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

Contents of /code/trunk/ChangeLog

Parent Directory Parent Directory | Revision Log Revision Log


Revision 388 - (show annotations) (download)
Wed Mar 11 17:03:17 2009 UTC (5 years, 4 months ago) by ph10
File size: 162384 byte(s)
Added REG_NOTEMPTY to the POSIX wrapper.

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

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