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

Contents of /code/trunk/ChangeLog

Parent Directory Parent Directory | Revision Log Revision Log


Revision 500 - (show annotations)
Sat Mar 6 19:00:29 2010 UTC (5 years, 3 months ago) by ph10
File size: 183722 byte(s)
Error occurred while calculating annotation data.
Fix bugs with \K in atomic groups, subroutines, and assertions.
1 ChangeLog for PCRE
2 ------------------
3
4 Version 8.02 01-Mar-2010
5 ------------------------
6
7 1. The Unicode data tables have been updated to Unicode 5.2.0.
8
9 2. Added the option --libs-cpp to pcre-config, but only when C++ support is
10 configured.
11
12 3. Updated the licensing terms in the pcregexp.pas file, as agreed with the
13 original author of that file, following a query about its status.
14
15 4. On systems that do not have stdint.h (e.g. Solaris), check for and include
16 inttypes.h instead. This fixes a bug that was introduced by change 8.01/8.
17
18 5. A pattern such as (?&t)*+(?(DEFINE)(?<t>.)) which has a possessive
19 quantifier applied to a forward-referencing subroutine call, could compile
20 incorrect code or give the error "internal error: previously-checked
21 referenced subpattern not found".
22
23 6. Both MS Visual Studio and Symbian OS have problems with initializing
24 variables to point to external functions. For these systems, therefore,
25 pcre_malloc etc. are now initialized to local functions that call the
26 relevant global functions.
27
28 7. There were two entries missing in the vectors called coptable and poptable
29 in pcre_dfa_exec.c. This could lead to memory accesses outsize the vectors.
30 I've fixed the data, and added a kludgy way of testing at compile time that
31 the lengths are correct (equal to the number of opcodes).
32
33 8. Following on from 7, I added a similar kludge to check the length of the
34 eint vector in pcreposix.c.
35
36 9. Error texts for pcre_compile() are held as one long string to avoid too
37 much relocation at load time. To find a text, the string is searched,
38 counting zeros. There was no check for running off the end of the string,
39 which could happen if a new error number was added without updating the
40 string.
41
42 10. \K gave a compile-time error if it appeared in a lookbehind assersion.
43
44 11. \K was not working if it appeared in an atomic group or in a group that
45 was called as a "subroutine", or in an assertion. Perl 5.11 documents that
46 \K is "not well defined" if used in an assertion. PCRE now accepts it if
47 the assertion is positive, but not if it is negative.
48
49
50 Version 8.01 19-Jan-2010
51 ------------------------
52
53 1. If a pattern contained a conditional subpattern with only one branch (in
54 particular, this includes all (*DEFINE) patterns), a call to pcre_study()
55 computed the wrong minimum data length (which is of course zero for such
56 subpatterns). This could cause incorrect "no match" results.
57
58 2. For patterns such as (?i)a(?-i)b|c where an option setting at the start of
59 the pattern is reset in the first branch, pcre_compile() failed with
60 "internal error: code overflow at offset...". This happened only when
61 the reset was to the original external option setting. (An optimization
62 abstracts leading options settings into an external setting, which was the
63 cause of this.)
64
65 3. A pattern such as ^(?!a(*SKIP)b) where a negative assertion contained one
66 of the verbs SKIP, PRUNE, or COMMIT, did not work correctly. When the
67 assertion pattern did not match (meaning that the assertion was true), it
68 was incorrectly treated as false if the SKIP had been reached during the
69 matching. This also applied to assertions used as conditions.
70
71 4. If an item that is not supported by pcre_dfa_exec() was encountered in an
72 assertion subpattern, including such a pattern used as a condition,
73 unpredictable results occurred, instead of the error return
74 PCRE_ERROR_DFA_UITEM.
75
76 5. The C++ GlobalReplace function was not working like Perl for the special
77 situation when an empty string is matched. It now does the fancy magic
78 stuff that is necessary.
79
80 6. In pcre_internal.h, obsolete includes to setjmp.h and stdarg.h have been
81 removed. (These were left over from very, very early versions of PCRE.)
82
83 7. Some cosmetic changes to the code to make life easier when compiling it
84 as part of something else:
85
86 (a) Change DEBUG to PCRE_DEBUG.
87
88 (b) In pcre_compile(), rename the member of the "branch_chain" structure
89 called "current" as "current_branch", to prevent a collision with the
90 Linux macro when compiled as a kernel module.
91
92 (c) In pcre_study(), rename the function set_bit() as set_table_bit(), to
93 prevent a collision with the Linux macro when compiled as a kernel
94 module.
95
96 8. In pcre_compile() there are some checks for integer overflows that used to
97 cast potentially large values to (double). This has been changed to that
98 when building, a check for int64_t is made, and if it is found, it is used
99 instead, thus avoiding the use of floating point arithmetic. (There is no
100 other use of FP in PCRE.) If int64_t is not found, the fallback is to
101 double.
102
103 9. Added two casts to avoid signed/unsigned warnings from VS Studio Express
104 2005 (difference between two addresses compared to an unsigned value).
105
106 10. Change the standard AC_CHECK_LIB test for libbz2 in configure.ac to a
107 custom one, because of the following reported problem in Windows:
108
109 - libbz2 uses the Pascal calling convention (WINAPI) for the functions
110 under Win32.
111 - The standard autoconf AC_CHECK_LIB fails to include "bzlib.h",
112 therefore missing the function definition.
113 - The compiler thus generates a "C" signature for the test function.
114 - The linker fails to find the "C" function.
115 - PCRE fails to configure if asked to do so against libbz2.
116
117 11. When running libtoolize from libtool-2.2.6b as part of autogen.sh, these
118 messages were output:
119
120 Consider adding `AC_CONFIG_MACRO_DIR([m4])' to configure.ac and
121 rerunning libtoolize, to keep the correct libtool macros in-tree.
122 Consider adding `-I m4' to ACLOCAL_AMFLAGS in Makefile.am.
123
124 I have done both of these things.
125
126 12. Although pcre_dfa_exec() does not use nearly as much stack as pcre_exec()
127 most of the time, it *can* run out if it is given a pattern that contains a
128 runaway infinite recursion. I updated the discussion in the pcrestack man
129 page.
130
131 13. Now that we have gone to the x.xx style of version numbers, the minor
132 version may start with zero. Using 08 or 09 is a bad idea because users
133 might check the value of PCRE_MINOR in their code, and 08 or 09 may be
134 interpreted as invalid octal numbers. I've updated the previous comment in
135 configure.ac, and also added a check that gives an error if 08 or 09 are
136 used.
137
138 14. Change 8.00/11 was not quite complete: code had been accidentally omitted,
139 causing partial matching to fail when the end of the subject matched \W
140 in a UTF-8 pattern where \W was quantified with a minimum of 3.
141
142 15. There were some discrepancies between the declarations in pcre_internal.h
143 of _pcre_is_newline(), _pcre_was_newline(), and _pcre_valid_utf8() and
144 their definitions. The declarations used "const uschar *" and the
145 definitions used USPTR. Even though USPTR is normally defined as "const
146 unsigned char *" (and uschar is typedeffed as "unsigned char"), it was
147 reported that: "This difference in casting confuses some C++ compilers, for
148 example, SunCC recognizes above declarations as different functions and
149 generates broken code for hbpcre." I have changed the declarations to use
150 USPTR.
151
152 16. GNU libtool is named differently on some systems. The autogen.sh script now
153 tries several variants such as glibtoolize (MacOSX) and libtoolize1x
154 (FreeBSD).
155
156 17. Applied Craig's patch that fixes an HP aCC compile error in pcre 8.00
157 (strtoXX undefined when compiling pcrecpp.cc). The patch contains this
158 comment: "Figure out how to create a longlong from a string: strtoll and
159 equivalent. It's not enough to call AC_CHECK_FUNCS: hpux has a strtoll, for
160 instance, but it only takes 2 args instead of 3!"
161
162 18. A subtle bug concerned with back references has been fixed by a change of
163 specification, with a corresponding code fix. A pattern such as
164 ^(xa|=?\1a)+$ which contains a back reference inside the group to which it
165 refers, was giving matches when it shouldn't. For example, xa=xaaa would
166 match that pattern. Interestingly, Perl (at least up to 5.11.3) has the
167 same bug. Such groups have to be quantified to be useful, or contained
168 inside another quantified group. (If there's no repetition, the reference
169 can never match.) The problem arises because, having left the group and
170 moved on to the rest of the pattern, a later failure that backtracks into
171 the group uses the captured value from the final iteration of the group
172 rather than the correct earlier one. I have fixed this in PCRE by forcing
173 any group that contains a reference to itself to be an atomic group; that
174 is, there cannot be any backtracking into it once it has completed. This is
175 similar to recursive and subroutine calls.
176
177
178 Version 8.00 19-Oct-09
179 ----------------------
180
181 1. The table for translating pcre_compile() error codes into POSIX error codes
182 was out-of-date, and there was no check on the pcre_compile() error code
183 being within the table. This could lead to an OK return being given in
184 error.
185
186 2. Changed the call to open a subject file in pcregrep from fopen(pathname,
187 "r") to fopen(pathname, "rb"), which fixed a problem with some of the tests
188 in a Windows environment.
189
190 3. The pcregrep --count option prints the count for each file even when it is
191 zero, as does GNU grep. However, pcregrep was also printing all files when
192 --files-with-matches was added. Now, when both options are given, it prints
193 counts only for those files that have at least one match. (GNU grep just
194 prints the file name in this circumstance, but including the count seems
195 more useful - otherwise, why use --count?) Also ensured that the
196 combination -clh just lists non-zero counts, with no names.
197
198 4. The long form of the pcregrep -F option was incorrectly implemented as
199 --fixed_strings instead of --fixed-strings. This is an incompatible change,
200 but it seems right to fix it, and I didn't think it was worth preserving
201 the old behaviour.
202
203 5. The command line items --regex=pattern and --regexp=pattern were not
204 recognized by pcregrep, which required --regex pattern or --regexp pattern
205 (with a space rather than an '='). The man page documented the '=' forms,
206 which are compatible with GNU grep; these now work.
207
208 6. No libpcreposix.pc file was created for pkg-config; there was just
209 libpcre.pc and libpcrecpp.pc. The omission has been rectified.
210
211 7. Added #ifndef SUPPORT_UCP into the pcre_ucd.c module, to reduce its size
212 when UCP support is not needed, by modifying the Python script that
213 generates it from Unicode data files. This should not matter if the module
214 is correctly used as a library, but I received one complaint about 50K of
215 unwanted data. My guess is that the person linked everything into his
216 program rather than using a library. Anyway, it does no harm.
217
218 8. A pattern such as /\x{123}{2,2}+/8 was incorrectly compiled; the trigger
219 was a minimum greater than 1 for a wide character in a possessive
220 repetition. The same bug could also affect patterns like /(\x{ff}{0,2})*/8
221 which had an unlimited repeat of a nested, fixed maximum repeat of a wide
222 character. Chaos in the form of incorrect output or a compiling loop could
223 result.
224
225 9. The restrictions on what a pattern can contain when partial matching is
226 requested for pcre_exec() have been removed. All patterns can now be
227 partially matched by this function. In addition, if there are at least two
228 slots in the offset vector, the offset of the earliest inspected character
229 for the match and the offset of the end of the subject are set in them when
230 PCRE_ERROR_PARTIAL is returned.
231
232 10. Partial matching has been split into two forms: PCRE_PARTIAL_SOFT, which is
233 synonymous with PCRE_PARTIAL, for backwards compatibility, and
234 PCRE_PARTIAL_HARD, which causes a partial match to supersede a full match,
235 and may be more useful for multi-segment matching.
236
237 11. Partial matching with pcre_exec() is now more intuitive. A partial match
238 used to be given if ever the end of the subject was reached; now it is
239 given only if matching could not proceed because another character was
240 needed. This makes a difference in some odd cases such as Z(*FAIL) with the
241 string "Z", which now yields "no match" instead of "partial match". In the
242 case of pcre_dfa_exec(), "no match" is given if every matching path for the
243 final character ended with (*FAIL).
244
245 12. Restarting a match using pcre_dfa_exec() after a partial match did not work
246 if the pattern had a "must contain" character that was already found in the
247 earlier partial match, unless partial matching was again requested. For
248 example, with the pattern /dog.(body)?/, the "must contain" character is
249 "g". If the first part-match was for the string "dog", restarting with
250 "sbody" failed. This bug has been fixed.
251
252 13. The string returned by pcre_dfa_exec() after a partial match has been
253 changed so that it starts at the first inspected character rather than the
254 first character of the match. This makes a difference only if the pattern
255 starts with a lookbehind assertion or \b or \B (\K is not supported by
256 pcre_dfa_exec()). It's an incompatible change, but it makes the two
257 matching functions compatible, and I think it's the right thing to do.
258
259 14. Added a pcredemo man page, created automatically from the pcredemo.c file,
260 so that the demonstration program is easily available in environments where
261 PCRE has not been installed from source.
262
263 15. Arranged to add -DPCRE_STATIC to cflags in libpcre.pc, libpcreposix.cp,
264 libpcrecpp.pc and pcre-config when PCRE is not compiled as a shared
265 library.
266
267 16. Added REG_UNGREEDY to the pcreposix interface, at the request of a user.
268 It maps to PCRE_UNGREEDY. It is not, of course, POSIX-compatible, but it
269 is not the first non-POSIX option to be added. Clearly some people find
270 these options useful.
271
272 17. If a caller to the POSIX matching function regexec() passes a non-zero
273 value for nmatch with a NULL value for pmatch, the value of
274 nmatch is forced to zero.
275
276 18. RunGrepTest did not have a test for the availability of the -u option of
277 the diff command, as RunTest does. It now checks in the same way as
278 RunTest, and also checks for the -b option.
279
280 19. If an odd number of negated classes containing just a single character
281 interposed, within parentheses, between a forward reference to a named
282 subpattern and the definition of the subpattern, compilation crashed with
283 an internal error, complaining that it could not find the referenced
284 subpattern. An example of a crashing pattern is /(?&A)(([^m])(?<A>))/.
285 [The bug was that it was starting one character too far in when skipping
286 over the character class, thus treating the ] as data rather than
287 terminating the class. This meant it could skip too much.]
288
289 20. Added PCRE_NOTEMPTY_ATSTART in order to be able to correctly implement the
290 /g option in pcretest when the pattern contains \K, which makes it possible
291 to have an empty string match not at the start, even when the pattern is
292 anchored. Updated pcretest and pcredemo to use this option.
293
294 21. If the maximum number of capturing subpatterns in a recursion was greater
295 than the maximum at the outer level, the higher number was returned, but
296 with unset values at the outer level. The correct (outer level) value is
297 now given.
298
299 22. If (*ACCEPT) appeared inside capturing parentheses, previous releases of
300 PCRE did not set those parentheses (unlike Perl). I have now found a way to
301 make it do so. The string so far is captured, making this feature
302 compatible with Perl.
303
304 23. The tests have been re-organized, adding tests 11 and 12, to make it
305 possible to check the Perl 5.10 features against Perl 5.10.
306
307 24. Perl 5.10 allows subroutine calls in lookbehinds, as long as the subroutine
308 pattern matches a fixed length string. PCRE did not allow this; now it
309 does. Neither allows recursion.
310
311 25. I finally figured out how to implement a request to provide the minimum
312 length of subject string that was needed in order to match a given pattern.
313 (It was back references and recursion that I had previously got hung up
314 on.) This code has now been added to pcre_study(); it finds a lower bound
315 to the length of subject needed. It is not necessarily the greatest lower
316 bound, but using it to avoid searching strings that are too short does give
317 some useful speed-ups. The value is available to calling programs via
318 pcre_fullinfo().
319
320 26. While implementing 25, I discovered to my embarrassment that pcretest had
321 not been passing the result of pcre_study() to pcre_dfa_exec(), so the
322 study optimizations had never been tested with that matching function.
323 Oops. What is worse, even when it was passed study data, there was a bug in
324 pcre_dfa_exec() that meant it never actually used it. Double oops. There
325 were also very few tests of studied patterns with pcre_dfa_exec().
326
327 27. If (?| is used to create subpatterns with duplicate numbers, they are now
328 allowed to have the same name, even if PCRE_DUPNAMES is not set. However,
329 on the other side of the coin, they are no longer allowed to have different
330 names, because these cannot be distinguished in PCRE, and this has caused
331 confusion. (This is a difference from Perl.)
332
333 28. When duplicate subpattern names are present (necessarily with different
334 numbers, as required by 27 above), and a test is made by name in a
335 conditional pattern, either for a subpattern having been matched, or for
336 recursion in such a pattern, all the associated numbered subpatterns are
337 tested, and the overall condition is true if the condition is true for any
338 one of them. This is the way Perl works, and is also more like the way
339 testing by number works.
340
341
342 Version 7.9 11-Apr-09
343 ---------------------
344
345 1. When building with support for bzlib/zlib (pcregrep) and/or readline
346 (pcretest), all targets were linked against these libraries. This included
347 libpcre, libpcreposix, and libpcrecpp, even though they do not use these
348 libraries. This caused unwanted dependencies to be created. This problem
349 has been fixed, and now only pcregrep is linked with bzlib/zlib and only
350 pcretest is linked with readline.
351
352 2. The "typedef int BOOL" in pcre_internal.h that was included inside the
353 "#ifndef FALSE" condition by an earlier change (probably 7.8/18) has been
354 moved outside it again, because FALSE and TRUE are already defined in AIX,
355 but BOOL is not.
356
357 3. The pcre_config() function was treating the PCRE_MATCH_LIMIT and
358 PCRE_MATCH_LIMIT_RECURSION values as ints, when they should be long ints.
359
360 4. The pcregrep documentation said spaces were inserted as well as colons (or
361 hyphens) following file names and line numbers when outputting matching
362 lines. This is not true; no spaces are inserted. I have also clarified the
363 wording for the --colour (or --color) option.
364
365 5. In pcregrep, when --colour was used with -o, the list of matching strings
366 was not coloured; this is different to GNU grep, so I have changed it to be
367 the same.
368
369 6. When --colo(u)r was used in pcregrep, only the first matching substring in
370 each matching line was coloured. Now it goes on to look for further matches
371 of any of the test patterns, which is the same behaviour as GNU grep.
372
373 7. A pattern that could match an empty string could cause pcregrep to loop; it
374 doesn't make sense to accept an empty string match in pcregrep, so I have
375 locked it out (using PCRE's PCRE_NOTEMPTY option). By experiment, this
376 seems to be how GNU grep behaves.
377
378 8. The pattern (?(?=.*b)b|^) was incorrectly compiled as "match must be at
379 start or after a newline", because the conditional assertion was not being
380 correctly handled. The rule now is that both the assertion and what follows
381 in the first alternative must satisfy the test.
382
383 9. If auto-callout was enabled in a pattern with a conditional group whose
384 condition was an assertion, PCRE could crash during matching, both with
385 pcre_exec() and pcre_dfa_exec().
386
387 10. The PCRE_DOLLAR_ENDONLY option was not working when pcre_dfa_exec() was
388 used for matching.
389
390 11. Unicode property support in character classes was not working for
391 characters (bytes) greater than 127 when not in UTF-8 mode.
392
393 12. Added the -M command line option to pcretest.
394
395 14. Added the non-standard REG_NOTEMPTY option to the POSIX interface.
396
397 15. Added the PCRE_NO_START_OPTIMIZE match-time option.
398
399 16. Added comments and documentation about mis-use of no_arg in the C++
400 wrapper.
401
402 17. Implemented support for UTF-8 encoding in EBCDIC environments, a patch
403 from Martin Jerabek that uses macro names for all relevant character and
404 string constants.
405
406 18. Added to pcre_internal.h two configuration checks: (a) If both EBCDIC and
407 SUPPORT_UTF8 are set, give an error; (b) If SUPPORT_UCP is set without
408 SUPPORT_UTF8, define SUPPORT_UTF8. The "configure" script handles both of
409 these, but not everybody uses configure.
410
411 19. A conditional group that had only one branch was not being correctly
412 recognized as an item that could match an empty string. This meant that an
413 enclosing group might also not be so recognized, causing infinite looping
414 (and probably a segfault) for patterns such as ^"((?(?=[a])[^"])|b)*"$
415 with the subject "ab", where knowledge that the repeated group can match
416 nothing is needed in order to break the loop.
417
418 20. If a pattern that was compiled with callouts was matched using pcre_dfa_
419 exec(), but without supplying a callout function, matching went wrong.
420
421 21. If PCRE_ERROR_MATCHLIMIT occurred during a recursion, there was a memory
422 leak if the size of the offset vector was greater than 30. When the vector
423 is smaller, the saved offsets during recursion go onto a local stack
424 vector, but for larger vectors malloc() is used. It was failing to free
425 when the recursion yielded PCRE_ERROR_MATCH_LIMIT (or any other "abnormal"
426 error, in fact).
427
428 22. There was a missing #ifdef SUPPORT_UTF8 round one of the variables in the
429 heapframe that is used only when UTF-8 support is enabled. This caused no
430 problem, but was untidy.
431
432 23. Steven Van Ingelgem's patch to CMakeLists.txt to change the name
433 CMAKE_BINARY_DIR to PROJECT_BINARY_DIR so that it works when PCRE is
434 included within another project.
435
436 24. Steven Van Ingelgem's patches to add more options to the CMake support,
437 slightly modified by me:
438
439 (a) PCRE_BUILD_TESTS can be set OFF not to build the tests, including
440 not building pcregrep.
441
442 (b) PCRE_BUILD_PCREGREP can be see OFF not to build pcregrep, but only
443 if PCRE_BUILD_TESTS is also set OFF, because the tests use pcregrep.
444
445 25. Forward references, both numeric and by name, in patterns that made use of
446 duplicate group numbers, could behave incorrectly or give incorrect errors,
447 because when scanning forward to find the reference group, PCRE was not
448 taking into account the duplicate group numbers. A pattern such as
449 ^X(?3)(a)(?|(b)|(q))(Y) is an example.
450
451 26. Changed a few more instances of "const unsigned char *" to USPTR, making
452 the feature of a custom pointer more persuasive (as requested by a user).
453
454 27. Wrapped the definitions of fileno and isatty for Windows, which appear in
455 pcretest.c, inside #ifndefs, because it seems they are sometimes already
456 pre-defined.
457
458 28. Added support for (*UTF8) at the start of a pattern.
459
460 29. Arrange for flags added by the "release type" setting in CMake to be shown
461 in the configuration summary.
462
463
464 Version 7.8 05-Sep-08
465 ---------------------
466
467 1. Replaced UCP searching code with optimized version as implemented for Ad
468 Muncher (http://www.admuncher.com/) by Peter Kankowski. This uses a two-
469 stage table and inline lookup instead of a function, giving speed ups of 2
470 to 5 times on some simple patterns that I tested. Permission was given to
471 distribute the MultiStage2.py script that generates the tables (it's not in
472 the tarball, but is in the Subversion repository).
473
474 2. Updated the Unicode datatables to Unicode 5.1.0. This adds yet more
475 scripts.
476
477 3. Change 12 for 7.7 introduced a bug in pcre_study() when a pattern contained
478 a group with a zero qualifier. The result of the study could be incorrect,
479 or the function might crash, depending on the pattern.
480
481 4. Caseless matching was not working for non-ASCII characters in back
482 references. For example, /(\x{de})\1/8i was not matching \x{de}\x{fe}.
483 It now works when Unicode Property Support is available.
484
485 5. In pcretest, an escape such as \x{de} in the data was always generating
486 a UTF-8 string, even in non-UTF-8 mode. Now it generates a single byte in
487 non-UTF-8 mode. If the value is greater than 255, it gives a warning about
488 truncation.
489
490 6. Minor bugfix in pcrecpp.cc (change "" == ... to NULL == ...).
491
492 7. Added two (int) casts to pcregrep when printing the difference of two
493 pointers, in case they are 64-bit values.
494
495 8. Added comments about Mac OS X stack usage to the pcrestack man page and to
496 test 2 if it fails.
497
498 9. Added PCRE_CALL_CONVENTION just before the names of all exported functions,
499 and a #define of that name to empty if it is not externally set. This is to
500 allow users of MSVC to set it if necessary.
501
502 10. The PCRE_EXP_DEFN macro which precedes exported functions was missing from
503 the convenience functions in the pcre_get.c source file.
504
505 11. An option change at the start of a pattern that had top-level alternatives
506 could cause overwriting and/or a crash. This command provoked a crash in
507 some environments:
508
509 printf "/(?i)[\xc3\xa9\xc3\xbd]|[\xc3\xa9\xc3\xbdA]/8\n" | pcretest
510
511 This potential security problem was recorded as CVE-2008-2371.
512
513 12. For a pattern where the match had to start at the beginning or immediately
514 after a newline (e.g /.*anything/ without the DOTALL flag), pcre_exec() and
515 pcre_dfa_exec() could read past the end of the passed subject if there was
516 no match. To help with detecting such bugs (e.g. with valgrind), I modified
517 pcretest so that it places the subject at the end of its malloc-ed buffer.
518
519 13. The change to pcretest in 12 above threw up a couple more cases when pcre_
520 exec() might read past the end of the data buffer in UTF-8 mode.
521
522 14. A similar bug to 7.3/2 existed when the PCRE_FIRSTLINE option was set and
523 the data contained the byte 0x85 as part of a UTF-8 character within its
524 first line. This applied both to normal and DFA matching.
525
526 15. Lazy qualifiers were not working in some cases in UTF-8 mode. For example,
527 /^[^d]*?$/8 failed to match "abc".
528
529 16. Added a missing copyright notice to pcrecpp_internal.h.
530
531 17. Make it more clear in the documentation that values returned from
532 pcre_exec() in ovector are byte offsets, not character counts.
533
534 18. Tidied a few places to stop certain compilers from issuing warnings.
535
536 19. Updated the Virtual Pascal + BCC files to compile the latest v7.7, as
537 supplied by Stefan Weber. I made a further small update for 7.8 because
538 there is a change of source arrangements: the pcre_searchfuncs.c module is
539 replaced by pcre_ucd.c.
540
541
542 Version 7.7 07-May-08
543 ---------------------
544
545 1. Applied Craig's patch to sort out a long long problem: "If we can't convert
546 a string to a long long, pretend we don't even have a long long." This is
547 done by checking for the strtoq, strtoll, and _strtoi64 functions.
548
549 2. Applied Craig's patch to pcrecpp.cc to restore ABI compatibility with
550 pre-7.6 versions, which defined a global no_arg variable instead of putting
551 it in the RE class. (See also #8 below.)
552
553 3. Remove a line of dead code, identified by coverity and reported by Nuno
554 Lopes.
555
556 4. Fixed two related pcregrep bugs involving -r with --include or --exclude:
557
558 (1) The include/exclude patterns were being applied to the whole pathnames
559 of files, instead of just to the final components.
560
561 (2) If there was more than one level of directory, the subdirectories were
562 skipped unless they satisfied the include/exclude conditions. This is
563 inconsistent with GNU grep (and could even be seen as contrary to the
564 pcregrep specification - which I improved to make it absolutely clear).
565 The action now is always to scan all levels of directory, and just
566 apply the include/exclude patterns to regular files.
567
568 5. Added the --include_dir and --exclude_dir patterns to pcregrep, and used
569 --exclude_dir in the tests to avoid scanning .svn directories.
570
571 6. Applied Craig's patch to the QuoteMeta function so that it escapes the
572 NUL character as backslash + 0 rather than backslash + NUL, because PCRE
573 doesn't support NULs in patterns.
574
575 7. Added some missing "const"s to declarations of static tables in
576 pcre_compile.c and pcre_dfa_exec.c.
577
578 8. Applied Craig's patch to pcrecpp.cc to fix a problem in OS X that was
579 caused by fix #2 above. (Subsequently also a second patch to fix the
580 first patch. And a third patch - this was a messy problem.)
581
582 9. Applied Craig's patch to remove the use of push_back().
583
584 10. Applied Alan Lehotsky's patch to add REG_STARTEND support to the POSIX
585 matching function regexec().
586
587 11. Added support for the Oniguruma syntax \g<name>, \g<n>, \g'name', \g'n',
588 which, however, unlike Perl's \g{...}, are subroutine calls, not back
589 references. PCRE supports relative numbers with this syntax (I don't think
590 Oniguruma does).
591
592 12. Previously, a group with a zero repeat such as (...){0} was completely
593 omitted from the compiled regex. However, this means that if the group
594 was called as a subroutine from elsewhere in the pattern, things went wrong
595 (an internal error was given). Such groups are now left in the compiled
596 pattern, with a new opcode that causes them to be skipped at execution
597 time.
598
599 13. Added the PCRE_JAVASCRIPT_COMPAT option. This makes the following changes
600 to the way PCRE behaves:
601
602 (a) A lone ] character is dis-allowed (Perl treats it as data).
603
604 (b) A back reference to an unmatched subpattern matches an empty string
605 (Perl fails the current match path).
606
607 (c) A data ] in a character class must be notated as \] because if the
608 first data character in a class is ], it defines an empty class. (In
609 Perl it is not possible to have an empty class.) The empty class []
610 never matches; it forces failure and is equivalent to (*FAIL) or (?!).
611 The negative empty class [^] matches any one character, independently
612 of the DOTALL setting.
613
614 14. A pattern such as /(?2)[]a()b](abc)/ which had a forward reference to a
615 non-existent subpattern following a character class starting with ']' and
616 containing () gave an internal compiling error instead of "reference to
617 non-existent subpattern". Fortunately, when the pattern did exist, the
618 compiled code was correct. (When scanning forwards to check for the
619 existencd of the subpattern, it was treating the data ']' as terminating
620 the class, so got the count wrong. When actually compiling, the reference
621 was subsequently set up correctly.)
622
623 15. The "always fail" assertion (?!) is optimzed to (*FAIL) by pcre_compile;
624 it was being rejected as not supported by pcre_dfa_exec(), even though
625 other assertions are supported. I have made pcre_dfa_exec() support
626 (*FAIL).
627
628 16. The implementation of 13c above involved the invention of a new opcode,
629 OP_ALLANY, which is like OP_ANY but doesn't check the /s flag. Since /s
630 cannot be changed at match time, I realized I could make a small
631 improvement to matching performance by compiling OP_ALLANY instead of
632 OP_ANY for "." when DOTALL was set, and then removing the runtime tests
633 on the OP_ANY path.
634
635 17. Compiling pcretest on Windows with readline support failed without the
636 following two fixes: (1) Make the unistd.h include conditional on
637 HAVE_UNISTD_H; (2) #define isatty and fileno as _isatty and _fileno.
638
639 18. Changed CMakeLists.txt and cmake/FindReadline.cmake to arrange for the
640 ncurses library to be included for pcretest when ReadLine support is
641 requested, but also to allow for it to be overridden. This patch came from
642 Daniel Bergström.
643
644 19. There was a typo in the file ucpinternal.h where f0_rangeflag was defined
645 as 0x00f00000 instead of 0x00800000. Luckily, this would not have caused
646 any errors with the current Unicode tables. Thanks to Peter Kankowski for
647 spotting this.
648
649
650 Version 7.6 28-Jan-08
651 ---------------------
652
653 1. A character class containing a very large number of characters with
654 codepoints greater than 255 (in UTF-8 mode, of course) caused a buffer
655 overflow.
656
657 2. Patch to cut out the "long long" test in pcrecpp_unittest when
658 HAVE_LONG_LONG is not defined.
659
660 3. Applied Christian Ehrlicher's patch to update the CMake build files to
661 bring them up to date and include new features. This patch includes:
662
663 - Fixed PH's badly added libz and libbz2 support.
664 - Fixed a problem with static linking.
665 - Added pcredemo. [But later removed - see 7 below.]
666 - Fixed dftables problem and added an option.
667 - Added a number of HAVE_XXX tests, including HAVE_WINDOWS_H and
668 HAVE_LONG_LONG.
669 - Added readline support for pcretest.
670 - Added an listing of the option settings after cmake has run.
671
672 4. A user submitted a patch to Makefile that makes it easy to create
673 "pcre.dll" under mingw when using Configure/Make. I added stuff to
674 Makefile.am that cause it to include this special target, without
675 affecting anything else. Note that the same mingw target plus all
676 the other distribution libraries and programs are now supported
677 when configuring with CMake (see 6 below) instead of with
678 Configure/Make.
679
680 5. Applied Craig's patch that moves no_arg into the RE class in the C++ code.
681 This is an attempt to solve the reported problem "pcrecpp::no_arg is not
682 exported in the Windows port". It has not yet been confirmed that the patch
683 solves the problem, but it does no harm.
684
685 6. Applied Sheri's patch to CMakeLists.txt to add NON_STANDARD_LIB_PREFIX and
686 NON_STANDARD_LIB_SUFFIX for dll names built with mingw when configured
687 with CMake, and also correct the comment about stack recursion.
688
689 7. Remove the automatic building of pcredemo from the ./configure system and
690 from CMakeLists.txt. The whole idea of pcredemo.c is that it is an example
691 of a program that users should build themselves after PCRE is installed, so
692 building it automatically is not really right. What is more, it gave
693 trouble in some build environments.
694
695 8. Further tidies to CMakeLists.txt from Sheri and Christian.
696
697
698 Version 7.5 10-Jan-08
699 ---------------------
700
701 1. Applied a patch from Craig: "This patch makes it possible to 'ignore'
702 values in parens when parsing an RE using the C++ wrapper."
703
704 2. Negative specials like \S did not work in character classes in UTF-8 mode.
705 Characters greater than 255 were excluded from the class instead of being
706 included.
707
708 3. The same bug as (2) above applied to negated POSIX classes such as
709 [:^space:].
710
711 4. PCRECPP_STATIC was referenced in pcrecpp_internal.h, but nowhere was it
712 defined or documented. It seems to have been a typo for PCRE_STATIC, so
713 I have changed it.
714
715 5. The construct (?&) was not diagnosed as a syntax error (it referenced the
716 first named subpattern) and a construct such as (?&a) would reference the
717 first named subpattern whose name started with "a" (in other words, the
718 length check was missing). Both these problems are fixed. "Subpattern name
719 expected" is now given for (?&) (a zero-length name), and this patch also
720 makes it give the same error for \k'' (previously it complained that that
721 was a reference to a non-existent subpattern).
722
723 6. The erroneous patterns (?+-a) and (?-+a) give different error messages;
724 this is right because (?- can be followed by option settings as well as by
725 digits. I have, however, made the messages clearer.
726
727 7. Patterns such as (?(1)a|b) (a pattern that contains fewer subpatterns
728 than the number used in the conditional) now cause a compile-time error.
729 This is actually not compatible with Perl, which accepts such patterns, but
730 treats the conditional as always being FALSE (as PCRE used to), but it
731 seems to me that giving a diagnostic is better.
732
733 8. Change "alphameric" to the more common word "alphanumeric" in comments
734 and messages.
735
736 9. Fix two occurrences of "backslash" in comments that should have been
737 "backspace".
738
739 10. Remove two redundant lines of code that can never be obeyed (their function
740 was moved elsewhere).
741
742 11. The program that makes PCRE's Unicode character property table had a bug
743 which caused it to generate incorrect table entries for sequences of
744 characters that have the same character type, but are in different scripts.
745 It amalgamated them into a single range, with the script of the first of
746 them. In other words, some characters were in the wrong script. There were
747 thirteen such cases, affecting characters in the following ranges:
748
749 U+002b0 - U+002c1
750 U+0060c - U+0060d
751 U+0061e - U+00612
752 U+0064b - U+0065e
753 U+0074d - U+0076d
754 U+01800 - U+01805
755 U+01d00 - U+01d77
756 U+01d9b - U+01dbf
757 U+0200b - U+0200f
758 U+030fc - U+030fe
759 U+03260 - U+0327f
760 U+0fb46 - U+0fbb1
761 U+10450 - U+1049d
762
763 12. The -o option (show only the matching part of a line) for pcregrep was not
764 compatible with GNU grep in that, if there was more than one match in a
765 line, it showed only the first of them. It now behaves in the same way as
766 GNU grep.
767
768 13. If the -o and -v options were combined for pcregrep, it printed a blank
769 line for every non-matching line. GNU grep prints nothing, and pcregrep now
770 does the same. The return code can be used to tell if there were any
771 non-matching lines.
772
773 14. Added --file-offsets and --line-offsets to pcregrep.
774
775 15. The pattern (?=something)(?R) was not being diagnosed as a potentially
776 infinitely looping recursion. The bug was that positive lookaheads were not
777 being skipped when checking for a possible empty match (negative lookaheads
778 and both kinds of lookbehind were skipped).
779
780 16. Fixed two typos in the Windows-only code in pcregrep.c, and moved the
781 inclusion of <windows.h> to before rather than after the definition of
782 INVALID_FILE_ATTRIBUTES (patch from David Byron).
783
784 17. Specifying a possessive quantifier with a specific limit for a Unicode
785 character property caused pcre_compile() to compile bad code, which led at
786 runtime to PCRE_ERROR_INTERNAL (-14). Examples of patterns that caused this
787 are: /\p{Zl}{2,3}+/8 and /\p{Cc}{2}+/8. It was the possessive "+" that
788 caused the error; without that there was no problem.
789
790 18. Added --enable-pcregrep-libz and --enable-pcregrep-libbz2.
791
792 19. Added --enable-pcretest-libreadline.
793
794 20. In pcrecpp.cc, the variable 'count' was incremented twice in
795 RE::GlobalReplace(). As a result, the number of replacements returned was
796 double what it should be. I removed one of the increments, but Craig sent a
797 later patch that removed the other one (the right fix) and added unit tests
798 that check the return values (which was not done before).
799
800 21. Several CMake things:
801
802 (1) Arranged that, when cmake is used on Unix, the libraries end up with
803 the names libpcre and libpcreposix, not just pcre and pcreposix.
804
805 (2) The above change means that pcretest and pcregrep are now correctly
806 linked with the newly-built libraries, not previously installed ones.
807
808 (3) Added PCRE_SUPPORT_LIBREADLINE, PCRE_SUPPORT_LIBZ, PCRE_SUPPORT_LIBBZ2.
809
810 22. In UTF-8 mode, with newline set to "any", a pattern such as .*a.*=.b.*
811 crashed when matching a string such as a\x{2029}b (note that \x{2029} is a
812 UTF-8 newline character). The key issue is that the pattern starts .*;
813 this means that the match must be either at the beginning, or after a
814 newline. The bug was in the code for advancing after a failed match and
815 checking that the new position followed a newline. It was not taking
816 account of UTF-8 characters correctly.
817
818 23. PCRE was behaving differently from Perl in the way it recognized POSIX
819 character classes. PCRE was not treating the sequence [:...:] as a
820 character class unless the ... were all letters. Perl, however, seems to
821 allow any characters between [: and :], though of course it rejects as
822 unknown any "names" that contain non-letters, because all the known class
823 names consist only of letters. Thus, Perl gives an error for [[:1234:]],
824 for example, whereas PCRE did not - it did not recognize a POSIX character
825 class. This seemed a bit dangerous, so the code has been changed to be
826 closer to Perl. The behaviour is not identical to Perl, because PCRE will
827 diagnose an unknown class for, for example, [[:l\ower:]] where Perl will
828 treat it as [[:lower:]]. However, PCRE does now give "unknown" errors where
829 Perl does, and where it didn't before.
830
831 24. Rewrite so as to remove the single use of %n from pcregrep because in some
832 Windows environments %n is disabled by default.
833
834
835 Version 7.4 21-Sep-07
836 ---------------------
837
838 1. Change 7.3/28 was implemented for classes by looking at the bitmap. This
839 means that a class such as [\s] counted as "explicit reference to CR or
840 LF". That isn't really right - the whole point of the change was to try to
841 help when there was an actual mention of one of the two characters. So now
842 the change happens only if \r or \n (or a literal CR or LF) character is
843 encountered.
844
845 2. The 32-bit options word was also used for 6 internal flags, but the numbers
846 of both had grown to the point where there were only 3 bits left.
847 Fortunately, there was spare space in the data structure, and so I have
848 moved the internal flags into a new 16-bit field to free up more option
849 bits.
850
851 3. The appearance of (?J) at the start of a pattern set the DUPNAMES option,
852 but did not set the internal JCHANGED flag - either of these is enough to
853 control the way the "get" function works - but the PCRE_INFO_JCHANGED
854 facility is supposed to tell if (?J) was ever used, so now (?J) at the
855 start sets both bits.
856
857 4. Added options (at build time, compile time, exec time) to change \R from
858 matching any Unicode line ending sequence to just matching CR, LF, or CRLF.
859
860 5. doc/pcresyntax.html was missing from the distribution.
861
862 6. Put back the definition of PCRE_ERROR_NULLWSLIMIT, for backward
863 compatibility, even though it is no longer used.
864
865 7. Added macro for snprintf to pcrecpp_unittest.cc and also for strtoll and
866 strtoull to pcrecpp.cc to select the available functions in WIN32 when the
867 windows.h file is present (where different names are used). [This was
868 reversed later after testing - see 16 below.]
869
870 8. Changed all #include <config.h> to #include "config.h". There were also
871 some further <pcre.h> cases that I changed to "pcre.h".
872
873 9. When pcregrep was used with the --colour option, it missed the line ending
874 sequence off the lines that it output.
875
876 10. It was pointed out to me that arrays of string pointers cause lots of
877 relocations when a shared library is dynamically loaded. A technique of
878 using a single long string with a table of offsets can drastically reduce
879 these. I have refactored PCRE in four places to do this. The result is
880 dramatic:
881
882 Originally: 290
883 After changing UCP table: 187
884 After changing error message table: 43
885 After changing table of "verbs" 36
886 After changing table of Posix names 22
887
888 Thanks to the folks working on Gregex for glib for this insight.
889
890 11. --disable-stack-for-recursion caused compiling to fail unless -enable-
891 unicode-properties was also set.
892
893 12. Updated the tests so that they work when \R is defaulted to ANYCRLF.
894
895 13. Added checks for ANY and ANYCRLF to pcrecpp.cc where it previously
896 checked only for CRLF.
897
898 14. Added casts to pcretest.c to avoid compiler warnings.
899
900 15. Added Craig's patch to various pcrecpp modules to avoid compiler warnings.
901
902 16. Added Craig's patch to remove the WINDOWS_H tests, that were not working,
903 and instead check for _strtoi64 explicitly, and avoid the use of snprintf()
904 entirely. This removes changes made in 7 above.
905
906 17. The CMake files have been updated, and there is now more information about
907 building with CMake in the NON-UNIX-USE document.
908
909
910 Version 7.3 28-Aug-07
911 ---------------------
912
913 1. In the rejigging of the build system that eventually resulted in 7.1, the
914 line "#include <pcre.h>" was included in pcre_internal.h. The use of angle
915 brackets there is not right, since it causes compilers to look for an
916 installed pcre.h, not the version that is in the source that is being
917 compiled (which of course may be different). I have changed it back to:
918
919 #include "pcre.h"
920
921 I have a vague recollection that the change was concerned with compiling in
922 different directories, but in the new build system, that is taken care of
923 by the VPATH setting the Makefile.
924
925 2. The pattern .*$ when run in not-DOTALL UTF-8 mode with newline=any failed
926 when the subject happened to end in the byte 0x85 (e.g. if the last
927 character was \x{1ec5}). *Character* 0x85 is one of the "any" newline
928 characters but of course it shouldn't be taken as a newline when it is part
929 of another character. The bug was that, for an unlimited repeat of . in
930 not-DOTALL UTF-8 mode, PCRE was advancing by bytes rather than by
931 characters when looking for a newline.
932
933 3. A small performance improvement in the DOTALL UTF-8 mode .* case.
934
935 4. Debugging: adjusted the names of opcodes for different kinds of parentheses
936 in debug output.
937
938 5. Arrange to use "%I64d" instead of "%lld" and "%I64u" instead of "%llu" for
939 long printing in the pcrecpp unittest when running under MinGW.
940
941 6. ESC_K was left out of the EBCDIC table.
942
943 7. Change 7.0/38 introduced a new limit on the number of nested non-capturing
944 parentheses; I made it 1000, which seemed large enough. Unfortunately, the
945 limit also applies to "virtual nesting" when a pattern is recursive, and in
946 this case 1000 isn't so big. I have been able to remove this limit at the
947 expense of backing off one optimization in certain circumstances. Normally,
948 when pcre_exec() would call its internal match() function recursively and
949 immediately return the result unconditionally, it uses a "tail recursion"
950 feature to save stack. However, when a subpattern that can match an empty
951 string has an unlimited repetition quantifier, it no longer makes this
952 optimization. That gives it a stack frame in which to save the data for
953 checking that an empty string has been matched. Previously this was taken
954 from the 1000-entry workspace that had been reserved. So now there is no
955 explicit limit, but more stack is used.
956
957 8. Applied Daniel's patches to solve problems with the import/export magic
958 syntax that is required for Windows, and which was going wrong for the
959 pcreposix and pcrecpp parts of the library. These were overlooked when this
960 problem was solved for the main library.
961
962 9. There were some crude static tests to avoid integer overflow when computing
963 the size of patterns that contain repeated groups with explicit upper
964 limits. As the maximum quantifier is 65535, the maximum group length was
965 set at 30,000 so that the product of these two numbers did not overflow a
966 32-bit integer. However, it turns out that people want to use groups that
967 are longer than 30,000 bytes (though not repeat them that many times).
968 Change 7.0/17 (the refactoring of the way the pattern size is computed) has
969 made it possible to implement the integer overflow checks in a much more
970 dynamic way, which I have now done. The artificial limitation on group
971 length has been removed - we now have only the limit on the total length of
972 the compiled pattern, which depends on the LINK_SIZE setting.
973
974 10. Fixed a bug in the documentation for get/copy named substring when
975 duplicate names are permitted. If none of the named substrings are set, the
976 functions return PCRE_ERROR_NOSUBSTRING (7); the doc said they returned an
977 empty string.
978
979 11. Because Perl interprets \Q...\E at a high level, and ignores orphan \E
980 instances, patterns such as [\Q\E] or [\E] or even [^\E] cause an error,
981 because the ] is interpreted as the first data character and the
982 terminating ] is not found. PCRE has been made compatible with Perl in this
983 regard. Previously, it interpreted [\Q\E] as an empty class, and [\E] could
984 cause memory overwriting.
985
986 10. Like Perl, PCRE automatically breaks an unlimited repeat after an empty
987 string has been matched (to stop an infinite loop). It was not recognizing
988 a conditional subpattern that could match an empty string if that
989 subpattern was within another subpattern. For example, it looped when
990 trying to match (((?(1)X|))*) but it was OK with ((?(1)X|)*) where the
991 condition was not nested. This bug has been fixed.
992
993 12. A pattern like \X?\d or \P{L}?\d in non-UTF-8 mode could cause a backtrack
994 past the start of the subject in the presence of bytes with the top bit
995 set, for example "\x8aBCD".
996
997 13. Added Perl 5.10 experimental backtracking controls (*FAIL), (*F), (*PRUNE),
998 (*SKIP), (*THEN), (*COMMIT), and (*ACCEPT).
999
1000 14. Optimized (?!) to (*FAIL).
1001
1002 15. Updated the test for a valid UTF-8 string to conform to the later RFC 3629.
1003 This restricts code points to be within the range 0 to 0x10FFFF, excluding
1004 the "low surrogate" sequence 0xD800 to 0xDFFF. Previously, PCRE allowed the
1005 full range 0 to 0x7FFFFFFF, as defined by RFC 2279. Internally, it still
1006 does: it's just the validity check that is more restrictive.
1007
1008 16. Inserted checks for integer overflows during escape sequence (backslash)
1009 processing, and also fixed erroneous offset values for syntax errors during
1010 backslash processing.
1011
1012 17. Fixed another case of looking too far back in non-UTF-8 mode (cf 12 above)
1013 for patterns like [\PPP\x8a]{1,}\x80 with the subject "A\x80".
1014
1015 18. An unterminated class in a pattern like (?1)\c[ with a "forward reference"
1016 caused an overrun.
1017
1018 19. A pattern like (?:[\PPa*]*){8,} which had an "extended class" (one with
1019 something other than just ASCII characters) inside a group that had an
1020 unlimited repeat caused a loop at compile time (while checking to see
1021 whether the group could match an empty string).
1022
1023 20. Debugging a pattern containing \p or \P could cause a crash. For example,
1024 [\P{Any}] did so. (Error in the code for printing property names.)
1025
1026 21. An orphan \E inside a character class could cause a crash.
1027
1028 22. A repeated capturing bracket such as (A)? could cause a wild memory
1029 reference during compilation.
1030
1031 23. There are several functions in pcre_compile() that scan along a compiled
1032 expression for various reasons (e.g. to see if it's fixed length for look
1033 behind). There were bugs in these functions when a repeated \p or \P was
1034 present in the pattern. These operators have additional parameters compared
1035 with \d, etc, and these were not being taken into account when moving along
1036 the compiled data. Specifically:
1037
1038 (a) A item such as \p{Yi}{3} in a lookbehind was not treated as fixed
1039 length.
1040
1041 (b) An item such as \pL+ within a repeated group could cause crashes or
1042 loops.
1043
1044 (c) A pattern such as \p{Yi}+(\P{Yi}+)(?1) could give an incorrect
1045 "reference to non-existent subpattern" error.
1046
1047 (d) A pattern like (\P{Yi}{2}\277)? could loop at compile time.
1048
1049 24. A repeated \S or \W in UTF-8 mode could give wrong answers when multibyte
1050 characters were involved (for example /\S{2}/8g with "A\x{a3}BC").
1051
1052 25. Using pcregrep in multiline, inverted mode (-Mv) caused it to loop.
1053
1054 26. Patterns such as [\P{Yi}A] which include \p or \P and just one other
1055 character were causing crashes (broken optimization).
1056
1057 27. Patterns such as (\P{Yi}*\277)* (group with possible zero repeat containing
1058 \p or \P) caused a compile-time loop.
1059
1060 28. More problems have arisen in unanchored patterns when CRLF is a valid line
1061 break. For example, the unstudied pattern [\r\n]A does not match the string
1062 "\r\nA" because change 7.0/46 below moves the current point on by two
1063 characters after failing to match at the start. However, the pattern \nA
1064 *does* match, because it doesn't start till \n, and if [\r\n]A is studied,
1065 the same is true. There doesn't seem any very clean way out of this, but
1066 what I have chosen to do makes the common cases work: PCRE now takes note
1067 of whether there can be an explicit match for \r or \n anywhere in the
1068 pattern, and if so, 7.0/46 no longer applies. As part of this change,
1069 there's a new PCRE_INFO_HASCRORLF option for finding out whether a compiled
1070 pattern has explicit CR or LF references.
1071
1072 29. Added (*CR) etc for changing newline setting at start of pattern.
1073
1074
1075 Version 7.2 19-Jun-07
1076 ---------------------
1077
1078 1. If the fr_FR locale cannot be found for test 3, try the "french" locale,
1079 which is apparently normally available under Windows.
1080
1081 2. Re-jig the pcregrep tests with different newline settings in an attempt
1082 to make them independent of the local environment's newline setting.
1083
1084 3. Add code to configure.ac to remove -g from the CFLAGS default settings.
1085
1086 4. Some of the "internals" tests were previously cut out when the link size
1087 was not 2, because the output contained actual offsets. The recent new
1088 "Z" feature of pcretest means that these can be cut out, making the tests
1089 usable with all link sizes.
1090
1091 5. Implemented Stan Switzer's goto replacement for longjmp() when not using
1092 stack recursion. This gives a massive performance boost under BSD, but just
1093 a small improvement under Linux. However, it saves one field in the frame
1094 in all cases.
1095
1096 6. Added more features from the forthcoming Perl 5.10:
1097
1098 (a) (?-n) (where n is a string of digits) is a relative subroutine or
1099 recursion call. It refers to the nth most recently opened parentheses.
1100
1101 (b) (?+n) is also a relative subroutine call; it refers to the nth next
1102 to be opened parentheses.
1103
1104 (c) Conditions that refer to capturing parentheses can be specified
1105 relatively, for example, (?(-2)... or (?(+3)...
1106
1107 (d) \K resets the start of the current match so that everything before
1108 is not part of it.
1109
1110 (e) \k{name} is synonymous with \k<name> and \k'name' (.NET compatible).
1111
1112 (f) \g{name} is another synonym - part of Perl 5.10's unification of
1113 reference syntax.
1114
1115 (g) (?| introduces a group in which the numbering of parentheses in each
1116 alternative starts with the same number.
1117
1118 (h) \h, \H, \v, and \V match horizontal and vertical whitespace.
1119
1120 7. Added two new calls to pcre_fullinfo(): PCRE_INFO_OKPARTIAL and
1121 PCRE_INFO_JCHANGED.
1122
1123 8. A pattern such as (.*(.)?)* caused pcre_exec() to fail by either not
1124 terminating or by crashing. Diagnosed by Viktor Griph; it was in the code
1125 for detecting groups that can match an empty string.
1126
1127 9. A pattern with a very large number of alternatives (more than several
1128 hundred) was running out of internal workspace during the pre-compile
1129 phase, where pcre_compile() figures out how much memory will be needed. A
1130 bit of new cunning has reduced the workspace needed for groups with
1131 alternatives. The 1000-alternative test pattern now uses 12 bytes of
1132 workspace instead of running out of the 4096 that are available.
1133
1134 10. Inserted some missing (unsigned int) casts to get rid of compiler warnings.
1135
1136 11. Applied patch from Google to remove an optimization that didn't quite work.
1137 The report of the bug said:
1138
1139 pcrecpp::RE("a*").FullMatch("aaa") matches, while
1140 pcrecpp::RE("a*?").FullMatch("aaa") does not, and
1141 pcrecpp::RE("a*?\\z").FullMatch("aaa") does again.
1142
1143 12. If \p or \P was used in non-UTF-8 mode on a character greater than 127
1144 it matched the wrong number of bytes.
1145
1146
1147 Version 7.1 24-Apr-07
1148 ---------------------
1149
1150 1. Applied Bob Rossi and Daniel G's patches to convert the build system to one
1151 that is more "standard", making use of automake and other Autotools. There
1152 is some re-arrangement of the files and adjustment of comments consequent
1153 on this.
1154
1155 2. Part of the patch fixed a problem with the pcregrep tests. The test of -r
1156 for recursive directory scanning broke on some systems because the files
1157 are not scanned in any specific order and on different systems the order
1158 was different. A call to "sort" has been inserted into RunGrepTest for the
1159 approprate test as a short-term fix. In the longer term there may be an
1160 alternative.
1161
1162 3. I had an email from Eric Raymond about problems translating some of PCRE's
1163 man pages to HTML (despite the fact that I distribute HTML pages, some
1164 people do their own conversions for various reasons). The problems
1165 concerned the use of low-level troff macros .br and .in. I have therefore
1166 removed all such uses from the man pages (some were redundant, some could
1167 be replaced by .nf/.fi pairs). The 132html script that I use to generate
1168 HTML has been updated to handle .nf/.fi and to complain if it encounters
1169 .br or .in.
1170
1171 4. Updated comments in configure.ac that get placed in config.h.in and also
1172 arranged for config.h to be included in the distribution, with the name
1173 config.h.generic, for the benefit of those who have to compile without
1174 Autotools (compare pcre.h, which is now distributed as pcre.h.generic).
1175
1176 5. Updated the support (such as it is) for Virtual Pascal, thanks to Stefan
1177 Weber: (1) pcre_internal.h was missing some function renames; (2) updated
1178 makevp.bat for the current PCRE, using the additional files
1179 makevp_c.txt, makevp_l.txt, and pcregexp.pas.
1180
1181 6. A Windows user reported a minor discrepancy with test 2, which turned out
1182 to be caused by a trailing space on an input line that had got lost in his
1183 copy. The trailing space was an accident, so I've just removed it.
1184
1185 7. Add -Wl,-R... flags in pcre-config.in for *BSD* systems, as I'm told
1186 that is needed.
1187
1188 8. Mark ucp_table (in ucptable.h) and ucp_gentype (in pcre_ucp_searchfuncs.c)
1189 as "const" (a) because they are and (b) because it helps the PHP
1190 maintainers who have recently made a script to detect big data structures
1191 in the php code that should be moved to the .rodata section. I remembered
1192 to update Builducptable as well, so it won't revert if ucptable.h is ever
1193 re-created.
1194
1195 9. Added some extra #ifdef SUPPORT_UTF8 conditionals into pcretest.c,
1196 pcre_printint.src, pcre_compile.c, pcre_study.c, and pcre_tables.c, in
1197 order to be able to cut out the UTF-8 tables in the latter when UTF-8
1198 support is not required. This saves 1.5-2K of code, which is important in
1199 some applications.
1200
1201 Later: more #ifdefs are needed in pcre_ord2utf8.c and pcre_valid_utf8.c
1202 so as not to refer to the tables, even though these functions will never be
1203 called when UTF-8 support is disabled. Otherwise there are problems with a
1204 shared library.
1205
1206 10. Fixed two bugs in the emulated memmove() function in pcre_internal.h:
1207
1208 (a) It was defining its arguments as char * instead of void *.
1209
1210 (b) It was assuming that all moves were upwards in memory; this was true
1211 a long time ago when I wrote it, but is no longer the case.
1212
1213 The emulated memove() is provided for those environments that have neither
1214 memmove() nor bcopy(). I didn't think anyone used it these days, but that
1215 is clearly not the case, as these two bugs were recently reported.
1216
1217 11. The script PrepareRelease is now distributed: it calls 132html, CleanTxt,
1218 and Detrail to create the HTML documentation, the .txt form of the man
1219 pages, and it removes trailing spaces from listed files. It also creates
1220 pcre.h.generic and config.h.generic from pcre.h and config.h. In the latter
1221 case, it wraps all the #defines with #ifndefs. This script should be run
1222 before "make dist".
1223
1224 12. Fixed two fairly obscure bugs concerned with quantified caseless matching
1225 with Unicode property support.
1226
1227 (a) For a maximizing quantifier, if the two different cases of the
1228 character were of different lengths in their UTF-8 codings (there are
1229 some cases like this - I found 11), and the matching function had to
1230 back up over a mixture of the two cases, it incorrectly assumed they
1231 were both the same length.
1232
1233 (b) When PCRE was configured to use the heap rather than the stack for
1234 recursion during matching, it was not correctly preserving the data for
1235 the other case of a UTF-8 character when checking ahead for a match
1236 while processing a minimizing repeat. If the check also involved
1237 matching a wide character, but failed, corruption could cause an
1238 erroneous result when trying to check for a repeat of the original
1239 character.
1240
1241 13. Some tidying changes to the testing mechanism:
1242
1243 (a) The RunTest script now detects the internal link size and whether there
1244 is UTF-8 and UCP support by running ./pcretest -C instead of relying on
1245 values substituted by "configure". (The RunGrepTest script already did
1246 this for UTF-8.) The configure.ac script no longer substitutes the
1247 relevant variables.
1248
1249 (b) The debugging options /B and /D in pcretest show the compiled bytecode
1250 with length and offset values. This means that the output is different
1251 for different internal link sizes. Test 2 is skipped for link sizes
1252 other than 2 because of this, bypassing the problem. Unfortunately,
1253 there was also a test in test 3 (the locale tests) that used /B and
1254 failed for link sizes other than 2. Rather than cut the whole test out,
1255 I have added a new /Z option to pcretest that replaces the length and
1256 offset values with spaces. This is now used to make test 3 independent
1257 of link size. (Test 2 will be tidied up later.)
1258
1259 14. If erroroffset was passed as NULL to pcre_compile, it provoked a
1260 segmentation fault instead of returning the appropriate error message.
1261
1262 15. In multiline mode when the newline sequence was set to "any", the pattern
1263 ^$ would give a match between the \r and \n of a subject such as "A\r\nB".
1264 This doesn't seem right; it now treats the CRLF combination as the line
1265 ending, and so does not match in that case. It's only a pattern such as ^$
1266 that would hit this one: something like ^ABC$ would have failed after \r
1267 and then tried again after \r\n.
1268
1269 16. Changed the comparison command for RunGrepTest from "diff -u" to "diff -ub"
1270 in an attempt to make files that differ only in their line terminators
1271 compare equal. This works on Linux.
1272
1273 17. Under certain error circumstances pcregrep might try to free random memory
1274 as it exited. This is now fixed, thanks to valgrind.
1275
1276 19. In pcretest, if the pattern /(?m)^$/g<any> was matched against the string
1277 "abc\r\n\r\n", it found an unwanted second match after the second \r. This
1278 was because its rules for how to advance for /g after matching an empty
1279 string at the end of a line did not allow for this case. They now check for
1280 it specially.
1281
1282 20. pcretest is supposed to handle patterns and data of any length, by
1283 extending its buffers when necessary. It was getting this wrong when the
1284 buffer for a data line had to be extended.
1285
1286 21. Added PCRE_NEWLINE_ANYCRLF which is like ANY, but matches only CR, LF, or
1287 CRLF as a newline sequence.
1288
1289 22. Code for handling Unicode properties in pcre_dfa_exec() wasn't being cut
1290 out by #ifdef SUPPORT_UCP. This did no harm, as it could never be used, but
1291 I have nevertheless tidied it up.
1292
1293 23. Added some casts to kill warnings from HP-UX ia64 compiler.
1294
1295 24. Added a man page for pcre-config.
1296
1297
1298 Version 7.0 19-Dec-06
1299 ---------------------
1300
1301 1. Fixed a signed/unsigned compiler warning in pcre_compile.c, shown up by
1302 moving to gcc 4.1.1.
1303
1304 2. The -S option for pcretest uses setrlimit(); I had omitted to #include
1305 sys/time.h, which is documented as needed for this function. It doesn't
1306 seem to matter on Linux, but it showed up on some releases of OS X.
1307
1308 3. It seems that there are systems where bytes whose values are greater than
1309 127 match isprint() in the "C" locale. The "C" locale should be the
1310 default when a C program starts up. In most systems, only ASCII printing
1311 characters match isprint(). This difference caused the output from pcretest
1312 to vary, making some of the tests fail. I have changed pcretest so that:
1313
1314 (a) When it is outputting text in the compiled version of a pattern, bytes
1315 other than 32-126 are always shown as hex escapes.
1316
1317 (b) When it is outputting text that is a matched part of a subject string,
1318 it does the same, unless a different locale has been set for the match
1319 (using the /L modifier). In this case, it uses isprint() to decide.
1320
1321 4. Fixed a major bug that caused incorrect computation of the amount of memory
1322 required for a compiled pattern when options that changed within the
1323 pattern affected the logic of the preliminary scan that determines the
1324 length. The relevant options are -x, and -i in UTF-8 mode. The result was
1325 that the computed length was too small. The symptoms of this bug were
1326 either the PCRE error "internal error: code overflow" from pcre_compile(),
1327 or a glibc crash with a message such as "pcretest: free(): invalid next
1328 size (fast)". Examples of patterns that provoked this bug (shown in
1329 pcretest format) are:
1330
1331 /(?-x: )/x
1332 /(?x)(?-x: \s*#\s*)/
1333 /((?i)[\x{c0}])/8
1334 /(?i:[\x{c0}])/8
1335
1336 HOWEVER: Change 17 below makes this fix obsolete as the memory computation
1337 is now done differently.
1338
1339 5. Applied patches from Google to: (a) add a QuoteMeta function to the C++
1340 wrapper classes; (b) implement a new function in the C++ scanner that is
1341 more efficient than the old way of doing things because it avoids levels of
1342 recursion in the regex matching; (c) add a paragraph to the documentation
1343 for the FullMatch() function.
1344
1345 6. The escape sequence \n was being treated as whatever was defined as
1346 "newline". Not only was this contrary to the documentation, which states
1347 that \n is character 10 (hex 0A), but it also went horribly wrong when
1348 "newline" was defined as CRLF. This has been fixed.
1349
1350 7. In pcre_dfa_exec.c the value of an unsigned integer (the variable called c)
1351 was being set to -1 for the "end of line" case (supposedly a value that no
1352 character can have). Though this value is never used (the check for end of
1353 line is "zero bytes in current character"), it caused compiler complaints.
1354 I've changed it to 0xffffffff.
1355
1356 8. In pcre_version.c, the version string was being built by a sequence of
1357 C macros that, in the event of PCRE_PRERELEASE being defined as an empty
1358 string (as it is for production releases) called a macro with an empty
1359 argument. The C standard says the result of this is undefined. The gcc
1360 compiler treats it as an empty string (which was what was wanted) but it is
1361 reported that Visual C gives an error. The source has been hacked around to
1362 avoid this problem.
1363
1364 9. On the advice of a Windows user, included <io.h> and <fcntl.h> in Windows
1365 builds of pcretest, and changed the call to _setmode() to use _O_BINARY
1366 instead of 0x8000. Made all the #ifdefs test both _WIN32 and WIN32 (not all
1367 of them did).
1368
1369 10. Originally, pcretest opened its input and output without "b"; then I was
1370 told that "b" was needed in some environments, so it was added for release
1371 5.0 to both the input and output. (It makes no difference on Unix-like
1372 systems.) Later I was told that it is wrong for the input on Windows. I've
1373 now abstracted the modes into two macros, to make it easier to fiddle with
1374 them, and removed "b" from the input mode under Windows.
1375
1376 11. Added pkgconfig support for the C++ wrapper library, libpcrecpp.
1377
1378 12. Added -help and --help to pcretest as an official way of being reminded
1379 of the options.
1380
1381 13. Removed some redundant semicolons after macro calls in pcrecpparg.h.in
1382 and pcrecpp.cc because they annoy compilers at high warning levels.
1383
1384 14. A bit of tidying/refactoring in pcre_exec.c in the main bumpalong loop.
1385
1386 15. Fixed an occurrence of == in configure.ac that should have been = (shell
1387 scripts are not C programs :-) and which was not noticed because it works
1388 on Linux.
1389
1390 16. pcretest is supposed to handle any length of pattern and data line (as one
1391 line or as a continued sequence of lines) by extending its input buffer if
1392 necessary. This feature was broken for very long pattern lines, leading to
1393 a string of junk being passed to pcre_compile() if the pattern was longer
1394 than about 50K.
1395
1396 17. I have done a major re-factoring of the way pcre_compile() computes the
1397 amount of memory needed for a compiled pattern. Previously, there was code
1398 that made a preliminary scan of the pattern in order to do this. That was
1399 OK when PCRE was new, but as the facilities have expanded, it has become
1400 harder and harder to keep it in step with the real compile phase, and there
1401 have been a number of bugs (see for example, 4 above). I have now found a
1402 cunning way of running the real compile function in a "fake" mode that
1403 enables it to compute how much memory it would need, while actually only
1404 ever using a few hundred bytes of working memory and without too many
1405 tests of the mode. This should make future maintenance and development
1406 easier. A side effect of this work is that the limit of 200 on the nesting
1407 depth of parentheses has been removed (though this was never a serious
1408 limitation, I suspect). However, there is a downside: pcre_compile() now
1409 runs more slowly than before (30% or more, depending on the pattern). I
1410 hope this isn't a big issue. There is no effect on runtime performance.
1411
1412 18. Fixed a minor bug in pcretest: if a pattern line was not terminated by a
1413 newline (only possible for the last line of a file) and it was a
1414 pattern that set a locale (followed by /Lsomething), pcretest crashed.
1415
1416 19. Added additional timing features to pcretest. (1) The -tm option now times
1417 matching only, not compiling. (2) Both -t and -tm can be followed, as a
1418 separate command line item, by a number that specifies the number of
1419 repeats to use when timing. The default is 50000; this gives better
1420 precision, but takes uncomfortably long for very large patterns.
1421
1422 20. Extended pcre_study() to be more clever in cases where a branch of a
1423 subpattern has no definite first character. For example, (a*|b*)[cd] would
1424 previously give no result from pcre_study(). Now it recognizes that the
1425 first character must be a, b, c, or d.
1426
1427 21. There was an incorrect error "recursive call could loop indefinitely" if
1428 a subpattern (or the entire pattern) that was being tested for matching an
1429 empty string contained only one non-empty item after a nested subpattern.
1430 For example, the pattern (?>\x{100}*)\d(?R) provoked this error
1431 incorrectly, because the \d was being skipped in the check.
1432
1433 22. The pcretest program now has a new pattern option /B and a command line
1434 option -b, which is equivalent to adding /B to every pattern. This causes
1435 it to show the compiled bytecode, without the additional information that
1436 -d shows. The effect of -d is now the same as -b with -i (and similarly, /D
1437 is the same as /B/I).
1438
1439 23. A new optimization is now able automatically to treat some sequences such
1440 as a*b as a*+b. More specifically, if something simple (such as a character
1441 or a simple class like \d) has an unlimited quantifier, and is followed by
1442 something that cannot possibly match the quantified thing, the quantifier
1443 is automatically "possessified".
1444
1445 24. A recursive reference to a subpattern whose number was greater than 39
1446 went wrong under certain circumstances in UTF-8 mode. This bug could also
1447 have affected the operation of pcre_study().
1448
1449 25. Realized that a little bit of performance could be had by replacing
1450 (c & 0xc0) == 0xc0 with c >= 0xc0 when processing UTF-8 characters.
1451
1452 26. Timing data from pcretest is now shown to 4 decimal places instead of 3.
1453
1454 27. Possessive quantifiers such as a++ were previously implemented by turning
1455 them into atomic groups such as ($>a+). Now they have their own opcodes,
1456 which improves performance. This includes the automatically created ones
1457 from 23 above.
1458
1459 28. A pattern such as (?=(\w+))\1: which simulates an atomic group using a
1460 lookahead was broken if it was not anchored. PCRE was mistakenly expecting
1461 the first matched character to be a colon. This applied both to named and
1462 numbered groups.
1463
1464 29. The ucpinternal.h header file was missing its idempotency #ifdef.
1465
1466 30. I was sent a "project" file called libpcre.a.dev which I understand makes
1467 building PCRE on Windows easier, so I have included it in the distribution.
1468
1469 31. There is now a check in pcretest against a ridiculously large number being
1470 returned by pcre_exec() or pcre_dfa_exec(). If this happens in a /g or /G
1471 loop, the loop is abandoned.
1472
1473 32. Forward references to subpatterns in conditions such as (?(2)...) where
1474 subpattern 2 is defined later cause pcre_compile() to search forwards in
1475 the pattern for the relevant set of parentheses. This search went wrong
1476 when there were unescaped parentheses in a character class, parentheses
1477 escaped with \Q...\E, or parentheses in a #-comment in /x mode.
1478
1479 33. "Subroutine" calls and backreferences were previously restricted to
1480 referencing subpatterns earlier in the regex. This restriction has now
1481 been removed.
1482
1483 34. Added a number of extra features that are going to be in Perl 5.10. On the
1484 whole, these are just syntactic alternatives for features that PCRE had
1485 previously implemented using the Python syntax or my own invention. The
1486 other formats are all retained for compatibility.
1487
1488 (a) Named groups can now be defined as (?<name>...) or (?'name'...) as well
1489 as (?P<name>...). The new forms, as well as being in Perl 5.10, are
1490 also .NET compatible.
1491
1492 (b) A recursion or subroutine call to a named group can now be defined as
1493 (?&name) as well as (?P>name).
1494
1495 (c) A backreference to a named group can now be defined as \k<name> or
1496 \k'name' as well as (?P=name). The new forms, as well as being in Perl
1497 5.10, are also .NET compatible.
1498
1499 (d) A conditional reference to a named group can now use the syntax
1500 (?(<name>) or (?('name') as well as (?(name).
1501
1502 (e) A "conditional group" of the form (?(DEFINE)...) can be used to define
1503 groups (named and numbered) that are never evaluated inline, but can be
1504 called as "subroutines" from elsewhere. In effect, the DEFINE condition
1505 is always false. There may be only one alternative in such a group.
1506
1507 (f) A test for recursion can be given as (?(R1).. or (?(R&name)... as well
1508 as the simple (?(R). The condition is true only if the most recent
1509 recursion is that of the given number or name. It does not search out
1510 through the entire recursion stack.
1511
1512 (g) The escape \gN or \g{N} has been added, where N is a positive or
1513 negative number, specifying an absolute or relative reference.
1514
1515 35. Tidied to get rid of some further signed/unsigned compiler warnings and
1516 some "unreachable code" warnings.
1517
1518 36. Updated the Unicode property tables to Unicode version 5.0.0. Amongst other
1519 things, this adds five new scripts.
1520
1521 37. Perl ignores orphaned \E escapes completely. PCRE now does the same.
1522 There were also incompatibilities regarding the handling of \Q..\E inside
1523 character classes, for example with patterns like [\Qa\E-\Qz\E] where the
1524 hyphen was adjacent to \Q or \E. I hope I've cleared all this up now.
1525
1526 38. Like Perl, PCRE detects when an indefinitely repeated parenthesized group
1527 matches an empty string, and forcibly breaks the loop. There were bugs in
1528 this code in non-simple cases. For a pattern such as ^(a()*)* matched
1529 against aaaa the result was just "a" rather than "aaaa", for example. Two
1530 separate and independent bugs (that affected different cases) have been
1531 fixed.
1532
1533 39. Refactored the code to abolish the use of different opcodes for small
1534 capturing bracket numbers. This is a tidy that I avoided doing when I
1535 removed the limit on the number of capturing brackets for 3.5 back in 2001.
1536 The new approach is not only tidier, it makes it possible to reduce the
1537 memory needed to fix the previous bug (38).
1538
1539 40. Implemented PCRE_NEWLINE_ANY to recognize any of the Unicode newline
1540 sequences (http://unicode.org/unicode/reports/tr18/) as "newline" when
1541 processing dot, circumflex, or dollar metacharacters, or #-comments in /x
1542 mode.
1543
1544 41. Add \R to match any Unicode newline sequence, as suggested in the Unicode
1545 report.
1546
1547 42. Applied patch, originally from Ari Pollak, modified by Google, to allow
1548 copy construction and assignment in the C++ wrapper.
1549
1550 43. Updated pcregrep to support "--newline=any". In the process, I fixed a
1551 couple of bugs that could have given wrong results in the "--newline=crlf"
1552 case.
1553
1554 44. Added a number of casts and did some reorganization of signed/unsigned int
1555 variables following suggestions from Dair Grant. Also renamed the variable
1556 "this" as "item" because it is a C++ keyword.
1557
1558 45. Arranged for dftables to add
1559
1560 #include "pcre_internal.h"
1561
1562 to pcre_chartables.c because without it, gcc 4.x may remove the array
1563 definition from the final binary if PCRE is built into a static library and
1564 dead code stripping is activated.
1565
1566 46. For an unanchored pattern, if a match attempt fails at the start of a
1567 newline sequence, and the newline setting is CRLF or ANY, and the next two
1568 characters are CRLF, advance by two characters instead of one.
1569
1570
1571 Version 6.7 04-Jul-06
1572 ---------------------
1573
1574 1. In order to handle tests when input lines are enormously long, pcretest has
1575 been re-factored so that it automatically extends its buffers when
1576 necessary. The code is crude, but this _is_ just a test program. The
1577 default size has been increased from 32K to 50K.
1578
1579 2. The code in pcre_study() was using the value of the re argument before
1580 testing it for NULL. (Of course, in any sensible call of the function, it
1581 won't be NULL.)
1582
1583 3. The memmove() emulation function in pcre_internal.h, which is used on
1584 systems that lack both memmove() and bcopy() - that is, hardly ever -
1585 was missing a "static" storage class specifier.
1586
1587 4. When UTF-8 mode was not set, PCRE looped when compiling certain patterns
1588 containing an extended class (one that cannot be represented by a bitmap
1589 because it contains high-valued characters or Unicode property items, e.g.
1590 [\pZ]). Almost always one would set UTF-8 mode when processing such a
1591 pattern, but PCRE should not loop if you do not (it no longer does).
1592 [Detail: two cases were found: (a) a repeated subpattern containing an
1593 extended class; (b) a recursive reference to a subpattern that followed a
1594 previous extended class. It wasn't skipping over the extended class
1595 correctly when UTF-8 mode was not set.]
1596
1597 5. A negated single-character class was not being recognized as fixed-length
1598 in lookbehind assertions such as (?<=[^f]), leading to an incorrect
1599 compile error "lookbehind assertion is not fixed length".
1600
1601 6. The RunPerlTest auxiliary script was showing an unexpected difference
1602 between PCRE and Perl for UTF-8 tests. It turns out that it is hard to
1603 write a Perl script that can interpret lines of an input file either as
1604 byte characters or as UTF-8, which is what "perltest" was being required to
1605 do for the non-UTF-8 and UTF-8 tests, respectively. Essentially what you
1606 can't do is switch easily at run time between having the "use utf8;" pragma
1607 or not. In the end, I fudged it by using the RunPerlTest script to insert
1608 "use utf8;" explicitly for the UTF-8 tests.
1609
1610 7. In multiline (/m) mode, PCRE was matching ^ after a terminating newline at
1611 the end of the subject string, contrary to the documentation and to what
1612 Perl does. This was true of both matching functions. Now it matches only at
1613 the start of the subject and immediately after *internal* newlines.
1614
1615 8. A call of pcre_fullinfo() from pcretest to get the option bits was passing
1616 a pointer to an int instead of a pointer to an unsigned long int. This
1617 caused problems on 64-bit systems.
1618
1619 9. Applied a patch from the folks at Google to pcrecpp.cc, to fix "another
1620 instance of the 'standard' template library not being so standard".
1621
1622 10. There was no check on the number of named subpatterns nor the maximum
1623 length of a subpattern name. The product of these values is used to compute
1624 the size of the memory block for a compiled pattern. By supplying a very
1625 long subpattern name and a large number of named subpatterns, the size
1626 computation could be caused to overflow. This is now prevented by limiting
1627 the length of names to 32 characters, and the number of named subpatterns
1628 to 10,000.
1629
1630 11. Subpatterns that are repeated with specific counts have to be replicated in
1631 the compiled pattern. The size of memory for this was computed from the
1632 length of the subpattern and the repeat count. The latter is limited to
1633 65535, but there was no limit on the former, meaning that integer overflow
1634 could in principle occur. The compiled length of a repeated subpattern is
1635 now limited to 30,000 bytes in order to prevent this.
1636
1637 12. Added the optional facility to have named substrings with the same name.
1638
1639 13. Added the ability to use a named substring as a condition, using the
1640 Python syntax: (?(name)yes|no). This overloads (?(R)... and names that
1641 are numbers (not recommended). Forward references are permitted.
1642
1643 14. Added forward references in named backreferences (if you see what I mean).
1644
1645 15. In UTF-8 mode, with the PCRE_DOTALL option set, a quantified dot in the
1646 pattern could run off the end of the subject. For example, the pattern
1647 "(?s)(.{1,5})"8 did this with the subject "ab".
1648
1649 16. If PCRE_DOTALL or PCRE_MULTILINE were set, pcre_dfa_exec() behaved as if
1650 PCRE_CASELESS was set when matching characters that were quantified with ?
1651 or *.
1652
1653 17. A character class other than a single negated character that had a minimum
1654 but no maximum quantifier - for example [ab]{6,} - was not handled
1655 correctly by pce_dfa_exec(). It would match only one character.
1656
1657 18. A valid (though odd) pattern that looked like a POSIX character
1658 class but used an invalid character after [ (for example [[,abc,]]) caused
1659 pcre_compile() to give the error "Failed: internal error: code overflow" or
1660 in some cases to crash with a glibc free() error. This could even happen if
1661 the pattern terminated after [[ but there just happened to be a sequence of
1662 letters, a binary zero, and a closing ] in the memory that followed.
1663
1664 19. Perl's treatment of octal escapes in the range \400 to \777 has changed
1665 over the years. Originally (before any Unicode support), just the bottom 8
1666 bits were taken. Thus, for example, \500 really meant \100. Nowadays the
1667 output from "man perlunicode" includes this:
1668
1669 The regular expression compiler produces polymorphic opcodes. That
1670 is, the pattern adapts to the data and automatically switches to
1671 the Unicode character scheme when presented with Unicode data--or
1672 instead uses a traditional byte scheme when presented with byte
1673 data.
1674
1675 Sadly, a wide octal escape does not cause a switch, and in a string with
1676 no other multibyte characters, these octal escapes are treated as before.
1677 Thus, in Perl, the pattern /\500/ actually matches \100 but the pattern
1678 /\500|\x{1ff}/ matches \500 or \777 because the whole thing is treated as a
1679 Unicode string.
1680
1681 I have not perpetrated such confusion in PCRE. Up till now, it took just
1682 the bottom 8 bits, as in old Perl. I have now made octal escapes with
1683 values greater than \377 illegal in non-UTF-8 mode. In UTF-8 mode they
1684 translate to the appropriate multibyte character.
1685
1686 29. Applied some refactoring to reduce the number of warnings from Microsoft
1687 and Borland compilers. This has included removing the fudge introduced
1688 seven years ago for the OS/2 compiler (see 2.02/2 below) because it caused
1689 a warning about an unused variable.
1690
1691 21. PCRE has not included VT (character 0x0b) in the set of whitespace
1692 characters since release 4.0, because Perl (from release 5.004) does not.
1693 [Or at least, is documented not to: some releases seem to be in conflict
1694 with the documentation.] However, when a pattern was studied with
1695 pcre_study() and all its branches started with \s, PCRE still included VT
1696 as a possible starting character. Of course, this did no harm; it just
1697 caused an unnecessary match attempt.
1698
1699 22. Removed a now-redundant internal flag bit that recorded the fact that case
1700 dependency changed within the pattern. This was once needed for "required
1701 byte" processing, but is no longer used. This recovers a now-scarce options
1702 bit. Also moved the least significant internal flag bit to the most-
1703 significant bit of the word, which was not previously used (hangover from
1704 the days when it was an int rather than a uint) to free up another bit for
1705 the future.
1706
1707 23. Added support for CRLF line endings as well as CR and LF. As well as the
1708 default being selectable at build time, it can now be changed at runtime
1709 via the PCRE_NEWLINE_xxx flags. There are now options for pcregrep to
1710 specify that it is scanning data with non-default line endings.
1711
1712 24. Changed the definition of CXXLINK to make it agree with the definition of
1713 LINK in the Makefile, by replacing LDFLAGS to CXXFLAGS.
1714
1715 25. Applied Ian Taylor's patches to avoid using another stack frame for tail
1716 recursions. This makes a big different to stack usage for some patterns.
1717
1718 26. If a subpattern containing a named recursion or subroutine reference such
1719 as (?P>B) was quantified, for example (xxx(?P>B)){3}, the calculation of
1720 the space required for the compiled pattern went wrong and gave too small a
1721 value. Depending on the environment, this could lead to "Failed: internal
1722 error: code overflow at offset 49" or "glibc detected double free or
1723 corruption" errors.
1724
1725 27. Applied patches from Google (a) to support the new newline modes and (b) to
1726 advance over multibyte UTF-8 characters in GlobalReplace.
1727
1728 28. Change free() to pcre_free() in pcredemo.c. Apparently this makes a
1729 difference for some implementation of PCRE in some Windows version.
1730
1731 29. Added some extra testing facilities to pcretest:
1732
1733 \q<number> in a data line sets the "match limit" value
1734 \Q<number> in a data line sets the "match recursion limt" value
1735 -S <number> sets the stack size, where <number> is in megabytes
1736
1737 The -S option isn't available for Windows.
1738
1739
1740 Version 6.6 06-Feb-06
1741 ---------------------
1742
1743 1. Change 16(a) for 6.5 broke things, because PCRE_DATA_SCOPE was not defined
1744 in pcreposix.h. I have copied the definition from pcre.h.
1745
1746 2. Change 25 for 6.5 broke compilation in a build directory out-of-tree
1747 because pcre.h is no longer a built file.
1748
1749 3. Added Jeff Friedl's additional debugging patches to pcregrep. These are
1750 not normally included in the compiled code.
1751
1752
1753 Version 6.5 01-Feb-06
1754 ---------------------
1755
1756 1. When using the partial match feature with pcre_dfa_exec(), it was not
1757 anchoring the second and subsequent partial matches at the new starting
1758 point. This could lead to incorrect results. For example, with the pattern
1759 /1234/, partially matching against "123" and then "a4" gave a match.
1760
1761 2. Changes to pcregrep:
1762
1763 (a) All non-match returns from pcre_exec() were being treated as failures
1764 to match the line. Now, unless the error is PCRE_ERROR_NOMATCH, an
1765 error message is output. Some extra information is given for the
1766 PCRE_ERROR_MATCHLIMIT and PCRE_ERROR_RECURSIONLIMIT errors, which are
1767 probably the only errors that are likely to be caused by users (by
1768 specifying a regex that has nested indefinite repeats, for instance).
1769 If there are more than 20 of these errors, pcregrep is abandoned.
1770
1771 (b) A binary zero was treated as data while matching, but terminated the
1772 output line if it was written out. This has been fixed: binary zeroes
1773 are now no different to any other data bytes.
1774
1775 (c) Whichever of the LC_ALL or LC_CTYPE environment variables is set is
1776 used to set a locale for matching. The --locale=xxxx long option has
1777 been added (no short equivalent) to specify a locale explicitly on the
1778 pcregrep command, overriding the environment variables.
1779
1780 (d) When -B was used with -n, some line numbers in the output were one less
1781 than they should have been.
1782
1783 (e) Added the -o (--only-matching) option.
1784
1785 (f) If -A or -C was used with -c (count only), some lines of context were
1786 accidentally printed for the final match.
1787
1788 (g) Added the -H (--with-filename) option.
1789
1790 (h) The combination of options -rh failed to suppress file names for files
1791 that were found from directory arguments.
1792
1793 (i) Added the -D (--devices) and -d (--directories) options.
1794
1795 (j) Added the -F (--fixed-strings) option.
1796
1797 (k) Allow "-" to be used as a file name for -f as well as for a data file.
1798
1799 (l) Added the --colo(u)r option.
1800
1801 (m) Added Jeffrey Friedl's -S testing option, but within #ifdefs so that it
1802 is not present by default.
1803
1804 3. A nasty bug was discovered in the handling of recursive patterns, that is,
1805 items such as (?R) or (?1), when the recursion could match a number of
1806 alternatives. If it matched one of the alternatives, but subsequently,
1807 outside the recursion, there was a failure, the code tried to back up into
1808 the recursion. However, because of the way PCRE is implemented, this is not
1809 possible, and the result was an incorrect result from the match.
1810
1811 In order to prevent this happening, the specification of recursion has
1812 been changed so that all such subpatterns are automatically treated as
1813 atomic groups. Thus, for example, (?R) is treated as if it were (?>(?R)).
1814
1815 4. I had overlooked the fact that, in some locales, there are characters for
1816 which isalpha() is true but neither isupper() nor islower() are true. In
1817 the fr_FR locale, for instance, the \xAA and \xBA characters (ordmasculine
1818 and ordfeminine) are like this. This affected the treatment of \w and \W
1819 when they appeared in character classes, but not when they appeared outside
1820 a character class. The bit map for "word" characters is now created
1821 separately from the results of isalnum() instead of just taking it from the
1822 upper, lower, and digit maps. (Plus the underscore character, of course.)
1823
1824 5. The above bug also affected the handling of POSIX character classes such as
1825 [[:alpha:]] and [[:alnum:]]. These do not have their own bit maps in PCRE's
1826 permanent tables. Instead, the bit maps for such a class were previously
1827 created as the appropriate unions of the upper, lower, and digit bitmaps.
1828 Now they are created by subtraction from the [[:word:]] class, which has
1829 its own bitmap.
1830
1831 6. The [[:blank:]] character class matches horizontal, but not vertical space.
1832 It is created by subtracting the vertical space characters (\x09, \x0a,
1833 \x0b, \x0c) from the [[:space:]] bitmap. Previously, however, the
1834 subtraction was done in the overall bitmap for a character class, meaning
1835 that a class such as [\x0c[:blank:]] was incorrect because \x0c would not
1836 be recognized. This bug has been fixed.
1837
1838 7. Patches from the folks at Google:
1839
1840 (a) pcrecpp.cc: "to handle a corner case that may or may not happen in
1841 real life, but is still worth protecting against".
1842
1843 (b) pcrecpp.cc: "corrects a bug when negative radixes are used with
1844 regular expressions".
1845
1846 (c) pcre_scanner.cc: avoid use of std::count() because not all systems
1847 have it.
1848
1849 (d) Split off pcrecpparg.h from pcrecpp.h and had the former built by
1850 "configure" and the latter not, in order to fix a problem somebody had
1851 with compiling the Arg class on HP-UX.
1852
1853 (e) Improve the error-handling of the C++ wrapper a little bit.
1854
1855 (f) New tests for checking recursion limiting.
1856
1857 8. The pcre_memmove() function, which is used only if the environment does not
1858 have a standard memmove() function (and is therefore rarely compiled),
1859 contained two bugs: (a) use of int instead of size_t, and (b) it was not
1860 returning a result (though PCRE never actually uses the result).
1861
1862 9. In the POSIX regexec() interface, if nmatch is specified as a ridiculously
1863 large number - greater than INT_MAX/(3*sizeof(int)) - REG_ESPACE is
1864 returned instead of calling malloc() with an overflowing number that would
1865 most likely cause subsequent chaos.
1866
1867 10. The debugging option of pcretest was not showing the NO_AUTO_CAPTURE flag.
1868
1869 11. The POSIX flag REG_NOSUB is now supported. When a pattern that was compiled
1870 with this option is matched, the nmatch and pmatch options of regexec() are
1871 ignored.
1872
1873 12. Added REG_UTF8 to the POSIX interface. This is not defined by POSIX, but is
1874 provided in case anyone wants to the the POSIX interface with UTF-8
1875 strings.
1876
1877 13. Added CXXLDFLAGS to the Makefile parameters to provide settings only on the
1878 C++ linking (needed for some HP-UX environments).
1879
1880 14. Avoid compiler warnings in get_ucpname() when compiled without UCP support
1881 (unused parameter) and in the pcre_printint() function (omitted "default"
1882 switch label when the default is to do nothing).
1883
1884 15. Added some code to make it possible, when PCRE is compiled as a C++
1885 library, to replace subject pointers for pcre_exec() with a smart pointer
1886 class, thus making it possible to process discontinuous strings.
1887
1888 16. The two macros PCRE_EXPORT and PCRE_DATA_SCOPE are confusing, and perform
1889 much the same function. They were added by different people who were trying
1890 to make PCRE easy to compile on non-Unix systems. It has been suggested
1891 that PCRE_EXPORT be abolished now that there is more automatic apparatus
1892 for compiling on Windows systems. I have therefore replaced it with
1893 PCRE_DATA_SCOPE. This is set automatically for Windows; if not set it
1894 defaults to "extern" for C or "extern C" for C++, which works fine on
1895 Unix-like systems. It is now possible to override the value of PCRE_DATA_
1896 SCOPE with something explicit in config.h. In addition:
1897
1898 (a) pcreposix.h still had just "extern" instead of either of these macros;
1899 I have replaced it with PCRE_DATA_SCOPE.
1900
1901 (b) Functions such as _pcre_xclass(), which are internal to the library,
1902 but external in the C sense, all had PCRE_EXPORT in their definitions.
1903 This is apparently wrong for the Windows case, so I have removed it.
1904 (It makes no difference on Unix-like systems.)
1905
1906 17. Added a new limit, MATCH_LIMIT_RECURSION, which limits the depth of nesting
1907 of recursive calls to match(). This is different to MATCH_LIMIT because
1908 that limits the total number of calls to match(), not all of which increase
1909 the depth of recursion. Limiting the recursion depth limits the amount of
1910 stack (or heap if NO_RECURSE is set) that is used. The default can be set
1911 when PCRE is compiled, and changed at run time. A patch from Google adds
1912 this functionality to the C++ interface.
1913
1914 18. Changes to the handling of Unicode character properties:
1915
1916 (a) Updated the table to Unicode 4.1.0.
1917
1918 (b) Recognize characters that are not in the table as "Cn" (undefined).
1919
1920 (c) I revised the way the table is implemented to a much improved format
1921 which includes recognition of ranges. It now supports the ranges that
1922 are defined in UnicodeData.txt, and it also amalgamates other
1923 characters into ranges. This has reduced the number of entries in the
1924 table from around 16,000 to around 3,000, thus reducing its size
1925 considerably. I realized I did not need to use a tree structure after
1926 all - a binary chop search is just as efficient. Having reduced the
1927 number of entries, I extended their size from 6 bytes to 8 bytes to
1928 allow for more data.
1929
1930 (d) Added support for Unicode script names via properties such as \p{Han}.
1931
1932 19. In UTF-8 mode, a backslash followed by a non-Ascii character was not
1933 matching that character.
1934
1935 20. When matching a repeated Unicode property with a minimum greater than zero,
1936 (for example \pL{2,}), PCRE could look past the end of the subject if it
1937 reached it while seeking the minimum number of characters. This could
1938 happen only if some of the characters were more than one byte long, because
1939 there is a check for at least the minimum number of bytes.
1940
1941 21. Refactored the implementation of \p and \P so as to be more general, to
1942 allow for more different types of property in future. This has changed the
1943 compiled form incompatibly. Anybody with saved compiled patterns that use
1944 \p or \P will have to recompile them.
1945
1946 22. Added "Any" and "L&" to the supported property types.
1947
1948 23. Recognize \x{...} as a code point specifier, even when not in UTF-8 mode,
1949 but give a compile time error if the value is greater than 0xff.
1950
1951 24. The man pages for pcrepartial, pcreprecompile, and pcre_compile2 were
1952 accidentally not being installed or uninstalled.
1953
1954 25. The pcre.h file was built from pcre.h.in, but the only changes that were
1955 made were to insert the current release number. This seemed silly, because
1956 it made things harder for people building PCRE on systems that don't run
1957 "configure". I have turned pcre.h into a distributed file, no longer built
1958 by "configure", with the version identification directly included. There is
1959 no longer a pcre.h.in file.
1960
1961 However, this change necessitated a change to the pcre-config script as
1962 well. It is built from pcre-config.in, and one of the substitutions was the
1963 release number. I have updated configure.ac so that ./configure now finds
1964 the release number by grepping pcre.h.
1965
1966 26. Added the ability to run the tests under valgrind.
1967
1968
1969 Version 6.4 05-Sep-05
1970 ---------------------
1971
1972 1. Change 6.0/10/(l) to pcregrep introduced a bug that caused separator lines
1973 "--" to be printed when multiple files were scanned, even when none of the
1974 -A, -B, or -C options were used. This is not compatible with Gnu grep, so I
1975 consider it to be a bug, and have restored the previous behaviour.
1976
1977 2. A couple of code tidies to get rid of compiler warnings.
1978
1979 3. The pcretest program used to cheat by referring to symbols in the library
1980 whose names begin with _pcre_. These are internal symbols that are not
1981 really supposed to be visible externally, and in some environments it is
1982 possible to suppress them. The cheating is now confined to including
1983 certain files from the library's source, which is a bit cleaner.
1984
1985 4. Renamed pcre.in as pcre.h.in to go with pcrecpp.h.in; it also makes the
1986 file's purpose clearer.
1987
1988 5. Reorganized pcre_ucp_findchar().
1989
1990
1991 Version 6.3 15-Aug-05
1992 ---------------------
1993
1994 1. The file libpcre.pc.in did not have general read permission in the tarball.
1995
1996 2. There were some problems when building without C++ support:
1997
1998 (a) If C++ support was not built, "make install" and "make test" still
1999 tried to test it.
2000
2001 (b) There were problems when the value of CXX was explicitly set. Some
2002 changes have been made to try to fix these, and ...
2003
2004 (c) --disable-cpp can now be used to explicitly disable C++ support.
2005
2006 (d) The use of @CPP_OBJ@ directly caused a blank line preceded by a
2007 backslash in a target when C++ was disabled. This confuses some
2008 versions of "make", apparently. Using an intermediate variable solves
2009 this. (Same for CPP_LOBJ.)
2010
2011 3. $(LINK_FOR_BUILD) now includes $(CFLAGS_FOR_BUILD) and $(LINK)
2012 (non-Windows) now includes $(CFLAGS) because these flags are sometimes
2013 necessary on certain architectures.
2014
2015 4. Added a setting of -export-symbols-regex to the link command to remove
2016 those symbols that are exported in the C sense, but actually are local
2017 within the library, and not documented. Their names all begin with
2018 "_pcre_". This is not a perfect job, because (a) we have to except some
2019 symbols that pcretest ("illegally") uses, and (b) the facility isn't always
2020 available (and never for static libraries). I have made a note to try to
2021 find a way round (a) in the future.
2022
2023
2024 Version 6.2 01-Aug-05
2025 ---------------------
2026
2027 1. There was no test for integer overflow of quantifier values. A construction
2028 such as {1111111111111111} would give undefined results. What is worse, if
2029 a minimum quantifier for a parenthesized subpattern overflowed and became
2030 negative, the calculation of the memory size went wrong. This could have
2031 led to memory overwriting.
2032
2033 2. Building PCRE using VPATH was broken. Hopefully it is now fixed.
2034
2035 3. Added "b" to the 2nd argument of fopen() in dftables.c, for non-Unix-like
2036 operating environments where this matters.
2037
2038 4. Applied Giuseppe Maxia's patch to add additional features for controlling
2039 PCRE options from within the C++ wrapper.
2040
2041 5. Named capturing subpatterns were not being correctly counted when a pattern
2042 was compiled. This caused two problems: (a) If there were more than 100
2043 such subpatterns, the calculation of the memory needed for the whole
2044 compiled pattern went wrong, leading to an overflow error. (b) Numerical
2045 back references of the form \12, where the number was greater than 9, were
2046 not recognized as back references, even though there were sufficient
2047 previous subpatterns.
2048
2049 6. Two minor patches to pcrecpp.cc in order to allow it to compile on older
2050 versions of gcc, e.g. 2.95.4.
2051
2052
2053 Version 6.1 21-Jun-05
2054 ---------------------
2055
2056 1. There was one reference to the variable "posix" in pcretest.c that was not
2057 surrounded by "#if !defined NOPOSIX".
2058
2059 2. Make it possible to compile pcretest without DFA support, UTF8 support, or
2060 the cross-check on the old pcre_info() function, for the benefit of the
2061 cut-down version of PCRE that is currently imported into Exim.
2062
2063 3. A (silly) pattern starting with (?i)(?-i) caused an internal space
2064 allocation error. I've done the easy fix, which wastes 2 bytes for sensible
2065 patterns that start (?i) but I don't think that matters. The use of (?i) is
2066 just an example; this all applies to the other options as well.
2067
2068 4. Since libtool seems to echo the compile commands it is issuing, the output
2069 from "make" can be reduced a bit by putting "@" in front of each libtool
2070 compile command.
2071
2072 5. Patch from the folks at Google for configure.in to be a bit more thorough
2073 in checking for a suitable C++ installation before trying to compile the
2074 C++ stuff. This should fix a reported problem when a compiler was present,
2075 but no suitable headers.
2076
2077 6. The man pages all had just "PCRE" as their title. I have changed them to
2078 be the relevant file name. I have also arranged that these names are
2079 retained in the file doc/pcre.txt, which is a concatenation in text format
2080 of all the man pages except the little individual ones for each function.
2081
2082 7. The NON-UNIX-USE file had not been updated for the different set of source
2083 files that come with release 6. I also added a few comments about the C++
2084 wrapper.
2085
2086
2087 Version 6.0 07-Jun-05
2088 ---------------------
2089
2090 1. Some minor internal re-organization to help with my DFA experiments.
2091
2092 2. Some missing #ifdef SUPPORT_UCP conditionals in pcretest and printint that
2093 didn't matter for the library itself when fully configured, but did matter
2094 when compiling without UCP support, or within Exim, where the ucp files are
2095 not imported.
2096
2097 3. Refactoring of the library code to split up the various functions into
2098 different source modules. The addition of the new DFA matching code (see
2099 below) to a single monolithic source would have made it really too
2100 unwieldy, quite apart from causing all the code to be include in a
2101 statically linked application, when only some functions are used. This is
2102 relevant even without the DFA addition now that patterns can be compiled in
2103 one application and matched in another.
2104
2105 The downside of splitting up is that there have to be some external
2106 functions and data tables that are used internally in different modules of
2107 the library but which are not part of the API. These have all had their
2108 names changed to start with "_pcre_" so that they are unlikely to clash
2109 with other external names.
2110
2111 4. Added an alternate matching function, pcre_dfa_exec(), which matches using
2112 a different (DFA) algorithm. Although it is slower than the original
2113 function, it does have some advantages for certain types of matching
2114 problem.
2115
2116 5. Upgrades to pcretest in order to test the features of pcre_dfa_exec(),
2117 including restarting after a partial match.
2118
2119 6. A patch for pcregrep that defines INVALID_FILE_ATTRIBUTES if it is not
2120 defined when compiling for Windows was sent to me. I have put it into the
2121 code, though I have no means of testing or verifying it.
2122
2123 7. Added the pcre_refcount() auxiliary function.
2124
2125 8. Added the PCRE_FIRSTLINE option. This constrains an unanchored pattern to
2126 match before or at the first newline in the subject string. In pcretest,
2127 the /f option on a pattern can be used to set this.
2128
2129 9. A repeated \w when used in UTF-8 mode with characters greater than 256
2130 would behave wrongly. This has been present in PCRE since release 4.0.
2131
2132 10. A number of changes to the pcregrep command:
2133
2134 (a) Refactored how -x works; insert ^(...)$ instead of setting
2135 PCRE_ANCHORED and checking the length, in preparation for adding
2136 something similar for -w.
2137
2138 (b) Added the -w (match as a word) option.
2139
2140 (c) Refactored the way lines are read and buffered so as to have more
2141 than one at a time available.
2142
2143 (d) Implemented a pcregrep test script.
2144
2145 (e) Added the -M (multiline match) option. This allows patterns to match
2146 over several lines of the subject. The buffering ensures that at least
2147 8K, or the rest of the document (whichever is the shorter) is available
2148 for matching (and similarly the previous 8K for lookbehind assertions).
2149
2150 (f) Changed the --help output so that it now says
2151
2152 -w, --word-regex(p)
2153
2154 instead of two lines, one with "regex" and the other with "regexp"
2155 because that confused at least one person since the short forms are the
2156 same. (This required a bit of code, as the output is generated
2157 automatically from a table. It wasn't just a text change.)
2158
2159 (g) -- can be used to terminate pcregrep options if the next thing isn't an
2160 option but starts with a hyphen. Could be a pattern or a path name
2161 starting with a hyphen, for instance.
2162
2163 (h) "-" can be given as a file name to represent stdin.
2164
2165 (i) When file names are being printed, "(standard input)" is used for
2166 the standard input, for compatibility with GNU grep. Previously
2167 "<stdin>" was used.
2168
2169 (j) The option --label=xxx can be used to supply a name to be used for
2170 stdin when file names are being printed. There is no short form.
2171
2172 (k) Re-factored the options decoding logic because we are going to add
2173 two more options that take data. Such options can now be given in four
2174 different ways, e.g. "-fname", "-f name", "--file=name", "--file name".
2175
2176 (l) Added the -A, -B, and -C options for requesting that lines of context
2177 around matches be printed.
2178
2179 (m) Added the -L option to print the names of files that do not contain
2180 any matching lines, that is, the complement of -l.
2181
2182 (n) The return code is 2 if any file cannot be opened, but pcregrep does
2183 continue to scan other files.
2184
2185 (o) The -s option was incorrectly implemented. For compatibility with other
2186 greps, it now suppresses the error message for a non-existent or non-
2187 accessible file (but not the return code). There is a new option called
2188 -q that suppresses the output of matching lines, which was what -s was
2189 previously doing.
2190
2191 (p) Added --include and --exclude options to specify files for inclusion
2192 and exclusion when recursing.
2193
2194 11. The Makefile was not using the Autoconf-supported LDFLAGS macro properly.
2195 Hopefully, it now does.
2196
2197 12. Missing cast in pcre_study().
2198
2199 13. Added an "uninstall" target to the makefile.
2200
2201 14. Replaced "extern" in the function prototypes in Makefile.in with
2202 "PCRE_DATA_SCOPE", which defaults to 'extern' or 'extern "C"' in the Unix
2203 world, but is set differently for Windows.
2204
2205 15. Added a second compiling function called pcre_compile2(). The only
2206 difference is that it has an extra argument, which is a pointer to an
2207 integer error code. When there is a compile-time failure, this is set
2208 non-zero, in addition to the error test pointer being set to point to an
2209 error message. The new argument may be NULL if no error number is required
2210 (but then you may as well call pcre_compile(), which is now just a
2211 wrapper). This facility is provided because some applications need a
2212 numeric error indication, but it has also enabled me to tidy up the way
2213 compile-time errors are handled in the POSIX wrapper.
2214
2215 16. Added VPATH=.libs to the makefile; this should help when building with one
2216 prefix path and installing with another. (Or so I'm told by someone who
2217 knows more about this stuff than I do.)
2218
2219 17. Added a new option, REG_DOTALL, to the POSIX function regcomp(). This
2220 passes PCRE_DOTALL to the pcre_compile() function, making the "." character
2221 match everything, including newlines. This is not POSIX-compatible, but
2222 somebody wanted the feature. From pcretest it can be activated by using
2223 both the P and the s flags.
2224
2225 18. AC_PROG_LIBTOOL appeared twice in Makefile.in. Removed one.
2226
2227 19. libpcre.pc was being incorrectly installed as executable.
2228
2229 20. A couple of places in pcretest check for end-of-line by looking for '\n';
2230 it now also looks for '\r' so that it will work unmodified on Windows.
2231
2232 21. Added Google's contributed C++ wrapper to the distribution.
2233
2234 22. Added some untidy missing memory free() calls in pcretest, to keep
2235 Electric Fence happy when testing.
2236
2237
2238
2239 Version 5.0 13-Sep-04
2240 ---------------------
2241
2242 1. Internal change: literal characters are no longer packed up into items
2243 containing multiple characters in a single byte-string. Each character
2244 is now matched using a separate opcode. However, there may be more than one
2245 byte in the character in UTF-8 mode.
2246
2247 2. The pcre_callout_block structure has two new fields: pattern_position and
2248 next_item_length. These contain the offset in the pattern to the next match
2249 item, and its length, respectively.
2250
2251 3. The PCRE_AUTO_CALLOUT option for pcre_compile() requests the automatic
2252 insertion of callouts before each pattern item. Added the /C option to
2253 pcretest to make use of this.
2254
2255 4. On the advice of a Windows user, the lines
2256
2257 #if defined(_WIN32) || defined(WIN32)
2258 _setmode( _fileno( stdout ), 0x8000 );
2259 #endif /* defined(_WIN32) || defined(WIN32) */
2260
2261 have been added to the source of pcretest. This apparently does useful
2262 magic in relation to line terminators.
2263
2264 5. Changed "r" and "w" in the calls to fopen() in pcretest to "rb" and "wb"
2265 for the benefit of those environments where the "b" makes a difference.
2266
2267 6. The icc compiler has the same options as gcc, but "configure" doesn't seem
2268 to know about it. I have put a hack into configure.in that adds in code
2269 to set GCC=yes if CC=icc. This seems to end up at a point in the
2270 generated configure script that is early enough to affect the setting of
2271 compiler options, which is what is needed, but I have no means of testing
2272 whether it really works. (The user who reported this had patched the
2273 generated configure script, which of course I cannot do.)
2274
2275 LATER: After change 22 below (new libtool files), the configure script
2276 seems to know about icc (and also ecc). Therefore, I have commented out
2277 this hack in configure.in.
2278
2279 7. Added support for pkg-config (2 patches were sent in).
2280
2281 8. Negated POSIX character classes that used a combination of internal tables
2282 were completely broken. These were [[:^alpha:]], [[:^alnum:]], and
2283 [[:^ascii]]. Typically, they would match almost any characters. The other
2284 POSIX classes were not broken in this way.
2285
2286 9. Matching the pattern "\b.*?" against "ab cd", starting at offset 1, failed
2287 to find the match, as PCRE was deluded into thinking that the match had to
2288 start at the start point or following a newline. The same bug applied to
2289 patterns with negative forward assertions or any backward assertions
2290 preceding ".*" at the start, unless the pattern required a fixed first
2291 character. This was a failing pattern: "(?!.bcd).*". The bug is now fixed.
2292
2293 10. In UTF-8 mode, when moving forwards in the subject after a failed match
2294 starting at the last subject character, bytes beyond the end of the subject
2295 string were read.
2296
2297 11. Renamed the variable "class" as "classbits" to make life easier for C++
2298 users. (Previously there was a macro definition, but it apparently wasn't
2299 enough.)
2300
2301 12. Added the new field "tables" to the extra data so that tables can be passed
2302 in at exec time, or the internal tables can be re-selected. This allows
2303 a compiled regex to be saved and re-used at a later time by a different
2304 program that might have everything at different addresses.
2305
2306 13. Modified the pcre-config script so that, when run on Solaris, it shows a
2307 -R library as well as a -L library.
2308
2309 14. The debugging options of pcretest (-d on the command line or D on a
2310 pattern) showed incorrect output for anything following an extended class
2311 that contained multibyte characters and which was followed by a quantifier.
2312
2313 15. Added optional support for general category Unicode character properties
2314 via the \p, \P, and \X escapes. Unicode property support implies UTF-8
2315 support. It adds about 90K to the size of the library. The meanings of the
2316 inbuilt class escapes such as \d and \s have NOT been changed.
2317
2318 16. Updated pcredemo.c to include calls to free() to release the memory for the
2319 compiled pattern.
2320
2321 17. The generated file chartables.c was being created in the source directory
2322 instead of in the building directory. This caused the build to fail if the
2323 source directory was different from the building directory, and was
2324 read-only.
2325
2326 18. Added some sample Win commands from Mark Tetrode into the NON-UNIX-USE
2327 file. No doubt somebody will tell me if they don't make sense... Also added
2328 Dan Mooney's comments about building on OpenVMS.
2329
2330 19. Added support for partial matching via the PCRE_PARTIAL option for
2331 pcre_exec() and the \P data escape in pcretest.
2332
2333 20. Extended pcretest with 3 new pattern features:
2334
2335 (i) A pattern option of the form ">rest-of-line" causes pcretest to
2336 write the compiled pattern to the file whose name is "rest-of-line".
2337 This is a straight binary dump of the data, with the saved pointer to
2338 the character tables forced to be NULL. The study data, if any, is
2339 written too. After writing, pcretest reads a new pattern.
2340
2341 (ii) If, instead of a pattern, "<rest-of-line" is given, pcretest reads a
2342 compiled pattern from the given file. There must not be any
2343 occurrences of "<" in the file name (pretty unlikely); if there are,
2344 pcretest will instead treat the initial "<" as a pattern delimiter.
2345 After reading in the pattern, pcretest goes on to read data lines as
2346 usual.
2347
2348 (iii) The F pattern option causes pcretest to flip the bytes in the 32-bit
2349 and 16-bit fields in a compiled pattern, to simulate a pattern that
2350 was compiled on a host of opposite endianness.
2351
2352 21. The pcre-exec() function can now cope with patterns that were compiled on
2353 hosts of opposite endianness, with this restriction:
2354
2355 As for any compiled expression that is saved and used later, the tables
2356 pointer field cannot be preserved; the extra_data field in the arguments
2357 to pcre_exec() should be used to pass in a tables address if a value
2358 other than the default internal tables were used at compile time.
2359
2360 22. Calling pcre_exec() with a negative value of the "ovecsize" parameter is
2361 now diagnosed as an error. Previously, most of the time, a negative number
2362 would have been treated as zero, but if in addition "ovector" was passed as
2363 NULL, a crash could occur.
2364
2365 23. Updated the files ltmain.sh, config.sub, config.guess, and aclocal.m4 with
2366 new versions from the libtool 1.5 distribution (the last one is a copy of
2367 a file called libtool.m4). This seems to have fixed the need to patch
2368 "configure" to support Darwin 1.3 (which I used to do). However, I still
2369 had to patch ltmain.sh to ensure that ${SED} is set (it isn't on my
2370 workstation).
2371
2372 24. Changed the PCRE licence to be the more standard "BSD" licence.
2373
2374
2375 Version 4.5 01-Dec-03
2376 ---------------------
2377
2378 1. There has been some re-arrangement of the code for the match() function so
2379 that it can be compiled in a version that does not call itself recursively.
2380 Instead, it keeps those local variables that need separate instances for
2381 each "recursion" in a frame on the heap, and gets/frees frames whenever it
2382 needs to "recurse". Keeping track of where control must go is done by means
2383 of setjmp/longjmp. The whole thing is implemented by a set of macros that
2384 hide most of the details from the main code, and operates only if
2385 NO_RECURSE is defined while compiling pcre.c. If PCRE is built using the
2386 "configure" mechanism, "--disable-stack-for-recursion" turns on this way of
2387 operating.
2388
2389 To make it easier for callers to provide specially tailored get/free
2390 functions for this usage, two new functions, pcre_stack_malloc, and
2391 pcre_stack_free, are used. They are always called in strict stacking order,
2392 and the size of block requested is always the same.
2393
2394 The PCRE_CONFIG_STACKRECURSE info parameter can be used to find out whether
2395 PCRE has been compiled to use the stack or the heap for recursion. The
2396 -C option of pcretest uses this to show which version is compiled.
2397
2398 A new data escape \S, is added to pcretest; it causes the amounts of store
2399 obtained and freed by both kinds of malloc/free at match time to be added
2400 to the output.
2401
2402 2. Changed the locale test to use "fr_FR" instead of "fr" because that's
2403 what's available on my current Linux desktop machine.
2404
2405 3. When matching a UTF-8 string, the test for a valid string at the start has
2406 been extended. If start_offset is not zero, PCRE now checks that it points
2407 to a byte that is the start of a UTF-8 character. If not, it returns
2408 PCRE_ERROR_BADUTF8_OFFSET (-11). Note: the whole string is still checked;
2409 this is necessary because there may be backward assertions in the pattern.
2410 When matching the same subject several times, it may save resources to use
2411 PCRE_NO_UTF8_CHECK on all but the first call if the string is long.
2412
2413 4. The code for checking the validity of UTF-8 strings has been tightened so
2414 that it rejects (a) strings containing 0xfe or 0xff bytes and (b) strings
2415 containing "overlong sequences".
2416
2417 5. Fixed a bug (appearing twice) that I could not find any way of exploiting!
2418 I had written "if ((digitab[*p++] && chtab_digit) == 0)" where the "&&"
2419 should have been "&", but it just so happened that all the cases this let
2420 through by mistake were picked up later in the function.
2421
2422 6. I had used a variable called "isblank" - this is a C99 function, causing
2423 some compilers to warn. To avoid this, I renamed it (as "blankclass").
2424
2425 7. Cosmetic: (a) only output another newline at the end of pcretest if it is
2426 prompting; (b) run "./pcretest /dev/null" at the start of the test script
2427 so the version is shown; (c) stop "make test" echoing "./RunTest".
2428
2429 8. Added patches from David Burgess to enable PCRE to run on EBCDIC systems.
2430
2431 9. The prototype for memmove() for systems that don't have it was using
2432 size_t, but the inclusion of the header that defines size_t was later. I've
2433 moved the #includes for the C headers earlier to avoid this.
2434
2435 10. Added some adjustments to the code to make it easier to compiler on certain
2436 special systems:
2437
2438 (a) Some "const" qualifiers were missing.
2439 (b) Added the macro EXPORT before all exported functions; by default this
2440 is defined to be empty.
2441 (c) Changed the dftables auxiliary program (that builds chartables.c) so
2442 that it reads its output file name as an argument instead of writing
2443 to the standard output and assuming this can be redirected.
2444
2445 11. In UTF-8 mode, if a recursive reference (e.g. (?1)) followed a character
2446 class containing characters with values greater than 255, PCRE compilation
2447 went into a loop.
2448
2449 12. A recursive reference to a subpattern that was within another subpattern
2450 that had a minimum quantifier of zero caused PCRE to crash. For example,
2451 (x(y(?2))z)? provoked this bug with a subject that got as far as the
2452 recursion. If the recursively-called subpattern itself had a zero repeat,
2453 that was OK.
2454
2455 13. In pcretest, the buffer for reading a data line was set at 30K, but the
2456 buffer into which it was copied (for escape processing) was still set at
2457 1024, so long lines caused crashes.
2458
2459 14. A pattern such as /[ab]{1,3}+/ failed to compile, giving the error
2460 "internal error: code overflow...". This applied to any character class
2461 that was followed by a possessive quantifier.
2462
2463 15. Modified the Makefile to add libpcre.la as a prerequisite for
2464 libpcreposix.la because I was told this is needed for a parallel build to
2465 work.
2466
2467 16. If a pattern that contained .* following optional items at the start was
2468 studied, the wrong optimizing data was generated, leading to matching
2469 errors. For example, studying /[ab]*.*c/ concluded, erroneously, that any
2470 matching string must start with a or b or c. The correct conclusion for
2471 this pattern is that a match can start with any character.
2472
2473
2474 Version 4.4 13-Aug-03
2475 ---------------------
2476
2477 1. In UTF-8 mode, a character class containing characters with values between
2478 127 and 255 was not handled correctly if the compiled pattern was studied.
2479 In fixing this, I have also improved the studying algorithm for such
2480 classes (slightly).
2481
2482 2. Three internal functions had redundant arguments passed to them. Removal
2483 might give a very teeny performance improvement.
2484
2485 3. Documentation bug: the value of the capture_top field in a callout is *one
2486 more than* the number of the hightest numbered captured substring.
2487
2488 4. The Makefile linked pcretest and pcregrep with -lpcre, which could result
2489 in incorrectly linking with a previously installed version. They now link
2490 explicitly with libpcre.la.
2491
2492 5. configure.in no longer needs to recognize Cygwin specially.
2493
2494 6. A problem in pcre.in for Windows platforms is fixed.
2495
2496 7. If a pattern was successfully studied, and the -d (or /D) flag was given to
2497 pcretest, it used to include the size of the study block as part of its
2498 output. Unfortunately, the structure contains a field that has a different
2499 size on different hardware architectures. This meant that the tests that
2500 showed this size failed. As the block is currently always of a fixed size,
2501 this information isn't actually particularly useful in pcretest output, so
2502 I have just removed it.
2503
2504 8. Three pre-processor statements accidentally did not start in column 1.
2505 Sadly, there are *still* compilers around that complain, even though
2506 standard C has not required this for well over a decade. Sigh.
2507
2508 9. In pcretest, the code for checking callouts passed small integers in the
2509 callout_data field, which is a void * field. However, some picky compilers
2510 complained about the casts involved for this on 64-bit systems. Now
2511 pcretest passes the address of the small integer instead, which should get
2512 rid of the warnings.
2513
2514 10. By default, when in UTF-8 mode, PCRE now checks for valid UTF-8 strings at
2515 both compile and run time, and gives an error if an invalid UTF-8 sequence
2516 is found. There is a option for disabling this check in cases where the
2517 string is known to be correct and/or the maximum performance is wanted.
2518
2519 11. In response to a bug report, I changed one line in Makefile.in from
2520
2521 -Wl,--out-implib,.libs/lib@WIN_PREFIX@pcreposix.dll.a \
2522 to
2523 -Wl,--out-implib,.libs/@WIN_PREFIX@libpcreposix.dll.a \
2524
2525 to look similar to other lines, but I have no way of telling whether this
2526 is the right thing to do, as I do not use Windows. No doubt I'll get told
2527 if it's wrong...
2528
2529
2530 Version 4.3 21-May-03
2531 ---------------------
2532
2533 1. Two instances of @WIN_PREFIX@ omitted from the Windows targets in the
2534 Makefile.
2535
2536 2. Some refactoring to improve the quality of the code:
2537
2538 (i) The utf8_table... variables are now declared "const".
2539
2540 (ii) The code for \cx, which used the "case flipping" table to upper case
2541 lower case letters, now just substracts 32. This is ASCII-specific,
2542 but the whole concept of \cx is ASCII-specific, so it seems
2543 reasonable.
2544
2545 (iii) PCRE was using its character types table to recognize decimal and
2546 hexadecimal digits in the pattern. This is silly, because it handles
2547 only 0-9, a-f, and A-F, but the character types table is locale-
2548 specific, which means strange things might happen. A private
2549 table is now used for this - though it costs 256 bytes, a table is
2550 much faster than multiple explicit tests. Of course, the standard
2551 character types table is still used for matching digits in subject
2552 strings against \d.
2553
2554 (iv) Strictly, the identifier ESC_t is reserved by POSIX (all identifiers
2555 ending in _t are). So I've renamed it as ESC_tee.
2556
2557 3. The first argument for regexec() in the POSIX wrapper should have been
2558 defined as "const".
2559
2560 4. Changed pcretest to use malloc() for its buffers so that they can be
2561 Electric Fenced for debugging.
2562
2563 5. There were several places in the code where, in UTF-8 mode, PCRE would try
2564 to read one or more bytes before the start of the subject string. Often this
2565 had no effect on PCRE's behaviour, but in some circumstances it could
2566 provoke a segmentation fault.
2567
2568 6. A lookbehind at the start of a pattern in UTF-8 mode could also cause PCRE
2569 to try to read one or more bytes before the start of the subject string.
2570
2571 7. A lookbehind in a pattern matched in non-UTF-8 mode on a PCRE compiled with
2572 UTF-8 support could misbehave in various ways if the subject string
2573 contained bytes with the 0x80 bit set and the 0x40 bit unset in a lookbehind
2574 area. (PCRE was not checking for the UTF-8 mode flag, and trying to move
2575 back over UTF-8 characters.)
2576
2577
2578 Version 4.2 14-Apr-03
2579 ---------------------
2580
2581 1. Typo "#if SUPPORT_UTF8" instead of "#ifdef SUPPORT_UTF8" fixed.
2582
2583 2. Changes to the building process, supplied by Ronald Landheer-Cieslak
2584 [ON_WINDOWS]: new variable, "#" on non-Windows platforms
2585 [NOT_ON_WINDOWS]: new variable, "#" on Windows platforms
2586 [WIN_PREFIX]: new variable, "cyg" for Cygwin
2587 * Makefile.in: use autoconf substitution for OBJEXT, EXEEXT, BUILD_OBJEXT
2588 and BUILD_EXEEXT
2589 Note: automatic setting of the BUILD variables is not yet working
2590 set CPPFLAGS and BUILD_CPPFLAGS (but don't use yet) - should be used at
2591 compile-time but not at link-time
2592 [LINK]: use for linking executables only
2593 make different versions for Windows and non-Windows
2594 [LINKLIB]: new variable, copy of UNIX-style LINK, used for linking
2595 libraries
2596 [LINK_FOR_BUILD]: new variable
2597 [OBJEXT]: use throughout
2598 [EXEEXT]: use throughout
2599 <winshared>: new target
2600 <wininstall>: new target
2601 <dftables.o>: use native compiler
2602 <dftables>: use native linker
2603 <install>: handle Windows platform correctly
2604 <clean>: ditto
2605 <check>: ditto
2606 copy DLL to top builddir before testing
2607
2608 As part of these changes, -no-undefined was removed again. This was reported
2609 to give trouble on HP-UX 11.0, so getting rid of it seems like a good idea
2610 in any case.
2611
2612 3. Some tidies to get rid of compiler warnings:
2613
2614 . In the match_data structure, match_limit was an unsigned long int, whereas
2615 match_call_count was an int. I've made them both unsigned long ints.
2616
2617 . In pcretest the fact that a const uschar * doesn't automatically cast to
2618 a void * provoked a warning.
2619
2620 . Turning on some more compiler warnings threw up some "shadow" variables
2621 and a few more missing casts.
2622
2623 4. If PCRE was complied with UTF-8 support, but called without the PCRE_UTF8
2624 option, a class that contained a single character with a value between 128
2625 and 255 (e.g. /[\xFF]/) caused PCRE to crash.
2626
2627 5. If PCRE was compiled with UTF-8 support, but called without the PCRE_UTF8
2628 option, a class that contained several characters, but with at least one
2629 whose value was between 128 and 255 caused PCRE to crash.
2630
2631
2632 Version 4.1 12-Mar-03
2633 ---------------------
2634
2635 1. Compiling with gcc -pedantic found a couple of places where casts were
2636 needed, and a string in dftables.c that was longer than standard compilers are
2637 required to support.
2638
2639 2. Compiling with Sun's compiler found a few more places where the code could
2640 be tidied up in order to avoid warnings.
2641
2642 3. The variables for cross-compiling were called HOST_CC and HOST_CFLAGS; the
2643 first of these names is deprecated in the latest Autoconf in favour of the name
2644 CC_FOR_BUILD, because "host" is typically used to mean the system on which the
2645 compiled code will be run. I can't find a reference for HOST_CFLAGS, but by
2646 analogy I have changed it to CFLAGS_FOR_BUILD.
2647
2648 4. Added -no-undefined to the linking command in the Makefile, because this is
2649 apparently helpful for Windows. To make it work, also added "-L. -lpcre" to the
2650 linking step for the pcreposix library.
2651
2652 5. PCRE was failing to diagnose the case of two named groups with the same
2653 name.
2654
2655 6. A problem with one of PCRE's optimizations was discovered. PCRE remembers a
2656 literal character that is needed in the subject for a match, and scans along to
2657 ensure that it is present before embarking on the full matching process. This
2658 saves time in cases of nested unlimited repeats that are never going to match.
2659 Problem: the scan can take a lot of time if the subject is very long (e.g.
2660 megabytes), thus penalizing straightforward matches. It is now done only if the
2661 amount of subject to be scanned is less than 1000 bytes.
2662
2663 7. A lesser problem with the same optimization is that it was recording the
2664 first character of an anchored pattern as "needed", thus provoking a search
2665 right along the subject, even when the first match of the pattern was going to
2666 fail. The "needed" character is now not set for anchored patterns, unless it
2667 follows something in the pattern that is of non-fixed length. Thus, it still
2668 fulfils its original purpose of finding quick non-matches in cases of nested
2669 unlimited repeats, but isn't used for simple anchored patterns such as /^abc/.
2670
2671
2672 Version 4.0 17-Feb-03
2673 ---------------------
2674
2675 1. If a comment in an extended regex that started immediately after a meta-item
2676 extended to the end of string, PCRE compiled incorrect data. This could lead to
2677 all kinds of weird effects. Example: /#/ was bad; /()#/ was bad; /a#/ was not.
2678
2679 2. Moved to autoconf 2.53 and libtool 1.4.2.
2680
2681 3. Perl 5.8 no longer needs "use utf8" for doing UTF-8 things. Consequently,
2682 the special perltest8 script is no longer needed - all the tests can be run
2683 from a single perltest script.
2684
2685 4. From 5.004, Perl has not included the VT character (0x0b) in the set defined
2686 by \s. It has now been removed in PCRE. This means it isn't recognized as
2687 whitespace in /x regexes too, which is the same as Perl. Note that the POSIX
2688 class [:space:] *does* include VT, thereby creating a mess.
2689
2690 5. Added the class [:blank:] (a GNU extension from Perl 5.8) to match only
2691 space and tab.
2692
2693 6. Perl 5.005 was a long time ago. It's time to amalgamate the tests that use
2694 its new features into the main test script, reducing the number of scripts.
2695
2696 7. Perl 5.8 has changed the meaning of patterns like /a(?i)b/. Earlier versions
2697 were backward compatible, and made the (?i) apply to the whole pattern, as if
2698 /i were given. Now it behaves more logically, and applies the option setting
2699 only to what follows. PCRE has been changed to follow suit. However, if it
2700 finds options settings right at the start of the pattern, it extracts them into
2701 the global options, as before. Thus, they show up in the info data.
2702
2703 8. Added support for the \Q...\E escape sequence. Characters in between are
2704 treated as literals. This is slightly different from Perl in that $ and @ are
2705 also handled as literals inside the quotes. In Perl, they will cause variable
2706 interpolation. Note the following examples:
2707
2708 Pattern PCRE matches Perl matches
2709
2710 \Qabc$xyz\E abc$xyz abc followed by the contents of $xyz
2711 \Qabc\$xyz\E abc\$xyz abc\$xyz
2712 \Qabc\E\$\Qxyz\E abc$xyz abc$xyz
2713
2714 For compatibility with Perl, \Q...\E sequences are recognized inside character
2715 classes as well as outside them.
2716
2717 9. Re-organized 3 code statements in pcretest to avoid "overflow in
2718 floating-point constant arithmetic" warnings from a Microsoft compiler. Added a
2719 (size_t) cast to one statement in pcretest and one in pcreposix to avoid
2720 signed/unsigned warnings.
2721
2722 10. SunOS4 doesn't have strtoul(). This was used only for unpicking the -o
2723 option for pcretest, so I've replaced it by a simple function that does just
2724 that job.
2725
2726 11. pcregrep was ending with code 0 instead of 2 for the commands "pcregrep" or
2727 "pcregrep -".
2728
2729 12. Added "possessive quantifiers" ?+, *+, ++, and {,}+ which come from Sun's
2730 Java package. This provides some syntactic sugar for simple cases of what my
2731 documentation calls "once-only subpatterns". A pattern such as x*+ is the same
2732 as (?>x*). In other words, if what is inside (?>...) is just a single repeated
2733 item, you can use this simplified notation. Note that only makes sense with
2734 greedy quantifiers. Consequently, the use of the possessive quantifier forces
2735 greediness, whatever the setting of the PCRE_UNGREEDY option.
2736
2737 13. A change of greediness default within a pattern was not taking effect at
2738 the current level for patterns like /(b+(?U)a+)/. It did apply to parenthesized
2739 subpatterns that followed. Patterns like /b+(?U)a+/ worked because the option
2740 was abstracted outside.
2741
2742 14. PCRE now supports the \G assertion. It is true when the current matching
2743 position is at the start point of the match. This differs from \A when the
2744 starting offset is non-zero. Used with the /g option of pcretest (or similar
2745 code), it works in the same way as it does for Perl's /g option. If all
2746 alternatives of a regex begin with \G, the expression is anchored to the start
2747 match position, and the "anchored" flag is set in the compiled expression.
2748
2749 15. Some bugs concerning the handling of certain option changes within patterns
2750 have been fixed. These applied to options other than (?ims). For example,
2751 "a(?x: b c )d" did not match "XabcdY" but did match "Xa b c dY". It should have
2752 been the other way round. Some of this was related to change 7 above.
2753
2754 16. PCRE now gives errors for /[.x.]/ and /[=x=]/ as unsupported POSIX
2755 features, as Perl does. Previously, PCRE gave the warnings only for /[[.x.]]/
2756 and /[[=x=]]/. PCRE now also gives an error for /[:name:]/ because it supports
2757 POSIX classes only within a class (e.g. /[[:alpha:]]/).
2758
2759 17. Added support for Perl's \C escape. This matches one byte, even in UTF8
2760 mode. Unlike ".", it always matches newline, whatever the setting of
2761 PCRE_DOTALL. However, PCRE does not permit \C to appear in lookbehind
2762 assertions. Perl allows it, but it doesn't (in general) work because it can't
2763 calculate the length of the lookbehind. At least, that's the case for Perl
2764 5.8.0 - I've been told they are going to document that it doesn't work in
2765 future.
2766
2767 18. Added an error diagnosis for escapes that PCRE does not support: these are
2768 \L, \l, \N, \P, \p, \U, \u, and \X.
2769
2770 19. Although correctly diagnosing a missing ']' in a character class, PCRE was
2771 reading past the end of the pattern in cases such as /[abcd/.
2772
2773 20. PCRE was getting more memory than necessary for patterns with classes that
2774 contained both POSIX named classes and other characters, e.g. /[[:space:]abc/.
2775
2776 21. Added some code, conditional on #ifdef VPCOMPAT, to make life easier for
2777 compiling PCRE for use with Virtual Pascal.
2778
2779 22. Small fix to the Makefile to make it work properly if the build is done
2780 outside the source tree.
2781
2782 23. Added a new extension: a condition to go with recursion. If a conditional
2783 subpattern starts with (?(R) the "true" branch is used if recursion has
2784 happened, whereas the "false" branch is used only at the top level.
2785
2786 24. When there was a very long string of literal characters (over 255 bytes
2787 without UTF support, over 250 bytes with UTF support), the computation of how
2788 much memory was required could be incorrect, leading to segfaults or other
2789 strange effects.
2790
2791 25. PCRE was incorrectly assuming anchoring (either to start of subject or to
2792 start of line for a non-DOTALL pattern) when a pattern started with (.*) and
2793 there was a subsequent back reference to those brackets. This meant that, for
2794 example, /(.*)\d+\1/ failed to match "abc123bc". Unfortunately, it isn't
2795 possible to check for precisely this case. All we can do is abandon the
2796 optimization if .* occurs inside capturing brackets when there are any back
2797 references whatsoever. (See below for a better fix that came later.)
2798
2799 26. The handling of the optimization for finding the first character of a
2800 non-anchored pattern, and for finding a character that is required later in the
2801 match were failing in some cases. This didn't break the matching; it just
2802 failed to optimize when it could. The way this is done has been re-implemented.
2803
2804 27. Fixed typo in error message for invalid (?R item (it said "(?p").
2805
2806 28. Added a new feature that provides some of the functionality that Perl
2807 provides with (?{...}). The facility is termed a "callout". The way it is done
2808 in PCRE is for the caller to provide an optional function, by setting
2809 pcre_callout to its entry point. Like pcre_malloc and pcre_free, this is a
2810 global variable. By default it is unset, which disables all calling out. To get
2811 the function called, the regex must include (?C) at appropriate points. This
2812 is, in fact, equivalent to (?C0), and any number <= 255 may be given with (?C).
2813 This provides a means of identifying different callout points. When PCRE
2814 reaches such a point in the regex, if pcre_callout has been set, the external
2815 function is called. It is provided with data in a structure called
2816 pcre_callout_block, which is defined in pcre.h. If the function returns 0,
2817 matching continues; if it returns a non-zero value, the match at the current
2818 point fails. However, backtracking will occur if possible. [This was changed
2819 later and other features added - see item 49 below.]
2820
2821 29. pcretest is upgraded to test the callout functionality. It provides a
2822 callout function that displays information. By default, it shows the start of
2823 the match and the current position in the text. There are some new data escapes
2824 to vary what happens:
2825
2826 \C+ in addition, show current contents of captured substrings
2827 \C- do not supply a callout function
2828 \C!n return 1 when callout number n is reached
2829 \C!n!m return 1 when callout number n is reached for the mth time
2830
2831 30. If pcregrep was called with the -l option and just a single file name, it
2832 output "<stdin>" if a match was found, instead of the file name.
2833
2834 31. Improve the efficiency of the POSIX API to PCRE. If the number of capturing
2835 slots is less than POSIX_MALLOC_THRESHOLD, use a block on the stack to pass to
2836 pcre_exec(). This saves a malloc/free per call. The default value of
2837 POSIX_MALLOC_THRESHOLD is 10; it can be changed by --with-posix-malloc-threshold
2838 when configuring.
2839
2840 32. The default maximum size of a compiled pattern is 64K. There have been a
2841 few cases of people hitting this limit. The code now uses macros to handle the
2842 storing of links as offsets within the compiled pattern. It defaults to 2-byte
2843 links, but this can be changed to 3 or 4 bytes by --with-link-size when
2844 configuring. Tests 2 and 5 work only with 2-byte links because they output
2845 debugging information about compiled patterns.
2846
2847 33. Internal code re-arrangements:
2848
2849 (a) Moved the debugging function for printing out a compiled regex into
2850 its own source file (printint.c) and used #include to pull it into
2851 pcretest.c and, when DEBUG is defined, into pcre.c, instead of having two
2852 separate copies.
2853
2854 (b) Defined the list of op-code names for debugging as a macro in
2855 internal.h so that it is next to the definition of the opcodes.
2856
2857 (c) Defined a table of op-code lengths for simpler skipping along compiled
2858 code. This is again a macro in internal.h so that it is next to the
2859 definition of the opcodes.
2860
2861 34. Added support for recursive calls to individual subpatterns, along the
2862 lines of Robin Houston's patch (but implemented somewhat differently).
2863
2864 35. Further mods to the Makefile to help Win32. Also, added code to pcregrep to
2865 allow it to read and process whole directories in Win32. This code was
2866 contributed by Lionel Fourquaux; it has not been tested by me.
2867
2868 36. Added support for named subpatterns. The Python syntax (?P<name>...) is
2869 used to name a group. Names consist of alphanumerics and underscores, and must
2870 be unique. Back references use the syntax (?P=name) and recursive calls use
2871 (?P>name) which is a PCRE extension to the Python extension. Groups still have
2872 numbers. The function pcre_fullinfo() can be used after compilation to extract
2873 a name/number map. There are three relevant calls:
2874
2875 PCRE_INFO_NAMEENTRYSIZE yields the size of each entry in the map
2876 PCRE_INFO_NAMECOUNT yields the number of entries
2877 PCRE_INFO_NAMETABLE yields a pointer to the map.
2878
2879 The map is a vector of fixed-size entries. The size of each entry depends on
2880 the length of the longest name used. The first two bytes of each entry are the
2881 group number, most significant byte first. There follows the corresponding
2882 name, zero terminated. The names are in alphabetical order.
2883
2884 37. Make the maximum literal string in the compiled code 250 for the non-UTF-8
2885 case instead of 255. Making it the same both with and without UTF-8 support
2886 means that the same test output works with both.
2887
2888 38. There was a case of malloc(0) in the POSIX testing code in pcretest. Avoid
2889 calling malloc() with a zero argument.
2890
2891 39. Change 25 above had to resort to a heavy-handed test for the .* anchoring
2892 optimization. I've improved things by keeping a bitmap of backreferences with
2893 numbers 1-31 so that if .* occurs inside capturing brackets that are not in
2894 fact referenced, the optimization can be applied. It is unlikely that a
2895 relevant occurrence of .* (i.e. one which might indicate anchoring or forcing
2896 the match to follow \n) will appear inside brackets with a number greater than
2897 31, but if it does, any back reference > 31 suppresses the optimization.
2898
2899 40. Added a new compile-time option PCRE_NO_AUTO_CAPTURE. This has the effect
2900 of disabling numbered capturing parentheses. Any opening parenthesis that is
2901 not followed by ? behaves as if it were followed by ?: but named parentheses
2902 can still be used for capturing (and they will acquire numbers in the usual
2903 way).
2904
2905 41. Redesigned the return codes from the match() function into yes/no/error so
2906 that errors can be passed back from deep inside the nested calls. A malloc
2907 failure while inside a recursive subpattern call now causes the
2908 PCRE_ERROR_NOMEMORY return instead of quietly going wrong.
2909
2910 42. It is now possible to set a limit on the number of times the match()
2911 function is called in a call to pcre_exec(). This facility makes it possible to
2912 limit the amount of recursion and backtracking, though not in a directly
2913 obvious way, because the match() function is used in a number of different
2914 circumstances. The count starts from zero for each position in the subject
2915 string (for non-anchored patterns). The default limit is, for compatibility, a
2916 large number, namely 10 000 000. You can change this in two ways:
2917
2918 (a) When configuring PCRE before making, you can use --with-match-limit=n
2919 to set a default value for the compiled library.
2920
2921 (b) For each call to pcre_exec(), you can pass a pcre_extra block in which
2922 a different value is set. See 45 below.
2923
2924 If the limit is exceeded, pcre_exec() returns PCRE_ERROR_MATCHLIMIT.
2925
2926 43. Added a new function pcre_config(int, void *) to enable run-time extraction
2927 of things that can be changed at compile time. The first argument specifies
2928 what is wanted and the second points to where the information is to be placed.
2929 The current list of available information is:
2930
2931 PCRE_CONFIG_UTF8
2932
2933 The output is an integer that is set to one if UTF-8 support is available;
2934 otherwise it is set to zero.
2935
2936 PCRE_CONFIG_NEWLINE
2937
2938 The output is an integer that it set to the value of the code that is used for
2939 newline. It is either LF (10) or CR (13).
2940
2941 PCRE_CONFIG_LINK_SIZE
2942
2943 The output is an integer that contains the number of bytes used for internal
2944 linkage in compiled expressions. The value is 2, 3, or 4. See item 32 above.
2945
2946 PCRE_CONFIG_POSIX_MALLOC_THRESHOLD
2947
2948 The output is an integer that contains the threshold above which the POSIX
2949 interface uses malloc() for output vectors. See item 31 above.
2950
2951 PCRE_CONFIG_MATCH_LIMIT
2952
2953 The output is an unsigned integer that contains the default limit of the number
2954 of match() calls in a pcre_exec() execution. See 42 above.
2955
2956 44. pcretest has been upgraded by the addition of the -C option. This causes it
2957 to extract all the available output from the new pcre_config() function, and to
2958 output it. The program then exits immediately.
2959
2960 45. A need has arisen to pass over additional data with calls to pcre_exec() in
2961 order to support additional features. One way would have been to define
2962 pcre_exec2() (for example) with extra arguments, but this would not have been
2963 extensible, and would also have required all calls to the original function to
2964 be mapped to the new one. Instead, I have chosen to extend the mechanism that
2965 is used for passing in "extra" data from pcre_study().
2966
2967 The pcre_extra structure is now exposed and defined in pcre.h. It currently
2968 contains the following fields:
2969
2970 flags a bitmap indicating which of the following fields are set
2971 study_data opaque data from pcre_study()
2972 match_limit a way of specifying a limit on match() calls for a specific
2973 call to pcre_exec()
2974 callout_data data for callouts (see 49 below)
2975
2976 The flag bits are also defined in pcre.h, and are
2977
2978 PCRE_EXTRA_STUDY_DATA
2979 PCRE_EXTRA_MATCH_LIMIT
2980 PCRE_EXTRA_CALLOUT_DATA
2981
2982 The pcre_study() function now returns one of these new pcre_extra blocks, with
2983 the actual study data pointed to by the study_data field, and the
2984 PCRE_EXTRA_STUDY_DATA flag set. This can be passed directly to pcre_exec() as
2985 before. That is, this change is entirely upwards-compatible and requires no
2986 change to existing code.
2987
2988 If you want to pass in additional data to pcre_exec(), you can either place it
2989 in a pcre_extra block provided by pcre_study(), or create your own pcre_extra
2990 block.
2991
2992 46. pcretest has been extended to test the PCRE_EXTRA_MATCH_LIMIT feature. If a
2993 data string contains the escape sequence \M, pcretest calls pcre_exec() several
2994 times with different match limits, until it finds the minimum value needed for
2995 pcre_exec() to complete. The value is then output. This can be instructive; for
2996 most simple matches the number is quite small, but for pathological cases it
2997 gets very large very quickly.
2998
2999 47. There's a new option for pcre_fullinfo() called PCRE_INFO_STUDYSIZE. It
3000 returns the size of the data block pointed to by the study_data field in a
3001 pcre_extra block, that is, the value that was passed as the argument to
3002 pcre_malloc() when PCRE was getting memory in which to place the information
3003 created by pcre_study(). The fourth argument should point to a size_t variable.
3004 pcretest has been extended so that this information is shown after a successful
3005 pcre_study() call when information about the compiled regex is being displayed.
3006
3007 48. Cosmetic change to Makefile: there's no need to have / after $(DESTDIR)
3008 because what follows is always an absolute path. (Later: it turns out that this
3009 is more than cosmetic for MinGW, because it doesn't like empty path
3010 components.)
3011
3012 49. Some changes have been made to the callout feature (see 28 above):
3013
3014 (i) A callout function now has three choices for what it returns:
3015
3016 0 => success, carry on matching
3017 > 0 => failure at this point, but backtrack if possible
3018 < 0 => serious error, return this value from pcre_exec()
3019
3020 Negative values should normally be chosen from the set of PCRE_ERROR_xxx
3021 values. In particular, returning PCRE_ERROR_NOMATCH forces a standard
3022 "match failed" error. The error number PCRE_ERROR_CALLOUT is reserved for
3023 use by callout functions. It will never be used by PCRE itself.
3024
3025 (ii) The pcre_extra structure (see 45 above) has a void * field called
3026 callout_data, with corresponding flag bit PCRE_EXTRA_CALLOUT_DATA. The
3027 pcre_callout_block structure has a field of the same name. The contents of
3028 the field passed in the pcre_extra structure are passed to the callout
3029 function in the corresponding field in the callout block. This makes it
3030 easier to use the same callout-containing regex from multiple threads. For
3031 testing, the pcretest program has a new data escape
3032
3033 \C*n pass the number n (may be negative) as callout_data
3034
3035 If the callout function in pcretest receives a non-zero value as
3036 callout_data, it returns that value.
3037
3038 50. Makefile wasn't handling CFLAGS properly when compiling dftables. Also,
3039 there were some redundant $(CFLAGS) in commands that are now specified as
3040 $(LINK), which already includes $(CFLAGS).
3041
3042 51. Extensions to UTF-8 support are listed below. These all apply when (a) PCRE
3043 has been compiled with UTF-8 support *and* pcre_compile() has been compiled
3044 with the PCRE_UTF8 flag. Patterns that are compiled without that flag assume
3045 one-byte characters throughout. Note that case-insensitive matching applies
3046 only to characters whose values are less than 256. PCRE doesn't support the
3047 notion of cases for higher-valued characters.
3048
3049 (i) A character class whose characters are all within 0-255 is handled as
3050 a bit map, and the map is inverted for negative classes. Previously, a
3051 character > 255 always failed to match such a class; however it should
3052 match if the class was a negative one (e.g. [^ab]). This has been fixed.
3053
3054 (ii) A negated character class with a single character < 255 is coded as
3055 "not this character" (OP_NOT). This wasn't working properly when the test
3056 character was multibyte, either singly or repeated.
3057
3058 (iii) Repeats of multibyte characters are now handled correctly in UTF-8
3059 mode, for example: \x{100}{2,3}.
3060
3061 (iv) The character escapes \b, \B, \d, \D, \s, \S, \w, and \W (either
3062 singly or repeated) now correctly test multibyte characters. However,
3063 PCRE doesn't recognize any characters with values greater than 255 as
3064 digits, spaces, or word characters. Such characters always match \D, \S,
3065 and \W, and never match \d, \s, or \w.
3066
3067 (v) Classes may now contain characters and character ranges with values
3068 greater than 255. For example: [ab\x{100}-\x{400}].
3069
3070 (vi) pcregrep now has a --utf-8 option (synonym -u) which makes it call
3071 PCRE in UTF-8 mode.
3072
3073 52. The info request value PCRE_INFO_FIRSTCHAR has been renamed
3074 PCRE_INFO_FIRSTBYTE because it is a byte value. However, the old name is
3075 retained for backwards compatibility. (Note that LASTLITERAL is also a byte
3076 value.)
3077
3078 53. The single man page has become too large. I have therefore split it up into
3079 a number of separate man pages. These also give rise to individual HTML pages;
3080 these are now put in a separate directory, and there is an index.html page that
3081 lists them all. Some hyperlinking between the pages has been installed.
3082
3083 54. Added convenience functions for handling named capturing parentheses.
3084
3085 55. Unknown escapes inside character classes (e.g. [\M]) and escapes that
3086 aren't interpreted therein (e.g. [\C]) are literals in Perl. This is now also
3087 true in PCRE, except when the PCRE_EXTENDED option is set, in which case they
3088 are faulted.
3089
3090 56. Introduced HOST_CC and HOST_CFLAGS which can be set in the environment when
3091 calling configure. These values are used when compiling the dftables.c program
3092 which is run to generate the source of the default character tables. They
3093 default to the values of CC and CFLAGS. If you are cross-compiling PCRE,
3094 you will need to set these values.
3095
3096 57. Updated the building process for Windows DLL, as provided by Fred Cox.
3097
3098
3099 Version 3.9 02-Jan-02
3100 ---------------------
3101
3102 1. A bit of extraneous text had somehow crept into the pcregrep documentation.
3103
3104 2. If --disable-static was given, the building process failed when trying to
3105 build pcretest and pcregrep. (For some reason it was using libtool to compile
3106 them, which is not right, as they aren't part of the library.)
3107
3108
3109 Version 3.8 18-Dec-01
3110 ---------------------
3111
3112 1. The experimental UTF-8 code was completely screwed up. It was packing the
3113 bytes in the wrong order. How dumb can you get?
3114
3115
3116 Version 3.7 29-Oct-01
3117 ---------------------
3118
3119 1. In updating pcretest to check change 1 of version 3.6, I screwed up.
3120 This caused pcretest, when used on the test data, to segfault. Unfortunately,
3121 this didn't happen under Solaris 8, where I normally test things.
3122
3123 2. The Makefile had to be changed to make it work on BSD systems, where 'make'
3124 doesn't seem to recognize that ./xxx and xxx are the same file. (This entry
3125 isn't in ChangeLog distributed with 3.7 because I forgot when I hastily made
3126 this fix an hour or so after the initial 3.7 release.)
3127
3128
3129 Version 3.6 23-Oct-01
3130 ---------------------
3131
3132 1. Crashed with /(sens|respons)e and \1ibility/ and "sense and sensibility" if
3133 offsets passed as NULL with zero offset count.
3134
3135 2. The config.guess and config.sub files had not been updated when I moved to
3136 the latest autoconf.
3137
3138
3139 Version 3.5 15-Aug-01
3140 ---------------------
3141
3142 1. Added some missing #if !defined NOPOSIX conditionals in pcretest.c that
3143 had been forgotten.
3144
3145 2. By using declared but undefined structures, we can avoid using "void"
3146 definitions in pcre.h while keeping the internal definitions of the structures
3147 private.
3148
3149 3. The distribution is now built using autoconf 2.50 and libtool 1.4. From a
3150 user point of view, this means that both static and shared libraries are built
3151 by default, but this can be individually controlled. More of the work of
3152 handling this static/shared cases is now inside libtool instead of PCRE's make
3153 file.
3154
3155 4. The pcretest utility is now installed along with pcregrep because it is
3156 useful for users (to test regexs) and by doing this, it automatically gets
3157 relinked by libtool. The documentation has been turned into a man page, so
3158 there are now .1, .txt, and .html versions in /doc.
3159
3160 5. Upgrades to pcregrep:
3161 (i) Added long-form option names like gnu grep.
3162 (ii) Added --help to list all options with an explanatory phrase.
3163 (iii) Added -r, --recursive to recurse into sub-directories.
3164 (iv) Added -f, --file to read patterns from a file.
3165
3166 6. pcre_exec() was referring to its "code" argument before testing that
3167 argument for NULL (and giving an error if it was NULL).
3168
3169 7. Upgraded Makefile.in to allow for compiling in a different directory from
3170 the source directory.
3171
3172 8. Tiny buglet in pcretest: when pcre_fullinfo() was called to retrieve the
3173 options bits, the pointer it was passed was to an int instead of to an unsigned
3174 long int. This mattered only on 64-bit systems.
3175
3176 9. Fixed typo (3.4/1) in pcre.h again. Sigh. I had changed pcre.h (which is
3177 generated) instead of pcre.in, which it its source. Also made the same change
3178 in several of the .c files.
3179
3180 10. A new release of gcc defines printf() as a macro, which broke pcretest
3181 because it had an ifdef in the middle of a string argument for printf(). Fixed
3182 by using separate calls to printf().
3183
3184 11. Added --enable-newline-is-cr and --enable-newline-is-lf to the configure
3185 script, to force use of CR or LF instead of \n in the source. On non-Unix
3186 systems, the value can be set in config.h.
3187
3188 12. The limit of 200 on non-capturing parentheses is a _nesting_ limit, not an
3189 absolute limit. Changed the text of the error message to make this clear, and
3190 likewise updated the man page.
3191
3192 13. The limit of 99 on the number of capturing subpatterns has been removed.
3193 The new limit is 65535, which I hope will not be a "real" limit.
3194
3195
3196 Version 3.4 22-Aug-00
3197 ---------------------
3198
3199 1. Fixed typo in pcre.h: unsigned const char * changed to const unsigned char *.
3200
3201 2. Diagnose condition (?(0) as an error instead of crashing on matching.
3202
3203
3204 Version 3.3 01-Aug-00
3205 ---------------------
3206
3207 1. If an octal character was given, but the value was greater than \377, it
3208 was not getting masked to the least significant bits, as documented. This could
3209 lead to crashes in some systems.
3210
3211 2. Perl 5.6 (if not earlier versions) accepts classes like [a-\d] and treats
3212 the hyphen as a literal. PCRE used to give an error; it now behaves like Perl.
3213
3214 3. Added the functions pcre_free_substring() and pcre_free_substring_list().
3215 These just pass their arguments on to (pcre_free)(), but they are provided
3216 because some uses of PCRE bind it to non-C systems that can call its functions,
3217 but cannot call free() or pcre_free() directly.
3218
3219 4. Add "make test" as a synonym for "make check". Corrected some comments in
3220 the Makefile.
3221
3222 5. Add $(DESTDIR)/ in front of all the paths in the "install" target in the
3223 Makefile.
3224
3225 6. Changed the name of pgrep to pcregrep, because Solaris has introduced a
3226 command called pgrep for grepping around the active processes.
3227
3228 7. Added the beginnings of support for UTF-8 character strings.
3229
3230 8. Arranged for the Makefile to pass over the settings of CC, CFLAGS, and
3231 RANLIB to ./ltconfig so that they are used by libtool. I think these are all
3232 the relevant ones. (AR is not passed because ./ltconfig does its own figuring
3233 out for the ar command.)
3234
3235
3236 Version 3.2 12-May-00
3237 ---------------------
3238
3239 This is purely a bug fixing release.
3240
3241 1. If the pattern /((Z)+|A)*/ was matched agained ZABCDEFG it matched Z instead
3242 of ZA. This was just one example of several cases that could provoke this bug,
3243 which was introduced by change 9 of version 2.00. The code for breaking
3244 infinite loops after an iteration that matches an empty string was't working
3245 correctly.
3246
3247 2. The pcretest program was not imitating Perl correctly for the pattern /a*/g
3248 when matched against abbab (for example). After matching an empty string, it
3249 wasn't forcing anchoring when setting PCRE_NOTEMPTY for the next attempt; this
3250 caused it to match further down the string than it should.
3251
3252 3. The code contained an inclusion of sys/types.h. It isn't clear why this
3253 was there because it doesn't seem to be needed, and it causes trouble on some
3254 systems, as it is not a Standard C header. It has been removed.
3255
3256 4. Made 4 silly changes to the source to avoid stupid compiler warnings that
3257 were reported on the Macintosh. The changes were from
3258
3259 while ((c = *(++ptr)) != 0 && c != '\n');
3260 to
3261 while ((c = *(++ptr)) != 0 && c != '\n') ;
3262
3263 Totally extraordinary, but if that's what it takes...
3264
3265 5. PCRE is being used in one environment where neither memmove() nor bcopy() is
3266 available. Added HAVE_BCOPY and an autoconf test for it; if neither
3267 HAVE_MEMMOVE nor HAVE_BCOPY is set, use a built-in emulation function which
3268 assumes the way PCRE uses memmove() (always moving upwards).
3269
3270 6. PCRE is being used in one environment where strchr() is not available. There
3271 was only one use in pcre.c, and writing it out to avoid strchr() probably gives
3272 faster code anyway.
3273
3274
3275 Version 3.1 09-Feb-00
3276 ---------------------
3277
3278 The only change in this release is the fixing of some bugs in Makefile.in for
3279 the "install" target:
3280
3281 (1) It was failing to install pcreposix.h.
3282
3283 (2) It was overwriting the pcre.3 man page with the pcreposix.3 man page.
3284
3285
3286 Version 3.0 01-Feb-00
3287 ---------------------
3288
3289 1. Add support for the /+ modifier to perltest (to output $` like it does in
3290 pcretest).
3291
3292 2. Add support for the /g modifier to perltest.
3293
3294 3. Fix pcretest so that it behaves even more like Perl for /g when the pattern
3295 matches null strings.
3296
3297 4. Fix perltest so that it doesn't do unwanted things when fed an empty
3298 pattern. Perl treats empty patterns specially - it reuses the most recent
3299 pattern, which is not what we want. Replace // by /(?#)/ in order to avoid this
3300 effect.
3301
3302 5. The POSIX interface was broken in that it was just handing over the POSIX
3303 captured string vector to pcre_exec(), but (since release 2.00) PCRE has
3304 required a bigger vector, with some working space on the end. This means that
3305 the POSIX wrapper now has to get and free some memory, and copy the results.
3306
3307 6. Added some simple autoconf support, placing the test data and the
3308 documentation in separate directories, re-organizing some of the
3309 information files, and making it build pcre-config (a GNU standard). Also added
3310 libtool support for building PCRE as a shared library, which is now the
3311 default.
3312
3313 7. Got rid of the leading zero in the definition of PCRE_MINOR because 08 and
3314 09 are not valid octal constants. Single digits will be used for minor values
3315 less than 10.
3316
3317 8. Defined REG_EXTENDED and REG_NOSUB as zero in the POSIX header, so that
3318 existing programs that set these in the POSIX interface can use PCRE without
3319 modification.
3320
3321 9. Added a new function, pcre_fullinfo() with an extensible interface. It can
3322 return all that pcre_info() returns, plus additional data. The pcre_info()
3323 function is retained for compatibility, but is considered to be obsolete.
3324
3325 10. Added experimental recursion feature (?R) to handle one common case that
3326 Perl 5.6 will be able to do with (?p{...}).
3327
3328 11. Added support for POSIX character classes like [:alpha:], which Perl is
3329 adopting.
3330
3331
3332 Version 2.08 31-Aug-99
3333 ----------------------
3334
3335 1. When startoffset was not zero and the pattern began with ".*", PCRE was not
3336 trying to match at the startoffset position, but instead was moving forward to
3337 the next newline as if a previous match had failed.
3338
3339 2. pcretest was not making use of PCRE_NOTEMPTY when repeating for /g and /G,
3340 and could get into a loop if a null string was matched other than at the start
3341 of the subject.
3342
3343 3. Added definitions of PCRE_MAJOR and PCRE_MINOR to pcre.h so the version can
3344 be distinguished at compile time, and for completeness also added PCRE_DATE.
3345
3346 5. Added Paul Sokolovsky's minor changes to make it easy to compile a Win32 DLL
3347 in GnuWin32 environments.
3348
3349
3350 Version 2.07 29-Jul-99
3351 ----------------------
3352
3353 1. The documentation is now supplied in plain text form and HTML as well as in
3354 the form of man page sources.
3355
3356 2. C++ compilers don't like assigning (void *) values to other pointer types.
3357 In particular this affects malloc(). Although there is no problem in Standard
3358 C, I've put in casts to keep C++ compilers happy.
3359
3360 3. Typo on pcretest.c; a cast of (unsigned char *) in the POSIX regexec() call
3361 should be (const char *).
3362
3363 4. If NOPOSIX is defined, pcretest.c compiles without POSIX support. This may
3364 be useful for non-Unix systems who don't want to bother with the POSIX stuff.
3365 However, I haven't made this a standard facility. The documentation doesn't
3366 mention it, and the Makefile doesn't support it.
3367
3368 5. The Makefile now contains an "install" target, with editable destinations at
3369 the top of the file. The pcretest program is not installed.
3370
3371 6. pgrep -V now gives the PCRE version number and date.
3372
3373 7. Fixed bug: a zero repetition after a literal string (e.g. /abcde{0}/) was
3374 causing the entire string to be ignored, instead of just the last character.
3375
3376 8. If a pattern like /"([^\\"]+|\\.)*"/ is applied in the normal way to a
3377 non-matching string, it can take a very, very long time, even for strings of
3378 quite modest length, because of the nested recursion. PCRE now does better in
3379 some of these cases. It does this by remembering the last required literal
3380 character in the pattern, and pre-searching the subject to ensure it is present
3381 before running the real match. In other words, it applies a heuristic to detect
3382 some types of certain failure quickly, and in the above example, if presented
3383 with a string that has no trailing " it gives "no match" very quickly.
3384
3385 9. A new runtime option PCRE_NOTEMPTY causes null string matches to be ignored;
3386 other alternatives are tried instead.
3387
3388
3389 Version 2.06 09-Jun-99
3390 ----------------------
3391
3392 1. Change pcretest's output for amount of store used to show just the code
3393 space, because the remainder (the data block) varies in size between 32-bit and
3394 64-bit systems.
3395
3396 2. Added an extra argument to pcre_exec() to supply an offset in the subject to
3397 start matching at. This allows lookbehinds to work when searching for multiple
3398 occurrences in a string.
3399
3400 3. Added additional options to pcretest for testing multiple occurrences:
3401
3402 /+ outputs the rest of the string that follows a match
3403 /g loops for multiple occurrences, using the new startoffset argument
3404 /G loops for multiple occurrences by passing an incremented pointer
3405
3406 4. PCRE wasn't doing the "first character" optimization for patterns starting
3407 with \b or \B, though it was doing it for other lookbehind assertions. That is,
3408 it wasn't noticing that a match for a pattern such as /\bxyz/ has to start with
3409 the letter 'x'. On long subject strings, this gives a significant speed-up.
3410
3411
3412 Version 2.05 21-Apr-99
3413 ----------------------
3414
3415 1. Changed the type of magic_number from int to long int so that it works
3416 properly on 16-bit systems.
3417
3418 2. Fixed a bug which caused patterns starting with .* not to work correctly
3419 when the subject string contained newline characters. PCRE was assuming
3420 anchoring for such patterns in all cases, which is not correct because .* will
3421 not pass a newline unless PCRE_DOTALL is set. It now assumes anchoring only if
3422 DOTALL is set at top level; otherwise it knows that patterns starting with .*
3423 must be retried after every newline in the subject.
3424
3425
3426 Version 2.04 18-Feb-99
3427 ----------------------
3428
3429 1. For parenthesized subpatterns with repeats whose minimum was zero, the
3430 computation of the store needed to hold the pattern was incorrect (too large).
3431 If such patterns were nested a few deep, this could multiply and become a real
3432 problem.
3433
3434 2. Added /M option to pcretest to show the memory requirement of a specific
3435 pattern. Made -m a synonym of -s (which does this globally) for compatibility.
3436
3437 3. Subpatterns of the form (regex){n,m} (i.e. limited maximum) were being
3438 compiled in such a way that the backtracking after subsequent failure was
3439 pessimal. Something like (a){0,3} was compiled as (a)?(a)?(a)? instead of
3440 ((a)((a)(a)?)?)? with disastrous performance if the maximum was of any size.
3441
3442
3443 Version 2.03 02-Feb-99
3444 ----------------------
3445
3446 1. Fixed typo and small mistake in man page.
3447
3448 2. Added 4th condition (GPL supersedes if conflict) and created separate
3449 LICENCE file containing the conditions.
3450
3451 3. Updated pcretest so that patterns such as /abc\/def/ work like they do in
3452 Perl, that is the internal \ allows the delimiter to be included in the
3453 pattern. Locked out the use of \ as a delimiter. If \ immediately follows
3454 the final delimiter, add \ to the end of the pattern (to test the error).
3455
3456 4. Added the convenience functions for extracting substrings after a successful
3457 match. Updated pcretest to make it able to test these functions.
3458
3459
3460 Version 2.02 14-Jan-99
3461 ----------------------
3462
3463 1. Initialized the working variables associated with each extraction so that
3464 their saving and restoring doesn't refer to uninitialized store.
3465
3466 2. Put dummy code into study.c in order to trick the optimizer of the IBM C
3467 compiler for OS/2 into generating correct code. Apparently IBM isn't going to
3468 fix the problem.
3469
3470 3. Pcretest: the timing code wasn't using LOOPREPEAT for timing execution
3471 calls, and wasn't printing the correct value for compiling calls. Increased the
3472 default value of LOOPREPEAT, and the number of significant figures in the
3473 times.
3474
3475 4. Changed "/bin/rm" in the Makefile to "-rm" so it works on Windows NT.
3476
3477 5. Renamed "deftables" as "dftables" to get it down to 8 characters, to avoid
3478 a building problem on Windows NT with a FAT file system.
3479
3480
3481 Version 2.01 21-Oct-98
3482 ----------------------
3483
3484 1. Changed the API for pcre_compile() to allow for the provision of a pointer
3485 to character tables built by pcre_maketables() in the current locale. If NULL
3486 is passed, the default tables are used.
3487
3488
3489 Version 2.00 24-Sep-98
3490 ----------------------
3491
3492 1. Since the (>?) facility is in Perl 5.005, don't require PCRE_EXTRA to enable
3493 it any more.
3494
3495 2. Allow quantification of (?>) groups, and make it work correctly.
3496
3497 3. The first character computation wasn't working for (?>) groups.
3498
3499 4. Correct the implementation of \Z (it is permitted to match on the \n at the
3500 end of the subject) and add 5.005's \z, which really does match only at the
3501 very end of the subject.
3502
3503 5. Remove the \X "cut" facility; Perl doesn't have it, and (?> is neater.
3504
3505 6. Remove the ability to specify CASELESS, MULTILINE, DOTALL, and
3506 DOLLAR_END_ONLY at runtime, to make it possible to implement the Perl 5.005
3507 localized options. All options to pcre_study() were also removed.
3508
3509 7. Add other new features from 5.005:
3510
3511 $(?<= positive lookbehind
3512 $(?<! negative lookbehind
3513 (?imsx-imsx) added the unsetting capability
3514 such a setting is global if at outer level; local otherwise
3515 (?imsx-imsx:) non-capturing groups with option setting
3516 (?(cond)re|re) conditional pattern matching
3517
3518 A backreference to itself in a repeated group matches the previous
3519 captured string.
3520
3521 8. General tidying up of studying (both automatic and via "study")
3522 consequential on the addition of new assertions.
3523
3524 9. As in 5.005, unlimited repeated groups that could match an empty substring
3525 are no longer faulted at compile time. Instead, the loop is forcibly broken at
3526 runtime if any iteration does actually match an empty substring.
3527
3528 10. Include the RunTest script in the distribution.
3529
3530 11. Added tests from the Perl 5.005_02 distribution. This showed up a few
3531 discrepancies, some of which were old and were also with respect to 5.004. They
3532 have now been fixed.
3533
3534
3535 Version 1.09 28-Apr-98
3536 ----------------------
3537
3538 1. A negated single character class followed by a quantifier with a minimum
3539 value of one (e.g. [^x]{1,6} ) was not compiled correctly. This could lead to
3540 program crashes, or just wrong answers. This did not apply to negated classes
3541 containing more than one character, or to minima other than one.
3542
3543
3544 Version 1.08 27-Mar-98
3545 ----------------------
3546
3547 1. Add PCRE_UNGREEDY to invert the greediness of quantifiers.
3548
3549 2. Add (?U) and (?X) to set PCRE_UNGREEDY and PCRE_EXTRA respectively. The
3550 latter must appear before anything that relies on it in the pattern.
3551
3552
3553 Version 1.07 16-Feb-98
3554 ----------------------
3555
3556 1. A pattern such as /((a)*)*/ was not being diagnosed as in error (unlimited
3557 repeat of a potentially empty string).
3558
3559
3560 Version 1.06 23-Jan-98
3561 ----------------------
3562
3563 1. Added Markus Oberhumer's little patches for C++.
3564
3565 2. Literal strings longer than 255 characters were broken.
3566
3567
3568 Version 1.05 23-Dec-97
3569 ----------------------
3570
3571 1. Negated character classes containing more than one character were failing if
3572 PCRE_CASELESS was set at run time.
3573
3574
3575 Version 1.04 19-Dec-97
3576 ----------------------
3577
3578 1. Corrected the man page, where some "const" qualifiers had been omitted.
3579
3580 2. Made debugging output print "{0,xxx}" instead of just "{,xxx}" to agree with
3581 input syntax.
3582
3583 3. Fixed memory leak which occurred when a regex with back references was
3584 matched with an offsets vector that wasn't big enough. The temporary memory
3585 that is used in this case wasn't being freed if the match failed.
3586
3587 4. Tidied pcretest to ensure it frees memory that it gets.
3588
3589 5. Temporary memory was being obtained in the case where the passed offsets
3590 vector was exactly big enough.
3591
3592 6. Corrected definition of offsetof() from change 5 below.
3593
3594 7. I had screwed up change 6 below and broken the rules for the use of
3595 setjmp(). Now fixed.
3596
3597
3598 Version 1.03 18-Dec-97
3599 ----------------------
3600
3601 1. A erroneous regex with a missing opening parenthesis was correctly
3602 diagnosed, but PCRE attempted to access brastack[-1], which could cause crashes
3603 on some systems.
3604
3605 2. Replaced offsetof(real_pcre, code) by offsetof(real_pcre, code[0]) because
3606 it was reported that one broken compiler failed on the former because "code" is
3607 also an independent variable.
3608
3609 3. The erroneous regex a[]b caused an array overrun reference.
3610
3611 4. A regex ending with a one-character negative class (e.g. /[^k]$/) did not
3612 fail on data ending with that character. (It was going on too far, and checking
3613 the next character, typically a binary zero.) This was specific to the
3614 optimized code for single-character negative classes.
3615
3616 5. Added a contributed patch from the TIN world which does the following:
3617
3618 + Add an undef for memmove, in case the the system defines a macro for it.
3619
3620 + Add a definition of offsetof(), in case there isn't one. (I don't know
3621 the reason behind this - offsetof() is part of the ANSI standard - but
3622 it does no harm).
3623
3624 + Reduce the ifdef's in pcre.c using macro DPRINTF, thereby eliminating
3625 most of the places where whitespace preceded '#'. I have given up and
3626 allowed the remaining 2 cases to be at the margin.
3627
3628 + Rename some variables in pcre to eliminate shadowing. This seems very
3629 pedantic, but does no harm, of course.
3630
3631 6. Moved the call to setjmp() into its own function, to get rid of warnings
3632 from gcc -Wall, and avoided calling it at all unless PCRE_EXTRA is used.
3633
3634 7. Constructs such as \d{8,} were compiling into the equivalent of
3635 \d{8}\d{0,65527} instead of \d{8}\d* which didn't make much difference to the
3636 outcome, but in this particular case used more store than had been allocated,
3637 which caused the bug to be discovered because it threw up an internal error.
3638
3639 8. The debugging code in both pcre and pcretest for outputting the compiled
3640 form of a regex was going wrong in the case of back references followed by
3641 curly-bracketed repeats.
3642
3643
3644 Version 1.02 12-Dec-97
3645 ----------------------
3646
3647 1. Typos in pcre.3 and comments in the source fixed.
3648
3649 2. Applied a contributed patch to get rid of places where it used to remove
3650 'const' from variables, and fixed some signed/unsigned and uninitialized
3651 variable warnings.
3652
3653 3. Added the "runtest" target to Makefile.
3654
3655 4. Set default compiler flag to -O2 rather than just -O.
3656
3657
3658 Version 1.01 19-Nov-97
3659 ----------------------
3660
3661 1. PCRE was failing to diagnose unlimited repeat of empty string for patterns
3662 like /([ab]*)*/, that is, for classes with more than one character in them.
3663
3664 2. Likewise, it wasn't diagnosing patterns with "once-only" subpatterns, such
3665 as /((?>a*))*/ (a PCRE_EXTRA facility).
3666
3667
3668 Version 1.00 18-Nov-97
3669 ----------------------
3670
3671 1. Added compile-time macros to support systems such as SunOS4 which don't have
3672 memmove() or strerror() but have other things that can be used instead.
3673
3674 2. Arranged that "make clean" removes the executables.
3675
3676
3677 Version 0.99 27-Oct-97
3678 ----------------------
3679
3680 1. Fixed bug in code for optimizing classes with only one character. It was
3681 initializing a 32-byte map regardless, which could cause it to run off the end
3682 of the memory it had got.
3683
3684 2. Added, conditional on PCRE_EXTRA, the proposed (?>REGEX) construction.
3685
3686
3687 Version 0.98 22-Oct-97
3688 ----------------------
3689
3690 1. Fixed bug in code for handling temporary memory usage when there are more
3691 back references than supplied space in the ovector. This could cause segfaults.
3692
3693
3694 Version 0.97 21-Oct-97
3695 ----------------------
3696
3697 1. Added the \X "cut" facility, conditional on PCRE_EXTRA.
3698
3699 2. Optimized negated single characters not to use a bit map.
3700
3701 3. Brought error texts together as macro definitions; clarified some of them;
3702 fixed one that was wrong - it said "range out of order" when it meant "invalid
3703 escape sequence".
3704
3705 4. Changed some char * arguments to const char *.
3706
3707 5. Added PCRE_NOTBOL and PCRE_NOTEOL (from POSIX).
3708
3709 6. Added the POSIX-style API wrapper in pcreposix.a and testing facilities in
3710 pcretest.
3711
3712
3713 Version 0.96 16-Oct-97
3714 ----------------------
3715
3716 1. Added a simple "pgrep" utility to the distribution.
3717
3718 2. Fixed an incompatibility with Perl: "{" is now treated as a normal character
3719 unless it appears in one of the precise forms "{ddd}", "{ddd,}", or "{ddd,ddd}"
3720 where "ddd" means "one or more decimal digits".
3721
3722 3. Fixed serious bug. If a pattern had a back reference, but the call to
3723 pcre_exec() didn't supply a large enough ovector to record the related
3724 identifying subpattern, the match always failed. PCRE now remembers the number
3725 of the largest back reference, and gets some temporary memory in which to save
3726 the offsets during matching if necessary, in order to ensure that
3727 backreferences always work.
3728
3729 4. Increased the compatibility with Perl in a number of ways:
3730
3731 (a) . no longer matches \n by default; an option PCRE_DOTALL is provided
3732 to request this handling. The option can be set at compile or exec time.
3733
3734 (b) $ matches before a terminating newline by default; an option
3735 PCRE_DOLLAR_ENDONLY is provided to override this (but not in multiline
3736 mode). The option can be set at compile or exec time.
3737
3738 (c) The handling of \ followed by a digit other than 0 is now supposed to be
3739 the same as Perl's. If the decimal number it represents is less than 10
3740 or there aren't that many previous left capturing parentheses, an octal
3741 escape is read. Inside a character class, it's always an octal escape,
3742 even if it is a single digit.
3743
3744 (d) An escaped but undefined alphabetic character is taken as a literal,
3745 unless PCRE_EXTRA is set. Currently this just reserves the remaining
3746 escapes.
3747
3748 (e) {0} is now permitted. (The previous item is removed from the compiled
3749 pattern).
3750
3751 5. Changed all the names of code files so that the basic parts are no longer
3752 than 10 characters, and abolished the teeny "globals.c" file.
3753
3754 6. Changed the handling of character classes; they are now done with a 32-byte
3755 bit map always.
3756
3757 7. Added the -d and /D options to pcretest to make it possible to look at the
3758 internals of compilation without having to recompile pcre.
3759
3760
3761 Version 0.95 23-Sep-97
3762 ----------------------
3763
3764 1. Fixed bug in pre-pass concerning escaped "normal" characters such as \x5c or
3765 \x20 at the start of a run of normal characters. These were being treated as
3766 real characters, instead of the source characters being re-checked.
3767
3768
3769 Version 0.94 18-Sep-97
3770 ----------------------
3771
3772 1. The functions are now thread-safe, with the caveat that the global variables
3773 containing pointers to malloc() and free() or alternative functions are the
3774 same for all threads.
3775
3776 2. Get pcre_study() to generate a bitmap of initial characters for non-
3777 anchored patterns when this is possible, and use it if passed to pcre_exec().
3778
3779
3780 Version 0.93 15-Sep-97
3781 ----------------------
3782
3783 1. /(b)|(:+)/ was computing an incorrect first character.
3784
3785 2. Add pcre_study() to the API and the passing of pcre_extra to pcre_exec(),
3786 but not actually doing anything yet.
3787
3788 3. Treat "-" characters in classes that cannot be part of ranges as literals,
3789 as Perl does (e.g. [-az] or [az-]).
3790
3791 4. Set the anchored flag if a branch starts with .* or .*? because that tests
3792 all possible positions.
3793
3794 5. Split up into different modules to avoid including unneeded functions in a
3795 compiled binary. However, compile and exec are still in one module. The "study"
3796 function is split off.
3797
3798 6. The character tables are now in a separate module whose source is generated
3799 by an auxiliary program - but can then be edited by hand if required. There are
3800 now no calls to isalnum(), isspace(), isdigit(), isxdigit(), tolower() or
3801 toupper() in the code.
3802
3803 7. Turn the malloc/free funtions variables into pcre_malloc and pcre_free and
3804 make them global. Abolish the function for setting them, as the caller can now
3805 set them directly.
3806
3807
3808 Version 0.92 11-Sep-97
3809 ----------------------
3810
3811 1. A repeat with a fixed maximum and a minimum of 1 for an ordinary character
3812 (e.g. /a{1,3}/) was broken (I mis-optimized it).
3813
3814 2. Caseless matching was not working in character classes if the characters in
3815 the pattern were in upper case.
3816
3817 3. Make ranges like [W-c] work in the same way as Perl for caseless matching.
3818
3819 4. Make PCRE_ANCHORED public and accept as a compile option.
3820
3821 5. Add an options word to pcre_exec() and accept PCRE_ANCHORED and
3822 PCRE_CASELESS at run time. Add escapes \A and \I to pcretest to cause it to
3823 pass them.
3824
3825 6. Give an error if bad option bits passed at compile or run time.
3826
3827 7. Add PCRE_MULTILINE at compile and exec time, and (?m) as well. Add \M to
3828 pcretest to cause it to pass that flag.
3829
3830 8. Add pcre_info(), to get the number of identifying subpatterns, the stored
3831 options, and the first character, if set.
3832
3833 9. Recognize C+ or C{n,m} where n >= 1 as providing a fixed starting character.
3834
3835
3836 Version 0.91 10-Sep-97
3837 ----------------------
3838
3839 1. PCRE was failing to diagnose unlimited repeats of subpatterns that could
3840 match the empty string as in /(a*)*/. It was looping and ultimately crashing.
3841
3842 2. PCRE was looping on encountering an indefinitely repeated back reference to
3843 a subpattern that had matched an empty string, e.g. /(a|)\1*/. It now does what
3844 Perl does - treats the match as successful.
3845
3846 ****

Properties

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

  ViewVC Help
Powered by ViewVC 1.1.5