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

Contents of /code/trunk/ChangeLog

Parent Directory Parent Directory | Revision Log Revision Log


Revision 528 - (show annotations) (download)
Sat May 29 16:40:22 2010 UTC (4 years, 5 months ago) by ph10
File size: 187681 byte(s)
Add knowledge of \R to auto-possessify feature.

1 ChangeLog for PCRE
2 ------------------
3
4 Version 8.10 03 May-2010
5 ------------------------
6
7 1. Added support for (*MARK:ARG) and for ARG additions to PRUNE, SKIP, and
8 THEN.
9
10 2. (*ACCEPT) was not working when inside an atomic group.
11
12 3. Inside a character class, \B is treated as a literal by default, but
13 faulted if PCRE_EXTRA is set. This mimics Perl's behaviour (the -w option
14 causes the error). The code is unchanged, but I tidied the documentation.
15
16 4. Inside a character class, PCRE always treated \R and \X as literals,
17 whereas Perl faults them if its -w option is set. I have changed PCRE so
18 that it faults them when PCRE_EXTRA is set.
19
20 5. Added support for \N, which always matches any character other than
21 newline. (It is the same as "." when PCRE_DOTALL is not set.)
22
23 6. When compiling pcregrep with newer versions of gcc which may have
24 FORTIFY_SOURCE set, several warnings "ignoring return value of 'fwrite',
25 declared with attribute warn_unused_result" were given. Just casting the
26 result to (void) does not stop the warnings; a more elaborate fudge is
27 needed. I've used a macro to implement this.
28
29 7. Minor change to pcretest.c to avoid a compiler warning.
30
31 8. Added four artifical Unicode properties to help with an option to make
32 \s etc use properties (see next item). The new properties are: Xan
33 (alphanumeric), Xsp (Perl space), Xps (POSIX space), and Xwd (word).
34
35 9. Added PCRE_UCP to make \b, \d, \s, \w, and certain POSIX character classes
36 use Unicode properties. (*UCP) at the start of a pattern can be used to set
37 this option. Modified pcretest to add /W to test this facility. Added
38 REG_UCP to make it available via the POSIX interface.
39
40 10. Added --line-buffered to pcregrep.
41
42 11. In UTF-8 mode, if a pattern that was compiled with PCRE_CASELESS was
43 studied, and the match started with a letter with a code point greater than
44 127 whose first byte was different to the first byte of the other case of
45 the letter, the other case of this starting letter was not recognized.
46
47 12. pcreposix.c included pcre.h before including pcre_internal.h. This caused a
48 conflict in the definition of PCRE_EXP_DECL. I have removed the include of
49 pcre.h as pcre_internal.h includes pcre.h itself. (This may be a bit of
50 historical tidying that never got done.)
51
52 13. If a pattern that was studied started with a repeated Unicode property
53 test, for example, \p{Nd}+, there was the theoretical possibility of
54 setting up an incorrect bitmap of starting bytes, but fortunately it could
55 not have actually happened in practice until change 8 above was made (it
56 added property types that matched character-matching opcodes).
57
58 14. pcre_study() now recognizes \h, \v, and \R when constructing a bit map of
59 possible starting bytes for non-anchored patterns.
60
61 15. The "auto-possessify" feature of pcre_compile() now recognizes \R.
62
63
64
65 Version 8.02 19-Mar-2010
66 ------------------------
67
68 1. The Unicode data tables have been updated to Unicode 5.2.0.
69
70 2. Added the option --libs-cpp to pcre-config, but only when C++ support is
71 configured.
72
73 3. Updated the licensing terms in the pcregexp.pas file, as agreed with the
74 original author of that file, following a query about its status.
75
76 4. On systems that do not have stdint.h (e.g. Solaris), check for and include
77 inttypes.h instead. This fixes a bug that was introduced by change 8.01/8.
78
79 5. A pattern such as (?&t)*+(?(DEFINE)(?<t>.)) which has a possessive
80 quantifier applied to a forward-referencing subroutine call, could compile
81 incorrect code or give the error "internal error: previously-checked
82 referenced subpattern not found".
83
84 6. Both MS Visual Studio and Symbian OS have problems with initializing
85 variables to point to external functions. For these systems, therefore,
86 pcre_malloc etc. are now initialized to local functions that call the
87 relevant global functions.
88
89 7. There were two entries missing in the vectors called coptable and poptable
90 in pcre_dfa_exec.c. This could lead to memory accesses outsize the vectors.
91 I've fixed the data, and added a kludgy way of testing at compile time that
92 the lengths are correct (equal to the number of opcodes).
93
94 8. Following on from 7, I added a similar kludge to check the length of the
95 eint vector in pcreposix.c.
96
97 9. Error texts for pcre_compile() are held as one long string to avoid too
98 much relocation at load time. To find a text, the string is searched,
99 counting zeros. There was no check for running off the end of the string,
100 which could happen if a new error number was added without updating the
101 string.
102
103 10. \K gave a compile-time error if it appeared in a lookbehind assersion.
104
105 11. \K was not working if it appeared in an atomic group or in a group that
106 was called as a "subroutine", or in an assertion. Perl 5.11 documents that
107 \K is "not well defined" if used in an assertion. PCRE now accepts it if
108 the assertion is positive, but not if it is negative.
109
110 12. Change 11 fortuitously reduced the size of the stack frame used in the
111 "match()" function of pcre_exec.c by one pointer. Forthcoming
112 implementation of support for (*MARK) will need an extra pointer on the
113 stack; I have reserved it now, so that the stack frame size does not
114 decrease.
115
116 13. A pattern such as (?P<L1>(?P<L2>0)|(?P>L2)(?P>L1)) in which the only other
117 item in branch that calls a recursion is a subroutine call - as in the
118 second branch in the above example - was incorrectly given the compile-
119 time error "recursive call could loop indefinitely" because pcre_compile()
120 was not correctly checking the subroutine for matching a non-empty string.
121
122 14. The checks for overrunning compiling workspace could trigger after an
123 overrun had occurred. This is a "should never occur" error, but it can be
124 triggered by pathological patterns such as hundreds of nested parentheses.
125 The checks now trigger 100 bytes before the end of the workspace.
126
127 15. Fix typo in configure.ac: "srtoq" should be "strtoq".
128
129
130 Version 8.01 19-Jan-2010
131 ------------------------
132
133 1. If a pattern contained a conditional subpattern with only one branch (in
134 particular, this includes all (*DEFINE) patterns), a call to pcre_study()
135 computed the wrong minimum data length (which is of course zero for such
136 subpatterns). This could cause incorrect "no match" results.
137
138 2. For patterns such as (?i)a(?-i)b|c where an option setting at the start of
139 the pattern is reset in the first branch, pcre_compile() failed with
140 "internal error: code overflow at offset...". This happened only when
141 the reset was to the original external option setting. (An optimization
142 abstracts leading options settings into an external setting, which was the
143 cause of this.)
144
145 3. A pattern such as ^(?!a(*SKIP)b) where a negative assertion contained one
146 of the verbs SKIP, PRUNE, or COMMIT, did not work correctly. When the
147 assertion pattern did not match (meaning that the assertion was true), it
148 was incorrectly treated as false if the SKIP had been reached during the
149 matching. This also applied to assertions used as conditions.
150
151 4. If an item that is not supported by pcre_dfa_exec() was encountered in an
152 assertion subpattern, including such a pattern used as a condition,
153 unpredictable results occurred, instead of the error return
154 PCRE_ERROR_DFA_UITEM.
155
156 5. The C++ GlobalReplace function was not working like Perl for the special
157 situation when an empty string is matched. It now does the fancy magic
158 stuff that is necessary.
159
160 6. In pcre_internal.h, obsolete includes to setjmp.h and stdarg.h have been
161 removed. (These were left over from very, very early versions of PCRE.)
162
163 7. Some cosmetic changes to the code to make life easier when compiling it
164 as part of something else:
165
166 (a) Change DEBUG to PCRE_DEBUG.
167
168 (b) In pcre_compile(), rename the member of the "branch_chain" structure
169 called "current" as "current_branch", to prevent a collision with the
170 Linux macro when compiled as a kernel module.
171
172 (c) In pcre_study(), rename the function set_bit() as set_table_bit(), to
173 prevent a collision with the Linux macro when compiled as a kernel
174 module.
175
176 8. In pcre_compile() there are some checks for integer overflows that used to
177 cast potentially large values to (double). This has been changed to that
178 when building, a check for int64_t is made, and if it is found, it is used
179 instead, thus avoiding the use of floating point arithmetic. (There is no
180 other use of FP in PCRE.) If int64_t is not found, the fallback is to
181 double.
182
183 9. Added two casts to avoid signed/unsigned warnings from VS Studio Express
184 2005 (difference between two addresses compared to an unsigned value).
185
186 10. Change the standard AC_CHECK_LIB test for libbz2 in configure.ac to a
187 custom one, because of the following reported problem in Windows:
188
189 - libbz2 uses the Pascal calling convention (WINAPI) for the functions
190 under Win32.
191 - The standard autoconf AC_CHECK_LIB fails to include "bzlib.h",
192 therefore missing the function definition.
193 - The compiler thus generates a "C" signature for the test function.
194 - The linker fails to find the "C" function.
195 - PCRE fails to configure if asked to do so against libbz2.
196
197 11. When running libtoolize from libtool-2.2.6b as part of autogen.sh, these
198 messages were output:
199
200 Consider adding `AC_CONFIG_MACRO_DIR([m4])' to configure.ac and
201 rerunning libtoolize, to keep the correct libtool macros in-tree.
202 Consider adding `-I m4' to ACLOCAL_AMFLAGS in Makefile.am.
203
204 I have done both of these things.
205
206 12. Although pcre_dfa_exec() does not use nearly as much stack as pcre_exec()
207 most of the time, it *can* run out if it is given a pattern that contains a
208 runaway infinite recursion. I updated the discussion in the pcrestack man
209 page.
210
211 13. Now that we have gone to the x.xx style of version numbers, the minor
212 version may start with zero. Using 08 or 09 is a bad idea because users
213 might check the value of PCRE_MINOR in their code, and 08 or 09 may be
214 interpreted as invalid octal numbers. I've updated the previous comment in
215 configure.ac, and also added a check that gives an error if 08 or 09 are
216 used.
217
218 14. Change 8.00/11 was not quite complete: code had been accidentally omitted,
219 causing partial matching to fail when the end of the subject matched \W
220 in a UTF-8 pattern where \W was quantified with a minimum of 3.
221
222 15. There were some discrepancies between the declarations in pcre_internal.h
223 of _pcre_is_newline(), _pcre_was_newline(), and _pcre_valid_utf8() and
224 their definitions. The declarations used "const uschar *" and the
225 definitions used USPTR. Even though USPTR is normally defined as "const
226 unsigned char *" (and uschar is typedeffed as "unsigned char"), it was
227 reported that: "This difference in casting confuses some C++ compilers, for
228 example, SunCC recognizes above declarations as different functions and
229 generates broken code for hbpcre." I have changed the declarations to use
230 USPTR.
231
232 16. GNU libtool is named differently on some systems. The autogen.sh script now
233 tries several variants such as glibtoolize (MacOSX) and libtoolize1x
234 (FreeBSD).
235
236 17. Applied Craig's patch that fixes an HP aCC compile error in pcre 8.00
237 (strtoXX undefined when compiling pcrecpp.cc). The patch contains this
238 comment: "Figure out how to create a longlong from a string: strtoll and
239 equivalent. It's not enough to call AC_CHECK_FUNCS: hpux has a strtoll, for
240 instance, but it only takes 2 args instead of 3!"
241
242 18. A subtle bug concerned with back references has been fixed by a change of
243 specification, with a corresponding code fix. A pattern such as
244 ^(xa|=?\1a)+$ which contains a back reference inside the group to which it
245 refers, was giving matches when it shouldn't. For example, xa=xaaa would
246 match that pattern. Interestingly, Perl (at least up to 5.11.3) has the
247 same bug. Such groups have to be quantified to be useful, or contained
248 inside another quantified group. (If there's no repetition, the reference
249 can never match.) The problem arises because, having left the group and
250 moved on to the rest of the pattern, a later failure that backtracks into
251 the group uses the captured value from the final iteration of the group
252 rather than the correct earlier one. I have fixed this in PCRE by forcing
253 any group that contains a reference to itself to be an atomic group; that
254 is, there cannot be any backtracking into it once it has completed. This is
255 similar to recursive and subroutine calls.
256
257
258 Version 8.00 19-Oct-09
259 ----------------------
260
261 1. The table for translating pcre_compile() error codes into POSIX error codes
262 was out-of-date, and there was no check on the pcre_compile() error code
263 being within the table. This could lead to an OK return being given in
264 error.
265
266 2. Changed the call to open a subject file in pcregrep from fopen(pathname,
267 "r") to fopen(pathname, "rb"), which fixed a problem with some of the tests
268 in a Windows environment.
269
270 3. The pcregrep --count option prints the count for each file even when it is
271 zero, as does GNU grep. However, pcregrep was also printing all files when
272 --files-with-matches was added. Now, when both options are given, it prints
273 counts only for those files that have at least one match. (GNU grep just
274 prints the file name in this circumstance, but including the count seems
275 more useful - otherwise, why use --count?) Also ensured that the
276 combination -clh just lists non-zero counts, with no names.
277
278 4. The long form of the pcregrep -F option was incorrectly implemented as
279 --fixed_strings instead of --fixed-strings. This is an incompatible change,
280 but it seems right to fix it, and I didn't think it was worth preserving
281 the old behaviour.
282
283 5. The command line items --regex=pattern and --regexp=pattern were not
284 recognized by pcregrep, which required --regex pattern or --regexp pattern
285 (with a space rather than an '='). The man page documented the '=' forms,
286 which are compatible with GNU grep; these now work.
287
288 6. No libpcreposix.pc file was created for pkg-config; there was just
289 libpcre.pc and libpcrecpp.pc. The omission has been rectified.
290
291 7. Added #ifndef SUPPORT_UCP into the pcre_ucd.c module, to reduce its size
292 when UCP support is not needed, by modifying the Python script that
293 generates it from Unicode data files. This should not matter if the module
294 is correctly used as a library, but I received one complaint about 50K of
295 unwanted data. My guess is that the person linked everything into his
296 program rather than using a library. Anyway, it does no harm.
297
298 8. A pattern such as /\x{123}{2,2}+/8 was incorrectly compiled; the trigger
299 was a minimum greater than 1 for a wide character in a possessive
300 repetition. The same bug could also affect patterns like /(\x{ff}{0,2})*/8
301 which had an unlimited repeat of a nested, fixed maximum repeat of a wide
302 character. Chaos in the form of incorrect output or a compiling loop could
303 result.
304
305 9. The restrictions on what a pattern can contain when partial matching is
306 requested for pcre_exec() have been removed. All patterns can now be
307 partially matched by this function. In addition, if there are at least two
308 slots in the offset vector, the offset of the earliest inspected character
309 for the match and the offset of the end of the subject are set in them when
310 PCRE_ERROR_PARTIAL is returned.
311
312 10. Partial matching has been split into two forms: PCRE_PARTIAL_SOFT, which is
313 synonymous with PCRE_PARTIAL, for backwards compatibility, and
314 PCRE_PARTIAL_HARD, which causes a partial match to supersede a full match,
315 and may be more useful for multi-segment matching.
316
317 11. Partial matching with pcre_exec() is now more intuitive. A partial match
318 used to be given if ever the end of the subject was reached; now it is
319 given only if matching could not proceed because another character was
320 needed. This makes a difference in some odd cases such as Z(*FAIL) with the
321 string "Z", which now yields "no match" instead of "partial match". In the
322 case of pcre_dfa_exec(), "no match" is given if every matching path for the
323 final character ended with (*FAIL).
324
325 12. Restarting a match using pcre_dfa_exec() after a partial match did not work
326 if the pattern had a "must contain" character that was already found in the
327 earlier partial match, unless partial matching was again requested. For
328 example, with the pattern /dog.(body)?/, the "must contain" character is
329 "g". If the first part-match was for the string "dog", restarting with
330 "sbody" failed. This bug has been fixed.
331
332 13. The string returned by pcre_dfa_exec() after a partial match has been
333 changed so that it starts at the first inspected character rather than the
334 first character of the match. This makes a difference only if the pattern
335 starts with a lookbehind assertion or \b or \B (\K is not supported by
336 pcre_dfa_exec()). It's an incompatible change, but it makes the two
337 matching functions compatible, and I think it's the right thing to do.
338
339 14. Added a pcredemo man page, created automatically from the pcredemo.c file,
340 so that the demonstration program is easily available in environments where
341 PCRE has not been installed from source.
342
343 15. Arranged to add -DPCRE_STATIC to cflags in libpcre.pc, libpcreposix.cp,
344 libpcrecpp.pc and pcre-config when PCRE is not compiled as a shared
345 library.
346
347 16. Added REG_UNGREEDY to the pcreposix interface, at the request of a user.
348 It maps to PCRE_UNGREEDY. It is not, of course, POSIX-compatible, but it
349 is not the first non-POSIX option to be added. Clearly some people find
350 these options useful.
351
352 17. If a caller to the POSIX matching function regexec() passes a non-zero
353 value for nmatch with a NULL value for pmatch, the value of
354 nmatch is forced to zero.
355
356 18. RunGrepTest did not have a test for the availability of the -u option of
357 the diff command, as RunTest does. It now checks in the same way as
358 RunTest, and also checks for the -b option.
359
360 19. If an odd number of negated classes containing just a single character
361 interposed, within parentheses, between a forward reference to a named
362 subpattern and the definition of the subpattern, compilation crashed with
363 an internal error, complaining that it could not find the referenced
364 subpattern. An example of a crashing pattern is /(?&A)(([^m])(?<A>))/.
365 [The bug was that it was starting one character too far in when skipping
366 over the character class, thus treating the ] as data rather than
367 terminating the class. This meant it could skip too much.]
368
369 20. Added PCRE_NOTEMPTY_ATSTART in order to be able to correctly implement the
370 /g option in pcretest when the pattern contains \K, which makes it possible
371 to have an empty string match not at the start, even when the pattern is
372 anchored. Updated pcretest and pcredemo to use this option.
373
374 21. If the maximum number of capturing subpatterns in a recursion was greater
375 than the maximum at the outer level, the higher number was returned, but
376 with unset values at the outer level. The correct (outer level) value is
377 now given.
378
379 22. If (*ACCEPT) appeared inside capturing parentheses, previous releases of
380 PCRE did not set those parentheses (unlike Perl). I have now found a way to
381 make it do so. The string so far is captured, making this feature
382 compatible with Perl.
383
384 23. The tests have been re-organized, adding tests 11 and 12, to make it
385 possible to check the Perl 5.10 features against Perl 5.10.
386
387 24. Perl 5.10 allows subroutine calls in lookbehinds, as long as the subroutine
388 pattern matches a fixed length string. PCRE did not allow this; now it
389 does. Neither allows recursion.
390
391 25. I finally figured out how to implement a request to provide the minimum
392 length of subject string that was needed in order to match a given pattern.
393 (It was back references and recursion that I had previously got hung up
394 on.) This code has now been added to pcre_study(); it finds a lower bound
395 to the length of subject needed. It is not necessarily the greatest lower
396 bound, but using it to avoid searching strings that are too short does give
397 some useful speed-ups. The value is available to calling programs via
398 pcre_fullinfo().
399
400 26. While implementing 25, I discovered to my embarrassment that pcretest had
401 not been passing the result of pcre_study() to pcre_dfa_exec(), so the
402 study optimizations had never been tested with that matching function.
403 Oops. What is worse, even when it was passed study data, there was a bug in
404 pcre_dfa_exec() that meant it never actually used it. Double oops. There
405 were also very few tests of studied patterns with pcre_dfa_exec().
406
407 27. If (?| is used to create subpatterns with duplicate numbers, they are now
408 allowed to have the same name, even if PCRE_DUPNAMES is not set. However,
409 on the other side of the coin, they are no longer allowed to have different
410 names, because these cannot be distinguished in PCRE, and this has caused
411 confusion. (This is a difference from Perl.)
412
413 28. When duplicate subpattern names are present (necessarily with different
414 numbers, as required by 27 above), and a test is made by name in a
415 conditional pattern, either for a subpattern having been matched, or for
416 recursion in such a pattern, all the associated numbered subpatterns are
417 tested, and the overall condition is true if the condition is true for any
418 one of them. This is the way Perl works, and is also more like the way
419 testing by number works.
420
421
422 Version 7.9 11-Apr-09
423 ---------------------
424
425 1. When building with support for bzlib/zlib (pcregrep) and/or readline
426 (pcretest), all targets were linked against these libraries. This included
427 libpcre, libpcreposix, and libpcrecpp, even though they do not use these
428 libraries. This caused unwanted dependencies to be created. This problem
429 has been fixed, and now only pcregrep is linked with bzlib/zlib and only
430 pcretest is linked with readline.
431
432 2. The "typedef int BOOL" in pcre_internal.h that was included inside the
433 "#ifndef FALSE" condition by an earlier change (probably 7.8/18) has been
434 moved outside it again, because FALSE and TRUE are already defined in AIX,
435 but BOOL is not.
436
437 3. The pcre_config() function was treating the PCRE_MATCH_LIMIT and
438 PCRE_MATCH_LIMIT_RECURSION values as ints, when they should be long ints.
439
440 4. The pcregrep documentation said spaces were inserted as well as colons (or
441 hyphens) following file names and line numbers when outputting matching
442 lines. This is not true; no spaces are inserted. I have also clarified the
443 wording for the --colour (or --color) option.
444
445 5. In pcregrep, when --colour was used with -o, the list of matching strings
446 was not coloured; this is different to GNU grep, so I have changed it to be
447 the same.
448
449 6. When --colo(u)r was used in pcregrep, only the first matching substring in
450 each matching line was coloured. Now it goes on to look for further matches
451 of any of the test patterns, which is the same behaviour as GNU grep.
452
453 7. A pattern that could match an empty string could cause pcregrep to loop; it
454 doesn't make sense to accept an empty string match in pcregrep, so I have
455 locked it out (using PCRE's PCRE_NOTEMPTY option). By experiment, this
456 seems to be how GNU grep behaves.
457
458 8. The pattern (?(?=.*b)b|^) was incorrectly compiled as "match must be at
459 start or after a newline", because the conditional assertion was not being
460 correctly handled. The rule now is that both the assertion and what follows
461 in the first alternative must satisfy the test.
462
463 9. If auto-callout was enabled in a pattern with a conditional group whose
464 condition was an assertion, PCRE could crash during matching, both with
465 pcre_exec() and pcre_dfa_exec().
466
467 10. The PCRE_DOLLAR_ENDONLY option was not working when pcre_dfa_exec() was
468 used for matching.
469
470 11. Unicode property support in character classes was not working for
471 characters (bytes) greater than 127 when not in UTF-8 mode.
472
473 12. Added the -M command line option to pcretest.
474
475 14. Added the non-standard REG_NOTEMPTY option to the POSIX interface.
476
477 15. Added the PCRE_NO_START_OPTIMIZE match-time option.
478
479 16. Added comments and documentation about mis-use of no_arg in the C++
480 wrapper.
481
482 17. Implemented support for UTF-8 encoding in EBCDIC environments, a patch
483 from Martin Jerabek that uses macro names for all relevant character and
484 string constants.
485
486 18. Added to pcre_internal.h two configuration checks: (a) If both EBCDIC and
487 SUPPORT_UTF8 are set, give an error; (b) If SUPPORT_UCP is set without
488 SUPPORT_UTF8, define SUPPORT_UTF8. The "configure" script handles both of
489 these, but not everybody uses configure.
490
491 19. A conditional group that had only one branch was not being correctly
492 recognized as an item that could match an empty string. This meant that an
493 enclosing group might also not be so recognized, causing infinite looping
494 (and probably a segfault) for patterns such as ^"((?(?=[a])[^"])|b)*"$
495 with the subject "ab", where knowledge that the repeated group can match
496 nothing is needed in order to break the loop.
497
498 20. If a pattern that was compiled with callouts was matched using pcre_dfa_
499 exec(), but without supplying a callout function, matching went wrong.
500
501 21. If PCRE_ERROR_MATCHLIMIT occurred during a recursion, there was a memory
502 leak if the size of the offset vector was greater than 30. When the vector
503 is smaller, the saved offsets during recursion go onto a local stack
504 vector, but for larger vectors malloc() is used. It was failing to free
505 when the recursion yielded PCRE_ERROR_MATCH_LIMIT (or any other "abnormal"
506 error, in fact).
507
508 22. There was a missing #ifdef SUPPORT_UTF8 round one of the variables in the
509 heapframe that is used only when UTF-8 support is enabled. This caused no
510 problem, but was untidy.
511
512 23. Steven Van Ingelgem's patch to CMakeLists.txt to change the name
513 CMAKE_BINARY_DIR to PROJECT_BINARY_DIR so that it works when PCRE is
514 included within another project.
515
516 24. Steven Van Ingelgem's patches to add more options to the CMake support,
517 slightly modified by me:
518
519 (a) PCRE_BUILD_TESTS can be set OFF not to build the tests, including
520 not building pcregrep.
521
522 (b) PCRE_BUILD_PCREGREP can be see OFF not to build pcregrep, but only
523 if PCRE_BUILD_TESTS is also set OFF, because the tests use pcregrep.
524
525 25. Forward references, both numeric and by name, in patterns that made use of
526 duplicate group numbers, could behave incorrectly or give incorrect errors,
527 because when scanning forward to find the reference group, PCRE was not
528 taking into account the duplicate group numbers. A pattern such as
529 ^X(?3)(a)(?|(b)|(q))(Y) is an example.
530
531 26. Changed a few more instances of "const unsigned char *" to USPTR, making
532 the feature of a custom pointer more persuasive (as requested by a user).
533
534 27. Wrapped the definitions of fileno and isatty for Windows, which appear in
535 pcretest.c, inside #ifndefs, because it seems they are sometimes already
536 pre-defined.
537
538 28. Added support for (*UTF8) at the start of a pattern.
539
540 29. Arrange for flags added by the "release type" setting in CMake to be shown
541 in the configuration summary.
542
543
544 Version 7.8 05-Sep-08
545 ---------------------
546
547 1. Replaced UCP searching code with optimized version as implemented for Ad
548 Muncher (http://www.admuncher.com/) by Peter Kankowski. This uses a two-
549 stage table and inline lookup instead of a function, giving speed ups of 2
550 to 5 times on some simple patterns that I tested. Permission was given to
551 distribute the MultiStage2.py script that generates the tables (it's not in
552 the tarball, but is in the Subversion repository).
553
554 2. Updated the Unicode datatables to Unicode 5.1.0. This adds yet more
555 scripts.
556
557 3. Change 12 for 7.7 introduced a bug in pcre_study() when a pattern contained
558 a group with a zero qualifier. The result of the study could be incorrect,
559 or the function might crash, depending on the pattern.
560
561 4. Caseless matching was not working for non-ASCII characters in back
562 references. For example, /(\x{de})\1/8i was not matching \x{de}\x{fe}.
563 It now works when Unicode Property Support is available.
564
565 5. In pcretest, an escape such as \x{de} in the data was always generating
566 a UTF-8 string, even in non-UTF-8 mode. Now it generates a single byte in
567 non-UTF-8 mode. If the value is greater than 255, it gives a warning about
568 truncation.
569
570 6. Minor bugfix in pcrecpp.cc (change "" == ... to NULL == ...).
571
572 7. Added two (int) casts to pcregrep when printing the difference of two
573 pointers, in case they are 64-bit values.
574
575 8. Added comments about Mac OS X stack usage to the pcrestack man page and to
576 test 2 if it fails.
577
578 9. Added PCRE_CALL_CONVENTION just before the names of all exported functions,
579 and a #define of that name to empty if it is not externally set. This is to
580 allow users of MSVC to set it if necessary.
581
582 10. The PCRE_EXP_DEFN macro which precedes exported functions was missing from
583 the convenience functions in the pcre_get.c source file.
584
585 11. An option change at the start of a pattern that had top-level alternatives
586 could cause overwriting and/or a crash. This command provoked a crash in
587 some environments:
588
589 printf "/(?i)[\xc3\xa9\xc3\xbd]|[\xc3\xa9\xc3\xbdA]/8\n" | pcretest
590
591 This potential security problem was recorded as CVE-2008-2371.
592
593 12. For a pattern where the match had to start at the beginning or immediately
594 after a newline (e.g /.*anything/ without the DOTALL flag), pcre_exec() and
595 pcre_dfa_exec() could read past the end of the passed subject if there was
596 no match. To help with detecting such bugs (e.g. with valgrind), I modified
597 pcretest so that it places the subject at the end of its malloc-ed buffer.
598
599 13. The change to pcretest in 12 above threw up a couple more cases when pcre_
600 exec() might read past the end of the data buffer in UTF-8 mode.
601
602 14. A similar bug to 7.3/2 existed when the PCRE_FIRSTLINE option was set and
603 the data contained the byte 0x85 as part of a UTF-8 character within its
604 first line. This applied both to normal and DFA matching.
605
606 15. Lazy qualifiers were not working in some cases in UTF-8 mode. For example,
607 /^[^d]*?$/8 failed to match "abc".
608
609 16. Added a missing copyright notice to pcrecpp_internal.h.
610
611 17. Make it more clear in the documentation that values returned from
612 pcre_exec() in ovector are byte offsets, not character counts.
613
614 18. Tidied a few places to stop certain compilers from issuing warnings.
615
616 19. Updated the Virtual Pascal + BCC files to compile the latest v7.7, as
617 supplied by Stefan Weber. I made a further small update for 7.8 because
618 there is a change of source arrangements: the pcre_searchfuncs.c module is
619 replaced by pcre_ucd.c.
620
621
622 Version 7.7 07-May-08
623 ---------------------
624
625 1. Applied Craig's patch to sort out a long long problem: "If we can't convert
626 a string to a long long, pretend we don't even have a long long." This is
627 done by checking for the strtoq, strtoll, and _strtoi64 functions.
628
629 2. Applied Craig's patch to pcrecpp.cc to restore ABI compatibility with
630 pre-7.6 versions, which defined a global no_arg variable instead of putting
631 it in the RE class. (See also #8 below.)
632
633 3. Remove a line of dead code, identified by coverity and reported by Nuno
634 Lopes.
635
636 4. Fixed two related pcregrep bugs involving -r with --include or --exclude:
637
638 (1) The include/exclude patterns were being applied to the whole pathnames
639 of files, instead of just to the final components.
640
641 (2) If there was more than one level of directory, the subdirectories were
642 skipped unless they satisfied the include/exclude conditions. This is
643 inconsistent with GNU grep (and could even be seen as contrary to the
644 pcregrep specification - which I improved to make it absolutely clear).
645 The action now is always to scan all levels of directory, and just
646 apply the include/exclude patterns to regular files.
647
648 5. Added the --include_dir and --exclude_dir patterns to pcregrep, and used
649 --exclude_dir in the tests to avoid scanning .svn directories.
650
651 6. Applied Craig's patch to the QuoteMeta function so that it escapes the
652 NUL character as backslash + 0 rather than backslash + NUL, because PCRE
653 doesn't support NULs in patterns.
654
655 7. Added some missing "const"s to declarations of static tables in
656 pcre_compile.c and pcre_dfa_exec.c.
657
658 8. Applied Craig's patch to pcrecpp.cc to fix a problem in OS X that was
659 caused by fix #2 above. (Subsequently also a second patch to fix the
660 first patch. And a third patch - this was a messy problem.)
661
662 9. Applied Craig's patch to remove the use of push_back().
663
664 10. Applied Alan Lehotsky's patch to add REG_STARTEND support to the POSIX
665 matching function regexec().
666
667 11. Added support for the Oniguruma syntax \g<name>, \g<n>, \g'name', \g'n',
668 which, however, unlike Perl's \g{...}, are subroutine calls, not back
669 references. PCRE supports relative numbers with this syntax (I don't think
670 Oniguruma does).
671
672 12. Previously, a group with a zero repeat such as (...){0} was completely
673 omitted from the compiled regex. However, this means that if the group
674 was called as a subroutine from elsewhere in the pattern, things went wrong
675 (an internal error was given). Such groups are now left in the compiled
676 pattern, with a new opcode that causes them to be skipped at execution
677 time.
678
679 13. Added the PCRE_JAVASCRIPT_COMPAT option. This makes the following changes
680 to the way PCRE behaves:
681
682 (a) A lone ] character is dis-allowed (Perl treats it as data).
683
684 (b) A back reference to an unmatched subpattern matches an empty string
685 (Perl fails the current match path).
686
687 (c) A data ] in a character class must be notated as \] because if the
688 first data character in a class is ], it defines an empty class. (In
689 Perl it is not possible to have an empty class.) The empty class []
690 never matches; it forces failure and is equivalent to (*FAIL) or (?!).
691 The negative empty class [^] matches any one character, independently
692 of the DOTALL setting.
693
694 14. A pattern such as /(?2)[]a()b](abc)/ which had a forward reference to a
695 non-existent subpattern following a character class starting with ']' and
696 containing () gave an internal compiling error instead of "reference to
697 non-existent subpattern". Fortunately, when the pattern did exist, the
698 compiled code was correct. (When scanning forwards to check for the
699 existencd of the subpattern, it was treating the data ']' as terminating
700 the class, so got the count wrong. When actually compiling, the reference
701 was subsequently set up correctly.)
702
703 15. The "always fail" assertion (?!) is optimzed to (*FAIL) by pcre_compile;
704 it was being rejected as not supported by pcre_dfa_exec(), even though
705 other assertions are supported. I have made pcre_dfa_exec() support
706 (*FAIL).
707
708 16. The implementation of 13c above involved the invention of a new opcode,
709 OP_ALLANY, which is like OP_ANY but doesn't check the /s flag. Since /s
710 cannot be changed at match time, I realized I could make a small
711 improvement to matching performance by compiling OP_ALLANY instead of
712 OP_ANY for "." when DOTALL was set, and then removing the runtime tests
713 on the OP_ANY path.
714
715 17. Compiling pcretest on Windows with readline support failed without the
716 following two fixes: (1) Make the unistd.h include conditional on
717 HAVE_UNISTD_H; (2) #define isatty and fileno as _isatty and _fileno.
718
719 18. Changed CMakeLists.txt and cmake/FindReadline.cmake to arrange for the
720 ncurses library to be included for pcretest when ReadLine support is
721 requested, but also to allow for it to be overridden. This patch came from
722 Daniel Bergström.
723
724 19. There was a typo in the file ucpinternal.h where f0_rangeflag was defined
725 as 0x00f00000 instead of 0x00800000. Luckily, this would not have caused
726 any errors with the current Unicode tables. Thanks to Peter Kankowski for
727 spotting this.
728
729
730 Version 7.6 28-Jan-08
731 ---------------------
732
733 1. A character class containing a very large number of characters with
734 codepoints greater than 255 (in UTF-8 mode, of course) caused a buffer
735 overflow.
736
737 2. Patch to cut out the "long long" test in pcrecpp_unittest when
738 HAVE_LONG_LONG is not defined.
739
740 3. Applied Christian Ehrlicher's patch to update the CMake build files to
741 bring them up to date and include new features. This patch includes:
742
743 - Fixed PH's badly added libz and libbz2 support.
744 - Fixed a problem with static linking.
745 - Added pcredemo. [But later removed - see 7 below.]
746 - Fixed dftables problem and added an option.
747 - Added a number of HAVE_XXX tests, including HAVE_WINDOWS_H and
748 HAVE_LONG_LONG.
749 - Added readline support for pcretest.
750 - Added an listing of the option settings after cmake has run.
751
752 4. A user submitted a patch to Makefile that makes it easy to create
753 "pcre.dll" under mingw when using Configure/Make. I added stuff to
754 Makefile.am that cause it to include this special target, without
755 affecting anything else. Note that the same mingw target plus all
756 the other distribution libraries and programs are now supported
757 when configuring with CMake (see 6 below) instead of with
758 Configure/Make.
759
760 5. Applied Craig's patch that moves no_arg into the RE class in the C++ code.
761 This is an attempt to solve the reported problem "pcrecpp::no_arg is not
762 exported in the Windows port". It has not yet been confirmed that the patch
763 solves the problem, but it does no harm.
764
765 6. Applied Sheri's patch to CMakeLists.txt to add NON_STANDARD_LIB_PREFIX and
766 NON_STANDARD_LIB_SUFFIX for dll names built with mingw when configured
767 with CMake, and also correct the comment about stack recursion.
768
769 7. Remove the automatic building of pcredemo from the ./configure system and
770 from CMakeLists.txt. The whole idea of pcredemo.c is that it is an example
771 of a program that users should build themselves after PCRE is installed, so
772 building it automatically is not really right. What is more, it gave
773 trouble in some build environments.
774
775 8. Further tidies to CMakeLists.txt from Sheri and Christian.
776
777
778 Version 7.5 10-Jan-08
779 ---------------------
780
781 1. Applied a patch from Craig: "This patch makes it possible to 'ignore'
782 values in parens when parsing an RE using the C++ wrapper."
783
784 2. Negative specials like \S did not work in character classes in UTF-8 mode.
785 Characters greater than 255 were excluded from the class instead of being
786 included.
787
788 3. The same bug as (2) above applied to negated POSIX classes such as
789 [:^space:].
790
791 4. PCRECPP_STATIC was referenced in pcrecpp_internal.h, but nowhere was it
792 defined or documented. It seems to have been a typo for PCRE_STATIC, so
793 I have changed it.
794
795 5. The construct (?&) was not diagnosed as a syntax error (it referenced the
796 first named subpattern) and a construct such as (?&a) would reference the
797 first named subpattern whose name started with "a" (in other words, the
798 length check was missing). Both these problems are fixed. "Subpattern name
799 expected" is now given for (?&) (a zero-length name), and this patch also
800 makes it give the same error for \k'' (previously it complained that that
801 was a reference to a non-existent subpattern).
802
803 6. The erroneous patterns (?+-a) and (?-+a) give different error messages;
804 this is right because (?- can be followed by option settings as well as by
805 digits. I have, however, made the messages clearer.
806
807 7. Patterns such as (?(1)a|b) (a pattern that contains fewer subpatterns
808 than the number used in the conditional) now cause a compile-time error.
809 This is actually not compatible with Perl, which accepts such patterns, but
810 treats the conditional as always being FALSE (as PCRE used to), but it
811 seems to me that giving a diagnostic is better.
812
813 8. Change "alphameric" to the more common word "alphanumeric" in comments
814 and messages.
815
816 9. Fix two occurrences of "backslash" in comments that should have been
817 "backspace".
818
819 10. Remove two redundant lines of code that can never be obeyed (their function
820 was moved elsewhere).
821
822 11. The program that makes PCRE's Unicode character property table had a bug
823 which caused it to generate incorrect table entries for sequences of
824 characters that have the same character type, but are in different scripts.
825 It amalgamated them into a single range, with the script of the first of
826 them. In other words, some characters were in the wrong script. There were
827 thirteen such cases, affecting characters in the following ranges:
828
829 U+002b0 - U+002c1
830 U+0060c - U+0060d
831 U+0061e - U+00612
832 U+0064b - U+0065e
833 U+0074d - U+0076d
834 U+01800 - U+01805
835 U+01d00 - U+01d77
836 U+01d9b - U+01dbf
837 U+0200b - U+0200f
838 U+030fc - U+030fe
839 U+03260 - U+0327f
840 U+0fb46 - U+0fbb1
841 U+10450 - U+1049d
842
843 12. The -o option (show only the matching part of a line) for pcregrep was not
844 compatible with GNU grep in that, if there was more than one match in a
845 line, it showed only the first of them. It now behaves in the same way as
846 GNU grep.
847
848 13. If the -o and -v options were combined for pcregrep, it printed a blank
849 line for every non-matching line. GNU grep prints nothing, and pcregrep now
850 does the same. The return code can be used to tell if there were any
851 non-matching lines.
852
853 14. Added --file-offsets and --line-offsets to pcregrep.
854
855 15. The pattern (?=something)(?R) was not being diagnosed as a potentially
856 infinitely looping recursion. The bug was that positive lookaheads were not
857 being skipped when checking for a possible empty match (negative lookaheads
858 and both kinds of lookbehind were skipped).
859
860 16. Fixed two typos in the Windows-only code in pcregrep.c, and moved the
861 inclusion of <windows.h> to before rather than after the definition of
862 INVALID_FILE_ATTRIBUTES (patch from David Byron).
863
864 17. Specifying a possessive quantifier with a specific limit for a Unicode
865 character property caused pcre_compile() to compile bad code, which led at
866 runtime to PCRE_ERROR_INTERNAL (-14). Examples of patterns that caused this
867 are: /\p{Zl}{2,3}+/8 and /\p{Cc}{2}+/8. It was the possessive "+" that
868 caused the error; without that there was no problem.
869
870 18. Added --enable-pcregrep-libz and --enable-pcregrep-libbz2.
871
872 19. Added --enable-pcretest-libreadline.
873
874 20. In pcrecpp.cc, the variable 'count' was incremented twice in
875 RE::GlobalReplace(). As a result, the number of replacements returned was
876 double what it should be. I removed one of the increments, but Craig sent a
877 later patch that removed the other one (the right fix) and added unit tests
878 that check the return values (which was not done before).
879
880 21. Several CMake things:
881
882 (1) Arranged that, when cmake is used on Unix, the libraries end up with
883 the names libpcre and libpcreposix, not just pcre and pcreposix.
884
885 (2) The above change means that pcretest and pcregrep are now correctly
886 linked with the newly-built libraries, not previously installed ones.
887
888 (3) Added PCRE_SUPPORT_LIBREADLINE, PCRE_SUPPORT_LIBZ, PCRE_SUPPORT_LIBBZ2.
889
890 22. In UTF-8 mode, with newline set to "any", a pattern such as .*a.*=.b.*
891 crashed when matching a string such as a\x{2029}b (note that \x{2029} is a
892 UTF-8 newline character). The key issue is that the pattern starts .*;
893 this means that the match must be either at the beginning, or after a
894 newline. The bug was in the code for advancing after a failed match and
895 checking that the new position followed a newline. It was not taking
896 account of UTF-8 characters correctly.
897
898 23. PCRE was behaving differently from Perl in the way it recognized POSIX
899 character classes. PCRE was not treating the sequence [:...:] as a
900 character class unless the ... were all letters. Perl, however, seems to
901 allow any characters between [: and :], though of course it rejects as
902 unknown any "names" that contain non-letters, because all the known class
903 names consist only of letters. Thus, Perl gives an error for [[:1234:]],
904 for example, whereas PCRE did not - it did not recognize a POSIX character
905 class. This seemed a bit dangerous, so the code has been changed to be
906 closer to Perl. The behaviour is not identical to Perl, because PCRE will
907 diagnose an unknown class for, for example, [[:l\ower:]] where Perl will
908 treat it as [[:lower:]]. However, PCRE does now give "unknown" errors where
909 Perl does, and where it didn't before.
910
911 24. Rewrite so as to remove the single use of %n from pcregrep because in some
912 Windows environments %n is disabled by default.
913
914
915 Version 7.4 21-Sep-07
916 ---------------------
917
918 1. Change 7.3/28 was implemented for classes by looking at the bitmap. This
919 means that a class such as [\s] counted as "explicit reference to CR or
920 LF". That isn't really right - the whole point of the change was to try to
921 help when there was an actual mention of one of the two characters. So now
922 the change happens only if \r or \n (or a literal CR or LF) character is
923 encountered.
924
925 2. The 32-bit options word was also used for 6 internal flags, but the numbers
926 of both had grown to the point where there were only 3 bits left.
927 Fortunately, there was spare space in the data structure, and so I have
928 moved the internal flags into a new 16-bit field to free up more option
929 bits.
930
931 3. The appearance of (?J) at the start of a pattern set the DUPNAMES option,
932 but did not set the internal JCHANGED flag - either of these is enough to
933 control the way the "get" function works - but the PCRE_INFO_JCHANGED
934 facility is supposed to tell if (?J) was ever used, so now (?J) at the
935 start sets both bits.
936
937 4. Added options (at build time, compile time, exec time) to change \R from
938 matching any Unicode line ending sequence to just matching CR, LF, or CRLF.
939
940 5. doc/pcresyntax.html was missing from the distribution.
941
942 6. Put back the definition of PCRE_ERROR_NULLWSLIMIT, for backward
943 compatibility, even though it is no longer used.
944
945 7. Added macro for snprintf to pcrecpp_unittest.cc and also for strtoll and
946 strtoull to pcrecpp.cc to select the available functions in WIN32 when the
947 windows.h file is present (where different names are used). [This was
948 reversed later after testing - see 16 below.]
949
950 8. Changed all #include <config.h> to #include "config.h". There were also
951 some further <pcre.h> cases that I changed to "pcre.h".
952
953 9. When pcregrep was used with the --colour option, it missed the line ending
954 sequence off the lines that it output.
955
956 10. It was pointed out to me that arrays of string pointers cause lots of
957 relocations when a shared library is dynamically loaded. A technique of
958 using a single long string with a table of offsets can drastically reduce
959 these. I have refactored PCRE in four places to do this. The result is
960 dramatic:
961
962 Originally: 290
963 After changing UCP table: 187
964 After changing error message table: 43
965 After changing table of "verbs" 36
966 After changing table of Posix names 22
967
968 Thanks to the folks working on Gregex for glib for this insight.
969
970 11. --disable-stack-for-recursion caused compiling to fail unless -enable-
971 unicode-properties was also set.
972
973 12. Updated the tests so that they work when \R is defaulted to ANYCRLF.
974
975 13. Added checks for ANY and ANYCRLF to pcrecpp.cc where it previously
976 checked only for CRLF.
977
978 14. Added casts to pcretest.c to avoid compiler warnings.
979
980 15. Added Craig's patch to various pcrecpp modules to avoid compiler warnings.
981
982 16. Added Craig's patch to remove the WINDOWS_H tests, that were not working,
983 and instead check for _strtoi64 explicitly, and avoid the use of snprintf()
984 entirely. This removes changes made in 7 above.
985
986 17. The CMake files have been updated, and there is now more information about
987 building with CMake in the NON-UNIX-USE document.
988
989
990 Version 7.3 28-Aug-07
991 ---------------------
992
993 1. In the rejigging of the build system that eventually resulted in 7.1, the
994 line "#include <pcre.h>" was included in pcre_internal.h. The use of angle
995 brackets there is not right, since it causes compilers to look for an
996 installed pcre.h, not the version that is in the source that is being
997 compiled (which of course may be different). I have changed it back to:
998
999 #include "pcre.h"
1000
1001 I have a vague recollection that the change was concerned with compiling in
1002 different directories, but in the new build system, that is taken care of
1003 by the VPATH setting the Makefile.
1004
1005 2. The pattern .*$ when run in not-DOTALL UTF-8 mode with newline=any failed
1006 when the subject happened to end in the byte 0x85 (e.g. if the last
1007 character was \x{1ec5}). *Character* 0x85 is one of the "any" newline
1008 characters but of course it shouldn't be taken as a newline when it is part
1009 of another character. The bug was that, for an unlimited repeat of . in
1010 not-DOTALL UTF-8 mode, PCRE was advancing by bytes rather than by
1011 characters when looking for a newline.
1012
1013 3. A small performance improvement in the DOTALL UTF-8 mode .* case.
1014
1015 4. Debugging: adjusted the names of opcodes for different kinds of parentheses
1016 in debug output.
1017
1018 5. Arrange to use "%I64d" instead of "%lld" and "%I64u" instead of "%llu" for
1019 long printing in the pcrecpp unittest when running under MinGW.
1020
1021 6. ESC_K was left out of the EBCDIC table.
1022
1023 7. Change 7.0/38 introduced a new limit on the number of nested non-capturing
1024 parentheses; I made it 1000, which seemed large enough. Unfortunately, the
1025 limit also applies to "virtual nesting" when a pattern is recursive, and in
1026 this case 1000 isn't so big. I have been able to remove this limit at the
1027 expense of backing off one optimization in certain circumstances. Normally,
1028 when pcre_exec() would call its internal match() function recursively and
1029 immediately return the result unconditionally, it uses a "tail recursion"
1030 feature to save stack. However, when a subpattern that can match an empty
1031 string has an unlimited repetition quantifier, it no longer makes this
1032 optimization. That gives it a stack frame in which to save the data for
1033 checking that an empty string has been matched. Previously this was taken
1034 from the 1000-entry workspace that had been reserved. So now there is no
1035 explicit limit, but more stack is used.
1036
1037 8. Applied Daniel's patches to solve problems with the import/export magic
1038 syntax that is required for Windows, and which was going wrong for the
1039 pcreposix and pcrecpp parts of the library. These were overlooked when this
1040 problem was solved for the main library.
1041
1042 9. There were some crude static tests to avoid integer overflow when computing
1043 the size of patterns that contain repeated groups with explicit upper
1044 limits. As the maximum quantifier is 65535, the maximum group length was
1045 set at 30,000 so that the product of these two numbers did not overflow a
1046 32-bit integer. However, it turns out that people want to use groups that
1047 are longer than 30,000 bytes (though not repeat them that many times).
1048 Change 7.0/17 (the refactoring of the way the pattern size is computed) has
1049 made it possible to implement the integer overflow checks in a much more
1050 dynamic way, which I have now done. The artificial limitation on group
1051 length has been removed - we now have only the limit on the total length of
1052 the compiled pattern, which depends on the LINK_SIZE setting.
1053
1054 10. Fixed a bug in the documentation for get/copy named substring when
1055 duplicate names are permitted. If none of the named substrings are set, the
1056 functions return PCRE_ERROR_NOSUBSTRING (7); the doc said they returned an
1057 empty string.
1058
1059 11. Because Perl interprets \Q...\E at a high level, and ignores orphan \E
1060 instances, patterns such as [\Q\E] or [\E] or even [^\E] cause an error,
1061 because the ] is interpreted as the first data character and the
1062 terminating ] is not found. PCRE has been made compatible with Perl in this
1063 regard. Previously, it interpreted [\Q\E] as an empty class, and [\E] could
1064 cause memory overwriting.
1065
1066 10. Like Perl, PCRE automatically breaks an unlimited repeat after an empty
1067 string has been matched (to stop an infinite loop). It was not recognizing
1068 a conditional subpattern that could match an empty string if that
1069 subpattern was within another subpattern. For example, it looped when
1070 trying to match (((?(1)X|))*) but it was OK with ((?(1)X|)*) where the
1071 condition was not nested. This bug has been fixed.
1072
1073 12. A pattern like \X?\d or \P{L}?\d in non-UTF-8 mode could cause a backtrack
1074 past the start of the subject in the presence of bytes with the top bit
1075 set, for example "\x8aBCD".
1076
1077 13. Added Perl 5.10 experimental backtracking controls (*FAIL), (*F), (*PRUNE),
1078 (*SKIP), (*THEN), (*COMMIT), and (*ACCEPT).
1079
1080 14. Optimized (?!) to (*FAIL).
1081
1082 15. Updated the test for a valid UTF-8 string to conform to the later RFC 3629.
1083 This restricts code points to be within the range 0 to 0x10FFFF, excluding
1084 the "low surrogate" sequence 0xD800 to 0xDFFF. Previously, PCRE allowed the
1085 full range 0 to 0x7FFFFFFF, as defined by RFC 2279. Internally, it still
1086 does: it's just the validity check that is more restrictive.
1087
1088 16. Inserted checks for integer overflows during escape sequence (backslash)
1089 processing, and also fixed erroneous offset values for syntax errors during
1090 backslash processing.
1091
1092 17. Fixed another case of looking too far back in non-UTF-8 mode (cf 12 above)
1093 for patterns like [\PPP\x8a]{1,}\x80 with the subject "A\x80".
1094
1095 18. An unterminated class in a pattern like (?1)\c[ with a "forward reference"
1096 caused an overrun.
1097
1098 19. A pattern like (?:[\PPa*]*){8,} which had an "extended class" (one with
1099 something other than just ASCII characters) inside a group that had an
1100 unlimited repeat caused a loop at compile time (while checking to see
1101 whether the group could match an empty string).
1102
1103 20. Debugging a pattern containing \p or \P could cause a crash. For example,
1104 [\P{Any}] did so. (Error in the code for printing property names.)
1105
1106 21. An orphan \E inside a character class could cause a crash.
1107
1108 22. A repeated capturing bracket such as (A)? could cause a wild memory
1109 reference during compilation.
1110
1111 23. There are several functions in pcre_compile() that scan along a compiled
1112 expression for various reasons (e.g. to see if it's fixed length for look
1113 behind). There were bugs in these functions when a repeated \p or \P was
1114 present in the pattern. These operators have additional parameters compared
1115 with \d, etc, and these were not being taken into account when moving along
1116 the compiled data. Specifically:
1117
1118 (a) A item such as \p{Yi}{3} in a lookbehind was not treated as fixed
1119 length.
1120
1121 (b) An item such as \pL+ within a repeated group could cause crashes or
1122 loops.
1123
1124 (c) A pattern such as \p{Yi}+(\P{Yi}+)(?1) could give an incorrect
1125 "reference to non-existent subpattern" error.
1126
1127 (d) A pattern like (\P{Yi}{2}\277)? could loop at compile time.
1128
1129 24. A repeated \S or \W in UTF-8 mode could give wrong answers when multibyte
1130 characters were involved (for example /\S{2}/8g with "A\x{a3}BC").
1131
1132 25. Using pcregrep in multiline, inverted mode (-Mv) caused it to loop.
1133
1134 26. Patterns such as [\P{Yi}A] which include \p or \P and just one other
1135 character were causing crashes (broken optimization).
1136
1137 27. Patterns such as (\P{Yi}*\277)* (group with possible zero repeat containing
1138 \p or \P) caused a compile-time loop.
1139
1140 28. More problems have arisen in unanchored patterns when CRLF is a valid line
1141 break. For example, the unstudied pattern [\r\n]A does not match the string
1142 "\r\nA" because change 7.0/46 below moves the current point on by two
1143 characters after failing to match at the start. However, the pattern \nA
1144 *does* match, because it doesn't start till \n, and if [\r\n]A is studied,
1145 the same is true. There doesn't seem any very clean way out of this, but
1146 what I have chosen to do makes the common cases work: PCRE now takes note
1147 of whether there can be an explicit match for \r or \n anywhere in the
1148 pattern, and if so, 7.0/46 no longer applies. As part of this change,
1149 there's a new PCRE_INFO_HASCRORLF option for finding out whether a compiled
1150 pattern has explicit CR or LF references.
1151
1152 29. Added (*CR) etc for changing newline setting at start of pattern.
1153
1154
1155 Version 7.2 19-Jun-07
1156 ---------------------
1157
1158 1. If the fr_FR locale cannot be found for test 3, try the "french" locale,
1159 which is apparently normally available under Windows.
1160
1161 2. Re-jig the pcregrep tests with different newline settings in an attempt
1162 to make them independent of the local environment's newline setting.
1163
1164 3. Add code to configure.ac to remove -g from the CFLAGS default settings.
1165
1166 4. Some of the "internals" tests were previously cut out when the link size
1167 was not 2, because the output contained actual offsets. The recent new
1168 "Z" feature of pcretest means that these can be cut out, making the tests
1169 usable with all link sizes.
1170
1171 5. Implemented Stan Switzer's goto replacement for longjmp() when not using
1172 stack recursion. This gives a massive performance boost under BSD, but just
1173 a small improvement under Linux. However, it saves one field in the frame
1174 in all cases.
1175
1176 6. Added more features from the forthcoming Perl 5.10:
1177
1178 (a) (?-n) (where n is a string of digits) is a relative subroutine or
1179 recursion call. It refers to the nth most recently opened parentheses.
1180
1181 (b) (?+n) is also a relative subroutine call; it refers to the nth next
1182 to be opened parentheses.
1183
1184 (c) Conditions that refer to capturing parentheses can be specified
1185 relatively, for example, (?(-2)... or (?(+3)...
1186
1187 (d) \K resets the start of the current match so that everything before
1188 is not part of it.
1189
1190 (e) \k{name} is synonymous with \k<name> and \k'name' (.NET compatible).
1191
1192 (f) \g{name} is another synonym - part of Perl 5.10's unification of
1193 reference syntax.
1194
1195 (g) (?| introduces a group in which the numbering of parentheses in each
1196 alternative starts with the same number.
1197
1198 (h) \h, \H, \v, and \V match horizontal and vertical whitespace.
1199
1200 7. Added two new calls to pcre_fullinfo(): PCRE_INFO_OKPARTIAL and
1201 PCRE_INFO_JCHANGED.
1202
1203 8. A pattern such as (.*(.)?)* caused pcre_exec() to fail by either not
1204 terminating or by crashing. Diagnosed by Viktor Griph; it was in the code
1205 for detecting groups that can match an empty string.
1206
1207 9. A pattern with a very large number of alternatives (more than several
1208 hundred) was running out of internal workspace during the pre-compile
1209 phase, where pcre_compile() figures out how much memory will be needed. A
1210 bit of new cunning has reduced the workspace needed for groups with
1211 alternatives. The 1000-alternative test pattern now uses 12 bytes of
1212 workspace instead of running out of the 4096 that are available.
1213
1214 10. Inserted some missing (unsigned int) casts to get rid of compiler warnings.
1215
1216 11. Applied patch from Google to remove an optimization that didn't quite work.
1217 The report of the bug said:
1218
1219 pcrecpp::RE("a*").FullMatch("aaa") matches, while
1220 pcrecpp::RE("a*?").FullMatch("aaa") does not, and
1221 pcrecpp::RE("a*?\\z").FullMatch("aaa") does again.
1222
1223 12. If \p or \P was used in non-UTF-8 mode on a character greater than 127
1224 it matched the wrong number of bytes.
1225
1226
1227 Version 7.1 24-Apr-07
1228 ---------------------
1229
1230 1. Applied Bob Rossi and Daniel G's patches to convert the build system to one
1231 that is more "standard", making use of automake and other Autotools. There
1232 is some re-arrangement of the files and adjustment of comments consequent
1233 on this.
1234
1235 2. Part of the patch fixed a problem with the pcregrep tests. The test of -r
1236 for recursive directory scanning broke on some systems because the files
1237 are not scanned in any specific order and on different systems the order
1238 was different. A call to "sort" has been inserted into RunGrepTest for the
1239 approprate test as a short-term fix. In the longer term there may be an
1240 alternative.
1241
1242 3. I had an email from Eric Raymond about problems translating some of PCRE's
1243 man pages to HTML (despite the fact that I distribute HTML pages, some
1244 people do their own conversions for various reasons). The problems
1245 concerned the use of low-level troff macros .br and .in. I have therefore
1246 removed all such uses from the man pages (some were redundant, some could
1247 be replaced by .nf/.fi pairs). The 132html script that I use to generate
1248 HTML has been updated to handle .nf/.fi and to complain if it encounters
1249 .br or .in.
1250
1251 4. Updated comments in configure.ac that get placed in config.h.in and also
1252 arranged for config.h to be included in the distribution, with the name
1253 config.h.generic, for the benefit of those who have to compile without
1254 Autotools (compare pcre.h, which is now distributed as pcre.h.generic).
1255
1256 5. Updated the support (such as it is) for Virtual Pascal, thanks to Stefan
1257 Weber: (1) pcre_internal.h was missing some function renames; (2) updated
1258 makevp.bat for the current PCRE, using the additional files
1259 makevp_c.txt, makevp_l.txt, and pcregexp.pas.
1260
1261 6. A Windows user reported a minor discrepancy with test 2, which turned out
1262 to be caused by a trailing space on an input line that had got lost in his
1263 copy. The trailing space was an accident, so I've just removed it.
1264
1265 7. Add -Wl,-R... flags in pcre-config.in for *BSD* systems, as I'm told
1266 that is needed.
1267
1268 8. Mark ucp_table (in ucptable.h) and ucp_gentype (in pcre_ucp_searchfuncs.c)
1269 as "const" (a) because they are and (b) because it helps the PHP
1270 maintainers who have recently made a script to detect big data structures
1271 in the php code that should be moved to the .rodata section. I remembered
1272 to update Builducptable as well, so it won't revert if ucptable.h is ever
1273 re-created.
1274
1275 9. Added some extra #ifdef SUPPORT_UTF8 conditionals into pcretest.c,
1276 pcre_printint.src, pcre_compile.c, pcre_study.c, and pcre_tables.c, in
1277 order to be able to cut out the UTF-8 tables in the latter when UTF-8
1278 support is not required. This saves 1.5-2K of code, which is important in
1279 some applications.
1280
1281 Later: more #ifdefs are needed in pcre_ord2utf8.c and pcre_valid_utf8.c
1282 so as not to refer to the tables, even though these functions will never be
1283 called when UTF-8 support is disabled. Otherwise there are problems with a
1284 shared library.
1285
1286 10. Fixed two bugs in the emulated memmove() function in pcre_internal.h:
1287
1288 (a) It was defining its arguments as char * instead of void *.
1289
1290 (b) It was assuming that all moves were upwards in memory; this was true
1291 a long time ago when I wrote it, but is no longer the case.
1292
1293 The emulated memove() is provided for those environments that have neither
1294 memmove() nor bcopy(). I didn't think anyone used it these days, but that
1295 is clearly not the case, as these two bugs were recently reported.
1296
1297 11. The script PrepareRelease is now distributed: it calls 132html, CleanTxt,
1298 and Detrail to create the HTML documentation, the .txt form of the man
1299 pages, and it removes trailing spaces from listed files. It also creates
1300 pcre.h.generic and config.h.generic from pcre.h and config.h. In the latter
1301 case, it wraps all the #defines with #ifndefs. This script should be run
1302 before "make dist".
1303
1304 12. Fixed two fairly obscure bugs concerned with quantified caseless matching
1305 with Unicode property support.
1306
1307 (a) For a maximizing quantifier, if the two different cases of the
1308 character were of different lengths in their UTF-8 codings (there are
1309 some cases like this - I found 11), and the matching function had to
1310 back up over a mixture of the two cases, it incorrectly assumed they
1311 were both the same length.
1312
1313 (b) When PCRE was configured to use the heap rather than the stack for
1314 recursion during matching, it was not correctly preserving the data for
1315 the other case of a UTF-8 character when checking ahead for a match
1316 while processing a minimizing repeat. If the check also involved
1317 matching a wide character, but failed, corruption could cause an
1318 erroneous result when trying to check for a repeat of the original
1319 character.
1320
1321 13. Some tidying changes to the testing mechanism:
1322
1323 (a) The RunTest script now detects the internal link size and whether there
1324 is UTF-8 and UCP support by running ./pcretest -C instead of relying on
1325 values substituted by "configure". (The RunGrepTest script already did
1326 this for UTF-8.) The configure.ac script no longer substitutes the
1327 relevant variables.
1328
1329 (b) The debugging options /B and /D in pcretest show the compiled bytecode
1330 with length and offset values. This means that the output is different
1331 for different internal link sizes. Test 2 is skipped for link sizes
1332 other than 2 because of this, bypassing the problem. Unfortunately,
1333 there was also a test in test 3 (the locale tests) that used /B and
1334 failed for link sizes other than 2. Rather than cut the whole test out,
1335 I have added a new /Z option to pcretest that replaces the length and
1336 offset values with spaces. This is now used to make test 3 independent
1337 of link size. (Test 2 will be tidied up later.)
1338
1339 14. If erroroffset was passed as NULL to pcre_compile, it provoked a
1340 segmentation fault instead of returning the appropriate error message.
1341
1342 15. In multiline mode when the newline sequence was set to "any", the pattern
1343 ^$ would give a match between the \r and \n of a subject such as "A\r\nB".
1344 This doesn't seem right; it now treats the CRLF combination as the line
1345 ending, and so does not match in that case. It's only a pattern such as ^$
1346 that would hit this one: something like ^ABC$ would have failed after \r
1347 and then tried again after \r\n.
1348
1349 16. Changed the comparison command for RunGrepTest from "diff -u" to "diff -ub"
1350 in an attempt to make files that differ only in their line terminators
1351 compare equal. This works on Linux.
1352
1353 17. Under certain error circumstances pcregrep might try to free random memory
1354 as it exited. This is now fixed, thanks to valgrind.
1355
1356 19. In pcretest, if the pattern /(?m)^$/g<any> was matched against the string
1357 "abc\r\n\r\n", it found an unwanted second match after the second \r. This
1358 was because its rules for how to advance for /g after matching an empty
1359 string at the end of a line did not allow for this case. They now check for
1360 it specially.
1361
1362 20. pcretest is supposed to handle patterns and data of any length, by
1363 extending its buffers when necessary. It was getting this wrong when the
1364 buffer for a data line had to be extended.
1365
1366 21. Added PCRE_NEWLINE_ANYCRLF which is like ANY, but matches only CR, LF, or
1367 CRLF as a newline sequence.
1368
1369 22. Code for handling Unicode properties in pcre_dfa_exec() wasn't being cut
1370 out by #ifdef SUPPORT_UCP. This did no harm, as it could never be used, but
1371 I have nevertheless tidied it up.
1372
1373 23. Added some casts to kill warnings from HP-UX ia64 compiler.
1374
1375 24. Added a man page for pcre-config.
1376
1377
1378 Version 7.0 19-Dec-06
1379 ---------------------
1380
1381 1. Fixed a signed/unsigned compiler warning in pcre_compile.c, shown up by
1382 moving to gcc 4.1.1.
1383
1384 2. The -S option for pcretest uses setrlimit(); I had omitted to #include
1385 sys/time.h, which is documented as needed for this function. It doesn't
1386 seem to matter on Linux, but it showed up on some releases of OS X.
1387
1388 3. It seems that there are systems where bytes whose values are greater than
1389 127 match isprint() in the "C" locale. The "C" locale should be the
1390 default when a C program starts up. In most systems, only ASCII printing
1391 characters match isprint(). This difference caused the output from pcretest
1392 to vary, making some of the tests fail. I have changed pcretest so that:
1393
1394 (a) When it is outputting text in the compiled version of a pattern, bytes
1395 other than 32-126 are always shown as hex escapes.
1396
1397 (b) When it is outputting text that is a matched part of a subject string,
1398 it does the same, unless a different locale has been set for the match
1399 (using the /L modifier). In this case, it uses isprint() to decide.
1400
1401 4. Fixed a major bug that caused incorrect computation of the amount of memory
1402 required for a compiled pattern when options that changed within the
1403 pattern affected the logic of the preliminary scan that determines the
1404 length. The relevant options are -x, and -i in UTF-8 mode. The result was
1405 that the computed length was too small. The symptoms of this bug were
1406 either the PCRE error "internal error: code overflow" from pcre_compile(),
1407 or a glibc crash with a message such as "pcretest: free(): invalid next
1408 size (fast)". Examples of patterns that provoked this bug (shown in
1409 pcretest format) are:
1410
1411 /(?-x: )/x
1412 /(?x)(?-x: \s*#\s*)/
1413 /((?i)[\x{c0}])/8
1414 /(?i:[\x{c0}])/8
1415
1416 HOWEVER: Change 17 below makes this fix obsolete as the memory computation
1417 is now done differently.
1418
1419 5. Applied patches from Google to: (a) add a QuoteMeta function to the C++
1420 wrapper classes; (b) implement a new function in the C++ scanner that is
1421 more efficient than the old way of doing things because it avoids levels of
1422 recursion in the regex matching; (c) add a paragraph to the documentation
1423 for the FullMatch() function.
1424
1425 6. The escape sequence \n was being treated as whatever was defined as
1426 "newline". Not only was this contrary to the documentation, which states
1427 that \n is character 10 (hex 0A), but it also went horribly wrong when
1428 "newline" was defined as CRLF. This has been fixed.
1429
1430 7. In pcre_dfa_exec.c the value of an unsigned integer (the variable called c)
1431 was being set to -1 for the "end of line" case (supposedly a value that no
1432 character can have). Though this value is never used (the check for end of
1433 line is "zero bytes in current character"), it caused compiler complaints.
1434 I've changed it to 0xffffffff.
1435
1436 8. In pcre_version.c, the version string was being built by a sequence of
1437 C macros that, in the event of PCRE_PRERELEASE being defined as an empty
1438 string (as it is for production releases) called a macro with an empty
1439 argument. The C standard says the result of this is undefined. The gcc
1440 compiler treats it as an empty string (which was what was wanted) but it is
1441 reported that Visual C gives an error. The source has been hacked around to
1442 avoid this problem.
1443
1444 9. On the advice of a Windows user, included <io.h> and <fcntl.h> in Windows
1445 builds of pcretest, and changed the call to _setmode() to use _O_BINARY
1446 instead of 0x8000. Made all the #ifdefs test both _WIN32 and WIN32 (not all
1447 of them did).
1448
1449 10. Originally, pcretest opened its input and output without "b"; then I was
1450 told that "b" was needed in some environments, so it was added for release
1451 5.0 to both the input and output. (It makes no difference on Unix-like
1452 systems.) Later I was told that it is wrong for the input on Windows. I've
1453 now abstracted the modes into two macros, to make it easier to fiddle with
1454 them, and removed "b" from the input mode under Windows.
1455
1456 11. Added pkgconfig support for the C++ wrapper library, libpcrecpp.
1457
1458 12. Added -help and --help to pcretest as an official way of being reminded
1459 of the options.
1460
1461 13. Removed some redundant semicolons after macro calls in pcrecpparg.h.in
1462 and pcrecpp.cc because they annoy compilers at high warning levels.
1463
1464 14. A bit of tidying/refactoring in pcre_exec.c in the main bumpalong loop.
1465
1466 15. Fixed an occurrence of == in configure.ac that should have been = (shell
1467 scripts are not C programs :-) and which was not noticed because it works
1468 on Linux.
1469
1470 16. pcretest is supposed to handle any length of pattern and data line (as one
1471 line or as a continued sequence of lines) by extending its input buffer if
1472 necessary. This feature was broken for very long pattern lines, leading to
1473 a string of junk being passed to pcre_compile() if the pattern was longer
1474 than about 50K.
1475
1476 17. I have done a major re-factoring of the way pcre_compile() computes the
1477 amount of memory needed for a compiled pattern. Previously, there was code
1478 that made a preliminary scan of the pattern in order to do this. That was
1479 OK when PCRE was new, but as the facilities have expanded, it has become
1480 harder and harder to keep it in step with the real compile phase, and there
1481 have been a number of bugs (see for example, 4 above). I have now found a
1482 cunning way of running the real compile function in a "fake" mode that
1483 enables it to compute how much memory it would need, while actually only
1484 ever using a few hundred bytes of working memory and without too many
1485 tests of the mode. This should make future maintenance and development
1486 easier. A side effect of this work is that the limit of 200 on the nesting
1487 depth of parentheses has been removed (though this was never a serious
1488 limitation, I suspect). However, there is a downside: pcre_compile() now
1489 runs more slowly than before (30% or more, depending on the pattern). I
1490 hope this isn't a big issue. There is no effect on runtime performance.
1491
1492 18. Fixed a minor bug in pcretest: if a pattern line was not terminated by a
1493 newline (only possible for the last line of a file) and it was a
1494 pattern that set a locale (followed by /Lsomething), pcretest crashed.
1495
1496 19. Added additional timing features to pcretest. (1) The -tm option now times
1497 matching only, not compiling. (2) Both -t and -tm can be followed, as a
1498 separate command line item, by a number that specifies the number of
1499 repeats to use when timing. The default is 50000; this gives better
1500 precision, but takes uncomfortably long for very large patterns.
1501
1502 20. Extended pcre_study() to be more clever in cases where a branch of a
1503 subpattern has no definite first character. For example, (a*|b*)[cd] would
1504 previously give no result from pcre_study(). Now it recognizes that the
1505 first character must be a, b, c, or d.
1506
1507 21. There was an incorrect error "recursive call could loop indefinitely" if
1508 a subpattern (or the entire pattern) that was being tested for matching an
1509 empty string contained only one non-empty item after a nested subpattern.
1510 For example, the pattern (?>\x{100}*)\d(?R) provoked this error
1511 incorrectly, because the \d was being skipped in the check.
1512
1513 22. The pcretest program now has a new pattern option /B and a command line
1514 option -b, which is equivalent to adding /B to every pattern. This causes
1515 it to show the compiled bytecode, without the additional information that
1516 -d shows. The effect of -d is now the same as -b with -i (and similarly, /D
1517 is the same as /B/I).
1518
1519 23. A new optimization is now able automatically to treat some sequences such
1520 as a*b as a*+b. More specifically, if something simple (such as a character
1521 or a simple class like \d) has an unlimited quantifier, and is followed by
1522 something that cannot possibly match the quantified thing, the quantifier
1523 is automatically "possessified".
1524
1525 24. A recursive reference to a subpattern whose number was greater than 39
1526 went wrong under certain circumstances in UTF-8 mode. This bug could also
1527 have affected the operation of pcre_study().
1528
1529 25. Realized that a little bit of performance could be had by replacing
1530 (c & 0xc0) == 0xc0 with c >= 0xc0 when processing UTF-8 characters.
1531
1532 26. Timing data from pcretest is now shown to 4 decimal places instead of 3.
1533
1534 27. Possessive quantifiers such as a++ were previously implemented by turning
1535 them into atomic groups such as ($>a+). Now they have their own opcodes,
1536 which improves performance. This includes the automatically created ones
1537 from 23 above.
1538
1539 28. A pattern such as (?=(\w+))\1: which simulates an atomic group using a
1540 lookahead was broken if it was not anchored. PCRE was mistakenly expecting
1541 the first matched character to be a colon. This applied both to named and
1542 numbered groups.
1543
1544 29. The ucpinternal.h header file was missing its idempotency #ifdef.
1545
1546 30. I was sent a "project" file called libpcre.a.dev which I understand makes
1547 building PCRE on Windows easier, so I have included it in the distribution.
1548
1549 31. There is now a check in pcretest against a ridiculously large number being
1550 returned by pcre_exec() or pcre_dfa_exec(). If this happens in a /g or /G
1551 loop, the loop is abandoned.
1552
1553 32. Forward references to subpatterns in conditions such as (?(2)...) where
1554 subpattern 2 is defined later cause pcre_compile() to search forwards in
1555 the pattern for the relevant set of parentheses. This search went wrong
1556 when there were unescaped parentheses in a character class, parentheses
1557 escaped with \Q...\E, or parentheses in a #-comment in /x mode.
1558
1559 33. "Subroutine" calls and backreferences were previously restricted to
1560 referencing subpatterns earlier in the regex. This restriction has now
1561 been removed.
1562
1563 34. Added a number of extra features that are going to be in Perl 5.10. On the
1564 whole, these are just syntactic alternatives for features that PCRE had
1565 previously implemented using the Python syntax or my own invention. The
1566 other formats are all retained for compatibility.
1567
1568 (a) Named groups can now be defined as (?<name>...) or (?'name'...) as well
1569 as (?P<name>...). The new forms, as well as being in Perl 5.10, are
1570 also .NET compatible.
1571
1572 (b) A recursion or subroutine call to a named group can now be defined as
1573 (?&name) as well as (?P>name).
1574
1575 (c) A backreference to a named group can now be defined as \k<name> or
1576 \k'name' as well as (?P=name). The new forms, as well as being in Perl
1577 5.10, are also .NET compatible.
1578
1579 (d) A conditional reference to a named group can now use the syntax
1580 (?(<name>) or (?('name') as well as (?(name).
1581
1582 (e) A "conditional group" of the form (?(DEFINE)...) can be used to define
1583 groups (named and numbered) that are never evaluated inline, but can be
1584 called as "subroutines" from elsewhere. In effect, the DEFINE condition
1585 is always false. There may be only one alternative in such a group.
1586
1587 (f) A test for recursion can be given as (?(R1).. or (?(R&name)... as well
1588 as the simple (?(R). The condition is true only if the most recent
1589 recursion is that of the given number or name. It does not search out
1590 through the entire recursion stack.
1591
1592 (g) The escape \gN or \g{N} has been added, where N is a positive or
1593 negative number, specifying an absolute or relative reference.
1594
1595 35. Tidied to get rid of some further signed/unsigned compiler warnings and
1596 some "unreachable code" warnings.
1597
1598 36. Updated the Unicode property tables to Unicode version 5.0.0. Amongst other
1599 things, this adds five new scripts.
1600
1601 37. Perl ignores orphaned \E escapes completely. PCRE now does the same.
1602 There were also incompatibilities regarding the handling of \Q..\E inside
1603 character classes, for example with patterns like [\Qa\E-\Qz\E] where the
1604 hyphen was adjacent to \Q or \E. I hope I've cleared all this up now.
1605
1606 38. Like Perl, PCRE detects when an indefinitely repeated parenthesized group
1607 matches an empty string, and forcibly breaks the loop. There were bugs in
1608 this code in non-simple cases. For a pattern such as ^(a()*)* matched
1609 against aaaa the result was just "a" rather than "aaaa", for example. Two
1610 separate and independent bugs (that affected different cases) have been
1611 fixed.
1612
1613 39. Refactored the code to abolish the use of different opcodes for small
1614 capturing bracket numbers. This is a tidy that I avoided doing when I
1615 removed the limit on the number of capturing brackets for 3.5 back in 2001.
1616 The new approach is not only tidier, it makes it possible to reduce the
1617 memory needed to fix the previous bug (38).
1618
1619 40. Implemented PCRE_NEWLINE_ANY to recognize any of the Unicode newline
1620 sequences (http://unicode.org/unicode/reports/tr18/) as "newline" when
1621 processing dot, circumflex, or dollar metacharacters, or #-comments in /x
1622 mode.
1623
1624 41. Add \R to match any Unicode newline sequence, as suggested in the Unicode
1625 report.
1626
1627 42. Applied patch, originally from Ari Pollak, modified by Google, to allow
1628 copy construction and assignment in the C++ wrapper.
1629
1630 43. Updated pcregrep to support "--newline=any". In the process, I fixed a
1631 couple of bugs that could have given wrong results in the "--newline=crlf"
1632 case.
1633
1634 44. Added a number of casts and did some reorganization of signed/unsigned int
1635 variables following suggestions from Dair Grant. Also renamed the variable
1636 "this" as "item" because it is a C++ keyword.
1637
1638 45. Arranged for dftables to add
1639
1640 #include "pcre_internal.h"
1641
1642 to pcre_chartables.c because without it, gcc 4.x may remove the array
1643 definition from the final binary if PCRE is built into a static library and
1644 dead code stripping is activated.
1645
1646 46. For an unanchored pattern, if a match attempt fails at the start of a
1647 newline sequence, and the newline setting is CRLF or ANY, and the next two
1648 characters are CRLF, advance by two characters instead of one.
1649
1650
1651 Version 6.7 04-Jul-06
1652 ---------------------
1653
1654 1. In order to handle tests when input lines are enormously long, pcretest has
1655 been re-factored so that it automatically extends its buffers when
1656 necessary. The code is crude, but this _is_ just a test program. The
1657 default size has been increased from 32K to 50K.
1658
1659 2. The code in pcre_study() was using the value of the re argument before
1660 testing it for NULL. (Of course, in any sensible call of the function, it
1661 won't be NULL.)
1662
1663 3. The memmove() emulation function in pcre_internal.h, which is used on
1664 systems that lack both memmove() and bcopy() - that is, hardly ever -
1665 was missing a "static" storage class specifier.
1666
1667 4. When UTF-8 mode was not set, PCRE looped when compiling certain patterns
1668 containing an extended class (one that cannot be represented by a bitmap
1669 because it contains high-valued characters or Unicode property items, e.g.
1670 [\pZ]). Almost always one would set UTF-8 mode when processing such a
1671 pattern, but PCRE should not loop if you do not (it no longer does).
1672 [Detail: two cases were found: (a) a repeated subpattern containing an
1673 extended class; (b) a recursive reference to a subpattern that followed a
1674 previous extended class. It wasn't skipping over the extended class
1675 correctly when UTF-8 mode was not set.]
1676
1677 5. A negated single-character class was not being recognized as fixed-length
1678 in lookbehind assertions such as (?<=[^f]), leading to an incorrect
1679 compile error "lookbehind assertion is not fixed length".
1680
1681 6. The RunPerlTest auxiliary script was showing an unexpected difference
1682 between PCRE and Perl for UTF-8 tests. It turns out that it is hard to
1683 write a Perl script that can interpret lines of an input file either as
1684 byte characters or as UTF-8, which is what "perltest" was being required to
1685 do for the non-UTF-8 and UTF-8 tests, respectively. Essentially what you
1686 can't do is switch easily at run time between having the "use utf8;" pragma
1687 or not. In the end, I fudged it by using the RunPerlTest script to insert
1688 "use utf8;" explicitly for the UTF-8 tests.
1689
1690 7. In multiline (/m) mode, PCRE was matching ^ after a terminating newline at
1691 the end of the subject string, contrary to the documentation and to what
1692 Perl does. This was true of both matching functions. Now it matches only at
1693 the start of the subject and immediately after *internal* newlines.
1694
1695 8. A call of pcre_fullinfo() from pcretest to get the option bits was passing
1696 a pointer to an int instead of a pointer to an unsigned long int. This
1697 caused problems on 64-bit systems.
1698
1699 9. Applied a patch from the folks at Google to pcrecpp.cc, to fix "another
1700 instance of the 'standard' template library not being so standard".
1701
1702 10. There was no check on the number of named subpatterns nor the maximum
1703 length of a subpattern name. The product of these values is used to compute
1704 the size of the memory block for a compiled pattern. By supplying a very
1705 long subpattern name and a large number of named subpatterns, the size
1706 computation could be caused to overflow. This is now prevented by limiting
1707 the length of names to 32 characters, and the number of named subpatterns
1708 to 10,000.
1709
1710 11. Subpatterns that are repeated with specific counts have to be replicated in
1711 the compiled pattern. The size of memory for this was computed from the
1712 length of the subpattern and the repeat count. The latter is limited to
1713 65535, but there was no limit on the former, meaning that integer overflow
1714 could in principle occur. The compiled length of a repeated subpattern is
1715 now limited to 30,000 bytes in order to prevent this.
1716
1717 12. Added the optional facility to have named substrings with the same name.
1718
1719 13. Added the ability to use a named substring as a condition, using the
1720 Python syntax: (?(name)yes|no). This overloads (?(R)... and names that
1721 are numbers (not recommended). Forward references are permitted.
1722
1723 14. Added forward references in named backreferences (if you see what I mean).
1724
1725 15. In UTF-8 mode, with the PCRE_DOTALL option set, a quantified dot in the
1726 pattern could run off the end of the subject. For example, the pattern
1727 "(?s)(.{1,5})"8 did this with the subject "ab".
1728
1729 16. If PCRE_DOTALL or PCRE_MULTILINE were set, pcre_dfa_exec() behaved as if
1730 PCRE_CASELESS was set when matching characters that were quantified with ?
1731 or *.
1732
1733 17. A character class other than a single negated character that had a minimum
1734 but no maximum quantifier - for example [ab]{6,} - was not handled
1735 correctly by pce_dfa_exec(). It would match only one character.
1736
1737 18. A valid (though odd) pattern that looked like a POSIX character
1738 class but used an invalid character after [ (for example [[,abc,]]) caused
1739 pcre_compile() to give the error "Failed: internal error: code overflow" or
1740 in some cases to crash with a glibc free() error. This could even happen if
1741 the pattern terminated after [[ but there just happened to be a sequence of
1742 letters, a binary zero, and a closing ] in the memory that followed.
1743
1744 19. Perl's treatment of octal escapes in the range \400 to \777 has changed
1745 over the years. Originally (before any Unicode support), just the bottom 8
1746 bits were taken. Thus, for example, \500 really meant \100. Nowadays the
1747 output from "man perlunicode" includes this:
1748
1749 The regular expression compiler produces polymorphic opcodes. That
1750 is, the pattern adapts to the data and automatically switches to
1751 the Unicode character scheme when presented with Unicode data--or
1752 instead uses a traditional byte scheme when presented with byte
1753 data.
1754
1755 Sadly, a wide octal escape does not cause a switch, and in a string with
1756 no other multibyte characters, these octal escapes are treated as before.
1757 Thus, in Perl, the pattern /\500/ actually matches \100 but the pattern
1758 /\500|\x{1ff}/ matches \500 or \777 because the whole thing is treated as a
1759 Unicode string.
1760
1761 I have not perpetrated such confusion in PCRE. Up till now, it took just
1762 the bottom 8 bits, as in old Perl. I have now made octal escapes with
1763 values greater than \377 illegal in non-UTF-8 mode. In UTF-8 mode they
1764 translate to the appropriate multibyte character.
1765
1766 29. Applied some refactoring to reduce the number of warnings from Microsoft
1767 and Borland compilers. This has included removing the fudge introduced
1768 seven years ago for the OS/2 compiler (see 2.02/2 below) because it caused
1769 a warning about an unused variable.
1770
1771 21. PCRE has not included VT (character 0x0b) in the set of whitespace
1772 characters since release 4.0, because Perl (from release 5.004) does not.
1773 [Or at least, is documented not to: some releases seem to be in conflict
1774 with the documentation.] However, when a pattern was studied with
1775 pcre_study() and all its branches started with \s, PCRE still included VT
1776 as a possible starting character. Of course, this did no harm; it just
1777 caused an unnecessary match attempt.
1778
1779 22. Removed a now-redundant internal flag bit that recorded the fact that case
1780 dependency changed within the pattern. This was once needed for "required
1781 byte" processing, but is no longer used. This recovers a now-scarce options
1782 bit. Also moved the least significant internal flag bit to the most-
1783 significant bit of the word, which was not previously used (hangover from
1784 the days when it was an int rather than a uint) to free up another bit for
1785 the future.
1786
1787 23. Added support for CRLF line endings as well as CR and LF. As well as the
1788 default being selectable at build time, it can now be changed at runtime
1789 via the PCRE_NEWLINE_xxx flags. There are now options for pcregrep to
1790 specify that it is scanning data with non-default line endings.
1791
1792 24. Changed the definition of CXXLINK to make it agree with the definition of
1793 LINK in the Makefile, by replacing LDFLAGS to CXXFLAGS.
1794
1795 25. Applied Ian Taylor's patches to avoid using another stack frame for tail
1796 recursions. This makes a big different to stack usage for some patterns.
1797
1798 26. If a subpattern containing a named recursion or subroutine reference such
1799 as (?P>B) was quantified, for example (xxx(?P>B)){3}, the calculation of
1800 the space required for the compiled pattern went wrong and gave too small a
1801 value. Depending on the environment, this could lead to "Failed: internal
1802 error: code overflow at offset 49" or "glibc detected double free or
1803 corruption" errors.
1804
1805 27. Applied patches from Google (a) to support the new newline modes and (b) to
1806 advance over multibyte UTF-8 characters in GlobalReplace.
1807
1808 28. Change free() to pcre_free() in pcredemo.c. Apparently this makes a
1809 difference for some implementation of PCRE in some Windows version.
1810
1811 29. Added some extra testing facilities to pcretest:
1812
1813 \q<number> in a data line sets the "match limit" value
1814 \Q<number> in a data line sets the "match recursion limt" value
1815 -S <number> sets the stack size, where <number> is in megabytes
1816
1817 The -S option isn't available for Windows.
1818
1819
1820 Version 6.6 06-Feb-06
1821 ---------------------
1822
1823 1. Change 16(a) for 6.5 broke things, because PCRE_DATA_SCOPE was not defined
1824 in pcreposix.h. I have copied the definition from pcre.h.
1825
1826 2. Change 25 for 6.5 broke compilation in a build directory out-of-tree
1827 because pcre.h is no longer a built file.
1828
1829 3. Added Jeff Friedl's additional debugging patches to pcregrep. These are
1830 not normally included in the compiled code.
1831
1832
1833 Version 6.5 01-Feb-06
1834 ---------------------
1835
1836 1. When using the partial match feature with pcre_dfa_exec(), it was not
1837 anchoring the second and subsequent partial matches at the new starting
1838 point. This could lead to incorrect results. For example, with the pattern
1839 /1234/, partially matching against "123" and then "a4" gave a match.
1840
1841 2. Changes to pcregrep:
1842
1843 (a) All non-match returns from pcre_exec() were being treated as failures
1844 to match the line. Now, unless the error is PCRE_ERROR_NOMATCH, an
1845 error message is output. Some extra information is given for the
1846 PCRE_ERROR_MATCHLIMIT and PCRE_ERROR_RECURSIONLIMIT errors, which are
1847 probably the only errors that are likely to be caused by users (by
1848 specifying a regex that has nested indefinite repeats, for instance).
1849 If there are more than 20 of these errors, pcregrep is abandoned.
1850
1851 (b) A binary zero was treated as data while matching, but terminated the
1852 output line if it was written out. This has been fixed: binary zeroes
1853 are now no different to any other data bytes.
1854
1855 (c) Whichever of the LC_ALL or LC_CTYPE environment variables is set is
1856 used to set a locale for matching. The --locale=xxxx long option has
1857 been added (no short equivalent) to specify a locale explicitly on the
1858 pcregrep command, overriding the environment variables.
1859
1860 (d) When -B was used with -n, some line numbers in the output were one less
1861 than they should have been.
1862
1863 (e) Added the -o (--only-matching) option.
1864
1865 (f) If -A or -C was used with -c (count only), some lines of context were
1866 accidentally printed for the final match.
1867
1868 (g) Added the -H (--with-filename) option.
1869
1870 (h) The combination of options -rh failed to suppress file names for files
1871 that were found from directory arguments.
1872
1873 (i) Added the -D (--devices) and -d (--directories) options.
1874
1875 (j) Added the -F (--fixed-strings) option.
1876
1877 (k) Allow "-" to be used as a file name for -f as well as for a data file.
1878
1879 (l) Added the --colo(u)r option.
1880
1881 (m) Added Jeffrey Friedl's -S testing option, but within #ifdefs so that it
1882 is not present by default.
1883
1884 3. A nasty bug was discovered in the handling of recursive patterns, that is,
1885 items such as (?R) or (?1), when the recursion could match a number of
1886 alternatives. If it matched one of the alternatives, but subsequently,
1887 outside the recursion, there was a failure, the code tried to back up into
1888 the recursion. However, because of the way PCRE is implemented, this is not
1889 possible, and the result was an incorrect result from the match.
1890
1891 In order to prevent this happening, the specification of recursion has
1892 been changed so that all such subpatterns are automatically treated as
1893 atomic groups. Thus, for example, (?R) is treated as if it were (?>(?R)).
1894
1895 4. I had overlooked the fact that, in some locales, there are characters for
1896 which isalpha() is true but neither isupper() nor islower() are true. In
1897 the fr_FR locale, for instance, the \xAA and \xBA characters (ordmasculine
1898 and ordfeminine) are like this. This affected the treatment of \w and \W
1899 when they appeared in character classes, but not when they appeared outside
1900 a character class. The bit map for "word" characters is now created
1901 separately from the results of isalnum() instead of just taking it from the
1902 upper, lower, and digit maps. (Plus the underscore character, of course.)
1903
1904 5. The above bug also affected the handling of POSIX character classes such as
1905 [[:alpha:]] and [[:alnum:]]. These do not have their own bit maps in PCRE's
1906 permanent tables. Instead, the bit maps for such a class were previously
1907 created as the appropriate unions of the upper, lower, and digit bitmaps.
1908 Now they are created by subtraction from the [[:word:]] class, which has
1909 its own bitmap.
1910
1911 6. The [[:blank:]] character class matches horizontal, but not vertical space.
1912 It is created by subtracting the vertical space characters (\x09, \x0a,
1913 \x0b, \x0c) from the [[:space:]] bitmap. Previously, however, the
1914 subtraction was done in the overall bitmap for a character class, meaning
1915 that a class such as [\x0c[:blank:]] was incorrect because \x0c would not
1916 be recognized. This bug has been fixed.
1917
1918 7. Patches from the folks at Google:
1919
1920 (a) pcrecpp.cc: "to handle a corner case that may or may not happen in
1921 real life, but is still worth protecting against".
1922
1923 (b) pcrecpp.cc: "corrects a bug when negative radixes are used with
1924 regular expressions".
1925
1926 (c) pcre_scanner.cc: avoid use of std::count() because not all systems
1927 have it.
1928
1929 (d) Split off pcrecpparg.h from pcrecpp.h and had the former built by
1930 "configure" and the latter not, in order to fix a problem somebody had
1931 with compiling the Arg class on HP-UX.
1932
1933 (e) Improve the error-handling of the C++ wrapper a little bit.
1934
1935 (f) New tests for checking recursion limiting.
1936
1937 8. The pcre_memmove() function, which is used only if the environment does not
1938 have a standard memmove() function (and is therefore rarely compiled),
1939 contained two bugs: (a) use of int instead of size_t, and (b) it was not
1940 returning a result (though PCRE never actually uses the result).
1941
1942 9. In the POSIX regexec() interface, if nmatch is specified as a ridiculously
1943 large number - greater than INT_MAX/(3*sizeof(int)) - REG_ESPACE is
1944 returned instead of calling malloc() with an overflowing number that would
1945 most likely cause subsequent chaos.
1946
1947 10. The debugging option of pcretest was not showing the NO_AUTO_CAPTURE flag.
1948
1949 11. The POSIX flag REG_NOSUB is now supported. When a pattern that was compiled
1950 with this option is matched, the nmatch and pmatch options of regexec() are
1951 ignored.
1952
1953 12. Added REG_UTF8 to the POSIX interface. This is not defined by POSIX, but is
1954 provided in case anyone wants to the the POSIX interface with UTF-8
1955 strings.
1956
1957 13. Added CXXLDFLAGS to the Makefile parameters to provide settings only on the
1958 C++ linking (needed for some HP-UX environments).
1959
1960 14. Avoid compiler warnings in get_ucpname() when compiled without UCP support
1961 (unused parameter) and in the pcre_printint() function (omitted "default"
1962 switch label when the default is to do nothing).
1963
1964 15. Added some code to make it possible, when PCRE is compiled as a C++
1965 library, to replace subject pointers for pcre_exec() with a smart pointer
1966 class, thus making it possible to process discontinuous strings.
1967
1968 16. The two macros PCRE_EXPORT and PCRE_DATA_SCOPE are confusing, and perform
1969 much the same function. They were added by different people who were trying
1970 to make PCRE easy to compile on non-Unix systems. It has been suggested
1971 that PCRE_EXPORT be abolished now that there is more automatic apparatus
1972 for compiling on Windows systems. I have therefore replaced it with
1973 PCRE_DATA_SCOPE. This is set automatically for Windows; if not set it
1974 defaults to "extern" for C or "extern C" for C++, which works fine on
1975 Unix-like systems. It is now possible to override the value of PCRE_DATA_
1976 SCOPE with something explicit in config.h. In addition:
1977
1978 (a) pcreposix.h still had just "extern" instead of either of these macros;
1979 I have replaced it with PCRE_DATA_SCOPE.
1980
1981 (b) Functions such as _pcre_xclass(), which are internal to the library,
1982 but external in the C sense, all had PCRE_EXPORT in their definitions.
1983 This is apparently wrong for the Windows case, so I have removed it.
1984 (It makes no difference on Unix-like systems.)
1985
1986 17. Added a new limit, MATCH_LIMIT_RECURSION, which limits the depth of nesting
1987 of recursive calls to match(). This is different to MATCH_LIMIT because
1988 that limits the total number of calls to match(), not all of which increase
1989 the depth of recursion. Limiting the recursion depth limits the amount of
1990 stack (or heap if NO_RECURSE is set) that is used. The default can be set
1991 when PCRE is compiled, and changed at run time. A patch from Google adds
1992 this functionality to the C++ interface.
1993
1994 18. Changes to the handling of Unicode character properties:
1995
1996 (a) Updated the table to Unicode 4.1.0.
1997
1998 (b) Recognize characters that are not in the table as "Cn" (undefined).
1999
2000 (c) I revised the way the table is implemented to a much improved format
2001 which includes recognition of ranges. It now supports the ranges that
2002 are defined in UnicodeData.txt, and it also amalgamates other
2003 characters into ranges. This has reduced the number of entries in the
2004 table from around 16,000 to around 3,000, thus reducing its size
2005 considerably. I realized I did not need to use a tree structure after
2006 all - a binary chop search is just as efficient. Having reduced the
2007 number of entries, I extended their size from 6 bytes to 8 bytes to
2008 allow for more data.
2009
2010 (d) Added support for Unicode script names via properties such as \p{Han}.
2011
2012 19. In UTF-8 mode, a backslash followed by a non-Ascii character was not
2013 matching that character.
2014
2015 20. When matching a repeated Unicode property with a minimum greater than zero,
2016 (for example \pL{2,}), PCRE could look past the end of the subject if it
2017 reached it while seeking the minimum number of characters. This could
2018 happen only if some of the characters were more than one byte long, because
2019 there is a check for at least the minimum number of bytes.
2020
2021 21. Refactored the implementation of \p and \P so as to be more general, to
2022 allow for more different types of property in future. This has changed the
2023 compiled form incompatibly. Anybody with saved compiled patterns that use
2024 \p or \P will have to recompile them.
2025
2026 22. Added "Any" and "L&" to the supported property types.
2027
2028 23. Recognize \x{...} as a code point specifier, even when not in UTF-8 mode,
2029 but give a compile time error if the value is greater than 0xff.
2030
2031 24. The man pages for pcrepartial, pcreprecompile, and pcre_compile2 were
2032 accidentally not being installed or uninstalled.
2033
2034 25. The pcre.h file was built from pcre.h.in, but the only changes that were
2035 made were to insert the current release number. This seemed silly, because
2036 it made things harder for people building PCRE on systems that don't run
2037 "configure". I have turned pcre.h into a distributed file, no longer built
2038 by "configure", with the version identification directly included. There is
2039 no longer a pcre.h.in file.
2040
2041 However, this change necessitated a change to the pcre-config script as
2042 well. It is built from pcre-config.in, and one of the substitutions was the
2043 release number. I have updated configure.ac so that ./configure now finds
2044 the release number by grepping pcre.h.
2045
2046 26. Added the ability to run the tests under valgrind.
2047
2048
2049 Version 6.4 05-Sep-05
2050 ---------------------
2051
2052 1. Change 6.0/10/(l) to pcregrep introduced a bug that caused separator lines
2053 "--" to be printed when multiple files were scanned, even when none of the
2054 -A, -B, or -C options were used. This is not compatible with Gnu grep, so I
2055 consider it to be a bug, and have restored the previous behaviour.
2056
2057 2. A couple of code tidies to get rid of compiler warnings.
2058
2059 3. The pcretest program used to cheat by referring to symbols in the library
2060 whose names begin with _pcre_. These are internal symbols that are not
2061 really supposed to be visible externally, and in some environments it is
2062 possible to suppress them. The cheating is now confined to including
2063 certain files from the library's source, which is a bit cleaner.
2064
2065 4. Renamed pcre.in as pcre.h.in to go with pcrecpp.h.in; it also makes the
2066 file's purpose clearer.
2067
2068 5. Reorganized pcre_ucp_findchar().
2069
2070
2071 Version 6.3 15-Aug-05
2072 ---------------------
2073
2074 1. The file libpcre.pc.in did not have general read permission in the tarball.
2075
2076 2. There were some problems when building without C++ support:
2077
2078 (a) If C++ support was not built, "make install" and "make test" still
2079 tried to test it.
2080
2081 (b) There were problems when the value of CXX was explicitly set. Some
2082 changes have been made to try to fix these, and ...
2083
2084 (c) --disable-cpp can now be used to explicitly disable C++ support.
2085
2086 (d) The use of @CPP_OBJ@ directly caused a blank line preceded by a
2087 backslash in a target when C++ was disabled. This confuses some
2088 versions of "make", apparently. Using an intermediate variable solves
2089 this. (Same for CPP_LOBJ.)
2090
2091 3. $(LINK_FOR_BUILD) now includes $(CFLAGS_FOR_BUILD) and $(LINK)
2092 (non-Windows) now includes $(CFLAGS) because these flags are sometimes
2093 necessary on certain architectures.
2094
2095 4. Added a setting of -export-symbols-regex to the link command to remove
2096 those symbols that are exported in the C sense, but actually are local
2097 within the library, and not documented. Their names all begin with
2098 "_pcre_". This is not a perfect job, because (a) we have to except some
2099 symbols that pcretest ("illegally") uses, and (b) the facility isn't always
2100 available (and never for static libraries). I have made a note to try to
2101 find a way round (a) in the future.
2102
2103
2104 Version 6.2 01-Aug-05
2105 ---------------------
2106
2107 1. There was no test for integer overflow of quantifier values. A construction
2108 such as {1111111111111111} would give undefined results. What is worse, if
2109 a minimum quantifier for a parenthesized subpattern overflowed and became
2110 negative, the calculation of the memory size went wrong. This could have
2111 led to memory overwriting.
2112
2113 2. Building PCRE using VPATH was broken. Hopefully it is now fixed.
2114
2115 3. Added "b" to the 2nd argument of fopen() in dftables.c, for non-Unix-like
2116 operating environments where this matters.
2117
2118 4. Applied Giuseppe Maxia's patch to add additional features for controlling
2119 PCRE options from within the C++ wrapper.
2120
2121 5. Named capturing subpatterns were not being correctly counted when a pattern
2122 was compiled. This caused two problems: (a) If there were more than 100
2123 such subpatterns, the calculation of the memory needed for the whole
2124 compiled pattern went wrong, leading to an overflow error. (b) Numerical
2125 back references of the form \12, where the number was greater than 9, were
2126 not recognized as back references, even though there were sufficient
2127 previous subpatterns.
2128
2129 6. Two minor patches to pcrecpp.cc in order to allow it to compile on older
2130 versions of gcc, e.g. 2.95.4.
2131
2132
2133 Version 6.1 21-Jun-05
2134 ---------------------
2135
2136 1. There was one reference to the variable "posix" in pcretest.c that was not
2137 surrounded by "#if !defined NOPOSIX".
2138
2139 2. Make it possible to compile pcretest without DFA support, UTF8 support, or
2140 the cross-check on the old pcre_info() function, for the benefit of the
2141 cut-down version of PCRE that is currently imported into Exim.
2142
2143 3. A (silly) pattern starting with (?i)(?-i) caused an internal space
2144 allocation error. I've done the easy fix, which wastes 2 bytes for sensible
2145 patterns that start (?i) but I don't think that matters. The use of (?i) is
2146 just an example; this all applies to the other options as well.
2147
2148 4. Since libtool seems to echo the compile commands it is issuing, the output
2149 from "make" can be reduced a bit by putting "@" in front of each libtool
2150 compile command.
2151
2152 5. Patch from the folks at Google for configure.in to be a bit more thorough
2153 in checking for a suitable C++ installation before trying to compile the
2154 C++ stuff. This should fix a reported problem when a compiler was present,
2155 but no suitable headers.
2156
2157 6. The man pages all had just "PCRE" as their title. I have changed them to
2158 be the relevant file name. I have also arranged that these names are
2159 retained in the file doc/pcre.txt, which is a concatenation in text format
2160 of all the man pages except the little individual ones for each function.
2161
2162 7. The NON-UNIX-USE file had not been updated for the different set of source
2163 files that come with release 6. I also added a few comments about the C++
2164 wrapper.
2165
2166
2167 Version 6.0 07-Jun-05
2168 ---------------------
2169
2170 1. Some minor internal re-organization to help with my DFA experiments.
2171
2172 2. Some missing #ifdef SUPPORT_UCP conditionals in pcretest and printint that
2173 didn't matter for the library itself when fully configured, but did matter
2174 when compiling without UCP support, or within Exim, where the ucp files are
2175 not imported.
2176
2177 3. Refactoring of the library code to split up the various functions into
2178 different source modules. The addition of the new DFA matching code (see
2179 below) to a single monolithic source would have made it really too
2180 unwieldy, quite apart from causing all the code to be include in a
2181 statically linked application, when only some functions are used. This is
2182 relevant even without the DFA addition now that patterns can be compiled in
2183 one application and matched in another.
2184
2185 The downside of splitting up is that there have to be some external
2186 functions and data tables that are used internally in different modules of
2187 the library but which are not part of the API. These have all had their
2188 names changed to start with "_pcre_" so that they are unlikely to clash
2189 with other external names.
2190
2191 4. Added an alternate matching function, pcre_dfa_exec(), which matches using
2192 a different (DFA) algorithm. Although it is slower than the original
2193 function, it does have some advantages for certain types of matching
2194 problem.
2195
2196 5. Upgrades to pcretest in order to test the features of pcre_dfa_exec(),
2197 including restarting after a partial match.
2198
2199 6. A patch for pcregrep that defines INVALID_FILE_ATTRIBUTES if it is not
2200 defined when compiling for Windows was sent to me. I have put it into the
2201 code, though I have no means of testing or verifying it.
2202
2203 7. Added the pcre_refcount() auxiliary function.
2204
2205 8. Added the PCRE_FIRSTLINE option. This constrains an unanchored pattern to
2206 match before or at the first newline in the subject string. In pcretest,
2207 the /f option on a pattern can be used to set this.
2208
2209 9. A repeated \w when used in UTF-8 mode with characters greater than 256
2210 would behave wrongly. This has been present in PCRE since release 4.0.
2211
2212 10. A number of changes to the pcregrep command:
2213
2214 (a) Refactored how -x works; insert ^(...)$ instead of setting
2215 PCRE_ANCHORED and checking the length, in preparation for adding
2216 something similar for -w.
2217
2218 (b) Added the -w (match as a word) option.
2219
2220 (c) Refactored the way lines are read and buffered so as to have more
2221 than one at a time available.
2222
2223 (d) Implemented a pcregrep test script.
2224
2225 (e) Added the -M (multiline match) option. This allows patterns to match
2226 over several lines of the subject. The buffering ensures that at least
2227 8K, or the rest of the document (whichever is the shorter) is available
2228 for matching (and similarly the previous 8K for lookbehind assertions).
2229
2230 (f) Changed the --help output so that it now says
2231
2232 -w, --word-regex(p)
2233
2234 instead of two lines, one with "regex" and the other with "regexp"
2235 because that confused at least one person since the short forms are the
2236 same. (This required a bit of code, as the output is generated
2237 automatically from a table. It wasn't just a text change.)
2238
2239 (g) -- can be used to terminate pcregrep options if the next thing isn't an
2240 option but starts with a hyphen. Could be a pattern or a path name
2241 starting with a hyphen, for instance.
2242
2243 (h) "-" can be given as a file name to represent stdin.
2244
2245 (i) When file names are being printed, "(standard input)" is used for
2246 the standard input, for compatibility with GNU grep. Previously
2247 "<stdin>" was used.
2248
2249 (j) The option --label=xxx can be used to supply a name to be used for
2250 stdin when file names are being printed. There is no short form.
2251
2252 (k) Re-factored the options decoding logic because we are going to add
2253 two more options that take data. Such options can now be given in four
2254 different ways, e.g. "-fname", "-f name", "--file=name", "--file name".
2255
2256 (l) Added the -A, -B, and -C options for requesting that lines of context
2257 around matches be printed.
2258
2259 (m) Added the -L option to print the names of files that do not contain
2260 any matching lines, that is, the complement of -l.
2261
2262 (n) The return code is 2 if any file cannot be opened, but pcregrep does
2263 continue to scan other files.
2264
2265 (o) The -s option was incorrectly implemented. For compatibility with other
2266 greps, it now suppresses the error message for a non-existent or non-
2267 accessible file (but not the return code). There is a new option called
2268 -q that suppresses the output of matching lines, which was what -s was
2269 previously doing.
2270
2271 (p) Added --include and --exclude options to specify files for inclusion
2272 and exclusion when recursing.
2273
2274 11. The Makefile was not using the Autoconf-supported LDFLAGS macro properly.
2275 Hopefully, it now does.
2276
2277 12. Missing cast in pcre_study().
2278
2279 13. Added an "uninstall" target to the makefile.
2280
2281 14. Replaced "extern" in the function prototypes in Makefile.in with
2282 "PCRE_DATA_SCOPE", which defaults to 'extern' or 'extern "C"' in the Unix
2283 world, but is set differently for Windows.
2284
2285 15. Added a second compiling function called pcre_compile2(). The only
2286 difference is that it has an extra argument, which is a pointer to an
2287 integer error code. When there is a compile-time failure, this is set
2288 non-zero, in addition to the error test pointer being set to point to an
2289 error message. The new argument may be NULL if no error number is required
2290 (but then you may as well call pcre_compile(), which is now just a
2291 wrapper). This facility is provided because some applications need a
2292 numeric error indication, but it has also enabled me to tidy up the way
2293 compile-time errors are handled in the POSIX wrapper.
2294
2295 16. Added VPATH=.libs to the makefile; this should help when building with one
2296 prefix path and installing with another. (Or so I'm told by someone who
2297 knows more about this stuff than I do.)
2298
2299 17. Added a new option, REG_DOTALL, to the POSIX function regcomp(). This
2300 passes PCRE_DOTALL to the pcre_compile() function, making the "." character
2301 match everything, including newlines. This is not POSIX-compatible, but
2302 somebody wanted the feature. From pcretest it can be activated by using
2303 both the P and the s flags.
2304
2305 18. AC_PROG_LIBTOOL appeared twice in Makefile.in. Removed one.
2306
2307 19. libpcre.pc was being incorrectly installed as executable.
2308
2309 20. A couple of places in pcretest check for end-of-line by looking for '\n';
2310 it now also looks for '\r' so that it will work unmodified on Windows.
2311
2312 21. Added Google's contributed C++ wrapper to the distribution.
2313
2314 22. Added some untidy missing memory free() calls in pcretest, to keep
2315 Electric Fence happy when testing.
2316
2317
2318
2319 Version 5.0 13-Sep-04
2320 ---------------------
2321
2322 1. Internal change: literal characters are no longer packed up into items
2323 containing multiple characters in a single byte-string. Each character
2324 is now matched using a separate opcode. However, there may be more than one
2325 byte in the character in UTF-8 mode.
2326
2327 2. The pcre_callout_block structure has two new fields: pattern_position and
2328 next_item_length. These contain the offset in the pattern to the next match
2329 item, and its length, respectively.
2330
2331 3. The PCRE_AUTO_CALLOUT option for pcre_compile() requests the automatic
2332 insertion of callouts before each pattern item. Added the /C option to
2333 pcretest to make use of this.
2334
2335 4. On the advice of a Windows user, the lines
2336
2337 #if defined(_WIN32) || defined(WIN32)
2338 _setmode( _fileno( stdout ), 0x8000 );
2339 #endif /* defined(_WIN32) || defined(WIN32) */
2340
2341 have been added to the source of pcretest. This apparently does useful
2342 magic in relation to line terminators.
2343
2344 5. Changed "r" and "w" in the calls to fopen() in pcretest to "rb" and "wb"
2345 for the benefit of those environments where the "b" makes a difference.
2346
2347 6. The icc compiler has the same options as gcc, but "configure" doesn't seem
2348 to know about it. I have put a hack into configure.in that adds in code
2349 to set GCC=yes if CC=icc. This seems to end up at a point in the
2350 generated configure script that is early enough to affect the setting of
2351 compiler options, which is what is needed, but I have no means of testing
2352 whether it really works. (The user who reported this had patched the
2353 generated configure script, which of course I cannot do.)
2354
2355 LATER: After change 22 below (new libtool files), the configure script
2356 seems to know about icc (and also ecc). Therefore, I have commented out
2357 this hack in configure.in.
2358
2359 7. Added support for pkg-config (2 patches were sent in).
2360
2361 8. Negated POSIX character classes that used a combination of internal tables
2362 were completely broken. These were [[:^alpha:]], [[:^alnum:]], and
2363 [[:^ascii]]. Typically, they would match almost any characters. The other
2364 POSIX classes were not broken in this way.
2365
2366 9. Matching the pattern "\b.*?" against "ab cd", starting at offset 1, failed
2367 to find the match, as PCRE was deluded into thinking that the match had to
2368 start at the start point or following a newline. The same bug applied to
2369 patterns with negative forward assertions or any backward assertions
2370 preceding ".*" at the start, unless the pattern required a fixed first
2371 character. This was a failing pattern: "(?!.bcd).*". The bug is now fixed.
2372
2373 10. In UTF-8 mode, when moving forwards in the subject after a failed match
2374 starting at the last subject character, bytes beyond the end of the subject
2375 string were read.
2376
2377 11. Renamed the variable "class" as "classbits" to make life easier for C++
2378 users. (Previously there was a macro definition, but it apparently wasn't
2379 enough.)
2380
2381 12. Added the new field "tables" to the extra data so that tables can be passed
2382 in at exec time, or the internal tables can be re-selected. This allows
2383 a compiled regex to be saved and re-used at a later time by a different
2384 program that might have everything at different addresses.
2385
2386 13. Modified the pcre-config script so that, when run on Solaris, it shows a
2387 -R library as well as a -L library.
2388
2389 14. The debugging options of pcretest (-d on the command line or D on a
2390 pattern) showed incorrect output for anything following an extended class
2391 that contained multibyte characters and which was followed by a quantifier.
2392
2393 15. Added optional support for general category Unicode character properties
2394 via the \p, \P, and \X escapes. Unicode property support implies UTF-8
2395 support. It adds about 90K to the size of the library. The meanings of the
2396 inbuilt class escapes such as \d and \s have NOT been changed.
2397
2398 16. Updated pcredemo.c to include calls to free() to release the memory for the
2399 compiled pattern.
2400
2401 17. The generated file chartables.c was being created in the source directory
2402 instead of in the building directory. This caused the build to fail if the
2403 source directory was different from the building directory, and was
2404 read-only.
2405
2406 18. Added some sample Win commands from Mark Tetrode into the NON-UNIX-USE
2407 file. No doubt somebody will tell me if they don't make sense... Also added
2408 Dan Mooney's comments about building on OpenVMS.
2409
2410 19. Added support for partial matching via the PCRE_PARTIAL option for
2411 pcre_exec() and the \P data escape in pcretest.
2412
2413 20. Extended pcretest with 3 new pattern features:
2414
2415 (i) A pattern option of the form ">rest-of-line" causes pcretest to
2416 write the compiled pattern to the file whose name is "rest-of-line".
2417 This is a straight binary dump of the data, with the saved pointer to
2418 the character tables forced to be NULL. The study data, if any, is
2419 written too. After writing, pcretest reads a new pattern.
2420
2421 (ii) If, instead of a pattern, "<rest-of-line" is given, pcretest reads a
2422 compiled pattern from the given file. There must not be any
2423 occurrences of "<" in the file name (pretty unlikely); if there are,
2424 pcretest will instead treat the initial "<" as a pattern delimiter.
2425 After reading in the pattern, pcretest goes on to read data lines as
2426 usual.
2427
2428 (iii) The F pattern option causes pcretest to flip the bytes in the 32-bit
2429 and 16-bit fields in a compiled pattern, to simulate a pattern that
2430 was compiled on a host of opposite endianness.
2431
2432 21. The pcre-exec() function can now cope with patterns that were compiled on
2433 hosts of opposite endianness, with this restriction:
2434
2435 As for any compiled expression that is saved and used later, the tables
2436 pointer field cannot be preserved; the extra_data field in the arguments
2437 to pcre_exec() should be used to pass in a tables address if a value
2438 other than the default internal tables were used at compile time.
2439
2440 22. Calling pcre_exec() with a negative value of the "ovecsize" parameter is
2441 now diagnosed as an error. Previously, most of the time, a negative number
2442 would have been treated as zero, but if in addition "ovector" was passed as
2443 NULL, a crash could occur.
2444
2445 23. Updated the files ltmain.sh, config.sub, config.guess, and aclocal.m4 with
2446 new versions from the libtool 1.5 distribution (the last one is a copy of
2447 a file called libtool.m4). This seems to have fixed the need to patch
2448 "configure" to support Darwin 1.3 (which I used to do). However, I still
2449 had to patch ltmain.sh to ensure that ${SED} is set (it isn't on my
2450 workstation).
2451
2452 24. Changed the PCRE licence to be the more standard "BSD" licence.
2453
2454
2455 Version 4.5 01-Dec-03
2456 ---------------------
2457
2458 1. There has been some re-arrangement of the code for the match() function so
2459 that it can be compiled in a version that does not call itself recursively.
2460 Instead, it keeps those local variables that need separate instances for
2461 each "recursion" in a frame on the heap, and gets/frees frames whenever it
2462 needs to "recurse". Keeping track of where control must go is done by means
2463 of setjmp/longjmp. The whole thing is implemented by a set of macros that
2464 hide most of the details from the main code, and operates only if
2465 NO_RECURSE is defined while compiling pcre.c. If PCRE is built using the
2466 "configure" mechanism, "--disable-stack-for-recursion" turns on this way of
2467 operating.
2468
2469 To make it easier for callers to provide specially tailored get/free
2470 functions for this usage, two new functions, pcre_stack_malloc, and
2471 pcre_stack_free, are used. They are always called in strict stacking order,
2472 and the size of block requested is always the same.
2473
2474 The PCRE_CONFIG_STACKRECURSE info parameter can be used to find out whether
2475 PCRE has been compiled to use the stack or the heap for recursion. The
2476 -C option of pcretest uses this to show which version is compiled.
2477
2478 A new data escape \S, is added to pcretest; it causes the amounts of store
2479 obtained and freed by both kinds of malloc/free at match time to be added
2480 to the output.
2481
2482 2. Changed the locale test to use "fr_FR" instead of "fr" because that's
2483 what's available on my current Linux desktop machine.
2484
2485 3. When matching a UTF-8 string, the test for a valid string at the start has
2486 been extended. If start_offset is not zero, PCRE now checks that it points
2487 to a byte that is the start of a UTF-8 character. If not, it returns
2488 PCRE_ERROR_BADUTF8_OFFSET (-11). Note: the whole string is still checked;
2489 this is necessary because there may be backward assertions in the pattern.
2490 When matching the same subject several times, it may save resources to use
2491 PCRE_NO_UTF8_CHECK on all but the first call if the string is long.
2492
2493 4. The code for checking the validity of UTF-8 strings has been tightened so
2494 that it rejects (a) strings containing 0xfe or 0xff bytes and (b) strings
2495 containing "overlong sequences".
2496
2497 5. Fixed a bug (appearing twice) that I could not find any way of exploiting!
2498 I had written "if ((digitab[*p++] && chtab_digit) == 0)" where the "&&"
2499 should have been "&", but it just so happened that all the cases this let
2500 through by mistake were picked up later in the function.
2501
2502 6. I had used a variable called "isblank" - this is a C99 function, causing
2503 some compilers to warn. To avoid this, I renamed it (as "blankclass").
2504
2505 7. Cosmetic: (a) only output another newline at the end of pcretest if it is
2506 prompting; (b) run "./pcretest /dev/null" at the start of the test script
2507 so the version is shown; (c) stop "make test" echoing "./RunTest".
2508
2509 8. Added patches from David Burgess to enable PCRE to run on EBCDIC systems.
2510
2511 9. The prototype for memmove() for systems that don't have it was using
2512 size_t, but the inclusion of the header that defines size_t was later. I've
2513 moved the #includes for the C headers earlier to avoid this.
2514
2515 10. Added some adjustments to the code to make it easier to compiler on certain
2516 special systems:
2517
2518 (a) Some "const" qualifiers were missing.
2519 (b) Added the macro EXPORT before all exported functions; by default this
2520 is defined to be empty.
2521 (c) Changed the dftables auxiliary program (that builds chartables.c) so
2522 that it reads its output file name as an argument instead of writing
2523 to the standard output and assuming this can be redirected.
2524
2525 11. In UTF-8 mode, if a recursive reference (e.g. (?1)) followed a character
2526 class containing characters with values greater than 255, PCRE compilation
2527 went into a loop.
2528
2529 12. A recursive reference to a subpattern that was within another subpattern
2530 that had a minimum quantifier of zero caused PCRE to crash. For example,
2531 (x(y(?2))z)? provoked this bug with a subject that got as far as the
2532 recursion. If the recursively-called subpattern itself had a zero repeat,
2533 that was OK.
2534
2535 13. In pcretest, the buffer for reading a data line was set at 30K, but the
2536 buffer into which it was copied (for escape processing) was still set at
2537 1024, so long lines caused crashes.
2538
2539 14. A pattern such as /[ab]{1,3}+/ failed to compile, giving the error
2540 "internal error: code overflow...". This applied to any character class
2541 that was followed by a possessive quantifier.
2542
2543 15. Modified the Makefile to add libpcre.la as a prerequisite for
2544 libpcreposix.la because I was told this is needed for a parallel build to
2545 work.
2546
2547 16. If a pattern that contained .* following optional items at the start was
2548 studied, the wrong optimizing data was generated, leading to matching
2549 errors. For example, studying /[ab]*.*c/ concluded, erroneously, that any
2550 matching string must start with a or b or c. The correct conclusion for
2551 this pattern is that a match can start with any character.
2552
2553
2554 Version 4.4 13-Aug-03
2555 ---------------------
2556
2557 1. In UTF-8 mode, a character class containing characters with values between
2558 127 and 255 was not handled correctly if the compiled pattern was studied.
2559 In fixing this, I have also improved the studying algorithm for such
2560 classes (slightly).
2561
2562 2. Three internal functions had redundant arguments passed to them. Removal
2563 might give a very teeny performance improvement.
2564
2565 3. Documentation bug: the value of the capture_top field in a callout is *one
2566 more than* the number of the hightest numbered captured substring.
2567
2568 4. The Makefile linked pcretest and pcregrep with -lpcre, which could result
2569 in incorrectly linking with a previously installed version. They now link
2570 explicitly with libpcre.la.
2571
2572 5. configure.in no longer needs to recognize Cygwin specially.
2573
2574 6. A problem in pcre.in for Windows platforms is fixed.
2575
2576 7. If a pattern was successfully studied, and the -d (or /D) flag was given to
2577 pcretest, it used to include the size of the study block as part of its
2578 output. Unfortunately, the structure contains a field that has a different
2579 size on different hardware architectures. This meant that the tests that
2580 showed this size failed. As the block is currently always of a fixed size,
2581 this information isn't actually particularly useful in pcretest output, so
2582 I have just removed it.
2583
2584 8. Three pre-processor statements accidentally did not start in column 1.
2585 Sadly, there are *still* compilers around that complain, even though
2586 standard C has not required this for well over a decade. Sigh.
2587
2588 9. In pcretest, the code for checking callouts passed small integers in the
2589 callout_data field, which is a void * field. However, some picky compilers
2590 complained about the casts involved for this on 64-bit systems. Now
2591 pcretest passes the address of the small integer instead, which should get
2592 rid of the warnings.
2593
2594 10. By default, when in UTF-8 mode, PCRE now checks for valid UTF-8 strings at
2595 both compile and run time, and gives an error if an invalid UTF-8 sequence
2596 is found. There is a option for disabling this check in cases where the
2597 string is known to be correct and/or the maximum performance is wanted.
2598
2599 11. In response to a bug report, I changed one line in Makefile.in from
2600
2601 -Wl,--out-implib,.libs/lib@WIN_PREFIX@pcreposix.dll.a \
2602 to
2603 -Wl,--out-implib,.libs/@WIN_PREFIX@libpcreposix.dll.a \
2604
2605 to look similar to other lines, but I have no way of telling whether this
2606 is the right thing to do, as I do not use Windows. No doubt I'll get told
2607 if it's wrong...
2608
2609
2610 Version 4.3 21-May-03
2611 ---------------------
2612
2613 1. Two instances of @WIN_PREFIX@ omitted from the Windows targets in the
2614 Makefile.
2615
2616 2. Some refactoring to improve the quality of the code:
2617
2618 (i) The utf8_table... variables are now declared "const".
2619
2620 (ii) The code for \cx, which used the "case flipping" table to upper case
2621 lower case letters, now just substracts 32. This is ASCII-specific,
2622 but the whole concept of \cx is ASCII-specific, so it seems
2623 reasonable.
2624
2625 (iii) PCRE was using its character types table to recognize decimal and
2626 hexadecimal digits in the pattern. This is silly, because it handles
2627 only 0-9, a-f, and A-F, but the character types table is locale-
2628 specific, which means strange things might happen. A private
2629 table is now used for this - though it costs 256 bytes, a table is
2630 much faster than multiple explicit tests. Of course, the standard
2631 character types table is still used for matching digits in subject
2632 strings against \d.
2633
2634 (iv) Strictly, the identifier ESC_t is reserved by POSIX (all identifiers
2635 ending in _t are). So I've renamed it as ESC_tee.
2636
2637 3. The first argument for regexec() in the POSIX wrapper should have been
2638 defined as "const".
2639
2640 4. Changed pcretest to use malloc() for its buffers so that they can be
2641 Electric Fenced for debugging.
2642
2643 5. There were several places in the code where, in UTF-8 mode, PCRE would try
2644 to read one or more bytes before the start of the subject string. Often this
2645 had no effect on PCRE's behaviour, but in some circumstances it could
2646 provoke a segmentation fault.
2647
2648 6. A lookbehind at the start of a pattern in UTF-8 mode could also cause PCRE
2649 to try to read one or more bytes before the start of the subject string.
2650
2651 7. A lookbehind in a pattern matched in non-UTF-8 mode on a PCRE compiled with
2652 UTF-8 support could misbehave in various ways if the subject string
2653 contained bytes with the 0x80 bit set and the 0x40 bit unset in a lookbehind
2654 area. (PCRE was not checking for the UTF-8 mode flag, and trying to move
2655 back over UTF-8 characters.)
2656
2657
2658 Version 4.2 14-Apr-03
2659 ---------------------
2660
2661 1. Typo "#if SUPPORT_UTF8" instead of "#ifdef SUPPORT_UTF8" fixed.
2662
2663 2. Changes to the building process, supplied by Ronald Landheer-Cieslak
2664 [ON_WINDOWS]: new variable, "#" on non-Windows platforms
2665 [NOT_ON_WINDOWS]: new variable, "#" on Windows platforms
2666 [WIN_PREFIX]: new variable, "cyg" for Cygwin
2667 * Makefile.in: use autoconf substitution for OBJEXT, EXEEXT, BUILD_OBJEXT
2668 and BUILD_EXEEXT
2669 Note: automatic setting of the BUILD variables is not yet working
2670 set CPPFLAGS and BUILD_CPPFLAGS (but don't use yet) - should be used at
2671 compile-time but not at link-time
2672 [LINK]: use for linking executables only
2673 make different versions for Windows and non-Windows
2674 [LINKLIB]: new variable, copy of UNIX-style LINK, used for linking
2675 libraries
2676 [LINK_FOR_BUILD]: new variable
2677 [OBJEXT]: use throughout
2678 [EXEEXT]: use throughout
2679 <winshared>: new target
2680 <wininstall>: new target
2681 <dftables.o>: use native compiler
2682 <dftables>: use native linker
2683 <install>: handle Windows platform correctly
2684 <clean>: ditto
2685 <check>: ditto
2686 copy DLL to top builddir before testing
2687
2688 As part of these changes, -no-undefined was removed again. This was reported
2689 to give trouble on HP-UX 11.0, so getting rid of it seems like a good idea
2690 in any case.
2691
2692 3. Some tidies to get rid of compiler warnings:
2693
2694 . In the match_data structure, match_limit was an unsigned long int, whereas
2695 match_call_count was an int. I've made them both unsigned long ints.
2696
2697 . In pcretest the fact that a const uschar * doesn't automatically cast to
2698 a void * provoked a warning.
2699
2700 . Turning on some more compiler warnings threw up some "shadow" variables
2701 and a few more missing casts.
2702
2703 4. If PCRE was complied with UTF-8 support, but called without the PCRE_UTF8
2704 option, a class that contained a single character with a value between 128
2705 and 255 (e.g. /[\xFF]/) caused PCRE to crash.
2706
2707 5. If PCRE was compiled with UTF-8 support, but called without the PCRE_UTF8
2708 option, a class that contained several characters, but with at least one
2709 whose value was between 128 and 255 caused PCRE to crash.
2710
2711
2712 Version 4.1 12-Mar-03
2713 ---------------------
2714
2715 1. Compiling with gcc -pedantic found a couple of places where casts were
2716 needed, and a string in dftables.c that was longer than standard compilers are
2717 required to support.
2718
2719 2. Compiling with Sun's compiler found a few more places where the code could
2720 be tidied up in order to avoid warnings.
2721
2722 3. The variables for cross-compiling were called HOST_CC and HOST_CFLAGS; the
2723 first of these names is deprecated in the latest Autoconf in favour of the name
2724 CC_FOR_BUILD, because "host" is typically used to mean the system on which the
2725 compiled code will be run. I can't find a reference for HOST_CFLAGS, but by
2726 analogy I have changed it to CFLAGS_FOR_BUILD.
2727
2728 4. Added -no-undefined to the linking command in the Makefile, because this is
2729 apparently helpful for Windows. To make it work, also added "-L. -lpcre" to the
2730 linking step for the pcreposix library.
2731
2732 5. PCRE was failing to diagnose the case of two named groups with the same
2733 name.
2734
2735 6. A problem with one of PCRE's optimizations was discovered. PCRE remembers a
2736 literal character that is needed in the subject for a match, and scans along to
2737 ensure that it is present before embarking on the full matching process. This
2738 saves time in cases of nested unlimited repeats that are never going to match.
2739 Problem: the scan can take a lot of time if the subject is very long (e.g.
2740 megabytes), thus penalizing straightforward matches. It is now done only if the
2741 amount of subject to be scanned is less than 1000 bytes.
2742
2743 7. A lesser problem with the same optimization is that it was recording the
2744 first character of an anchored pattern as "needed", thus provoking a search
2745 right along the subject, even when the first match of the pattern was going to
2746 fail. The "needed" character is now not set for anchored patterns, unless it
2747 follows something in the pattern that is of non-fixed length. Thus, it still
2748 fulfils its original purpose of finding quick non-matches in cases of nested
2749 unlimited repeats, but isn't used for simple anchored patterns such as /^abc/.
2750
2751
2752 Version 4.0 17-Feb-03
2753 ---------------------
2754
2755 1. If a comment in an extended regex that started immediately after a meta-item
2756 extended to the end of string, PCRE compiled incorrect data. This could lead to
2757 all kinds of weird effects. Example: /#/ was bad; /()#/ was bad; /a#/ was not.
2758
2759 2. Moved to autoconf 2.53 and libtool 1.4.2.
2760
2761 3. Perl 5.8 no longer needs "use utf8" for doing UTF-8 things. Consequently,
2762 the special perltest8 script is no longer needed - all the tests can be run
2763 from a single perltest script.
2764
2765 4. From 5.004, Perl has not included the VT character (0x0b) in the set defined
2766 by \s. It has now been removed in PCRE. This means it isn't recognized as
2767 whitespace in /x regexes too, which is the same as Perl. Note that the POSIX
2768 class [:space:] *does* include VT, thereby creating a mess.
2769
2770 5. Added the class [:blank:] (a GNU extension from Perl 5.8) to match only
2771 space and tab.
2772
2773 6. Perl 5.005 was a long time ago. It's time to amalgamate the tests that use
2774 its new features into the main test script, reducing the number of scripts.
2775
2776 7. Perl 5.8 has changed the meaning of patterns like /a(?i)b/. Earlier versions
2777 were backward compatible, and made the (?i) apply to the whole pattern, as if
2778 /i were given. Now it behaves more logically, and applies the option setting
2779 only to what follows. PCRE has been changed to follow suit. However, if it
2780 finds options settings right at the start of the pattern, it extracts them into
2781 the global options, as before. Thus, they show up in the info data.
2782
2783 8. Added support for the \Q...\E escape sequence. Characters in between are
2784 treated as literals. This is slightly different from Perl in that $ and @ are
2785 also handled as literals inside the quotes. In Perl, they will cause variable
2786 interpolation. Note the following examples:
2787
2788 Pattern PCRE matches Perl matches
2789
2790 \Qabc$xyz\E abc$xyz abc followed by the contents of $xyz
2791 \Qabc\$xyz\E abc\$xyz abc\$xyz
2792 \Qabc\E\$\Qxyz\E abc$xyz abc$xyz
2793
2794 For compatibility with Perl, \Q...\E sequences are recognized inside character
2795 classes as well as outside them.
2796
2797 9. Re-organized 3 code statements in pcretest to avoid "overflow in
2798 floating-point constant arithmetic" warnings from a Microsoft compiler. Added a
2799 (size_t) cast to one statement in pcretest and one in pcreposix to avoid
2800 signed/unsigned warnings.
2801
2802 10. SunOS4 doesn't have strtoul(). This was used only for unpicking the -o
2803 option for pcretest, so I've replaced it by a simple function that does just
2804 that job.
2805
2806 11. pcregrep was ending with code 0 instead of 2 for the commands "pcregrep" or
2807 "pcregrep -".
2808
2809 12. Added "possessive quantifiers" ?+, *+, ++, and {,}+ which come from Sun's
2810 Java package. This provides some syntactic sugar for simple cases of what my
2811 documentation calls "once-only subpatterns". A pattern such as x*+ is the same
2812 as (?>x*). In other words, if what is inside (?>...) is just a single repeated
2813 item, you can use this simplified notation. Note that only makes sense with
2814 greedy quantifiers. Consequently, the use of the possessive quantifier forces
2815 greediness, whatever the setting of the PCRE_UNGREEDY option.
2816
2817 13. A change of greediness default within a pattern was not taking effect at
2818 the current level for patterns like /(b+(?U)a+)/. It did apply to parenthesized
2819 subpatterns that followed. Patterns like /b+(?U)a+/ worked because the option
2820 was abstracted outside.
2821
2822 14. PCRE now supports the \G assertion. It is true when the current matching
2823 position is at the start point of the match. This differs from \A when the
2824 starting offset is non-zero. Used with the /g option of pcretest (or similar
2825 code), it works in the same way as it does for Perl's /g option. If all
2826 alternatives of a regex begin with \G, the expression is anchored to the start
2827 match position, and the "anchored" flag is set in the compiled expression.
2828
2829 15. Some bugs concerning the handling of certain option changes within patterns
2830 have been fixed. These applied to options other than (?ims). For example,
2831 "a(?x: b c )d" did not match "XabcdY" but did match "Xa b c dY". It should have
2832 been the other way round. Some of this was related to change 7 above.
2833
2834 16. PCRE now gives errors for /[.x.]/ and /[=x=]/ as unsupported POSIX
2835 features, as Perl does. Previously, PCRE gave the warnings only for /[[.x.]]/
2836 and /[[=x=]]/. PCRE now also gives an error for /[:name:]/ because it supports
2837 POSIX classes only within a class (e.g. /[[:alpha:]]/).
2838
2839 17. Added support for Perl's \C escape. This matches one byte, even in UTF8
2840 mode. Unlike ".", it always matches newline, whatever the setting of
2841 PCRE_DOTALL. However, PCRE does not permit \C to appear in lookbehind
2842 assertions. Perl allows it, but it doesn't (in general) work because it can't
2843 calculate the length of the lookbehind. At least, that's the case for Perl
2844 5.8.0 - I've been told they are going to document that it doesn't work in
2845 future.
2846
2847 18. Added an error diagnosis for escapes that PCRE does not support: these are
2848 \L, \l, \N, \P, \p, \U, \u, and \X.
2849
2850 19. Although correctly diagnosing a missing ']' in a character class, PCRE was
2851 reading past the end of the pattern in cases such as /[abcd/.
2852
2853 20. PCRE was getting more memory than necessary for patterns with classes that
2854 contained both POSIX named classes and other characters, e.g. /[[:space:]abc/.
2855
2856 21. Added some code, conditional on #ifdef VPCOMPAT, to make life easier for
2857 compiling PCRE for use with Virtual Pascal.
2858
2859 22. Small fix to the Makefile to make it work properly if the build is done
2860 outside the source tree.
2861
2862 23. Added a new extension: a condition to go with recursion. If a conditional
2863 subpattern starts with (?(R) the "true" branch is used if recursion has
2864 happened, whereas the "false" branch is used only at the top level.
2865
2866 24. When there was a very long string of literal characters (over 255 bytes
2867 without UTF support, over 250 bytes with UTF support), the computation of how
2868 much memory was required could be incorrect, leading to segfaults or other
2869 strange effects.
2870
2871 25. PCRE was incorrectly assuming anchoring (either to start of subject or to
2872 start of line for a non-DOTALL pattern) when a pattern started with (.*) and
2873 there was a subsequent back reference to those brackets. This meant that, for
2874 example, /(.*)\d+\1/ failed to match "abc123bc". Unfortunately, it isn't
2875 possible to check for precisely this case. All we can do is abandon the
2876 optimization if .* occurs inside capturing brackets when there are any back
2877 references whatsoever. (See below for a better fix that came later.)
2878
2879 26. The handling of the optimization for finding the first character of a
2880 non-anchored pattern, and for finding a character that is required later in the
2881 match were failing in some cases. This didn't break the matching; it just
2882 failed to optimize when it could. The way this is done has been re-implemented.
2883
2884 27. Fixed typo in error message for invalid (?R item (it said "(?p").
2885
2886 28. Added a new feature that provides some of the functionality that Perl
2887 provides with (?{...}). The facility is termed a "callout". The way it is done
2888 in PCRE is for the caller to provide an optional function, by setting
2889 pcre_callout to its entry point. Like pcre_malloc and pcre_free, this is a
2890 global variable. By default it is unset, which disables all calling out. To get
2891 the function called, the regex must include (?C) at appropriate points. This
2892 is, in fact, equivalent to (?C0), and any number <= 255 may be given with (?C).
2893 This provides a means of identifying different callout points. When PCRE
2894 reaches such a point in the regex, if pcre_callout has been set, the external
2895 function is called. It is provided with data in a structure called
2896 pcre_callout_block, which is defined in pcre.h. If the function returns 0,
2897 matching continues; if it returns a non-zero value, the match at the current
2898 point fails. However, backtracking will occur if possible. [This was changed
2899 later and other features added - see item 49 below.]
2900
2901 29. pcretest is upgraded to test the callout functionality. It provides a
2902 callout function that displays information. By default, it shows the start of
2903 the match and the current position in the text. There are some new data escapes
2904 to vary what happens:
2905
2906 \C+ in addition, show current contents of captured substrings
2907 \C- do not supply a callout function
2908 \C!n return 1 when callout number n is reached
2909 \C!n!m return 1 when callout number n is reached for the mth time
2910
2911 30. If pcregrep was called with the -l option and just a single file name, it
2912 output "<stdin>" if a match was found, instead of the file name.
2913
2914 31. Improve the efficiency of the POSIX API to PCRE. If the number of capturing
2915 slots is less than POSIX_MALLOC_THRESHOLD, use a block on the stack to pass to
2916 pcre_exec(). This saves a malloc/free per call. The default value of
2917 POSIX_MALLOC_THRESHOLD is 10; it can be changed by --with-posix-malloc-threshold
2918 when configuring.
2919
2920 32. The default maximum size of a compiled pattern is 64K. There have been a
2921 few cases of people hitting this limit. The code now uses macros to handle the
2922 storing of links as offsets within the compiled pattern. It defaults to 2-byte
2923 links, but this can be changed to 3 or 4 bytes by --with-link-size when
2924 configuring. Tests 2 and 5 work only with 2-byte links because they output
2925 debugging information about compiled patterns.
2926
2927 33. Internal code re-arrangements:
2928
2929 (a) Moved the debugging function for printing out a compiled regex into
2930 its own source file (printint.c) and used #include to pull it into
2931 pcretest.c and, when DEBUG is defined, into pcre.c, instead of having two
2932 separate copies.
2933
2934 (b) Defined the list of op-code names for debugging as a macro in
2935 internal.h so that it is next to the definition of the opcodes.
2936
2937 (c) Defined a table of op-code lengths for simpler skipping along compiled
2938 code. This is again a macro in internal.h so that it is next to the
2939 definition of the opcodes.
2940
2941 34. Added support for recursive calls to individual subpatterns, along the
2942 lines of Robin Houston's patch (but implemented somewhat differently).
2943
2944 35. Further mods to the Makefile to help Win32. Also, added code to pcregrep to
2945 allow it to read and process whole directories in Win32. This code was
2946 contributed by Lionel Fourquaux; it has not been tested by me.
2947
2948 36. Added support for named subpatterns. The Python syntax (?P<name>...) is
2949 used to name a group. Names consist of alphanumerics and underscores, and must
2950 be unique. Back references use the syntax (?P=name) and recursive calls use
2951 (?P>name) which is a PCRE extension to the Python extension. Groups still have
2952 numbers. The function pcre_fullinfo() can be used after compilation to extract
2953 a name/number map. There are three relevant calls:
2954
2955 PCRE_INFO_NAMEENTRYSIZE yields the size of each entry in the map
2956 PCRE_INFO_NAMECOUNT yields the number of entries
2957 PCRE_INFO_NAMETABLE yields a pointer to the map.
2958
2959 The map is a vector of fixed-size entries. The size of each entry depends on
2960 the length of the longest name used. The first two bytes of each entry are the
2961 group number, most significant byte first. There follows the corresponding
2962 name, zero terminated. The names are in alphabetical order.
2963
2964 37. Make the maximum literal string in the compiled code 250 for the non-UTF-8
2965 case instead of 255. Making it the same both with and without UTF-8 support
2966 means that the same test output works with both.
2967
2968 38. There was a case of malloc(0) in the POSIX testing code in pcretest. Avoid
2969 calling malloc() with a zero argument.
2970
2971 39. Change 25 above had to resort to a heavy-handed test for the .* anchoring
2972 optimization. I've improved things by keeping a bitmap of backreferences with
2973 numbers 1-31 so that if .* occurs inside capturing brackets that are not in
2974 fact referenced, the optimization can be applied. It is unlikely that a
2975 relevant occurrence of .* (i.e. one which might indicate anchoring or forcing
2976 the match to follow \n) will appear inside brackets with a number greater than
2977 31, but if it does, any back reference > 31 suppresses the optimization.
2978
2979 40. Added a new compile-time option PCRE_NO_AUTO_CAPTURE. This has the effect
2980 of disabling numbered capturing parentheses. Any opening parenthesis that is
2981 not followed by ? behaves as if it were followed by ?: but named parentheses
2982 can still be used for capturing (and they will acquire numbers in the usual
2983 way).
2984
2985 41. Redesigned the return codes from the match() function into yes/no/error so
2986 that errors can be passed back from deep inside the nested calls. A malloc
2987 failure while inside a recursive subpattern call now causes the
2988 PCRE_ERROR_NOMEMORY return instead of quietly going wrong.
2989
2990 42. It is now possible to set a limit on the number of times the match()
2991 function is called in a call to pcre_exec(). This facility makes it possible to
2992 limit the amount of recursion and backtracking, though not in a directly
2993 obvious way, because the match() function is used in a number of different
2994 circumstances. The count starts from zero for each position in the subject
2995 string (for non-anchored patterns). The default limit is, for compatibility, a
2996 large number, namely 10 000 000. You can change this in two ways:
2997
2998 (a) When configuring PCRE before making, you can use --with-match-limit=n
2999 to set a default value for the compiled library.
3000
3001 (b) For each call to pcre_exec(), you can pass a pcre_extra block in which
3002 a different value is set. See 45 below.
3003
3004 If the limit is exceeded, pcre_exec() returns PCRE_ERROR_MATCHLIMIT.
3005
3006 43. Added a new function pcre_config(int, void *) to enable run-time extraction
3007 of things that can be changed at compile time. The first argument specifies
3008 what is wanted and the second points to where the information is to be placed.
3009 The current list of available information is:
3010
3011 PCRE_CONFIG_UTF8
3012
3013 The output is an integer that is set to one if UTF-8 support is available;
3014 otherwise it is set to zero.
3015
3016 PCRE_CONFIG_NEWLINE
3017
3018 The output is an integer that it set to the value of the code that is used for
3019 newline. It is either LF (10) or CR (13).
3020
3021 PCRE_CONFIG_LINK_SIZE
3022
3023 The output is an integer that contains the number of bytes used for internal
3024 linkage in compiled expressions. The value is 2, 3, or 4. See item 32 above.
3025
3026 PCRE_CONFIG_POSIX_MALLOC_THRESHOLD
3027
3028 The output is an integer that contains the threshold above which the POSIX
3029 interface uses malloc() for output vectors. See item 31 above.
3030
3031 PCRE_CONFIG_MATCH_LIMIT
3032
3033 The output is an unsigned integer that contains the default limit of the number
3034 of match() calls in a pcre_exec() execution. See 42 above.
3035
3036 44. pcretest has been upgraded by the addition of the -C option. This causes it
3037 to extract all the available output from the new pcre_config() function, and to
3038 output it. The program then exits immediately.
3039
3040 45. A need has arisen to pass over additional data with calls to pcre_exec() in
3041 order to support additional features. One way would have been to define
3042 pcre_exec2() (for example) with extra arguments, but this would not have been
3043 extensible, and would also have required all calls to the original function to
3044 be mapped to the new one. Instead, I have chosen to extend the mechanism that
3045 is used for passing in "extra" data from pcre_study().
3046
3047 The pcre_extra structure is now exposed and defined in pcre.h. It currently
3048 contains the following fields:
3049
3050 flags a bitmap indicating which of the following fields are set
3051 study_data opaque data from pcre_study()
3052 match_limit a way of specifying a limit on match() calls for a specific
3053 call to pcre_exec()
3054 callout_data data for callouts (see 49 below)
3055
3056 The flag bits are also defined in pcre.h, and are
3057
3058 PCRE_EXTRA_STUDY_DATA
3059 PCRE_EXTRA_MATCH_LIMIT
3060 PCRE_EXTRA_CALLOUT_DATA
3061
3062 The pcre_study() function now returns one of these new pcre_extra blocks, with
3063 the actual study data pointed to by the study_data field, and the
3064 PCRE_EXTRA_STUDY_DATA flag set. This can be passed directly to pcre_exec() as
3065 before. That is, this change is entirely upwards-compatible and requires no
3066 change to existing code.
3067
3068 If you want to pass in additional data to pcre_exec(), you can either place it
3069 in a pcre_extra block provided by pcre_study(), or create your own pcre_extra
3070 block.
3071
3072 46. pcretest has been extended to test the PCRE_EXTRA_MATCH_LIMIT feature. If a
3073 data string contains the escape sequence \M, pcretest calls pcre_exec() several
3074 times with different match limits, until it finds the minimum value needed for
3075 pcre_exec() to complete. The value is then output. This can be instructive; for
3076 most simple matches the number is quite small, but for pathological cases it
3077 gets very large very quickly.
3078
3079 47. There's a new option for pcre_fullinfo() called PCRE_INFO_STUDYSIZE. It
3080 returns the size of the data block pointed to by the study_data field in a
3081 pcre_extra block, that is, the value that was passed as the argument to
3082 pcre_malloc() when PCRE was getting memory in which to place the information
3083 created by pcre_study(). The fourth argument should point to a size_t variable.
3084 pcretest has been extended so that this information is shown after a successful
3085 pcre_study() call when information about the compiled regex is being displayed.
3086
3087 48. Cosmetic change to Makefile: there's no need to have / after $(DESTDIR)
3088 because what follows is always an absolute path. (Later: it turns out that this
3089 is more than cosmetic for MinGW, because it doesn't like empty path
3090 components.)
3091
3092 49. Some changes have been made to the callout feature (see 28 above):
3093
3094 (i) A callout function now has three choices for what it returns:
3095
3096 0 => success, carry on matching
3097 > 0 => failure at this point, but backtrack if possible
3098 < 0 => serious error, return this value from pcre_exec()
3099
3100 Negative values should normally be chosen from the set of PCRE_ERROR_xxx
3101 values. In particular, returning PCRE_ERROR_NOMATCH forces a standard
3102 "match failed" error. The error number PCRE_ERROR_CALLOUT is reserved for
3103 use by callout functions. It will never be used by PCRE itself.
3104
3105 (ii) The pcre_extra structure (see 45 above) has a void * field called
3106 callout_data, with corresponding flag bit PCRE_EXTRA_CALLOUT_DATA. The
3107 pcre_callout_block structure has a field of the same name. The contents of
3108 the field passed in the pcre_extra structure are passed to the callout
3109 function in the corresponding field in the callout block. This makes it
3110 easier to use the same callout-containing regex from multiple threads. For
3111 testing, the pcretest program has a new data escape
3112
3113 \C*n pass the number n (may be negative) as callout_data
3114
3115 If the callout function in pcretest receives a non-zero value as
3116 callout_data, it returns that value.
3117
3118 50. Makefile wasn't handling CFLAGS properly when compiling dftables. Also,
3119 there were some redundant $(CFLAGS) in commands that are now specified as
3120 $(LINK), which already includes $(CFLAGS).
3121
3122 51. Extensions to UTF-8 support are listed below. These all apply when (a) PCRE
3123 has been compiled with UTF-8 support *and* pcre_compile() has been compiled
3124 with the PCRE_UTF8 flag. Patterns that are compiled without that flag assume
3125 one-byte characters throughout. Note that case-insensitive matching applies
3126 only to characters whose values are less than 256. PCRE doesn't support the
3127 notion of cases for higher-valued characters.
3128
3129 (i) A character class whose characters are all within 0-255 is handled as
3130 a bit map, and the map is inverted for negative classes. Previously, a
3131 character > 255 always failed to match such a class; however it should
3132 match if the class was a negative one (e.g. [^ab]). This has been fixed.
3133
3134 (ii) A negated character class with a single character < 255 is coded as
3135 "not this character" (OP_NOT). This wasn't working properly when the test
3136 character was multibyte, either singly or repeated.
3137
3138 (iii) Repeats of multibyte characters are now handled correctly in UTF-8
3139 mode, for example: \x{100}{2,3}.
3140
3141 (iv) The character escapes \b, \B, \d, \D, \s, \S, \w, and \W (either
3142 singly or repeated) now correctly test multibyte characters. However,
3143 PCRE doesn't recognize any characters with values greater than 255 as
3144 digits, spaces, or word characters. Such characters always match \D, \S,
3145 and \W, and never match \d, \s, or \w.
3146
3147 (v) Classes may now contain characters and character ranges with values
3148 greater than 255. For example: [ab\x{100}-\x{400}].
3149
3150 (vi) pcregrep now has a --utf-8 option (synonym -u) which makes it call
3151 PCRE in UTF-8 mode.
3152
3153 52. The info request value PCRE_INFO_FIRSTCHAR has been renamed
3154 PCRE_INFO_FIRSTBYTE because it is a byte value. However, the old name is
3155 retained for backwards compatibility. (Note that LASTLITERAL is also a byte
3156 value.)
3157
3158 53. The single man page has become too large. I have therefore split it up into
3159 a number of separate man pages. These also give rise to individual HTML pages;
3160 these are now put in a separate directory, and there is an index.html page that
3161 lists them all. Some hyperlinking between the pages has been installed.
3162
3163 54. Added convenience functions for handling named capturing parentheses.
3164
3165 55. Unknown escapes inside character classes (e.g. [\M]) and escapes that
3166 aren't interpreted therein (e.g. [\C]) are literals in Perl. This is now also
3167 true in PCRE, except when the PCRE_EXTENDED option is set, in which case they
3168 are faulted.
3169
3170 56. Introduced HOST_CC and HOST_CFLAGS which can be set in the environment when
3171 calling configure. These values are used when compiling the dftables.c program
3172 which is run to generate the source of the default character tables. They
3173 default to the values of CC and CFLAGS. If you are cross-compiling PCRE,
3174 you will need to set these values.
3175
3176 57. Updated the building process for Windows DLL, as provided by Fred Cox.
3177
3178
3179 Version 3.9 02-Jan-02
3180 ---------------------
3181
3182 1. A bit of extraneous text had somehow crept into the pcregrep documentation.
3183
3184 2. If --disable-static was given, the building process failed when trying to
3185 build pcretest and pcregrep. (For some reason it was using libtool to compile
3186 them, which is not right, as they aren't part of the library.)
3187
3188
3189 Version 3.8 18-Dec-01
3190 ---------------------
3191
3192 1. The experimental UTF-8 code was completely screwed up. It was packing the
3193 bytes in the wrong order. How dumb can you get?
3194
3195
3196 Version 3.7 29-Oct-01
3197 ---------------------
3198
3199 1. In updating pcretest to check change 1 of version 3.6, I screwed up.
3200 This caused pcretest, when used on the test data, to segfault. Unfortunately,
3201 this didn't happen under Solaris 8, where I normally test things.
3202
3203 2. The Makefile had to be changed to make it work on BSD systems, where 'make'
3204 doesn't seem to recognize that ./xxx and xxx are the same file. (This entry
3205 isn't in ChangeLog distributed with 3.7 because I forgot when I hastily made
3206 this fix an hour or so after the initial 3.7 release.)
3207
3208
3209 Version 3.6 23-Oct-01
3210 ---------------------
3211
3212 1. Crashed with /(sens|respons)e and \1ibility/ and "sense and sensibility" if
3213 offsets passed as NULL with zero offset count.
3214
3215 2. The config.guess and config.sub files had not been updated when I moved to
3216 the latest autoconf.
3217
3218
3219 Version 3.5 15-Aug-01
3220 ---------------------
3221
3222 1. Added some missing #if !defined NOPOSIX conditionals in pcretest.c that
3223 had been forgotten.
3224
3225 2. By using declared but undefined structures, we can avoid using "void"
3226 definitions in pcre.h while keeping the internal definitions of the structures
3227 private.
3228
3229 3. The distribution is now built using autoconf 2.50 and libtool 1.4. From a
3230 user point of view, this means that both static and shared libraries are built
3231 by default, but this can be individually controlled. More of the work of
3232 handling this static/shared cases is now inside libtool instead of PCRE's make
3233 file.
3234
3235 4. The pcretest utility is now installed along with pcregrep because it is
3236 useful for users (to test regexs) and by doing this, it automatically gets
3237 relinked by libtool. The documentation has been turned into a man page, so
3238 there are now .1, .txt, and .html versions in /doc.
3239
3240 5. Upgrades to pcregrep:
3241 (i) Added long-form option names like gnu grep.
3242 (ii) Added --help to list all options with an explanatory phrase.
3243 (iii) Added -r, --recursive to recurse into sub-directories.
3244 (iv) Added -f, --file to read patterns from a file.
3245
3246 6. pcre_exec() was referring to its "code" argument before testing that
3247 argument for NULL (and giving an error if it was NULL).
3248
3249 7. Upgraded Makefile.in to allow for compiling in a different directory from
3250 the source directory.
3251
3252 8. Tiny buglet in pcretest: when pcre_fullinfo() was called to retrieve the
3253 options bits, the pointer it was passed was to an int instead of to an unsigned
3254 long int. This mattered only on 64-bit systems.
3255
3256 9. Fixed typo (3.4/1) in pcre.h again. Sigh. I had changed pcre.h (which is
3257 generated) instead of pcre.in, which it its source. Also made the same change
3258 in several of the .c files.
3259
3260 10. A new release of gcc defines printf() as a macro, which broke pcretest
3261 because it had an ifdef in the middle of a string argument for printf(). Fixed
3262 by using separate calls to printf().
3263
3264 11. Added --enable-newline-is-cr and --enable-newline-is-lf to the configure
3265 script, to force use of CR or LF instead of \n in the source. On non-Unix
3266 systems, the value can be set in config.h.
3267
3268 12. The limit of 200 on non-capturing parentheses is a _nesting_ limit, not an
3269 absolute limit. Changed the text of the error message to make this clear, and
3270 likewise updated the man page.
3271
3272 13. The limit of 99 on the number of capturing subpatterns has been removed.
3273 The new limit is 65535, which I hope will not be a "real" limit.
3274
3275
3276 Version 3.4 22-Aug-00
3277 ---------------------
3278
3279 1. Fixed typo in pcre.h: unsigned const char * changed to const unsigned char *.
3280
3281 2. Diagnose condition (?(0) as an error instead of crashing on matching.
3282
3283
3284 Version 3.3 01-Aug-00
3285 ---------------------
3286
3287 1. If an octal character was given, but the value was greater than \377, it
3288 was not getting masked to the least significant bits, as documented. This could
3289 lead to crashes in some systems.
3290
3291 2. Perl 5.6 (if not earlier versions) accepts classes like [a-\d] and treats
3292 the hyphen as a literal. PCRE used to give an error; it now behaves like Perl.
3293
3294 3. Added the functions pcre_free_substring() and pcre_free_substring_list().
3295 These just pass their arguments on to (pcre_free)(), but they are provided
3296 because some uses of PCRE bind it to non-C systems that can call its functions,
3297 but cannot call free() or pcre_free() directly.
3298
3299 4. Add "make test" as a synonym for "make check". Corrected some comments in
3300 the Makefile.
3301
3302 5. Add $(DESTDIR)/ in front of all the paths in the "install" target in the
3303 Makefile.
3304
3305 6. Changed the name of pgrep to pcregrep, because Solaris has introduced a
3306 command called pgrep for grepping around the active processes.
3307
3308 7. Added the beginnings of support for UTF-8 character strings.
3309
3310 8. Arranged for the Makefile to pass over the settings of CC, CFLAGS, and
3311 RANLIB to ./ltconfig so that they are used by libtool. I think these are all
3312 the relevant ones. (AR is not passed because ./ltconfig does its own figuring
3313 out for the ar command.)
3314
3315
3316 Version 3.2 12-May-00
3317 ---------------------
3318
3319 This is purely a bug fixing release.
3320
3321 1. If the pattern /((Z)+|A)*/ was matched agained ZABCDEFG it matched Z instead
3322 of ZA. This was just one example of several cases that could provoke this bug,
3323 which was introduced by change 9 of version 2.00. The code for breaking
3324 infinite loops after an iteration that matches an empty string was't working
3325 correctly.
3326
3327 2. The pcretest program was not imitating Perl correctly for the pattern /a*/g
3328 when matched against abbab (for example). After matching an empty string, it
3329 wasn't forcing anchoring when setting PCRE_NOTEMPTY for the next attempt; this
3330 caused it to match further down the string than it should.
3331
3332 3. The code contained an inclusion of sys/types.h. It isn't clear why this
3333 was there because it doesn't seem to be needed, and it causes trouble on some
3334 systems, as it is not a Standard C header. It has been removed.
3335
3336 4. Made 4 silly changes to the source to avoid stupid compiler warnings that
3337 were reported on the Macintosh. The changes were from
3338
3339 while ((c = *(++ptr)) != 0 && c != '\n');
3340 to
3341 while ((c = *(++ptr)) != 0 && c != '\n') ;
3342
3343 Totally extraordinary, but if that's what it takes...
3344
3345 5. PCRE is being used in one environment where neither memmove() nor bcopy() is
3346 available. Added HAVE_BCOPY and an autoconf test for it; if neither
3347 HAVE_MEMMOVE nor HAVE_BCOPY is set, use a built-in emulation function which
3348 assumes the way PCRE uses memmove() (always moving upwards).
3349
3350 6. PCRE is being used in one environment where strchr() is not available. There
3351 was only one use in pcre.c, and writing it out to avoid strchr() probably gives
3352 faster code anyway.
3353
3354
3355 Version 3.1 09-Feb-00
3356 ---------------------
3357
3358 The only change in this release is the fixing of some bugs in Makefile.in for
3359 the "install" target:
3360
3361 (1) It was failing to install pcreposix.h.
3362
3363 (2) It was overwriting the pcre.3 man page with the pcreposix.3 man page.
3364
3365
3366 Version 3.0 01-Feb-00
3367 ---------------------
3368
3369 1. Add support for the /+ modifier to perltest (to output $` like it does in
3370 pcretest).
3371
3372 2. Add support for the /g modifier to perltest.
3373
3374 3. Fix pcretest so that it behaves even more like Perl for /g when the pattern
3375 matches null strings.
3376
3377 4. Fix perltest so that it doesn't do unwanted things when fed an empty
3378 pattern. Perl treats empty patterns specially - it reuses the most recent
3379 pattern, which is not what we want. Replace // by /(?#)/ in order to avoid this
3380 effect.
3381
3382 5. The POSIX interface was broken in that it was just handing over the POSIX
3383 captured string vector to pcre_exec(), but (since release 2.00) PCRE has
3384 required a bigger vector, with some working space on the end. This means that
3385 the POSIX wrapper now has to get and free some memory, and copy the results.
3386
3387 6. Added some simple autoconf support, placing the test data and the
3388 documentation in separate directories, re-organizing some of the
3389 information files, and making it build pcre-config (a GNU standard). Also added
3390 libtool support for building PCRE as a shared library, which is now the
3391 default.
3392
3393 7. Got rid of the leading zero in the definition of PCRE_MINOR because 08 and
3394 09 are not valid octal constants. Single digits will be used for minor values
3395 less than 10.
3396
3397 8. Defined REG_EXTENDED and REG_NOSUB as zero in the POSIX header, so that
3398 existing programs that set these in the POSIX interface can use PCRE without
3399 modification.
3400
3401 9. Added a new function, pcre_fullinfo() with an extensible interface. It can
3402 return all that pcre_info() returns, plus additional data. The pcre_info()
3403 function is retained for compatibility, but is considered to be obsolete.
3404
3405 10. Added experimental recursion feature (?R) to handle one common case that
3406 Perl 5.6 will be able to do with (?p{...}).
3407
3408 11. Added support for POSIX character classes like [:alpha:], which Perl is
3409 adopting.
3410
3411
3412 Version 2.08 31-Aug-99
3413 ----------------------
3414
3415 1. When startoffset was not zero and the pattern began with ".*", PCRE was not
3416 trying to match at the startoffset position, but instead was moving forward to
3417 the next newline as if a previous match had failed.
3418
3419 2. pcretest was not making use of PCRE_NOTEMPTY when repeating for /g and /G,
3420 and could get into a loop if a null string was matched other than at the start
3421 of the subject.
3422
3423 3. Added definitions of PCRE_MAJOR and PCRE_MINOR to pcre.h so the version can
3424 be distinguished at compile time, and for completeness also added PCRE_DATE.
3425
3426 5. Added Paul Sokolovsky's minor changes to make it easy to compile a Win32 DLL
3427 in GnuWin32 environments.
3428
3429
3430 Version 2.07 29-Jul-99
3431 ----------------------
3432
3433 1. The documentation is now supplied in plain text form and HTML as well as in
3434 the form of man page sources.
3435
3436 2. C++ compilers don't like assigning (void *) values to other pointer types.
3437 In particular this affects malloc(). Although there is no problem in Standard
3438 C, I've put in casts to keep C++ compilers happy.
3439
3440 3. Typo on pcretest.c; a cast of (unsigned char *) in the POSIX regexec() call
3441 should be (const char *).
3442
3443 4. If NOPOSIX is defined, pcretest.c compiles without POSIX support. This may
3444 be useful for non-Unix systems who don't want to bother with the POSIX stuff.
3445 However, I haven't made this a standard facility. The documentation doesn't
3446 mention it, and the Makefile doesn't support it.
3447
3448 5. The Makefile now contains an "install" target, with editable destinations at
3449 the top of the file. The pcretest program is not installed.
3450
3451 6. pgrep -V now gives the PCRE version number and date.
3452
3453 7. Fixed bug: a zero repetition after a literal string (e.g. /abcde{0}/) was
3454 causing the entire string to be ignored, instead of just the last character.
3455
3456 8. If a pattern like /"([^\\"]+|\\.)*"/ is applied in the normal way to a
3457 non-matching string, it can take a very, very long time, even for strings of
3458 quite modest length, because of the nested recursion. PCRE now does better in
3459 some of these cases. It does this by remembering the last required literal
3460 character in the pattern, and pre-searching the subject to ensure it is present
3461 before running the real match. In other words, it applies a heuristic to detect
3462 some types of certain failure quickly, and in the above example, if presented
3463 with a string that has no trailing " it gives "no match" very quickly.
3464
3465 9. A new runtime option PCRE_NOTEMPTY causes null string matches to be ignored;
3466 other alternatives are tried instead.
3467
3468
3469 Version 2.06 09-Jun-99
3470 ----------------------
3471
3472 1. Change pcretest's output for amount of store used to show just the code
3473 space, because the remainder (the data block) varies in size between 32-bit and
3474 64-bit systems.
3475
3476 2. Added an extra argument to pcre_exec() to supply an offset in the subject to
3477 start matching at. This allows lookbehinds to work when searching for multiple
3478 occurrences in a string.
3479
3480 3. Added additional options to pcretest for testing multiple occurrences:
3481
3482 /+ outputs the rest of the string that follows a match
3483 /g loops for multiple occurrences, using the new startoffset argument
3484 /G loops for multiple occurrences by passing an incremented pointer
3485
3486 4. PCRE wasn't doing the "first character" optimization for patterns starting
3487 with \b or \B, though it was doing it for other lookbehind assertions. That is,
3488 it wasn't noticing that a match for a pattern such as /\bxyz/ has to start with
3489 the letter 'x'. On long subject strings, this gives a significant speed-up.
3490
3491
3492 Version 2.05 21-Apr-99
3493 ----------------------
3494
3495 1. Changed the type of magic_number from int to long int so that it works
3496 properly on 16-bit systems.
3497
3498 2. Fixed a bug which caused patterns starting with .* not to work correctly
3499 when the subject string contained newline characters. PCRE was assuming
3500 anchoring for such patterns in all cases, which is not correct because .* will
3501 not pass a newline unless PCRE_DOTALL is set. It now assumes anchoring only if
3502 DOTALL is set at top level; otherwise it knows that patterns starting with .*
3503 must be retried after every newline in the subject.
3504
3505
3506 Version 2.04 18-Feb-99
3507 ----------------------
3508
3509 1. For parenthesized subpatterns with repeats whose minimum was zero, the
3510 computation of the store needed to hold the pattern was incorrect (too large).
3511 If such patterns were nested a few deep, this could multiply and become a real
3512 problem.
3513
3514 2. Added /M option to pcretest to show the memory requirement of a specific
3515 pattern. Made -m a synonym of -s (which does this globally) for compatibility.
3516
3517 3. Subpatterns of the form (regex){n,m} (i.e. limited maximum) were being
3518 compiled in such a way that the backtracking after subsequent failure was
3519 pessimal. Something like (a){0,3} was compiled as (a)?(a)?(a)? instead of
3520 ((a)((a)(a)?)?)? with disastrous performance if the maximum was of any size.
3521
3522
3523 Version 2.03 02-Feb-99
3524 ----------------------
3525
3526 1. Fixed typo and small mistake in man page.
3527
3528 2. Added 4th condition (GPL supersedes if conflict) and created separate
3529 LICENCE file containing the conditions.
3530
3531 3. Updated pcretest so that patterns such as /abc\/def/ work like they do in
3532 Perl, that is the internal \ allows the delimiter to be included in the
3533 pattern. Locked out the use of \ as a delimiter. If \ immediately follows
3534 the final delimiter, add \ to the end of the pattern (to test the error).
3535
3536 4. Added the convenience functions for extracting substrings after a successful
3537 match. Updated pcretest to make it able to test these functions.
3538
3539
3540 Version 2.02 14-Jan-99
3541 ----------------------
3542
3543 1. Initialized the working variables associated with each extraction so that
3544 their saving and restoring doesn't refer to uninitialized store.
3545
3546 2. Put dummy code into study.c in order to trick the optimizer of the IBM C
3547 compiler for OS/2 into generating correct code. Apparently IBM isn't going to
3548 fix the problem.
3549
3550 3. Pcretest: the timing code wasn't using LOOPREPEAT for timing execution
3551 calls, and wasn't printing the correct value for compiling calls. Increased the
3552 default value of LOOPREPEAT, and the number of significant figures in the
3553 times.
3554
3555 4. Changed "/bin/rm" in the Makefile to "-rm" so it works on Windows NT.
3556
3557 5. Renamed "deftables" as "dftables" to get it down to 8 characters, to avoid
3558 a building problem on Windows NT with a FAT file system.
3559
3560
3561 Version 2.01 21-Oct-98
3562 ----------------------
3563
3564 1. Changed the API for pcre_compile() to allow for the provision of a pointer
3565 to character tables built by pcre_maketables() in the current locale. If NULL
3566 is passed, the default tables are used.
3567
3568
3569 Version 2.00 24-Sep-98
3570 ----------------------
3571
3572 1. Since the (>?) facility is in Perl 5.005, don't require PCRE_EXTRA to enable
3573 it any more.
3574
3575 2. Allow quantification of (?>) groups, and make it work correctly.
3576
3577 3. The first character computation wasn't working for (?>) groups.
3578
3579 4. Correct the implementation of \Z (it is permitted to match on the \n at the
3580 end of the subject) and add 5.005's \z, which really does match only at the
3581 very end of the subject.
3582
3583 5. Remove the \X "cut" facility; Perl doesn't have it, and (?> is neater.
3584
3585 6. Remove the ability to specify CASELESS, MULTILINE, DOTALL, and
3586 DOLLAR_END_ONLY at runtime, to make it possible to implement the Perl 5.005
3587 localized options. All options to pcre_study() were also removed.
3588
3589 7. Add other new features from 5.005:
3590
3591 $(?<= positive lookbehind
3592 $(?<! negative lookbehind
3593 (?imsx-imsx) added the unsetting capability
3594 such a setting is global if at outer level; local otherwise
3595 (?imsx-imsx:) non-capturing groups with option setting
3596 (?(cond)re|re) conditional pattern matching
3597
3598 A backreference to itself in a repeated group matches the previous
3599 captured string.
3600
3601 8. General tidying up of studying (both automatic and via "study")
3602 consequential on the addition of new assertions.
3603
3604 9. As in 5.005, unlimited repeated groups that could match an empty substring
3605 are no longer faulted at compile time. Instead, the loop is forcibly broken at
3606 runtime if any iteration does actually match an empty substring.
3607
3608 10. Include the RunTest script in the distribution.
3609
3610 11. Added tests from the Perl 5.005_02 distribution. This showed up a few
3611 discrepancies, some of which were old and were also with respect to 5.004. They
3612 have now been fixed.
3613
3614
3615 Version 1.09 28-Apr-98
3616 ----------------------
3617
3618 1. A negated single character class followed by a quantifier with a minimum
3619 value of one (e.g. [^x]{1,6} ) was not compiled correctly. This could lead to
3620 program crashes, or just wrong answers. This did not apply to negated classes
3621 containing more than one character, or to minima other than one.
3622
3623
3624 Version 1.08 27-Mar-98
3625 ----------------------
3626
3627 1. Add PCRE_UNGREEDY to invert the greediness of quantifiers.
3628
3629 2. Add (?U) and (?X) to set PCRE_UNGREEDY and PCRE_EXTRA respectively. The
3630 latter must appear before anything that relies on it in the pattern.
3631
3632
3633 Version 1.07 16-Feb-98
3634 ----------------------
3635
3636 1. A pattern such as /((a)*)*/ was not being diagnosed as in error (unlimited
3637 repeat of a potentially empty string).
3638
3639
3640 Version 1.06 23-Jan-98
3641 ----------------------
3642
3643 1. Added Markus Oberhumer's little patches for C++.
3644
3645 2. Literal strings longer than 255 characters were broken.
3646
3647
3648 Version 1.05 23-Dec-97
3649 ----------------------
3650
3651 1. Negated character classes containing more than one character were failing if
3652 PCRE_CASELESS was set at run time.
3653
3654
3655 Version 1.04 19-Dec-97
3656 ----------------------
3657
3658 1. Corrected the man page, where some "const" qualifiers had been omitted.
3659
3660 2. Made debugging output print "{0,xxx}" instead of just "{,xxx}" to agree with
3661 input syntax.
3662
3663 3. Fixed memory leak which occurred when a regex with back references was
3664 matched with an offsets vector that wasn't big enough. The temporary memory
3665 that is used in this case wasn't being freed if the match failed.
3666
3667 4. Tidied pcretest to ensure it frees memory that it gets.
3668
3669 5. Temporary memory was being obtained in the case where the passed offsets
3670 vector was exactly big enough.
3671
3672 6. Corrected definition of offsetof() from change 5 below.
3673
3674 7. I had screwed up change 6 below and broken the rules for the use of
3675 setjmp(). Now fixed.
3676
3677
3678 Version 1.03 18-Dec-97
3679 ----------------------
3680
3681 1. A erroneous regex with a missing opening parenthesis was correctly
3682 diagnosed, but PCRE attempted to access brastack[-1], which could cause crashes
3683 on some systems.
3684
3685 2. Replaced offsetof(real_pcre, code) by offsetof(real_pcre, code[0]) because
3686 it was reported that one broken compiler failed on the former because "code" is
3687 also an independent variable.
3688
3689 3. The erroneous regex a[]b caused an array overrun reference.
3690
3691 4. A regex ending with a one-character negative class (e.g. /[^k]$/) did not
3692 fail on data ending with that character. (It was going on too far, and checking
3693 the next character, typically a binary zero.) This was specific to the
3694 optimized code for single-character negative classes.
3695
3696 5. Added a contributed patch from the TIN world which does the following:
3697
3698 + Add an undef for memmove, in case the the system defines a macro for it.
3699
3700 + Add a definition of offsetof(), in case there isn't one. (I don't know
3701 the reason behind this - offsetof() is part of the ANSI standard - but
3702 it does no harm).
3703
3704 + Reduce the ifdef's in pcre.c using macro DPRINTF, thereby eliminating
3705 most of the places where whitespace preceded '#'. I have given up and
3706 allowed the remaining 2 cases to be at the margin.
3707
3708 + Rename some variables in pcre to eliminate shadowing. This seems very
3709 pedantic, but does no harm, of course.
3710
3711 6. Moved the call to setjmp() into its own function, to get rid of warnings
3712 from gcc -Wall, and avoided calling it at all unless PCRE_EXTRA is used.
3713
3714 7. Constructs such as \d{8,} were compiling into the equivalent of
3715 \d{8}\d{0,65527} instead of \d{8}\d* which didn't make much difference to the
3716 outcome, but in this particular case used more store than had been allocated,
3717 which caused the bug to be discovered because it threw up an internal error.
3718
3719 8. The debugging code in both pcre and pcretest for outputting the compiled
3720 form of a regex was going wrong in the case of back references followed by
3721 curly-bracketed repeats.
3722
3723
3724 Version 1.02 12-Dec-97
3725 ----------------------
3726
3727 1. Typos in pcre.3 and comments in the source fixed.
3728
3729 2. Applied a contributed patch to get rid of places where it used to remove
3730 'const' from variables, and fixed some signed/unsigned and uninitialized
3731 variable warnings.
3732
3733 3. Added the "runtest" target to Makefile.
3734
3735 4. Set default compiler flag to -O2 rather than just -O.
3736
3737
3738 Version 1.01 19-Nov-97
3739 ----------------------
3740
3741 1. PCRE was failing to diagnose unlimited repeat of empty string for patterns
3742 like /([ab]*)*/, that is, for classes with more than one character in them.
3743
3744 2. Likewise, it wasn't diagnosing patterns with "once-only" subpatterns, such
3745 as /((?>a*))*/ (a PCRE_EXTRA facility).
3746
3747
3748 Version 1.00 18-Nov-97
3749 ----------------------
3750
3751 1. Added compile-time macros to support systems such as SunOS4 which don't have
3752 memmove() or strerror() but have other things that can be used instead.
3753
3754 2. Arranged that "make clean" removes the executables.
3755
3756
3757 Version 0.99 27-Oct-97
3758 ----------------------
3759
3760 1. Fixed bug in code for optimizing classes with only one character. It was
3761 initializing a 32-byte map regardless, which could cause it to run off the end
3762 of the memory it had got.
3763
3764 2. Added, conditional on PCRE_EXTRA, the proposed (?>REGEX) construction.
3765
3766
3767 Version 0.98 22-Oct-97
3768 ----------------------
3769
3770 1. Fixed bug in code for handling temporary memory usage when there are more
3771 back references than supplied space in the ovector. This could cause segfaults.
3772
3773
3774 Version 0.97 21-Oct-97
3775 ----------------------
3776
3777 1. Added the \X "cut" facility, conditional on PCRE_EXTRA.
3778
3779 2. Optimized negated single characters not to use a bit map.
3780
3781 3. Brought error texts together as macro definitions; clarified some of them;
3782 fixed one that was wrong - it said "range out of order" when it meant "invalid
3783 escape sequence".
3784
3785 4. Changed some char * arguments to const char *.
3786
3787 5. Added PCRE_NOTBOL and PCRE_NOTEOL (from POSIX).
3788
3789 6. Added the POSIX-style API wrapper in pcreposix.a and testing facilities in
3790 pcretest.
3791
3792
3793 Version 0.96 16-Oct-97
3794 ----------------------
3795
3796 1. Added a simple "pgrep" utility to the distribution.
3797
3798 2. Fixed an incompatibility with Perl: "{" is now treated as a normal character
3799 unless it appears in one of the precise forms "{ddd}", "{ddd,}", or "{ddd,ddd}"
3800 where "ddd" means "one or more decimal digits".
3801
3802 3. Fixed serious bug. If a pattern had a back reference, but the call to
3803 pcre_exec() didn't supply a large enough ovector to record the related
3804 identifying subpattern, the match always failed. PCRE now remembers the number
3805 of the largest back reference, and gets some temporary memory in which to save
3806 the offsets during matching if necessary, in order to ensure that
3807 backreferences always work.
3808
3809 4. Increased the compatibility with Perl in a number of ways:
3810
3811 (a) . no longer matches \n by default; an option PCRE_DOTALL is provided
3812 to request this handling. The option can be set at compile or exec time.
3813
3814 (b) $ matches before a terminating newline by default; an option
3815 PCRE_DOLLAR_ENDONLY is provided to override this (but not in multiline
3816 mode). The option can be set at compile or exec time.
3817
3818 (c) The handling of \ followed by a digit other than 0 is now supposed to be
3819 the same as Perl's. If the decimal number it represents is less than 10
3820 or there aren't that many previous left capturing parentheses, an octal
3821 escape is read. Inside a character class, it's always an octal escape,
3822 even if it is a single digit.
3823
3824 (d) An escaped but undefined alphabetic character is taken as a literal,
3825 unless PCRE_EXTRA is set. Currently this just reserves the remaining
3826 escapes.
3827
3828 (e) {0} is now permitted. (The previous item is removed from the compiled
3829 pattern).
3830
3831 5. Changed all the names of code files so that the basic parts are no longer
3832 than 10 characters, and abolished the teeny "globals.c" file.
3833
3834 6. Changed the handling of character classes; they are now done with a 32-byte
3835 bit map always.
3836
3837 7. Added the -d and /D options to pcretest to make it possible to look at the
3838 internals of compilation without having to recompile pcre.
3839
3840
3841 Version 0.95 23-Sep-97
3842 ----------------------
3843
3844 1. Fixed bug in pre-pass concerning escaped "normal" characters such as \x5c or
3845 \x20 at the start of a run of normal characters. These were being treated as
3846 real characters, instead of the source characters being re-checked.
3847
3848
3849 Version 0.94 18-Sep-97
3850 ----------------------
3851
3852 1. The functions are now thread-safe, with the caveat that the global variables
3853 containing pointers to malloc() and free() or alternative functions are the
3854 same for all threads.
3855
3856 2. Get pcre_study() to generate a bitmap of initial characters for non-
3857 anchored patterns when this is possible, and use it if passed to pcre_exec().
3858
3859
3860 Version 0.93 15-Sep-97
3861 ----------------------
3862
3863 1. /(b)|(:+)/ was computing an incorrect first character.
3864
3865 2. Add pcre_study() to the API and the passing of pcre_extra to pcre_exec(),
3866 but not actually doing anything yet.
3867
3868 3. Treat "-" characters in classes that cannot be part of ranges as literals,
3869 as Perl does (e.g. [-az] or [az-]).
3870
3871 4. Set the anchored flag if a branch starts with .* or .*? because that tests
3872 all possible positions.
3873
3874 5. Split up into different modules to avoid including unneeded functions in a
3875 compiled binary. However, compile and exec are still in one module. The "study"
3876 function is split off.
3877
3878 6. The character tables are now in a separate module whose source is generated
3879 by an auxiliary program - but can then be edited by hand if required. There are
3880 now no calls to isalnum(), isspace(), isdigit(), isxdigit(), tolower() or
3881 toupper() in the code.
3882
3883 7. Turn the malloc/free funtions variables into pcre_malloc and pcre_free and
3884 make them global. Abolish the function for setting them, as the caller can now
3885 set them directly.
3886
3887
3888 Version 0.92 11-Sep-97
3889 ----------------------
3890
3891 1. A repeat with a fixed maximum and a minimum of 1 for an ordinary character
3892 (e.g. /a{1,3}/) was broken (I mis-optimized it).
3893
3894 2. Caseless matching was not working in character classes if the characters in
3895 the pattern were in upper case.
3896
3897 3. Make ranges like [W-c] work in the same way as Perl for caseless matching.
3898
3899 4. Make PCRE_ANCHORED public and accept as a compile option.
3900
3901 5. Add an options word to pcre_exec() and accept PCRE_ANCHORED and
3902 PCRE_CASELESS at run time. Add escapes \A and \I to pcretest to cause it to
3903 pass them.
3904
3905 6. Give an error if bad option bits passed at compile or run time.
3906
3907 7. Add PCRE_MULTILINE at compile and exec time, and (?m) as well. Add \M to
3908 pcretest to cause it to pass that flag.
3909
3910 8. Add pcre_info(), to get the number of identifying subpatterns, the stored
3911 options, and the first character, if set.
3912
3913 9. Recognize C+ or C{n,m} where n >= 1 as providing a fixed starting character.
3914
3915
3916 Version 0.91 10-Sep-97
3917 ----------------------
3918
3919 1. PCRE was failing to diagnose unlimited repeats of subpatterns that could
3920 match the empty string as in /(a*)*/. It was looping and ultimately crashing.
3921
3922 2. PCRE was looping on encountering an indefinitely repeated back reference to
3923 a subpattern that had matched an empty string, e.g. /(a|)\1*/. It now does what
3924 Perl does - treats the match as successful.
3925
3926 ****

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