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

Contents of /code/trunk/ChangeLog

Parent Directory Parent Directory | Revision Log Revision Log


Revision 384 - (show annotations) (download)
Sun Mar 8 16:27:43 2009 UTC (5 years, 8 months ago) by ph10
File size: 162256 byte(s)
Fix Unicode property support in character classes for chars > 127 in non-UTF-8 
mode.

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

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