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

Contents of /code/trunk/ChangeLog

Parent Directory Parent Directory | Revision Log Revision Log


Revision 524 - (show annotations) (download)
Mon May 24 17:06:28 2010 UTC (4 years, 5 months ago) by ph10
File size: 187604 byte(s)
Make pcre_study() recognize \h, \v, and \R.

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

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