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

Contents of /code/trunk/ChangeLog

Parent Directory Parent Directory | Revision Log Revision Log


Revision 383 - (show annotations) (download)
Sun Mar 8 15:26:59 2009 UTC (5 years, 9 months ago) by ph10
File size: 162116 byte(s)
Fix PCRE_DOLLAR_ENDONLY bug in pcre_dfa_exec().

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

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