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

Contents of /code/trunk/ChangeLog

Parent Directory Parent Directory | Revision Log Revision Log


Revision 478 - (show annotations) (download)
Sun Jan 3 16:05:13 2010 UTC (4 years, 6 months ago) by ph10
File size: 178075 byte(s)
Update configure.ac to solve libbz2 problem under Win32.

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76 was out-of-date, and there was no check on the pcre_compile() error code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

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