ViewVC logotype

Contents of /code/trunk/ChangeLog

Parent Directory Parent Directory | Revision Log Revision Log

Revision 446 - (hide annotations) (download)
Tue Sep 15 10:49:50 2009 UTC (5 years, 7 months ago) by ph10
File size: 172499 byte(s)
Correct returned capture count after recursion has matched more than outer.

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8     was out-of-date, and there was no check on the pcre_compile() error code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


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

ViewVC Help
Powered by ViewVC 1.1.12