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

Contents of /code/trunk/ChangeLog

Parent Directory Parent Directory | Revision Log Revision Log


Revision 425 - (show annotations) (download)
Mon Aug 17 14:48:48 2009 UTC (5 years ago) by ph10
File size: 167752 byte(s)
Add #ifndef SUPPORT_UCP to pcre_ucd.c

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8 was out-of-date, and there was no check on the pcre_compile() error code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

Properties

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

webmaster@exim.org
ViewVC Help
Powered by ViewVC 1.1.12