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

Contents of /code/trunk/ChangeLog

Parent Directory Parent Directory | Revision Log Revision Log


Revision 488 - (show annotations) (download)
Mon Jan 11 15:29:42 2010 UTC (4 years, 6 months ago) by ph10
File size: 181484 byte(s)
Fix #947, recursive back reference bug.

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

Properties

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

webmaster@exim.org
ViewVC Help
Powered by ViewVC 1.1.12