bytes to 8 bytes to
1902 allow for more data.
1903
1904 (d) Added support for Unicode script names via properties such as \p{Han}.
1905
1906 19. In UTF-8 mode, a backslash followed by a non-Ascii character was not
1907 matching that character.
1908
1909 20. When matching a repeated Unicode property with a minimum greater than zero,
1910 (for example \pL{2,}), PCRE could look past the end of the subject if it
1911 reached it while seeking the minimum number of characters. This could
1912 happen only if some of the characters were more than one byte long, because
1913 there is a check for at least the minimum number of bytes.
1914
1915 21. Refactored the implementation of \p and \P so as to be more general, to
1916 allow for more different types of property in future. This has changed the
1917 compiled form incompatibly. Anybody with saved compiled patterns that use
1918 \p or \P will have to recompile them.
1919
1920 22. Added "Any" and "L&" to the supported property types.
1921
1922 23. Recognize \x{...} as a code point specifier, even when not in UTF-8 mode,
1923 but give a compile time error if the value is greater than 0xff.
1924
1925 24. The man pages for pcrepartial, pcreprecompile, and pcre_compile2 were
1926 accidentally not being installed or uninstalled.
1927
1928 25. The pcre.h file was built from pcre.h.in, but the only changes that were
1929 made were to insert the current release number. This seemed silly, because
1930 it made things harder for people building PCRE on systems that don't run
1931 "configure". I have turned pcre.h into a distributed file, no longer built
1932 by "configure", with the version identification directly included. There is
1933 no longer a pcre.h.in file.
1934
1935 However, this change necessitated a change to the pcre-config script as
1936 well. It is built from pcre-config.in, and one of the substitutions was the
1937 release number. I have updated configure.ac so that ./configure now finds
1938 the release number by grepping pcre.h.
1939
1940 26. Added the ability to run the tests under valgrind.
1941
1942
1943 Version 6.4 05-Sep-05
1944 ---------------------
1945
1946 1. Change 6.0/10/(l) to pcregrep introduced a bug that caused separator lines
1947 "--" to be printed when multiple files were scanned, even when none of the
1948 -A, -B, or -C options were used. This is not compatible with Gnu grep, so I
1949 consider it to be a bug, and have restored the previous behaviour.
1950
1951 2. A couple of code tidies to get rid of compiler warnings.
1952
1953 3. The pcretest program used to cheat by referring to symbols in the library
1954 whose names begin with _pcre_. These are internal symbols that are not
1955 really supposed to be visible externally, and in some environments it is
1956 possible to suppress them. The cheating is now confined to including
1957 certain files from the library's source, which is a bit cleaner.
1958
1959 4. Renamed pcre.in as pcre.h.in to go with pcrecpp.h.in; it also makes the
1960 file's purpose clearer.
1961
1962 5. Reorganized pcre_ucp_findchar().
1963
1964
1965 Version 6.3 15-Aug-05
1966 ---------------------
1967
1968 1. The file libpcre.pc.in did not have general read permission in the tarball.
1969
1970 2. There were some problems when building without C++ support:
1971
1972 (a) If C++ support was not built, "make install" and "make test" still
1973 tried to test it.
1974
1975 (b) There were problems when the value of CXX was explicitly set. Some
1976 changes have been made to try to fix these, and ...
1977
1978 (c) --disable-cpp can now be used to explicitly disable C++ support.
1979
1980 (d) The use of @CPP_OBJ@ directly caused a blank line preceded by a
1981 backslash in a target when C++ was disabled. This confuses some
1982 versions of "make", apparently. Using an intermediate variable solves
1983 this. (Same for CPP_LOBJ.)
1984
1985 3. $(LINK_FOR_BUILD) now includes $(CFLAGS_FOR_BUILD) and $(LINK)
1986 (non-Windows) now includes $(CFLAGS) because these flags are sometimes
1987 necessary on certain architectures.
1988
1989 4. Added a setting of -export-symbols-regex to the link command to remove
1990 those symbols that are exported in the C sense, but actually are local
1991 within the library, and not documented. Their names all begin with
1992 "_pcre_". This is not a perfect job, because (a) we have to except some
1993 symbols that pcretest ("illegally") uses, and (b) the facility isn't always
1994 available (and never for static libraries). I have made a note to try to
1995 find a way round (a) in the future.
1996
1997
1998 Version 6.2 01-Aug-05
1999 ---------------------
2000
2001 1. There was no test for integer overflow of quantifier values. A construction
2002 such as {1111111111111111} would give undefined results. What is worse, if
2003 a minimum quantifier for a parenthesized subpattern overflowed and became
2004 negative, the calculation of the memory size went wrong. This could have
2005 led to memory overwriting.
2006
2007 2. Building PCRE using VPATH was broken. Hopefully it is now fixed.
2008
2009 3. Added "b" to the 2nd argument of fopen() in dftables.c, for non-Unix-like
2010 operating environments where this matters.
2011
2012 4. Applied Giuseppe Maxia's patch to add additional features for controlling
2013 PCRE options from within the C++ wrapper.
2014
2015 5. Named capturing subpatterns were not being correctly counted when a pattern
2016 was compiled. This caused two problems: (a) If there were more than 100
2017 such subpatterns, the calculation of the memory needed for the whole
2018 compiled pattern went wrong, leading to an overflow error. (b) Numerical
2019 back references of the form \12, where the number was greater than 9, were
2020 not recognized as back references, even though there were sufficient
2021 previous subpatterns.
2022
2023 6. Two minor patches to pcrecpp.cc in order to allow it to compile on older
2024 versions of gcc, e.g. 2.95.4.
2025
2026
2027 Version 6.1 21-Jun-05
2028 ---------------------
2029
2030 1. There was one reference to the variable "posix" in pcretest.c that was not
2031 surrounded by "#if !defined NOPOSIX".
2032
2033 2. Make it possible to compile pcretest without DFA support, UTF8 support, or
2034 the cross-check on the old pcre_info() function, for the benefit of the
2035 cut-down version of PCRE that is currently imported into Exim.
2036
2037 3. A (silly) pattern starting with (?i)(?-i) caused an internal space
2038 allocation error. I've done the easy fix, which wastes 2 bytes for sensible
2039 patterns that start (?i) but I don't think that matters. The use of (?i) is
2040 just an example; this all applies to the other options as well.
2041
2042 4. Since libtool seems to echo the compile commands it is issuing, the output
2043 from "make" can be reduced a bit by putting "@" in front of each libtool
2044 compile command.
2045
2046 5. Patch from the folks at Google for configure.in to be a bit more thorough
2047 in checking for a suitable C++ installation before trying to compile the
2048 C++ stuff. This should fix a reported problem when a compiler was present,
2049 but no suitable headers.
2050
2051 6. The man pages all had just "PCRE" as their title. I have changed them to
2052 be the relevant file name. I have also arranged that these names are
2053 retained in the file doc/pcre.txt, which is a concatenation in text format
2054 of all the man pages except the little individual ones for each function.
2055
2056 7. The NON-UNIX-USE file had not been updated for the different set of source
2057 files that come with release 6. I also added a few comments about the C++
2058 wrapper.
2059
2060
2061 Version 6.0 07-Jun-05
2062 ---------------------
2063
2064 1. Some minor internal re-organization to help with my DFA experiments.
2065
2066 2. Some missing #ifdef SUPPORT_UCP conditionals in pcretest and printint that
2067 didn't matter for the library itself when fully configured, but did matter
2068 when compiling without UCP support, or within Exim, where the ucp files are
2069 not imported.
2070
2071 3. Refactoring of the library code to split up the various functions into
2072 different source modules. The addition of the new DFA matching code (see
2073 below) to a single monolithic source would have made it really too
2074 unwieldy, quite apart from causing all the code to be include in a
2075 statically linked application, when only some functions are used. This is
2076 relevant even without the DFA addition now that patterns can be compiled in
2077 one application and matched in another.
2078
2079 The downside of splitting up is that there have to be some external
2080 functions and data tables that are used internally in different modules of
2081 the library but which are not part of the API. These have all had their
2082 names changed to start with "_pcre_" so that they are unlikely to clash
2083 with other external names.
2084
2085 4. Added an alternate matching function, pcre_dfa_exec(), which matches using
2086 a different (DFA) algorithm. Although it is slower than the original
2087 function, it does have some advantages for certain types of matching
2088 problem.
2089
2090 5. Upgrades to pcretest in order to test the features of pcre_dfa_exec(),
2091 including restarting after a partial match.
2092
2093 6. A patch for pcregrep that defines INVALID_FILE_ATTRIBUTES if it is not
2094 defined when compiling for Windows was sent to me. I have put it into the
2095 code, though I have no means of testing or verifying it.
2096
2097 7. Added the pcre_refcount() auxiliary function.
2098
2099 8. Added the PCRE_FIRSTLINE option. This constrains an unanchored pattern to
2100 match before or at the first newline in the subject string. In pcretest,
2101 the /f option on a pattern can be used to set this.
2102
2103 9. A repeated \w when used in UTF-8 mode with characters greater than 256
2104 would behave wrongly. This has been present in PCRE since release 4.0.
2105
2106 10. A number of changes to the pcregrep command:
2107
2108 (a) Refactored how -x works; insert ^(...)$ instead of setting
2109 PCRE_ANCHORED and checking the length, in preparation for adding
2110 something similar for -w.
2111
2112 (b) Added the -w (match as a word) option.
2113
2114 (c) Refactored the way lines are read and buffered so as to have more
2115 than one at a time available.
2116
2117 (d) Implemented a pcregrep test script.
2118
2119 (e) Added the -M (multiline match) option. This allows patterns to match
2120 over several lines of the subject. The buffering ensures that at least
2121 8K, or the rest of the document (whichever is the shorter) is available
2122 for matching (and similarly the previous 8K for lookbehind assertions).
2123
2124 (f) Changed the --help output so that it now says
2125
2126 -w, --word-regex(p)
2127
2128 instead of two lines, one with "regex" and the other with "regexp"
2129 because that confused at least one person since the short forms are the
2130 same. (This required a bit of code, as the output is generated
2131 automatically from a table. It wasn't just a text change.)
2132
2133 (g) -- can be used to terminate pcregrep options if the next thing isn't an
2134 option but starts with a hyphen. Could be a pattern or a path name
2135 starting with a hyphen, for instance.
2136
2137 (h) "-" can be given as a file name to represent stdin.
2138
2139 (i) When file names are being printed, "(standard input)" is used for
2140 the standard input, for compatibility with GNU grep. Previously
2141 "<stdin>" was used.
2142
2143 (j) The option --label=xxx can be used to supply a name to be used for
2144 stdin when file names are being printed. There is no short form.
2145
2146 (k) Re-factored the options decoding logic because we are going to add
2147 two more options that take data. Such options can now be given in four
2148 different ways, e.g. "-fname", "-f name", "--file=name", "--file name".
2149
2150 (l) Added the -A, -B, and -C options for requesting that lines of context
2151 around matches be printed.
2152
2153 (m) Added the -L option to print the names of files that do not contain
2154 any matching lines, that is, the complement of -l.
2155
2156 (n) The return code is 2 if any file cannot be opened, but pcregrep does
2157 continue to scan other files.
2158
2159 (o) The -s option was incorrectly implemented. For compatibility with other
2160 greps, it now suppresses the error message for a non-existent or non-
2161 accessible file (but not the return code). There is a new option called
2162 -q that suppresses the output of matching lines, which was what -s was
2163 previously doing.
2164
2165 (p) Added --include and --exclude options to specify files for inclusion
2166 and exclusion when recursing.
2167
2168 11. The Makefile was not using the Autoconf-supported LDFLAGS macro properly.
2169 Hopefully, it now does.
2170
2171 12. Missing cast in pcre_study().
2172
2173 13. Added an "uninstall" target to the makefile.
2174
2175 14. Replaced "extern" in the function prototypes in Makefile.in with
2176 "PCRE_DATA_SCOPE", which defaults to 'extern' or 'extern "C"' in the Unix
2177 world, but is set differently for Windows.
2178
2179 15. Added a second compiling function called pcre_compile2(). The only
2180 difference is that it has an extra argument, which is a pointer to an
2181 integer error code. When there is a compile-time failure, this is set
2182 non-zero, in addition to the error test pointer being set to point to an
2183 error message. The new argument may be NULL if no error number is required
2184 (but then you may as well call pcre_compile(), which is now just a
2185 wrapper). This facility is provided because some applications need a
2186 numeric error indication, but it has also enabled me to tidy up the way
2187 compile-time errors are handled in the POSIX wrapper.
2188
2189 16. Added VPATH=.libs to the makefile; this should help when building with one
2190 prefix path and installing with another. (Or so I'm told by someone who
2191 knows more about this stuff than I do.)
2192
2193 17. Added a new option, REG_DOTALL, to the POSIX function regcomp(). This
2194 passes PCRE_DOTALL to the pcre_compile() function, making the "." character
2195 match everything, including newlines. This is not POSIX-compatible, but
2196 somebody wanted the feature. From pcretest it can be activated by using
2197 both the P and the s flags.
2198
2199 18. AC_PROG_LIBTOOL appeared twice in Makefile.in. Removed one.
2200
2201 19. libpcre.pc was being incorrectly installed as executable.
2202
2203 20. A couple of places in pcretest check for end-of-line by looking for '\n';
2204 it now also looks for '\r' so that it will work unmodified on Windows.
2205
2206 21. Added Google's contributed C++ wrapper to the distribution.
2207
2208 22. Added some untidy missing memory free() calls in pcretest, to keep
2209 Electric Fence happy when testing.
2210
2211
2212
2213 Version 5.0 13-Sep-04
2214 ---------------------
2215
2216 1. Internal change: literal characters are no longer packed up into items
2217 containing multiple characters in a single byte-string. Each character
2218 is now matched using a separate opcode. However, there may be more than one
2219 byte in the character in UTF-8 mode.
2220
2221 2. The pcre_callout_block structure has two new fields: pattern_position and
2222 next_item_length. These contain the offset in the pattern to the next match
2223 item, and its length, respectively.
2224
2225 3. The PCRE_AUTO_CALLOUT option for pcre_compile() requests the automatic
2226 insertion of callouts before each pattern item. Added the /C option to
2227 pcretest to make use of this.
2228
2229 4. On the advice of a Windows user, the lines
2230
2231 #if defined(_WIN32) || defined(WIN32)
2232 _setmode( _fileno( stdout ), 0x8000 );
2233 #endif /* defined(_WIN32) || defined(WIN32) */
2234
2235 have been added to the source of pcretest. This apparently does useful
2236 magic in relation to line terminators.
2237
2238 5. Changed "r" and "w" in the calls to fopen() in pcretest to "rb" and "wb"
2239 for the benefit of those environments where the "b" makes a difference.
2240
2241 6. The icc compiler has the same options as gcc, but "configure" doesn't seem
2242 to know about it. I have put a hack into configure.in that adds in code
2243 to set GCC=yes if CC=icc. This seems to end up at a point in the
2244 generated configure script that is early enough to affect the setting of
2245 compiler options, which is what is needed, but I have no means of testing
2246 whether it really works. (The user who reported this had patched the
2247 generated configure script, which of course I cannot do.)
2248
2249 LATER: After change 22 below (new libtool files), the configure script
2250 seems to know about icc (and also ecc). Therefore, I have commented out
2251 this hack in configure.in.
2252
2253 7. Added support for pkg-config (2 patches were sent in).
2254
2255 8. Negated POSIX character classes that used a combination of internal tables
2256 were completely broken. These were [[:^alpha:]], [[:^alnum:]], and
2257 [[:^ascii]]. Typically, they would match almost any characters. The other
2258 POSIX classes were not broken in this way.
2259
2260 9. Matching the pattern "\b.*?" against "ab cd", starting at offset 1, failed
2261 to find the match, as PCRE was deluded into thinking that the match had to
2262 start at the start point or following a newline. The same bug applied to
2263 patterns with negative forward assertions or any backward assertions
2264 preceding ".*" at the start, unless the pattern required a fixed first
2265 character. This was a failing pattern: "(?!.bcd).*". The bug is now fixed.
2266
2267 10. In UTF-8 mode, when moving forwards in the subject after a failed match
2268 starting at the last subject character, bytes beyond the end of the subject
2269 string were read.
2270
2271 11. Renamed the variable "class" as "classbits" to make life easier for C++
2272 users. (Previously there was a macro definition, but it apparently wasn't
2273 enough.)
2274
2275 12. Added the new field "tables" to the extra data so that tables can be passed
2276 in at exec time, or the internal tables can be re-selected. This allows
2277 a compiled regex to be saved and re-used at a later time by a different
2278 program that might have everything at different addresses.
2279
2280 13. Modified the pcre-config script so that, when run on Solaris, it shows a
2281 -R library as well as a -L library.
2282
2283 14. The debugging options of pcretest (-d on the command line or D on a
2284 pattern) showed incorrect output for anything following an extended class
2285 that contained multibyte characters and which was followed by a quantifier.
2286
2287 15. Added optional support for general category Unicode character properties
2288 via the \p, \P, and \X escapes. Unicode property support implies UTF-8
2289 support. It adds about 90K to the size of the library. The meanings of the
2290 inbuilt class escapes such as \d and \s have NOT been changed.
2291
2292 16. Updated pcredemo.c to include calls to free() to release the memory for the
2293 compiled pattern.
2294
2295 17. The generated file chartables.c was being created in the source directory
2296 instead of in the building directory. This caused the build to fail if the
2297 source directory was different from the building directory, and was
2298 read-only.
2299
2300 18. Added some sample Win commands from Mark Tetrode into the NON-UNIX-USE
2301 file. No doubt somebody will tell me if they don't make sense... Also added
2302 Dan Mooney's comments about building on OpenVMS.
2303
2304 19. Added support for partial matching via the PCRE_PARTIAL option for
2305 pcre_exec() and the \P data escape in pcretest.
2306
2307 20. Extended pcretest with 3 new pattern features:
2308
2309 (i) A pattern option of the form ">rest-of-line" causes pcretest to
2310 write the compiled pattern to the file whose name is "rest-of-line".
2311 This is a straight binary dump of the data, with the saved pointer to
2312 the character tables forced to be NULL. The study data, if any, is
2313 written too. After writing, pcretest reads a new pattern.
2314
2315 (ii) If, instead of a pattern, "<rest-of-line" is given, pcretest reads a
2316 compiled pattern from the given file. There must not be any
2317 occurrences of "<" in the file name (pretty unlikely); if there are,
2318 pcretest will instead treat the initial "<" as a pattern delimiter.
2319 After reading in the pattern, pcretest goes on to read data lines as
2320 usual.
2321
2322 (iii) The F pattern option causes pcretest to flip the bytes in the 32-bit
2323 and 16-bit fields in a compiled pattern, to simulate a pattern that
2324 was compiled on a host of opposite endianness.
2325
2326 21. The pcre-exec() function can now cope with patterns that were compiled on
2327 hosts of opposite endianness, with this restriction:
2328
2329 As for any compiled expression that is saved and used later, the tables
2330 pointer field cannot be preserved; the extra_data field in the arguments
2331 to pcre_exec() should be used to pass in a tables address if a value
2332 other than the default internal tables were used at compile time.
2333
2334 22. Calling pcre_exec() with a negative value of the "ovecsize" parameter is
2335 now diagnosed as an error. Previously, most of the time, a negative number
2336 would have been treated as zero, but if in addition "ovector" was passed as
2337 NULL, a crash could occur.
2338
2339 23. Updated the files ltmain.sh, config.sub, config.guess, and aclocal.m4 with
2340 new versions from the libtool 1.5 distribution (the last one is a copy of
2341 a file called libtool.m4). This seems to have fixed the need to patch
2342 "configure" to support Darwin 1.3 (which I used to do). However, I still
2343 had to patch ltmain.sh to ensure that ${SED} is set (it isn't on my
2344 workstation).
2345
2346 24. Changed the PCRE licence to be the more standard "BSD" licence.
2347
2348
2349 Version 4.5 01-Dec-03
2350 ---------------------
2351
2352 1. There has been some re-arrangement of the code for the match() function so
2353 that it can be compiled in a version that does not call itself recursively.
2354 Instead, it keeps those local variables that need separate instances for
2355 each "recursion" in a frame on the heap, and gets/frees frames whenever it
2356 needs to "recurse". Keeping track of where control must go is done by means
2357 of setjmp/longjmp. The whole thing is implemented by a set of macros that
2358 hide most of the details from the main code, and operates only if
2359 NO_RECURSE is defined while compiling pcre.c. If PCRE is built using the
2360 "configure" mechanism, "--disable-stack-for-recursion" turns on this way of
2361 operating.
2362
2363 To make it easier for callers to provide specially tailored get/free
2364 functions for this usage, two new functions, pcre_stack_malloc, and
2365 pcre_stack_free, are used. They are always called in strict stacking order,
2366 and the size of block requested is always the same.
2367
2368 The PCRE_CONFIG_STACKRECURSE info parameter can be used to find out whether
2369 PCRE has been compiled to use the stack or the heap for recursion. The
2370 -C option of pcretest uses this to show which version is compiled.
2371
2372 A new data escape \S, is added to pcretest; it causes the amounts of store
2373 obtained and freed by both kinds of malloc/free at match time to be added
2374 to the output.
2375
2376 2. Changed the locale test to use "fr_FR" instead of "fr" because that's
2377 what's available on my current Linux desktop machine.
2378
2379 3. When matching a UTF-8 string, the test for a valid string at the start has
2380 been extended. If start_offset is not zero, PCRE now checks that it points
2381 to a byte that is the start of a UTF-8 character. If not, it returns
2382 PCRE_ERROR_BADUTF8_OFFSET (-11). Note: the whole string is still checked;
2383 this is necessary because there may be backward assertions in the pattern.
2384 When matching the same subject several times, it may save resources to use
2385 PCRE_NO_UTF8_CHECK on all but the first call if the string is long.
2386
2387 4. The code for checking the validity of UTF-8 strings has been tightened so
2388 that it rejects (a) strings containing 0xfe or 0xff bytes and (b) strings
2389 containing "overlong sequences".
2390
2391 5. Fixed a bug (appearing twice) that I could not find any way of exploiting!
2392 I had written "if ((digitab[*p++] && chtab_digit) == 0)" where the "&&"
2393 should have been "&", but it just so happened that all the cases this let
2394 through by mistake were picked up later in the function.
2395
2396 6. I had used a variable called "isblank" - this is a C99 function, causing
2397 some compilers to warn. To avoid this, I renamed it (as "blankclass").
2398
2399 7. Cosmetic: (a) only output another newline at the end of pcretest if it is
2400 prompting; (b) run "./pcretest /dev/null" at the start of the test script
2401 so the version is shown; (c) stop "make test" echoing "./RunTest".
2402
2403 8. Added patches from David Burgess to enable PCRE to run on EBCDIC systems.
2404
2405 9. The prototype for memmove() for systems that don't have it was using
2406 size_t, but the inclusion of the header that defines size_t was later. I've
2407 moved the #includes for the C headers earlier to avoid this.
2408
2409 10. Added some adjustments to the code to make it easier to compiler on certain
2410 special systems:
2411
2412 (a) Some "const" qualifiers were missing.
2413 (b) Added the macro EXPORT before all exported functions; by default this
2414 is defined to be empty.
2415 (c) Changed the dftables auxiliary program (that builds chartables.c) so
2416 that it reads its output file name as an argument instead of writing
2417 to the standard output and assuming this can be redirected.
2418
2419 11. In UTF-8 mode, if a recursive reference (e.g. (?1)) followed a character
2420 class containing characters with values greater than 255, PCRE compilation
2421 went into a loop.
2422
2423 12. A recursive reference to a subpattern that was within another subpattern
2424 that had a minimum quantifier of zero caused PCRE to crash. For example,
2425 (x(y(?2))z)? provoked this bug with a subject that got as far as the
2426 recursion. If the recursively-called subpattern itself had a zero repeat,
2427 that was OK.
2428
2429 13. In pcretest, the buffer for reading a data line was set at 30K, but the
2430 buffer into which it was copied (for escape processing) was still set at
2431 1024, so long lines caused crashes.
2432
2433 14. A pattern such as /[ab]{1,3}+/ failed to compile, giving the error
2434 "internal error: code overflow...". This applied to any character class
2435 that was followed by a possessive quantifier.
2436
2437 15. Modified the Makefile to add libpcre.la as a prerequisite for
2438 libpcreposix.la because I was told this is needed for a parallel build to
2439 work.
2440
2441 16. If a pattern that contained .* following optional items at the start was
2442 studied, the wrong optimizing data was generated, leading to matching
2443 errors. For example, studying /[ab]*.*c/ concluded, erroneously, that any
2444 matching string must start with a or b or c. The correct conclusion for
2445 this pattern is that a match can start with any character.
2446
2447
2448 Version 4.4 13-Aug-03
2449 ---------------------
2450
2451 1. In UTF-8 mode, a character class containing characters with values between
2452 127 and 255 was not handled correctly if the compiled pattern was studied.
2453 In fixing this, I have also improved the studying algorithm for such
2454 classes (slightly).
2455
2456 2. Three internal functions had redundant arguments passed to them. Removal
2457 might give a very teeny performance improvement.
2458
2459 3. Documentation bug: the value of the capture_top field in a callout is *one
2460 more than* the number of the hightest numbered captured substring.
2461
2462 4. The Makefile linked pcretest and pcregrep with -lpcre, which could result
2463 in incorrectly linking with a previously installed version. They now link
2464 explicitly with libpcre.la.
2465
2466 5. configure.in no longer needs to recognize Cygwin specially.
2467
2468 6. A problem in pcre.in for Windows platforms is fixed.
2469
2470 7. If a pattern was successfully studied, and the -d (or /D) flag was given to
2471 pcretest, it used to include the size of the study block as part of its
2472 output. Unfortunately, the structure contains a field that has a different
2473 size on different hardware architectures. This meant that the tests that
2474 showed this size failed. As the block is currently always of a fixed size,
2475 this information isn't actually particularly useful in pcretest output, so
2476 I have just removed it.
2477
2478 8. Three pre-processor statements accidentally did not start in column 1.
2479 Sadly, there are *still* compilers around that complain, even though
2480 standard C has not required this for well over a decade. Sigh.
2481
2482 9. In pcretest, the code for checking callouts passed small integers in the
2483 callout_data field, which is a void * field. However, some picky compilers
2484 complained about the casts involved for this on 64-bit systems. Now
2485 pcretest passes the address of the small integer instead, which should get
2486 rid of the warnings.
2487
2488 10. By default, when in UTF-8 mode, PCRE now checks for valid UTF-8 strings at
2489 both compile and run time, and gives an error if an invalid UTF-8 sequence
2490 is found. There is a option for disabling this check in cases where the
2491 string is known to be correct and/or the maximum performance is wanted.
2492
2493 11. In response to a bug report, I changed one line in Makefile.in from
2494
2495 -Wl,--out-implib,.libs/lib@WIN_PREFIX@pcreposix.dll.a \
2496 to
2497 -Wl,--out-implib,.libs/@WIN_PREFIX@libpcreposix.dll.a \
2498
2499 to look similar to other lines, but I have no way of telling whether this
2500 is the right thing to do, as I do not use Windows. No doubt I'll get told
2501 if it's wrong...
2502
2503
2504 Version 4.3 21-May-03
2505 ---------------------
2506
2507 1. Two instances of @WIN_PREFIX@ omitted from the Windows targets in the
2508 Makefile.
2509
2510 2. Some refactoring to improve the quality of the code:
2511
2512 (i) The utf8_table... variables are now declared "const".
2513
2514 (ii) The code for \cx, which used the "case flipping" table to upper case
2515 lower case letters, now just substracts 32. This is ASCII-specific,
2516 but the whole concept of \cx is ASCII-specific, so it seems
2517 reasonable.
2518
2519 (iii) PCRE was using its character types table to recognize decimal and
2520 hexadecimal digits in the pattern. This is silly, because it handles
2521 only 0-9, a-f, and A-F, but the character types table is locale-
2522 specific, which means strange things might happen. A private
2523 table is now used for this - though it costs 256 bytes, a table is
2524 much faster than multiple explicit tests. Of course, the standard
2525 character types table is still used for matching digits in subject
2526 strings against \d.
2527
2528 (iv) Strictly, the identifier ESC_t is reserved by POSIX (all identifiers
2529 ending in _t are). So I've renamed it as ESC_tee.
2530
2531 3. The first argument for regexec() in the POSIX wrapper should have been
2532 defined as "const".
2533
2534 4. Changed pcretest to use malloc() for its buffers so that they can be
2535 Electric Fenced for debugging.
2536
2537 5. There were several places in the code where, in UTF-8 mode, PCRE would try
2538 to read one or more bytes before the start of the subject string. Often this
2539 had no effect on PCRE's behaviour, but in some circumstances it could
2540 provoke a segmentation fault.
2541
2542 6. A lookbehind at the start of a pattern in UTF-8 mode could also cause PCRE
2543 to try to read one or more bytes before the start of the subject string.
2544
2545 7. A lookbehind in a pattern matched in non-UTF-8 mode on a PCRE compiled with
2546 UTF-8 support could misbehave in various ways if the subject string
2547 contained bytes with the 0x80 bit set and the 0x40 bit unset in a lookbehind
2548 area. (PCRE was not checking for the UTF-8 mode flag, and trying to move
2549 back over UTF-8 characters.)
2550
2551
2552 Version 4.2 14-Apr-03
2553 ---------------------
2554
2555 1. Typo "#if SUPPORT_UTF8" instead of "#ifdef SUPPORT_UTF8" fixed.
2556
2557 2. Changes to the building process, supplied by Ronald Landheer-Cieslak
2558 [ON_WINDOWS]: new variable, "#" on non-Windows platforms
2559 [NOT_ON_WINDOWS]: new variable, "#" on Windows platforms
2560 [WIN_PREFIX]: new variable, "cyg" for Cygwin
2561 * Makefile.in: use autoconf substitution for OBJEXT, EXEEXT, BUILD_OBJEXT
2562 and BUILD_EXEEXT
2563 Note: automatic setting of the BUILD variables is not yet working
2564 set CPPFLAGS and BUILD_CPPFLAGS (but don't use yet) - should be used at
2565 compile-time but not at link-time
2566 [LINK]: use for linking executables only
2567 make different versions for Windows and non-Windows
2568 [LINKLIB]: new variable, copy of UNIX-style LINK, used for linking
2569 libraries
2570 [LINK_FOR_BUILD]: new variable
2571 [OBJEXT]: use throughout
2572 [EXEEXT]: use throughout
2573 <winshared>: new target
2574 <wininstall>: new target
2575 <dftables.o>: use native compiler
2576 <dftables>: use native linker
2577 <install>: handle Windows platform correctly
2578 <clean>: ditto
2579 <check>: ditto
2580 copy DLL to top builddir before testing
2581
2582 As part of these changes, -no-undefined was removed again. This was reported
2583 to give trouble on HP-UX 11.0, so getting rid of it seems like a good idea
2584 in any case.
2585
2586 3. Some tidies to get rid of compiler warnings:
2587
2588 . In the match_data structure, match_limit was an unsigned long int, whereas
2589 match_call_count was an int. I've made them both unsigned long ints.
2590
2591 . In pcretest the fact that a const uschar * doesn't automatically cast to
2592 a void * provoked a warning.
2593
2594 . Turning on some more compiler warnings threw up some "shadow" variables
2595 and a few more missing casts.
2596
2597 4. If PCRE was complied with UTF-8 support, but called without the PCRE_UTF8
2598 option, a class that contained a single character with a value between 128
2599 and 255 (e.g. /[\xFF]/) caused PCRE to crash.
2600
2601 5. If PCRE was compiled with UTF-8 support, but called without the PCRE_UTF8
2602 option, a class that contained several characters, but with at least one
2603 whose value was between 128 and 255 caused PCRE to crash.
2604
2605
2606 Version 4.1 12-Mar-03
2607 ---------------------
2608
2609 1. Compiling with gcc -pedantic found a couple of places where casts were
2610 needed, and a string in dftables.c that was longer than standard compilers are
2611 required to support.
2612
2613 2. Compiling with Sun's compiler found a few more places where the code could
2614 be tidied up in order to avoid warnings.
2615
2616 3. The variables for cross-compiling were called HOST_CC and HOST_CFLAGS; the
2617 first of these names is deprecated in the latest Autoconf in favour of the name
2618 CC_FOR_BUILD, because "host" is typically used to mean the system on which the
2619 compiled code will be run. I can't find a reference for HOST_CFLAGS, but by
2620 analogy I have changed it to CFLAGS_FOR_BUILD.
2621
2622 4. Added -no-undefined to the linking command in the Makefile, because this is
2623 apparently helpful for Windows. To make it work, also added "-L. -lpcre" to the
2624 linking step for the pcreposix library.
2625
2626 5. PCRE was failing to diagnose the case of two named groups with the same
2627 name.
2628
2629 6. A problem with one of PCRE's optimizations was discovered. PCRE remembers a
2630 literal character that is needed in the subject for a match, and scans along to
2631 ensure that it is present before embarking on the full matching process. This
2632 saves time in cases of nested unlimited repeats that are never going to match.
2633 Problem: the scan can take a lot of time if the subject is very long (e.g.
2634 megabytes), thus penalizing straightforward matches. It is now done only if the
2635 amount of subject to be scanned is less than 1000 bytes.
2636
2637 7. A lesser problem with the same optimization is that it was recording the
2638 first character of an anchored pattern as "needed", thus provoking a search
2639 right along the subject, even when the first match of the pattern was going to
2640 fail. The "needed" character is now not set for anchored patterns, unless it
2641 follows something in the pattern that is of non-fixed length. Thus, it still
2642 fulfils its original purpose of finding quick non-matches in cases of nested
2643 unlimited repeats, but isn't used for simple anchored patterns such as /^abc/.
2644
2645
2646 Version 4.0 17-Feb-03
2647 ---------------------
2648
2649 1. If a comment in an extended regex that started immediately after a meta-item
2650 extended to the end of string, PCRE compiled incorrect data. This could lead to
2651 all kinds of weird effects. Example: /#/ was bad; /()#/ was bad; /a#/ was not.
2652
2653 2. Moved to autoconf 2.53 and libtool 1.4.2.
2654
2655 3. Perl 5.8 no longer needs "use utf8" for doing UTF-8 things. Consequently,
2656 the special perltest8 script is no longer needed - all the tests can be run
2657 from a single perltest script.
2658
2659 4. From 5.004, Perl has not included the VT character (0x0b) in the set defined
2660 by \s. It has now been removed in PCRE. This means it isn't recognized as
2661 whitespace in /x regexes too, which is the same as Perl. Note that the POSIX
2662 class [:space:] *does* include VT, thereby creating a mess.
2663
2664 5. Added the class [:blank:] (a GNU extension from Perl 5.8) to match only
2665 space and tab.
2666
2667 6. Perl 5.005 was a long time ago. It's time to amalgamate the tests that use
2668 its new features into the main test script, reducing the number of scripts.
2669
2670 7. Perl 5.8 has changed the meaning of patterns like /a(?i)b/. Earlier versions
2671 were backward compatible, and made the (?i) apply to the whole pattern, as if
2672 /i were given. Now it behaves more logically, and applies the option setting
2673 only to what follows. PCRE has been changed to follow suit. However, if it
2674 finds options settings right at the start of the pattern, it extracts them into
2675 the global options, as before. Thus, they show up in the info data.
2676
2677 8. Added support for the \Q...\E escape sequence. Characters in between are
2678 treated as literals. This is slightly different from Perl in that $ and @ are
2679 also handled as literals inside the quotes. In Perl, they will cause variable
2680 interpolation. Note the following examples:
2681
2682 Pattern PCRE matches Perl matches
2683
2684 \Qabc$xyz\E abc$xyz abc followed by the contents of $xyz
2685 \Qabc\$xyz\E abc\$xyz abc\$xyz
2686 \Qabc\E\$\Qxyz\E abc$xyz abc$xyz
2687
2688 For compatibility with Perl, \Q...\E sequences are recognized inside character
2689 classes as well as outside them.
2690
2691 9. Re-organized 3 code statements in pcretest to avoid "overflow in
2692 floating-point constant arithmetic" warnings from a Microsoft compiler. Added a
2693 (size_t) cast to one statement in pcretest and one in pcreposix to avoid
2694 signed/unsigned warnings.
2695
2696 10. SunOS4 doesn't have strtoul(). This was used only for unpicking the -o
2697 option for pcretest, so I've replaced it by a simple function that does just
2698 that job.
2699
2700 11. pcregrep was ending with code 0 instead of 2 for the commands "pcregrep" or
2701 "pcregrep -".
2702
2703 12. Added "possessive quantifiers" ?+, *+, ++, and {,}+ which come from Sun's
2704 Java package. This provides some syntactic sugar for simple cases of what my
2705 documentation calls "once-only subpatterns". A pattern such as x*+ is the same
2706 as (?>x*). In other words, if what is inside (?>...) is just a single repeated
2707 item, you can use this simplified notation. Note that only makes sense with
2708 greedy quantifiers. Consequently, the use of the possessive quantifier forces
2709 greediness, whatever the setting of the PCRE_UNGREEDY option.
2710
2711 13. A change of greediness default within a pattern was not taking effect at
2712 the current level for patterns like /(b+(?U)a+)/. It did apply to parenthesized
2713 subpatterns that followed. Patterns like /b+(?U)a+/ worked because the option
2714 was abstracted outside.
2715
2716 14. PCRE now supports the \G assertion. It is true when the current matching
2717 position is at the start point of the match. This differs from \A when the
2718 starting offset is non-zero. Used with the /g option of pcretest (or similar
2719 code), it works in the same way as it does for Perl's /g option. If all
2720 alternatives of a regex begin with \G, the expression is anchored to the start
2721 match position, and the "anchored" flag is set in the compiled expression.
2722
2723 15. Some bugs concerning the handling of certain option changes within patterns
2724 have been fixed. These applied to options other than (?ims). For example,
2725 "a(?x: b c )d" did not match "XabcdY" but did match "Xa b c dY". It should have
2726 been the other way round. Some of this was related to change 7 above.
2727
2728 16. PCRE now gives errors for /[.x.]/ and /[=x=]/ as unsupported POSIX
2729 features, as Perl does. Previously, PCRE gave the warnings only for /[[.x.]]/
2730 and /[[=x=]]/. PCRE now also gives an error for /[:name:]/ because it supports
2731 POSIX classes only within a class (e.g. /[[:alpha:]]/).
2732
2733 17. Added support for Perl's \C escape. This matches one byte, even in UTF8
2734 mode. Unlike ".", it always matches newline, whatever the setting of
2735 PCRE_DOTALL. However, PCRE does not permit \C to appear in lookbehind
2736 assertions. Perl allows it, but it doesn't (in general) work because it can't
2737 calculate the length of the lookbehind. At least, that's the case for Perl
2738 5.8.0 - I've been told they are going to document that it doesn't work in
2739 future.
2740
2741 18. Added an error diagnosis for escapes that PCRE does not support: these are
2742 \L, \l, \N, \P, \p, \U, \u, and \X.
2743
2744 19. Although correctly diagnosing a missing ']' in a character class, PCRE was
2745 reading past the end of the pattern in cases such as /[abcd/.
2746
2747 20. PCRE was getting more memory than necessary for patterns with classes that
2748 contained both POSIX named classes and other characters, e.g. /[[:space:]abc/.
2749
2750 21. Added some code, conditional on #ifdef VPCOMPAT, to make life easier for
2751 compiling PCRE for use with Virtual Pascal.
2752
2753 22. Small fix to the Makefile to make it work properly if the build is done
2754 outside the source tree.
2755
2756 23. Added a new extension: a condition to go with recursion. If a conditional
2757 subpattern starts with (?(R) the "true" branch is used if recursion has
2758 happened, whereas the "false" branch is used only at the top level.
2759
2760 24. When there was a very long string of literal characters (over 255 bytes
2761 without UTF support, over 250 bytes with UTF support), the computation of how
2762 much memory was required could be incorrect, leading to segfaults or other
2763 strange effects.
2764
2765 25. PCRE was incorrectly assuming anchoring (either to start of subject or to
2766 start of line for a non-DOTALL pattern) when a pattern started with (.*) and
2767 there was a subsequent back reference to those brackets. This meant that, for
2768 example, /(.*)\d+\1/ failed to match "abc123bc". Unfortunately, it isn't
2769 possible to check for precisely this case. All we can do is abandon the
2770 optimization if .* occurs inside capturing brackets when there are any back
2771 references whatsoever. (See below for a better fix that came later.)
2772
2773 26. The handling of the optimization for finding the first character of a
2774 non-anchored pattern, and for finding a character that is required later in the
2775 match were failing in some cases. This didn't break the matching; it just
2776 failed to optimize when it could. The way this is done has been re-implemented.
2777
2778 27. Fixed typo in error message for invalid (?R item (it said "(?p").
2779
2780 28. Added a new feature that provides some of the functionality that Perl
2781 provides with (?{...}). The facility is termed a "callout". The way it is done
2782 in PCRE is for the caller to provide an optional function, by setting
2783 pcre_callout to its entry point. Like pcre_malloc and pcre_free, this is a
2784 global variable. By default it is unset, which disables all calling out. To get
2785 the function called, the regex must include (?C) at appropriate points. This
2786 is, in fact, equivalent to (?C0), and any number <= 255 may be given with (?C).
2787 This provides a means of identifying different callout points. When PCRE
2788 reaches such a point in the regex, if pcre_callout has been set, the external
2789 function is called. It is provided with data in a structure called
2790 pcre_callout_block, which is defined in pcre.h. If the function returns 0,
2791 matching continues; if it returns a non-zero value, the match at the current
2792 point fails. However, backtracking will occur if possible. [This was changed
2793 later and other features added - see item 49 below.]
2794
2795 29. pcretest is upgraded to test the callout functionality. It provides a
2796 callout function that displays information. By default, it shows the start of
2797 the match and the current position in the text. There are some new data escapes
2798 to vary what happens:
2799
2800 \C+ in addition, show current contents of captured substrings
2801 \C- do not supply a callout function
2802 \C!n return 1 when callout number n is reached
2803 \C!n!m return 1 when callout number n is reached for the mth time
2804
2805 30. If pcregrep was called with the -l option and just a single file name, it
2806 output "<stdin>" if a match was found, instead of the file name.
2807
2808 31. Improve the efficiency of the POSIX API to PCRE. If the number of capturing
2809 slots is less than POSIX_MALLOC_THRESHOLD, use a block on the stack to pass to
2810 pcre_exec(). This saves a malloc/free per call. The default value of
2811 POSIX_MALLOC_THRESHOLD is 10; it can be changed by --with-posix-malloc-threshold
2812 when configuring.
2813
2814 32. The default maximum size of a compiled pattern is 64K. There have been a
2815 few cases of people hitting this limit. The code now uses macros to handle the
2816 storing of links as offsets within the compiled pattern. It defaults to 2-byte
2817 links, but this can be changed to 3 or 4 bytes by --with-link-size when
2818 configuring. Tests 2 and 5 work only with 2-byte links because they output
2819 debugging information about compiled patterns.
2820
2821 33. Internal code re-arrangements:
2822
2823 (a) Moved the debugging function for printing out a compiled regex into
2824 its own source file (printint.c) and used #include to pull it into
2825 pcretest.c and, when DEBUG is defined, into pcre.c, instead of having two
2826 separate copies.
2827
2828 (b) Defined the list of op-code names for debugging as a macro in
2829 internal.h so that it is next to the definition of the opcodes.
2830
2831 (c) Defined a table of op-code lengths for simpler skipping along compiled
2832 code. This is again a macro in internal.h so that it is next to the
2833 definition of the opcodes.
2834
2835 34. Added support for recursive calls to individual subpatterns, along the
2836 lines of Robin Houston's patch (but implemented somewhat differently).
2837
2838 35. Further mods to the Makefile to help Win32. Also, added code to pcregrep to
2839 allow it to read and process whole directories in Win32. This code was
2840 contributed by Lionel Fourquaux; it has not been tested by me.
2841
2842 36. Added support for named subpatterns. The Python syntax (?P<name>...) is
2843 used to name a group. Names consist of alphanumerics and underscores, and must
2844 be unique. Back references use the syntax (?P=name) and recursive calls use
2845 (?P>name) which is a PCRE extension to the Python extension. Groups still have
2846 numbers. The function pcre_fullinfo() can be used after compilation to extract
2847 a name/number map. There are three relevant calls:
2848
2849 PCRE_INFO_NAMEENTRYSIZE yields the size of each entry in the map
2850 PCRE_INFO_NAMECOUNT yields the number of entries
2851 PCRE_INFO_NAMETABLE yields a pointer to the map.
2852
2853 The map is a vector of fixed-size entries. The size of each entry depends on
2854 the length of the longest name used. The first two bytes of each entry are the
2855 group number, most significant byte first. There follows the corresponding
2856 name, zero terminated. The names are in alphabetical order.
2857
2858 37. Make the maximum literal string in the compiled code 250 for the non-UTF-8
2859 case instead of 255. Making it the same both with and without UTF-8 support
2860 means that the same test output works with both.
2861
2862 38. There was a case of malloc(0) in the POSIX testing code in pcretest. Avoid
2863 calling malloc() with a zero argument.
2864
2865 39. Change 25 above had to resort to a heavy-handed test for the .* anchoring
2866 optimization. I've improved things by keeping a bitmap of backreferences with
2867 numbers 1-31 so that if .* occurs inside capturing brackets that are not in
2868 fact referenced, the optimization can be applied. It is unlikely that a
2869 relevant occurrence of .* (i.e. one which might indicate anchoring or forcing
2870 the match to follow \n) will appear inside brackets with a number greater than
2871 31, but if it does, any back reference > 31 suppresses the optimization.
2872
2873 40. Added a new compile-time option PCRE_NO_AUTO_CAPTURE. This has the effect
2874 of disabling numbered capturing parentheses. Any opening parenthesis that is
2875 not followed by ? behaves as if it were followed by ?: but named parentheses
2876 can still be used for capturing (and they will acquire numbers in the usual
2877 way).
2878
2879 41. Redesigned the return codes from the match() function into yes/no/error so
2880 that errors can be passed back from deep inside the nested calls. A malloc
2881 failure while inside a recursive subpattern call now causes the
2882 PCRE_ERROR_NOMEMORY return instead of quietly going wrong.
2883
2884 42. It is now possible to set a limit on the number of times the match()
2885 function is called in a call to pcre_exec(). This facility makes it possible to
2886 limit the amount of recursion and backtracking, though not in a directly
2887 obvious way, because the match() function is used in a number of different
2888 circumstances. The count starts from zero for each position in the subject
2889 string (for non-anchored patterns). The default limit is, for compatibility, a
2890 large number, namely 10 000 000. You can change this in two ways:
2891
2892 (a) When configuring PCRE before making, you can use --with-match-limit=n
2893 to set a default value for the compiled library.
2894
2895 (b) For each call to pcre_exec(), you can pass a pcre_extra block in which
2896 a different value is set. See 45 below.
2897
2898 If the limit is exceeded, pcre_exec() returns PCRE_ERROR_MATCHLIMIT.
2899
2900 43. Added a new function pcre_config(int, void *) to enable run-time extraction
2901 of things that can be changed at compile time. The first argument specifies
2902 what is wanted and the second points to where the information is to be placed.
2903 The current list of available information is:
2904
2905 PCRE_CONFIG_UTF8
2906
2907 The output is an integer that is set to one if UTF-8 support is available;
2908 otherwise it is set to zero.
2909
2910 PCRE_CONFIG_NEWLINE
2911
2912 The output is an integer that it set to the value of the code that is used for
2913 newline. It is either LF (10) or CR (13).
2914
2915 PCRE_CONFIG_LINK_SIZE
2916
2917 The output is an integer that contains the number of bytes used for internal
2918 linkage in compiled expressions. The value is 2, 3, or 4. See item 32 above.
2919
2920 PCRE_CONFIG_POSIX_MALLOC_THRESHOLD
2921
2922 The output is an integer that contains the threshold above which the POSIX
2923 interface uses malloc() for output vectors. See item 31 above.
2924
2925 PCRE_CONFIG_MATCH_LIMIT
2926
2927 The output is an unsigned integer that contains the default limit of the number
2928 of match() calls in a pcre_exec() execution. See 42 above.
2929
2930 44. pcretest has been upgraded by the addition of the -C option. This causes it
2931 to extract all the available output from the new pcre_config() function, and to
2932 output it. The program then exits immediately.
2933
2934 45. A need has arisen to pass over additional data with calls to pcre_exec() in
2935 order to support additional features. One way would have been to define
2936 pcre_exec2() (for example) with extra arguments, but this would not have been
2937 extensible, and would also have required all calls to the original function to
2938 be mapped to the new one. Instead, I have chosen to extend the mechanism that
2939 is used for passing in "extra" data from pcre_study().
2940
2941 The pcre_extra structure is now exposed and defined in pcre.h. It currently
2942 contains the following fields:
2943
2944 flags a bitmap indicating which of the following fields are set
2945 study_data opaque data from pcre_study()
2946 match_limit a way of specifying a limit on match() calls for a specific
2947 call to pcre_exec()
2948 callout_data data for callouts (see 49 below)
2949
2950 The flag bits are also defined in pcre.h, and are
2951
2952 PCRE_EXTRA_STUDY_DATA
2953 PCRE_EXTRA_MATCH_LIMIT
2954 PCRE_EXTRA_CALLOUT_DATA
2955
2956 The pcre_study() function now returns one of these new pcre_extra blocks, with
2957 the actual study data pointed to by the study_data field, and the
2958 PCRE_EXTRA_STUDY_DATA flag set. This can be passed directly to pcre_exec() as
2959 before. That is, this change is entirely upwards-compatible and requires no
2960 change to existing code.
2961
2962 If you want to pass in additional data to pcre_exec(), you can either place it
2963 in a pcre_extra block provided by pcre_study(), or create your own pcre_extra
2964 block.
2965
2966 46. pcretest has been extended to test the PCRE_EXTRA_MATCH_LIMIT feature. If a
2967 data string contains the escape sequence \M, pcretest calls pcre_exec() several
2968 times with different match limits, until it finds the minimum value needed for
2969 pcre_exec() to complete. The value is then output. This can be instructive; for
2970 most simple matches the number is quite small, but for pathological cases it
2971 gets very large very quickly.
2972
2973 47. There's a new option for pcre_fullinfo() called PCRE_INFO_STUDYSIZE. It
2974 returns the size of the data block pointed to by the study_data field in a
2975 pcre_extra block, that is, the value that was passed as the argument to
2976 pcre_malloc() when PCRE was getting memory in which to place the information
2977 created by pcre_study(). The fourth argument should point to a size_t variable.
2978 pcretest has been extended so that this information is shown after a successful
2979 pcre_study() call when information about the compiled regex is being displayed.
2980
2981 48. Cosmetic change to Makefile: there's no need to have / after $(DESTDIR)
2982 because what follows is always an absolute path. (Later: it turns out that this
2983 is more than cosmetic for MinGW, because it doesn't like empty path
2984 components.)
2985
2986 49. Some changes have been made to the callout feature (see 28 above):
2987
2988 (i) A callout function now has three choices for what it returns:
2989
2990 0 => success, carry on matching
2991 > 0 => failure at this point, but backtrack if possible
2992 < 0 => serious error, return this value from pcre_exec()
2993
2994 Negative values should normally be chosen from the set of PCRE_ERROR_xxx
2995 values. In particular, returning PCRE_ERROR_NOMATCH forces a standard
2996 "match failed" error. The error number PCRE_ERROR_CALLOUT is reserved for
2997 use by callout functions. It will never be used by PCRE itself.
2998
2999 (ii) The pcre_extra structure (see 45 above) has a void * field called
3000 callout_data, with corresponding flag bit PCRE_EXTRA_CALLOUT_DATA. The
3001 pcre_callout_block structure has a field of the same name. The contents of
3002 the field passed in the pcre_extra structure are passed to the callout
3003 function in the corresponding field in the callout block. This makes it
3004 easier to use the same callout-containing regex from multiple threads. For
3005 testing, the pcretest program has a new data escape
3006
3007 \C*n pass the number n (may be negative) as callout_data
3008
3009 If the callout function in pcretest receives a non-zero value as
3010 callout_data, it returns that value.
3011
3012 50. Makefile wasn't handling CFLAGS properly when compiling dftables. Also,
3013 there were some redundant $(CFLAGS) in commands that are now specified as
3014 $(LINK), which already includes $(CFLAGS).
3015
3016 51. Extensions to UTF-8 support are listed below. These all apply when (a) PCRE
3017 has been compiled with UTF-8 support *and* pcre_compile() has been compiled
3018 with the PCRE_UTF8 flag. Patterns that are compiled without that flag assume
3019 one-byte characters throughout. Note that case-insensitive matching applies
3020 only to characters whose values are less than 256. PCRE doesn't support the
3021 notion of cases for higher-valued characters.
3022
3023 (i) A character class whose characters are all within 0-255 is handled as
3024 a bit map, and the map is inverted for negative classes. Previously, a
3025 character > 255 always failed to match such a class; however it should
3026 match if the class was a negative one (e.g. [^ab]). This has been fixed.
3027
3028 (ii) A negated character class with a single character < 255 is coded as
3029 "not this character" (OP_NOT). This wasn't working properly when the test
3030 character was multibyte, either singly or repeated.
3031
3032 (iii) Repeats of multibyte characters are now handled correctly in UTF-8
3033 mode, for example: \x{100}{2,3}.
3034
3035 (iv) The character escapes \b, \B, \d, \D, \s, \S, \w, and \W (either
3036 singly or repeated) now correctly test multibyte characters. However,
3037 PCRE doesn't recognize any characters with values greater than 255 as
3038 digits, spaces, or word characters. Such characters always match \D, \S,
3039 and \W, and never match \d, \s, or \w.
3040
3041 (v) Classes may now contain characters and character ranges with values
3042 greater than 255. For example: [ab\x{100}-\x{400}].
3043
3044 (vi) pcregrep now has a --utf-8 option (synonym -u) which makes it call
3045 PCRE in UTF-8 mode.
3046
3047 52. The info request value PCRE_INFO_FIRSTCHAR has been renamed
3048 PCRE_INFO_FIRSTBYTE because it is a byte value. However, the old name is
3049 retained for backwards compatibility. (Note that LASTLITERAL is also a byte
3050 value.)
3051
3052 53. The single man page has become too large. I have therefore split it up into
3053 a number of separate man pages. These also give rise to individual HTML pages;
3054 these are now put in a separate directory, and there is an index.html page that
3055 lists them all. Some hyperlinking between the pages has been installed.
3056
3057 54. Added convenience functions for handling named capturing parentheses.
3058
3059 55. Unknown escapes inside character classes (e.g. [\M]) and escapes that
3060 aren't interpreted therein (e.g. [\C]) are literals in Perl. This is now also
3061 true in PCRE, except when the PCRE_EXTENDED option is set, in which case they
3062 are faulted.
3063
3064 56. Introduced HOST_CC and HOST_CFLAGS which can be set in the environment when
3065 calling configure. These values are used when compiling the dftables.c program
3066 which is run to generate the source of the default character tables. They
3067 default to the values of CC and CFLAGS. If you are cross-compiling PCRE,
3068 you will need to set these values.
3069
3070 57. Updated the building process for Windows DLL, as provided by Fred Cox.
3071
3072
3073 Version 3.9 02-Jan-02
3074 ---------------------
3075
3076 1. A bit of extraneous text had somehow crept into the pcregrep documentation.
3077
3078 2. If --disable-static was given, the building process failed when trying to
3079 build pcretest and pcregrep. (For some reason it was using libtool to compile
3080 them, which is not right, as they aren't part of the library.)
3081
3082
3083 Version 3.8 18-Dec-01
3084 ---------------------
3085
3086 1. The experimental UTF-8 code was completely screwed up. It was packing the
3087 bytes in the wrong order. How dumb can you get?
3088
3089
3090 Version 3.7 29-Oct-01
3091 ---------------------
3092
3093 1. In updating pcretest to check change 1 of version 3.6, I screwed up.
3094 This caused pcretest, when used on the test data, to segfault. Unfortunately,
3095 this didn't happen under Solaris 8, where I normally test things.
3096
3097 2. The Makefile had to be changed to make it work on BSD systems, where 'make'
3098 doesn't seem to recognize that ./xxx and xxx are the same file. (This entry
3099 isn't in ChangeLog distributed with 3.7 because I forgot when I hastily made
3100 this fix an hour or so after the initial 3.7 release.)
3101
3102
3103 Version 3.6 23-Oct-01
3104 ---------------------
3105
3106 1. Crashed with /(sens|respons)e and \1ibility/ and "sense and sensibility" if
3107 offsets passed as NULL with zero offset count.
3108
3109 2. The config.guess and config.sub files had not been updated when I moved to
3110 the latest autoconf.
3111
3112
3113 Version 3.5 15-Aug-01
3114 ---------------------
3115
3116 1. Added some missing #if !defined NOPOSIX conditionals in pcretest.c that
3117 had been forgotten.
3118
3119 2. By using declared but undefined structures, we can avoid using "void"
3120 definitions in pcre.h while keeping the internal definitions of the structures
3121 private.
3122
3123 3. The distribution is now built using autoconf 2.50 and libtool 1.4. From a
3124 user point of view, this means that both static and shared libraries are built
3125 by default, but this can be individually controlled. More of the work of
3126 handling this static/shared cases is now inside libtool instead of PCRE's make
3127 file.
3128
3129 4. The pcretest utility is now installed along with pcregrep because it is
3130 useful for users (to test regexs) and by doing this, it automatically gets
3131 relinked by libtool. The documentation has been turned into a man page, so
3132 there are now .1, .txt, and .html versions in /doc.
3133
3134 5. Upgrades to pcregrep:
3135 (i) Added long-form option names like gnu grep.
3136 (ii) Added --help to list all options with an explanatory phrase.
3137 (iii) Added -r, --recursive to recurse into sub-directories.
3138 (iv) Added -f, --file to read patterns from a file.
3139
3140 6. pcre_exec() was referring to its "code" argument before testing that
3141 argument for NULL (and giving an error if it was NULL).
3142
3143 7. Upgraded Makefile.in to allow for compiling in a different directory from
3144 the source directory.
3145
3146 8. Tiny buglet in pcretest: when pcre_fullinfo() was called to retrieve the
3147 options bits, the pointer it was passed was to an int instead of to an unsigned
3148 long int. This mattered only on 64-bit systems.
3149
3150 9. Fixed typo (3.4/1) in pcre.h again. Sigh. I had changed pcre.h (which is
3151 generated) instead of pcre.in, which it its source. Also made the same change
3152 in several of the .c files.
3153
3154 10. A new release of gcc defines printf() as a macro, which broke pcretest
3155 because it had an ifdef in the middle of a string argument for printf(). Fixed
3156 by using separate calls to printf().
3157
3158 11. Added --enable-newline-is-cr and --enable-newline-is-lf to the configure
3159 script, to force use of CR or LF instead of \n in the source. On non-Unix
3160 systems, the value can be set in config.h.
3161
3162 12. The limit of 200 on non-capturing parentheses is a _nesting_ limit, not an
3163 absolute limit. Changed the text of the error message to make this clear, and
3164 likewise updated the man page.
3165
3166 13. The limit of 99 on the number of capturing subpatterns has been removed.
3167 The new limit is 65535, which I hope will not be a "real" limit.
3168
3169
3170 Version 3.4 22-Aug-00
3171 ---------------------
3172
3173 1. Fixed typo in pcre.h: unsigned const char * changed to const unsigned char *.
3174
3175 2. Diagnose condition (?(0) as an error instead of crashing on matching.
3176
3177
3178 Version 3.3 01-Aug-00
3179 ---------------------
3180
3181 1. If an octal character was given, but the value was greater than \377, it
3182 was not getting masked to the least significant bits, as documented. This could
3183 lead to crashes in some systems.
3184
3185 2. Perl 5.6 (if not earlier versions) accepts classes like [a-\d] and treats
3186 the hyphen as a literal. PCRE used to give an error; it now behaves like Perl.
3187
3188 3. Added the functions pcre_free_substring() and pcre_free_substring_list().
3189 These just pass their arguments on to (pcre_free)(), but they are provided
3190 because some uses of PCRE bind it to non-C systems that can call its functions,
3191 but cannot call free() or pcre_free() directly.
3192
3193 4. Add "make test" as a synonym for "make check". Corrected some comments in
3194 the Makefile.
3195
3196 5. Add $(DESTDIR)/ in front of all the paths in the "install" target in the
3197 Makefile.
3198
3199 6. Changed the name of pgrep to pcregrep, because Solaris has introduced a
3200 command called pgrep for grepping around the active processes.
3201
3202 7. Added the beginnings of support for UTF-8 character strings.
3203
3204 8. Arranged for the Makefile to pass over the settings of CC, CFLAGS, and
3205 RANLIB to ./ltconfig so that they are used by libtool. I think these are all
3206 the relevant ones. (AR is not passed because ./ltconfig does its own figuring
3207 out for the ar command.)
3208
3209
3210 Version 3.2 12-May-00
3211 ---------------------
3212
3213 This is purely a bug fixing release.
3214
3215 1. If the pattern /((Z)+|A)*/ was matched agained ZABCDEFG it matched Z instead
3216 of ZA. This was just one example of several cases that could provoke this bug,
3217 which was introduced by change 9 of version 2.00. The code for breaking
3218 infinite loops after an iteration that matches an empty string was't working
3219 correctly.
3220
3221 2. The pcretest program was not imitating Perl correctly for the pattern /a*/g
3222 when matched against abbab (for example). After matching an empty string, it
3223 wasn't forcing anchoring when setting PCRE_NOTEMPTY for the next attempt; this
3224 caused it to match further down the string than it should.
3225
3226 3. The code contained an inclusion of sys/types.h. It isn't clear why this
3227 was there because it doesn't seem to be needed, and it causes trouble on some
3228 systems, as it is not a Standard C header. It has been removed.
3229
3230 4. Made 4 silly changes to the source to avoid stupid compiler warnings that
3231 were reported on the Macintosh. The changes were from
3232
3233 while ((c = *(++ptr)) != 0 && c != '\n');
3234 to
3235 while ((c = *(++ptr)) != 0 && c != '\n') ;
3236
3237 Totally extraordinary, but if that's what it takes...
3238
3239 5. PCRE is being used in one environment where neither memmove() nor bcopy() is
3240 available. Added HAVE_BCOPY and an autoconf test for it; if neither
3241 HAVE_MEMMOVE nor HAVE_BCOPY is set, use a built-in emulation function which
3242 assumes the way PCRE uses memmove() (always moving upwards).
3243
3244 6. PCRE is being used in one environment where strchr() is not available. There
3245 was only one use in pcre.c, and writing it out to avoid strchr() probably gives
3246 faster code anyway.
3247
3248
3249 Version 3.1 09-Feb-00
3250 ---------------------
3251
3252 The only change in this release is the fixing of some bugs in Makefile.in for
3253 the "install" target:
3254
3255 (1) It was failing to install pcreposix.h.
3256
3257 (2) It was overwriting the pcre.3 man page with the pcreposix.3 man page.
3258
3259
3260 Version 3.0 01-Feb-00
3261 ---------------------
3262
3263 1. Add support for the /+ modifier to perltest (to output $` like it does in
3264 pcretest).
3265
3266 2. Add support for the /g modifier to perltest.
3267
3268 3. Fix pcretest so that it behaves even more like Perl for /g when the pattern
3269 matches null strings.
3270
3271 4. Fix perltest so that it doesn't do unwanted things when fed an empty
3272 pattern. Perl treats empty patterns specially - it reuses the most recent
3273 pattern, which is not what we want. Replace // by /(?#)/ in order to avoid this
3274 effect.
3275
3276 5. The POSIX interface was broken in that it was just handing over the POSIX
3277 captured string vector to pcre_exec(), but (since release 2.00) PCRE has
3278 required a bigger vector, with some working space on the end. This means that
3279 the POSIX wrapper now has to get and free some memory, and copy the results.
3280
3281 6. Added some simple autoconf support, placing the test data and the
3282 documentation in separate directories, re-organizing some of the
3283 information files, and making it build pcre-config (a GNU standard). Also added
3284 libtool support for building PCRE as a shared library, which is now the
3285 default.
3286
3287 7. Got rid of the leading zero in the definition of PCRE_MINOR because 08 and
3288 09 are not valid octal constants. Single digits will be used for minor values
3289 less than 10.
3290
3291 8. Defined REG_EXTENDED and REG_NOSUB as zero in the POSIX header, so that
3292 existing programs that set these in the POSIX interface can use PCRE without
3293 modification.
3294
3295 9. Added a new function, pcre_fullinfo() with an extensible interface. It can
3296 return all that pcre_info() returns, plus additional data. The pcre_info()
3297 function is retained for compatibility, but is considered to be obsolete.
3298
3299 10. Added experimental recursion feature (?R) to handle one common case that
3300 Perl 5.6 will be able to do with (?p{...}).
3301
3302 11. Added support for POSIX character classes like [:alpha:], which Perl is
3303 adopting.
3304
3305
3306 Version 2.08 31-Aug-99
3307 ----------------------
3308
3309 1. When startoffset was not zero and the pattern began with ".*", PCRE was not
3310 trying to match at the startoffset position, but instead was moving forward to
3311 the next newline as if a previous match had failed.
3312
3313 2. pcretest was not making use of PCRE_NOTEMPTY when repeating for /g and /G,
3314 and could get into a loop if a null string was matched other than at the start
3315 of the subject.
3316
3317 3. Added definitions of PCRE_MAJOR and PCRE_MINOR to pcre.h so the version can
3318 be distinguished at compile time, and for completeness also added PCRE_DATE.
3319
3320 5. Added Paul Sokolovsky's minor changes to make it easy to compile a Win32 DLL
3321 in GnuWin32 environments.
3322
3323
3324 Version 2.07 29-Jul-99
3325 ----------------------
3326
3327 1. The documentation is now supplied in plain text form and HTML as well as in
3328 the form of man page sources.
3329
3330 2. C++ compilers don't like assigning (void *) values to other pointer types.
3331 In particular this affects malloc(). Although there is no problem in Standard
3332 C, I've put in casts to keep C++ compilers happy.
3333
3334 3. Typo on pcretest.c; a cast of (unsigned char *) in the POSIX regexec() call
3335 should be (const char *).
3336
3337 4. If NOPOSIX is defined, pcretest.c compiles without POSIX support. This may
3338 be useful for non-Unix systems who don't want to bother with the POSIX stuff.
3339 However, I haven't made this a standard facility. The documentation doesn't
3340 mention it, and the Makefile doesn't support it.
3341
3342 5. The Makefile now contains an "install" target, with editable destinations at
3343 the top of the file. The pcretest program is not installed.
3344
3345 6. pgrep -V now gives the PCRE version number and date.
3346
3347 7. Fixed bug: a zero repetition after a literal string (e.g. /abcde{0}/) was
3348 causing the entire string to be ignored, instead of just the last character.
3349
3350 8. If a pattern like /"([^\\"]+|\\.)*"/ is applied in the normal way to a
3351 non-matching string, it can take a very, very long time, even for strings of
3352 quite modest length, because of the nested recursion. PCRE now does better in
3353 some of these cases. It does this by remembering the last required literal
3354 character in the pattern, and pre-searching the subject to ensure it is present
3355 before running the real match. In other words, it applies a heuristic to detect
3356 some types of certain failure quickly, and in the above example, if presented
3357 with a string that has no trailing " it gives "no match" very quickly.
3358
3359 9. A new runtime option PCRE_NOTEMPTY causes null string matches to be ignored;
3360 other alternatives are tried instead.
3361
3362
3363 Version 2.06 09-Jun-99
3364 ----------------------
3365
3366 1. Change pcretest's output for amount of store used to show just the code
3367 space, because the remainder (the data block) varies in size between 32-bit and
3368 64-bit systems.
3369
3370 2. Added an extra argument to pcre_exec() to supply an offset in the subject to
3371 start matching at. This allows lookbehinds to work when searching for multiple
3372 occurrences in a string.
3373
3374 3. Added additional options to pcretest for testing multiple occurrences:
3375
3376 /+ outputs the rest of the string that follows a match
3377 /g loops for multiple occurrences, using the new startoffset argument
3378 /G loops for multiple occurrences by passing an incremented pointer
3379
3380 4. PCRE wasn't doing the "first character" optimization for patterns starting
3381 with \b or \B, though it was doing it for other lookbehind assertions. That is,
3382 it wasn't noticing that a match for a pattern such as /\bxyz/ has to start with
3383 the letter 'x'. On long subject strings, this gives a significant speed-up.
3384
3385
3386 Version 2.05 21-Apr-99
3387 ----------------------
3388
3389 1. Changed the type of magic_number from int to long int so that it works
3390 properly on 16-bit systems.
3391
3392 2. Fixed a bug which caused patterns starting with .* not to work correctly
3393 when the subject string contained newline characters. PCRE was assuming
3394 anchoring for such patterns in all cases, which is not correct because .* will
3395 not pass a newline unless PCRE_DOTALL is set. It now assumes anchoring only if
3396 DOTALL is set at top level; otherwise it knows that patterns starting with .*
3397 must be retried after every newline in the subject.
3398
3399
3400 Version 2.04 18-Feb-99
3401 ----------------------
3402
3403 1. For parenthesized subpatterns with repeats whose minimum was zero, the
3404 computation of the store needed to hold the pattern was incorrect (too large).
3405 If such patterns were nested a few deep, this could multiply and become a real
3406 problem.
3407
3408 2. Added /M option to pcretest to show the memory requirement of a specific
3409 pattern. Made -m a synonym of -s (which does this globally) for compatibility.
3410
3411 3. Subpatterns of the form (regex){n,m} (i.e. limited maximum) were being
3412 compiled in such a way that the backtracking after subsequent failure was
3413 pessimal. Something like (a){0,3} was compiled as (a)?(a)?(a)? instead of
3414 ((a)((a)(a)?)?)? with disastrous performance if the maximum was of any size.
3415
3416
3417 Version 2.03 02-Feb-99
3418 ----------------------
3419
3420 1. Fixed typo and small mistake in man page.
3421
3422 2. Added 4th condition (GPL supersedes if conflict) and created separate
3423 LICENCE file containing the conditions.
3424
3425 3. Updated pcretest so that patterns such as /abc\/def/ work like they do in
3426 Perl, that is the internal \ allows the delimiter to be included in the
3427 pattern. Locked out the use of \ as a delimiter. If \ immediately follows
3428 the final delimiter, add \ to the end of the pattern (to test the error).
3429
3430 4. Added the convenience functions for extracting substrings after a successful
3431 match. Updated pcretest to make it able to test these functions.
3432
3433
3434 Version 2.02 14-Jan-99
3435 ----------------------
3436
3437 1. Initialized the working variables associated with each extraction so that
3438 their saving and restoring doesn't refer to uninitialized store.
3439
3440 2. Put dummy code into study.c in order to trick the optimizer of the IBM C
3441 compiler for OS/2 into generating correct code. Apparently IBM isn't going to
3442 fix the problem.
3443
3444 3. Pcretest: the timing code wasn't using LOOPREPEAT for timing execution
3445 calls, and wasn't printing the correct value for compiling calls. Increased the
3446 default value of LOOPREPEAT, and the number of significant figures in the
3447 times.
3448
3449 4. Changed "/bin/rm" in the Makefile to "-rm" so it works on Windows NT.
3450
3451 5. Renamed "deftables" as "dftables" to get it down to 8 characters, to avoid
3452 a building problem on Windows NT with a FAT file system.
3453
3454
3455 Version 2.01 21-Oct-98
3456 ----------------------
3457
3458 1. Changed the API for pcre_compile() to allow for the provision of a pointer
3459 to character tables built by pcre_maketables() in the current locale. If NULL
3460 is passed, the default tables are used.
3461
3462
3463 Version 2.00 24-Sep-98
3464 ----------------------
3465
3466 1. Since the (>?) facility is in Perl 5.005, don't require PCRE_EXTRA to enable
3467 it any more.
3468
3469 2. Allow quantification of (?>) groups, and make it work correctly.
3470
3471 3. The first character computation wasn't working for (?>) groups.
3472
3473 4. Correct the implementation of \Z (it is permitted to match on the \n at the
3474 end of the subject) and add 5.005's \z, which really does match only at the
3475 very end of the subject.
3476
3477 5. Remove the \X "cut" facility; Perl doesn't have it, and (?> is neater.
3478
3479 6. Remove the ability to specify CASELESS, MULTILINE, DOTALL, and
3480 DOLLAR_END_ONLY at runtime, to make it possible to implement the Perl 5.005
3481 localized options. All options to pcre_study() were also removed.
3482
3483 7. Add other new features from 5.005:
3484
3485 $(?<= positive lookbehind
3486 $(?<! negative lookbehind
3487 (?imsx-imsx) added the unsetting capability
3488 such a setting is global if at outer level; local otherwise
3489 (?imsx-imsx:) non-capturing groups with option setting
3490 (?(cond)re|re) conditional pattern matching
3491
3492 A backreference to itself in a repeated group matches the previous
3493 captured string.
3494
3495 8. General tidying up of studying (both automatic and via "study")
3496 consequential on the addition of new assertions.
3497
3498 9. As in 5.005, unlimited repeated groups that could match an empty substring
3499 are no longer faulted at compile time. Instead, the loop is forcibly broken at
3500 runtime if any iteration does actually match an empty substring.
3501
3502 10. Include the RunTest script in the distribution.
3503
3504 11. Added tests from the Perl 5.005_02 distribution. This showed up a few
3505 discrepancies, some of which were old and were also with respect to 5.004. They
3506 have now been fixed.
3507
3508
3509 Version 1.09 28-Apr-98
3510 ----------------------
3511
3512 1. A negated single character class followed by a quantifier with a minimum
3513 value of one (e.g. [^x]{1,6} ) was not compiled correctly. This could lead to
3514 program crashes, or just wrong answers. This did not apply to negated classes
3515 containing more than one character, or to minima other than one.
3516
3517
3518 Version 1.08 27-Mar-98
3519 ----------------------
3520
3521 1. Add PCRE_UNGREEDY to invert the greediness of quantifiers.
3522
3523 2. Add (?U) and (?X) to set PCRE_UNGREEDY and PCRE_EXTRA respectively. The
3524 latter must appear before anything that relies on it in the pattern.
3525
3526
3527 Version 1.07 16-Feb-98
3528 ----------------------
3529
3530 1. A pattern such as /((a)*)*/ was not being diagnosed as in error (unlimited
3531 repeat of a potentially empty string).
3532
3533
3534 Version 1.06 23-Jan-98
3535 ----------------------
3536
3537 1. Added Markus Oberhumer's little patches for C++.
3538
3539 2. Literal strings longer than 255 characters were broken.
3540
3541
3542 Version 1.05 23-Dec-97
3543 ----------------------
3544
3545 1. Negated character classes containing more than one character were failing if
3546 PCRE_CASELESS was set at run time.
3547
3548
3549 Version 1.04 19-Dec-97
3550 ----------------------
3551
3552 1. Corrected the man page, where some "const" qualifiers had been omitted.
3553
3554 2. Made debugging output print "{0,xxx}" instead of just "{,xxx}" to agree with
3555 input syntax.
3556
3557 3. Fixed memory leak which occurred when a regex with back references was
3558 matched with an offsets vector that wasn't big enough. The temporary memory
3559 that is used in this case wasn't being freed if the match failed.
3560
3561 4. Tidied pcretest to ensure it frees memory that it gets.
3562
3563 5. Temporary memory was being obtained in the case where the passed offsets
3564 vector was exactly big enough.
3565
3566 6. Corrected definition of offsetof() from change 5 below.
3567
3568 7. I had screwed up change 6 below and broken the rules for the use of
3569 setjmp(). Now fixed.
3570
3571
3572 Version 1.03 18-Dec-97
3573 ----------------------
3574
3575 1. A erroneous regex with a missing opening parenthesis was correctly
3576 diagnosed, but PCRE attempted to access brastack[-1], which could cause crashes
3577 on some systems.
3578
3579 2. Replaced offsetof(real_pcre, code) by offsetof(real_pcre, code[0]) because
3580 it was reported that one broken compiler failed on the former because "code" is
3581 also an independent variable.
3582
3583 3. The erroneous regex a[]b caused an array overrun reference.
3584
3585 4. A regex ending with a one-character negative class (e.g. /[^k]$/) did not
3586 fail on data ending with that character. (It was going on too far, and checking
3587 the next character, typically a binary zero.) This was specific to the
3588 optimized code for single-character negative classes.
3589
3590 5. Added a contributed patch from the TIN world which does the following:
3591
3592 + Add an undef for memmove, in case the the system defines a macro for it.
3593
3594 + Add a definition of offsetof(), in case there isn't one. (I don't know
3595 the reason behind this - offsetof() is part of the ANSI standard - but
3596 it does no harm).
3597
3598 + Reduce the ifdef's in pcre.c using macro DPRINTF, thereby eliminating
3599 most of the places where whitespace preceded '#'. I have given up and
3600 allowed the remaining 2 cases to be at the margin.
3601
3602 + Rename some variables in pcre to eliminate shadowing. This seems very
3603 pedantic, but does no harm, of course.
3604
3605 6. Moved the call to setjmp() into its own function, to get rid of warnings
3606 from gcc -Wall, and avoided calling it at all unless PCRE_EXTRA is used.
3607
3608 7. Constructs such as \d{8,} were compiling into the equivalent of
3609 \d{8}\d{0,65527} instead of \d{8}\d* which didn't make much difference to the
3610 outcome, but in this particular case used more store than had been allocated,
3611 which caused the bug to be discovered because it threw up an internal error.
3612
3613 8. The debugging code in both pcre and pcretest for outputting the compiled
3614 form of a regex was going wrong in the case of back references followed by
3615 curly-bracketed repeats.
3616
3617
3618 Version 1.02 12-Dec-97
3619 ----------------------
3620
3621 1. Typos in pcre.3 and comments in the source fixed.
3622
3623 2. Applied a contributed patch to get rid of places where it used to remove
3624 'const' from variables, and fixed some signed/unsigned and uninitialized
3625 variable warnings.
3626
3627 3. Added the "runtest" target to Makefile.
3628
3629 4. Set default compiler flag to -O2 rather than just -O.
3630
3631
3632 Version 1.01 19-Nov-97
3633 ----------------------
3634
3635 1. PCRE was failing to diagnose unlimited repeat of empty string for patterns
3636 like /([ab]*)*/, that is, for classes with more than one character in them.
3637
3638 2. Likewise, it wasn't diagnosing patterns with "once-only" subpatterns, such
3639 as /((?>a*))*/ (a PCRE_EXTRA facility).
3640
3641
3642 Version 1.00 18-Nov-97
3643 ----------------------
3644
3645 1. Added compile-time macros to support systems such as SunOS4 which don't have
3646 memmove() or strerror() but have other things that can be used instead.
3647
3648 2. Arranged that "make clean" removes the executables.
3649
3650
3651 Version 0.99 27-Oct-97
3652 ----------------------
3653
3654 1. Fixed bug in code for optimizing classes with only one character. It was
3655 initializing a 32-byte map regardless, which could cause it to run off the end
3656 of the memory it had got.
3657
3658 2. Added, conditional on PCRE_EXTRA, the proposed (?>REGEX) construction.
3659
3660
3661 Version 0.98 22-Oct-97
3662 ----------------------
3663
3664 1. Fixed bug in code for handling temporary memory usage when there are more
3665 back references than supplied space in the ovector. This could cause segfaults.
3666
3667
3668 Version 0.97 21-Oct-97
3669 ----------------------
3670
3671 1. Added the \X "cut" facility, conditional on PCRE_EXTRA.
3672
3673 2. Optimized negated single characters not to use a bit map.
3674
3675 3. Brought error texts together as macro definitions; clarified some of them;
3676 fixed one that was wrong - it said "range out of order" when it meant "invalid
3677 escape sequence".
3678
3679 4. Changed some char * arguments to const char *.
3680
3681 5. Added PCRE_NOTBOL and PCRE_NOTEOL (from POSIX).
3682
3683 6. Added the POSIX-style API wrapper in pcreposix.a and testing facilities in
3684 pcretest.
3685
3686
3687 Version 0.96 16-Oct-97
3688 ----------------------
3689
3690 1. Added a simple "pgrep" utility to the distribution.
3691
3692 2. Fixed an incompatibility with Perl: "{" is now treated as a normal character
3693 unless it appears in one of the precise forms "{ddd}", "{ddd,}", or "{ddd,ddd}"
3694 where "ddd" means "one or more decimal digits".
3695
3696 3. Fixed serious bug. If a pattern had a back reference, but the call to
3697 pcre_exec() didn't supply a large enough ovector to record the related
3698 identifying subpattern, the match always failed. PCRE now remembers the number
3699 of the largest back reference, and gets some temporary memory in which to save
3700 the offsets during matching if necessary, in order to ensure that
3701 backreferences always work.
3702
3703 4. Increased the compatibility with Perl in a number of ways:
3704
3705 (a) . no longer matches \n by default; an option PCRE_DOTALL is provided
3706 to request this handling. The option can be set at compile or exec time.
3707
3708 (b) $ matches before a terminating newline by default; an option
3709 PCRE_DOLLAR_ENDONLY is provided to override this (but not in multiline
3710 mode). The option can be set at compile or exec time.
3711
3712 (c) The handling of \ followed by a digit other than 0 is now supposed to be
3713 the same as Perl's. If the decimal number it represents is less than 10
3714 or there aren't that many previous left capturing parentheses, an octal
3715 escape is read. Inside a character class, it's always an octal escape,
3716 even if it is a single digit.
3717
3718 (d) An escaped but undefined alphabetic character is taken as a literal,
3719 unless PCRE_EXTRA is set. Currently this just reserves the remaining
3720 escapes.
3721
3722 (e) {0} is now permitted. (The previous item is removed from the compiled
3723 pattern).
3724
3725 5. Changed all the names of code files so that the basic parts are no longer
3726 than 10 characters, and abolished the teeny "globals.c" file.
3727
3728 6. Changed the handling of character classes; they are now done with a 32-byte
3729 bit map always.
3730
3731 7. Added the -d and /D options to pcretest to make it possible to look at the
3732 internals of compilation without having to recompile pcre.
3733
3734
3735 Version 0.95 23-Sep-97
3736 ----------------------
3737
3738 1. Fixed bug in pre-pass concerning escaped "normal" characters such as \x5c or
3739 \x20 at the start of a run of normal characters. These were being treated as
3740 real characters, instead of the source characters being re-checked.
3741
3742
3743 Version 0.94 18-Sep-97
3744 ----------------------
3745
3746 1. The functions are now thread-safe, with the caveat that the global variables
3747 containing pointers to malloc() and free() or alternative functions are the
3748 same for all threads.
3749
3750 2. Get pcre_study() to generate a bitmap of initial characters for non-
3751 anchored patterns when this is possible, and use it if passed to pcre_exec().
3752
3753
3754 Version 0.93 15-Sep-97
3755 ----------------------
3756
3757 1. /(b)|(:+)/ was computing an incorrect first character.
3758
3759 2. Add pcre_study() to the API and the passing of pcre_extra to pcre_exec(),
3760 but not actually doing anything yet.
3761
3762 3. Treat "-" characters in classes that cannot be part of ranges as literals,
3763 as Perl does (e.g. [-az] or [az-]).
3764
3765 4. Set the anchored flag if a branch starts with .* or .*? because that tests
3766 all possible positions.
3767
3768 5. Split up into different modules to avoid including unneeded functions in a
3769 compiled binary. However, compile and exec are still in one module. The "study"
3770 function is split off.
3771
3772 6. The character tables are now in a separate module whose source is generated
3773 by an auxiliary program - but can then be edited by hand if required. There are
3774 now no calls to isalnum(), isspace(), isdigit(), isxdigit(), tolower() or
3775 toupper() in the code.
3776
3777 7. Turn the malloc/free funtions variables into pcre_malloc and pcre_free and
3778 make them global. Abolish the function for setting them, as the caller can now
3779 set them directly.
3780
3781
3782 Version 0.92 11-Sep-97
3783 ----------------------
3784
3785 1. A repeat with a fixed maximum and a minimum of 1 for an ordinary character
3786 (e.g. /a{1,3}/) was broken (I mis-optimized it).
3787
3788 2. Caseless matching was not working in character classes if the characters in
3789 the pattern were in upper case.
3790
3791 3. Make ranges like [W-c] work in the same way as Perl for caseless matching.
3792
3793 4. Make PCRE_ANCHORED public and accept as a compile option.
3794
3795 5. Add an options word to pcre_exec() and accept PCRE_ANCHORED and
3796 PCRE_CASELESS at run time. Add escapes \A and \I to pcretest to cause it to
3797 pass them.
3798
3799 6. Give an error if bad option bits passed at compile or run time.
3800
3801 7. Add PCRE_MULTILINE at compile and exec time, and (?m) as well. Add \M to
3802 pcretest to cause it to pass that flag.
3803
3804 8. Add pcre_info(), to get the number of identifying subpatterns, the stored
3805 options, and the first character, if set.
3806
3807 9. Recognize C+ or C{n,m} where n >= 1 as providing a fixed starting character.
3808
3809
3810 Version 0.91 10-Sep-97
3811 ----------------------
3812
3813 1. PCRE was failing to diagnose unlimited repeats of subpatterns that could
3814 match the empty string as in /(a*)*/. It was looping and ultimately crashing.
3815
3816 2. PCRE was looping on encountering an indefinitely repeated back reference to
3817 a subpattern that had matched an empty string, e.g. /(a|)\1*/. It now does what
3818 Perl does - treats the match as successful.
3819
3820 ****

Properties

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

webmaster@exim.org
ViewVC Help
Powered by ViewVC 1.1.12