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

Contents of /code/trunk/ChangeLog

Parent Directory Parent Directory | Revision Log Revision Log


Revision 336 - (show annotations)
Sat Apr 12 15:59:03 2008 UTC (7 years, 1 month ago) by ph10
File size: 153781 byte(s)
Error occurred while calculating annotation data.
Added PCRE_JAVASCRIPT_COMPAT option.
1 ChangeLog for PCRE
2 ------------------
3
4 Version 7.7 05-Mar-08
5 ---------------------
6
7 1. Applied Craig's patch to sort out a long long problem: "If we can't convert
8 a string to a long long, pretend we don't even have a long long." This is
9 done by checking for the strtoq, strtoll, and _strtoi64 functions.
10
11 2. Applied Craig's patch to pcrecpp.cc to restore ABI compatibility with
12 pre-7.6 versions, which defined a global no_arg variable instead of putting
13 it in the RE class. (See also #8 below.)
14
15 3. Remove a line of dead code, identified by coverity and reported by Nuno
16 Lopes.
17
18 4. Fixed two related pcregrep bugs involving -r with --include or --exclude:
19
20 (1) The include/exclude patterns were being applied to the whole pathnames
21 of files, instead of just to the final components.
22
23 (2) If there was more than one level of directory, the subdirectories were
24 skipped unless they satisfied the include/exclude conditions. This is
25 inconsistent with GNU grep (and could even be seen as contrary to the
26 pcregrep specification - which I improved to make it absolutely clear).
27 The action now is always to scan all levels of directory, and just
28 apply the include/exclude patterns to regular files.
29
30 5. Added the --include_dir and --exclude_dir patterns to pcregrep, and used
31 --exclude_dir in the tests to avoid scanning .svn directories.
32
33 6. Applied Craig's patch to the QuoteMeta function so that it escapes the
34 NUL character as backslash + 0 rather than backslash + NUL, because PCRE
35 doesn't support NULs in patterns.
36
37 7. Added some missing "const"s to declarations of static tables in
38 pcre_compile.c and pcre_dfa_exec.c.
39
40 8. Applied Craig's patch to pcrecpp.cc to fix a problem in OS X that was
41 caused by fix #2 above. (Subsequently also a second patch to fix the
42 first patch. And a third patch - this was a messy problem.)
43
44 9. Applied Craig's patch to remove the use of push_back().
45
46 10. Applied Alan Lehotsky's patch to add REG_STARTEND support to the POSIX
47 matching function regexec().
48
49 11. Added support for the Oniguruma syntax \g<name>, \g<n>, \g'name', \g'n',
50 which, however, unlike Perl's \g{...}, are subroutine calls, not back
51 references. PCRE supports relative numbers with this syntax (I don't think
52 Oniguruma does).
53
54 12. Previously, a group with a zero repeat such as (...){0} was completely
55 omitted from the compiled regex. However, this means that if the group
56 was called as a subroutine from elsewhere in the pattern, things went wrong
57 (an internal error was given). Such groups are now left in the compiled
58 pattern, with a new opcode that causes them to be skipped at execution
59 time.
60
61 13. Added the PCRE_JAVASCRIPT_COMPAT option. This currently does two things:
62 (a) A lone ] character is dis-allowed (Perl treats it as data).
63 (b) A back reference to an unmatched subpattern matches an empty string
64 (Perl fails the current match path).
65
66
67 Version 7.6 28-Jan-08
68 ---------------------
69
70 1. A character class containing a very large number of characters with
71 codepoints greater than 255 (in UTF-8 mode, of course) caused a buffer
72 overflow.
73
74 2. Patch to cut out the "long long" test in pcrecpp_unittest when
75 HAVE_LONG_LONG is not defined.
76
77 3. Applied Christian Ehrlicher's patch to update the CMake build files to
78 bring them up to date and include new features. This patch includes:
79
80 - Fixed PH's badly added libz and libbz2 support.
81 - Fixed a problem with static linking.
82 - Added pcredemo. [But later removed - see 7 below.]
83 - Fixed dftables problem and added an option.
84 - Added a number of HAVE_XXX tests, including HAVE_WINDOWS_H and
85 HAVE_LONG_LONG.
86 - Added readline support for pcretest.
87 - Added an listing of the option settings after cmake has run.
88
89 4. A user submitted a patch to Makefile that makes it easy to create
90 "pcre.dll" under mingw when using Configure/Make. I added stuff to
91 Makefile.am that cause it to include this special target, without
92 affecting anything else. Note that the same mingw target plus all
93 the other distribution libraries and programs are now supported
94 when configuring with CMake (see 6 below) instead of with
95 Configure/Make.
96
97 5. Applied Craig's patch that moves no_arg into the RE class in the C++ code.
98 This is an attempt to solve the reported problem "pcrecpp::no_arg is not
99 exported in the Windows port". It has not yet been confirmed that the patch
100 solves the problem, but it does no harm.
101
102 6. Applied Sheri's patch to CMakeLists.txt to add NON_STANDARD_LIB_PREFIX and
103 NON_STANDARD_LIB_SUFFIX for dll names built with mingw when configured
104 with CMake, and also correct the comment about stack recursion.
105
106 7. Remove the automatic building of pcredemo from the ./configure system and
107 from CMakeLists.txt. The whole idea of pcredemo.c is that it is an example
108 of a program that users should build themselves after PCRE is installed, so
109 building it automatically is not really right. What is more, it gave
110 trouble in some build environments.
111
112 8. Further tidies to CMakeLists.txt from Sheri and Christian.
113
114
115 Version 7.5 10-Jan-08
116 ---------------------
117
118 1. Applied a patch from Craig: "This patch makes it possible to 'ignore'
119 values in parens when parsing an RE using the C++ wrapper."
120
121 2. Negative specials like \S did not work in character classes in UTF-8 mode.
122 Characters greater than 255 were excluded from the class instead of being
123 included.
124
125 3. The same bug as (2) above applied to negated POSIX classes such as
126 [:^space:].
127
128 4. PCRECPP_STATIC was referenced in pcrecpp_internal.h, but nowhere was it
129 defined or documented. It seems to have been a typo for PCRE_STATIC, so
130 I have changed it.
131
132 5. The construct (?&) was not diagnosed as a syntax error (it referenced the
133 first named subpattern) and a construct such as (?&a) would reference the
134 first named subpattern whose name started with "a" (in other words, the
135 length check was missing). Both these problems are fixed. "Subpattern name
136 expected" is now given for (?&) (a zero-length name), and this patch also
137 makes it give the same error for \k'' (previously it complained that that
138 was a reference to a non-existent subpattern).
139
140 6. The erroneous patterns (?+-a) and (?-+a) give different error messages;
141 this is right because (?- can be followed by option settings as well as by
142 digits. I have, however, made the messages clearer.
143
144 7. Patterns such as (?(1)a|b) (a pattern that contains fewer subpatterns
145 than the number used in the conditional) now cause a compile-time error.
146 This is actually not compatible with Perl, which accepts such patterns, but
147 treats the conditional as always being FALSE (as PCRE used to), but it
148 seems to me that giving a diagnostic is better.
149
150 8. Change "alphameric" to the more common word "alphanumeric" in comments
151 and messages.
152
153 9. Fix two occurrences of "backslash" in comments that should have been
154 "backspace".
155
156 10. Remove two redundant lines of code that can never be obeyed (their function
157 was moved elsewhere).
158
159 11. The program that makes PCRE's Unicode character property table had a bug
160 which caused it to generate incorrect table entries for sequences of
161 characters that have the same character type, but are in different scripts.
162 It amalgamated them into a single range, with the script of the first of
163 them. In other words, some characters were in the wrong script. There were
164 thirteen such cases, affecting characters in the following ranges:
165
166 U+002b0 - U+002c1
167 U+0060c - U+0060d
168 U+0061e - U+00612
169 U+0064b - U+0065e
170 U+0074d - U+0076d
171 U+01800 - U+01805
172 U+01d00 - U+01d77
173 U+01d9b - U+01dbf
174 U+0200b - U+0200f
175 U+030fc - U+030fe
176 U+03260 - U+0327f
177 U+0fb46 - U+0fbb1
178 U+10450 - U+1049d
179
180 12. The -o option (show only the matching part of a line) for pcregrep was not
181 compatible with GNU grep in that, if there was more than one match in a
182 line, it showed only the first of them. It now behaves in the same way as
183 GNU grep.
184
185 13. If the -o and -v options were combined for pcregrep, it printed a blank
186 line for every non-matching line. GNU grep prints nothing, and pcregrep now
187 does the same. The return code can be used to tell if there were any
188 non-matching lines.
189
190 14. Added --file-offsets and --line-offsets to pcregrep.
191
192 15. The pattern (?=something)(?R) was not being diagnosed as a potentially
193 infinitely looping recursion. The bug was that positive lookaheads were not
194 being skipped when checking for a possible empty match (negative lookaheads
195 and both kinds of lookbehind were skipped).
196
197 16. Fixed two typos in the Windows-only code in pcregrep.c, and moved the
198 inclusion of <windows.h> to before rather than after the definition of
199 INVALID_FILE_ATTRIBUTES (patch from David Byron).
200
201 17. Specifying a possessive quantifier with a specific limit for a Unicode
202 character property caused pcre_compile() to compile bad code, which led at
203 runtime to PCRE_ERROR_INTERNAL (-14). Examples of patterns that caused this
204 are: /\p{Zl}{2,3}+/8 and /\p{Cc}{2}+/8. It was the possessive "+" that
205 caused the error; without that there was no problem.
206
207 18. Added --enable-pcregrep-libz and --enable-pcregrep-libbz2.
208
209 19. Added --enable-pcretest-libreadline.
210
211 20. In pcrecpp.cc, the variable 'count' was incremented twice in
212 RE::GlobalReplace(). As a result, the number of replacements returned was
213 double what it should be. I removed one of the increments, but Craig sent a
214 later patch that removed the other one (the right fix) and added unit tests
215 that check the return values (which was not done before).
216
217 21. Several CMake things:
218
219 (1) Arranged that, when cmake is used on Unix, the libraries end up with
220 the names libpcre and libpcreposix, not just pcre and pcreposix.
221
222 (2) The above change means that pcretest and pcregrep are now correctly
223 linked with the newly-built libraries, not previously installed ones.
224
225 (3) Added PCRE_SUPPORT_LIBREADLINE, PCRE_SUPPORT_LIBZ, PCRE_SUPPORT_LIBBZ2.
226
227 22. In UTF-8 mode, with newline set to "any", a pattern such as .*a.*=.b.*
228 crashed when matching a string such as a\x{2029}b (note that \x{2029} is a
229 UTF-8 newline character). The key issue is that the pattern starts .*;
230 this means that the match must be either at the beginning, or after a
231 newline. The bug was in the code for advancing after a failed match and
232 checking that the new position followed a newline. It was not taking
233 account of UTF-8 characters correctly.
234
235 23. PCRE was behaving differently from Perl in the way it recognized POSIX
236 character classes. PCRE was not treating the sequence [:...:] as a
237 character class unless the ... were all letters. Perl, however, seems to
238 allow any characters between [: and :], though of course it rejects as
239 unknown any "names" that contain non-letters, because all the known class
240 names consist only of letters. Thus, Perl gives an error for [[:1234:]],
241 for example, whereas PCRE did not - it did not recognize a POSIX character
242 class. This seemed a bit dangerous, so the code has been changed to be
243 closer to Perl. The behaviour is not identical to Perl, because PCRE will
244 diagnose an unknown class for, for example, [[:l\ower:]] where Perl will
245 treat it as [[:lower:]]. However, PCRE does now give "unknown" errors where
246 Perl does, and where it didn't before.
247
248 24. Rewrite so as to remove the single use of %n from pcregrep because in some
249 Windows environments %n is disabled by default.
250
251
252 Version 7.4 21-Sep-07
253 ---------------------
254
255 1. Change 7.3/28 was implemented for classes by looking at the bitmap. This
256 means that a class such as [\s] counted as "explicit reference to CR or
257 LF". That isn't really right - the whole point of the change was to try to
258 help when there was an actual mention of one of the two characters. So now
259 the change happens only if \r or \n (or a literal CR or LF) character is
260 encountered.
261
262 2. The 32-bit options word was also used for 6 internal flags, but the numbers
263 of both had grown to the point where there were only 3 bits left.
264 Fortunately, there was spare space in the data structure, and so I have
265 moved the internal flags into a new 16-bit field to free up more option
266 bits.
267
268 3. The appearance of (?J) at the start of a pattern set the DUPNAMES option,
269 but did not set the internal JCHANGED flag - either of these is enough to
270 control the way the "get" function works - but the PCRE_INFO_JCHANGED
271 facility is supposed to tell if (?J) was ever used, so now (?J) at the
272 start sets both bits.
273
274 4. Added options (at build time, compile time, exec time) to change \R from
275 matching any Unicode line ending sequence to just matching CR, LF, or CRLF.
276
277 5. doc/pcresyntax.html was missing from the distribution.
278
279 6. Put back the definition of PCRE_ERROR_NULLWSLIMIT, for backward
280 compatibility, even though it is no longer used.
281
282 7. Added macro for snprintf to pcrecpp_unittest.cc and also for strtoll and
283 strtoull to pcrecpp.cc to select the available functions in WIN32 when the
284 windows.h file is present (where different names are used). [This was
285 reversed later after testing - see 16 below.]
286
287 8. Changed all #include <config.h> to #include "config.h". There were also
288 some further <pcre.h> cases that I changed to "pcre.h".
289
290 9. When pcregrep was used with the --colour option, it missed the line ending
291 sequence off the lines that it output.
292
293 10. It was pointed out to me that arrays of string pointers cause lots of
294 relocations when a shared library is dynamically loaded. A technique of
295 using a single long string with a table of offsets can drastically reduce
296 these. I have refactored PCRE in four places to do this. The result is
297 dramatic:
298
299 Originally: 290
300 After changing UCP table: 187
301 After changing error message table: 43
302 After changing table of "verbs" 36
303 After changing table of Posix names 22
304
305 Thanks to the folks working on Gregex for glib for this insight.
306
307 11. --disable-stack-for-recursion caused compiling to fail unless -enable-
308 unicode-properties was also set.
309
310 12. Updated the tests so that they work when \R is defaulted to ANYCRLF.
311
312 13. Added checks for ANY and ANYCRLF to pcrecpp.cc where it previously
313 checked only for CRLF.
314
315 14. Added casts to pcretest.c to avoid compiler warnings.
316
317 15. Added Craig's patch to various pcrecpp modules to avoid compiler warnings.
318
319 16. Added Craig's patch to remove the WINDOWS_H tests, that were not working,
320 and instead check for _strtoi64 explicitly, and avoid the use of snprintf()
321 entirely. This removes changes made in 7 above.
322
323 17. The CMake files have been updated, and there is now more information about
324 building with CMake in the NON-UNIX-USE document.
325
326
327 Version 7.3 28-Aug-07
328 ---------------------
329
330 1. In the rejigging of the build system that eventually resulted in 7.1, the
331 line "#include <pcre.h>" was included in pcre_internal.h. The use of angle
332 brackets there is not right, since it causes compilers to look for an
333 installed pcre.h, not the version that is in the source that is being
334 compiled (which of course may be different). I have changed it back to:
335
336 #include "pcre.h"
337
338 I have a vague recollection that the change was concerned with compiling in
339 different directories, but in the new build system, that is taken care of
340 by the VPATH setting the Makefile.
341
342 2. The pattern .*$ when run in not-DOTALL UTF-8 mode with newline=any failed
343 when the subject happened to end in the byte 0x85 (e.g. if the last
344 character was \x{1ec5}). *Character* 0x85 is one of the "any" newline
345 characters but of course it shouldn't be taken as a newline when it is part
346 of another character. The bug was that, for an unlimited repeat of . in
347 not-DOTALL UTF-8 mode, PCRE was advancing by bytes rather than by
348 characters when looking for a newline.
349
350 3. A small performance improvement in the DOTALL UTF-8 mode .* case.
351
352 4. Debugging: adjusted the names of opcodes for different kinds of parentheses
353 in debug output.
354
355 5. Arrange to use "%I64d" instead of "%lld" and "%I64u" instead of "%llu" for
356 long printing in the pcrecpp unittest when running under MinGW.
357
358 6. ESC_K was left out of the EBCDIC table.
359
360 7. Change 7.0/38 introduced a new limit on the number of nested non-capturing
361 parentheses; I made it 1000, which seemed large enough. Unfortunately, the
362 limit also applies to "virtual nesting" when a pattern is recursive, and in
363 this case 1000 isn't so big. I have been able to remove this limit at the
364 expense of backing off one optimization in certain circumstances. Normally,
365 when pcre_exec() would call its internal match() function recursively and
366 immediately return the result unconditionally, it uses a "tail recursion"
367 feature to save stack. However, when a subpattern that can match an empty
368 string has an unlimited repetition quantifier, it no longer makes this
369 optimization. That gives it a stack frame in which to save the data for
370 checking that an empty string has been matched. Previously this was taken
371 from the 1000-entry workspace that had been reserved. So now there is no
372 explicit limit, but more stack is used.
373
374 8. Applied Daniel's patches to solve problems with the import/export magic
375 syntax that is required for Windows, and which was going wrong for the
376 pcreposix and pcrecpp parts of the library. These were overlooked when this
377 problem was solved for the main library.
378
379 9. There were some crude static tests to avoid integer overflow when computing
380 the size of patterns that contain repeated groups with explicit upper
381 limits. As the maximum quantifier is 65535, the maximum group length was
382 set at 30,000 so that the product of these two numbers did not overflow a
383 32-bit integer. However, it turns out that people want to use groups that
384 are longer than 30,000 bytes (though not repeat them that many times).
385 Change 7.0/17 (the refactoring of the way the pattern size is computed) has
386 made it possible to implement the integer overflow checks in a much more
387 dynamic way, which I have now done. The artificial limitation on group
388 length has been removed - we now have only the limit on the total length of
389 the compiled pattern, which depends on the LINK_SIZE setting.
390
391 10. Fixed a bug in the documentation for get/copy named substring when
392 duplicate names are permitted. If none of the named substrings are set, the
393 functions return PCRE_ERROR_NOSUBSTRING (7); the doc said they returned an
394 empty string.
395
396 11. Because Perl interprets \Q...\E at a high level, and ignores orphan \E
397 instances, patterns such as [\Q\E] or [\E] or even [^\E] cause an error,
398 because the ] is interpreted as the first data character and the
399 terminating ] is not found. PCRE has been made compatible with Perl in this
400 regard. Previously, it interpreted [\Q\E] as an empty class, and [\E] could
401 cause memory overwriting.
402
403 10. Like Perl, PCRE automatically breaks an unlimited repeat after an empty
404 string has been matched (to stop an infinite loop). It was not recognizing
405 a conditional subpattern that could match an empty string if that
406 subpattern was within another subpattern. For example, it looped when
407 trying to match (((?(1)X|))*) but it was OK with ((?(1)X|)*) where the
408 condition was not nested. This bug has been fixed.
409
410 12. A pattern like \X?\d or \P{L}?\d in non-UTF-8 mode could cause a backtrack
411 past the start of the subject in the presence of bytes with the top bit
412 set, for example "\x8aBCD".
413
414 13. Added Perl 5.10 experimental backtracking controls (*FAIL), (*F), (*PRUNE),
415 (*SKIP), (*THEN), (*COMMIT), and (*ACCEPT).
416
417 14. Optimized (?!) to (*FAIL).
418
419 15. Updated the test for a valid UTF-8 string to conform to the later RFC 3629.
420 This restricts code points to be within the range 0 to 0x10FFFF, excluding
421 the "low surrogate" sequence 0xD800 to 0xDFFF. Previously, PCRE allowed the
422 full range 0 to 0x7FFFFFFF, as defined by RFC 2279. Internally, it still
423 does: it's just the validity check that is more restrictive.
424
425 16. Inserted checks for integer overflows during escape sequence (backslash)
426 processing, and also fixed erroneous offset values for syntax errors during
427 backslash processing.
428
429 17. Fixed another case of looking too far back in non-UTF-8 mode (cf 12 above)
430 for patterns like [\PPP\x8a]{1,}\x80 with the subject "A\x80".
431
432 18. An unterminated class in a pattern like (?1)\c[ with a "forward reference"
433 caused an overrun.
434
435 19. A pattern like (?:[\PPa*]*){8,} which had an "extended class" (one with
436 something other than just ASCII characters) inside a group that had an
437 unlimited repeat caused a loop at compile time (while checking to see
438 whether the group could match an empty string).
439
440 20. Debugging a pattern containing \p or \P could cause a crash. For example,
441 [\P{Any}] did so. (Error in the code for printing property names.)
442
443 21. An orphan \E inside a character class could cause a crash.
444
445 22. A repeated capturing bracket such as (A)? could cause a wild memory
446 reference during compilation.
447
448 23. There are several functions in pcre_compile() that scan along a compiled
449 expression for various reasons (e.g. to see if it's fixed length for look
450 behind). There were bugs in these functions when a repeated \p or \P was
451 present in the pattern. These operators have additional parameters compared
452 with \d, etc, and these were not being taken into account when moving along
453 the compiled data. Specifically:
454
455 (a) A item such as \p{Yi}{3} in a lookbehind was not treated as fixed
456 length.
457
458 (b) An item such as \pL+ within a repeated group could cause crashes or
459 loops.
460
461 (c) A pattern such as \p{Yi}+(\P{Yi}+)(?1) could give an incorrect
462 "reference to non-existent subpattern" error.
463
464 (d) A pattern like (\P{Yi}{2}\277)? could loop at compile time.
465
466 24. A repeated \S or \W in UTF-8 mode could give wrong answers when multibyte
467 characters were involved (for example /\S{2}/8g with "A\x{a3}BC").
468
469 25. Using pcregrep in multiline, inverted mode (-Mv) caused it to loop.
470
471 26. Patterns such as [\P{Yi}A] which include \p or \P and just one other
472 character were causing crashes (broken optimization).
473
474 27. Patterns such as (\P{Yi}*\277)* (group with possible zero repeat containing
475 \p or \P) caused a compile-time loop.
476
477 28. More problems have arisen in unanchored patterns when CRLF is a valid line
478 break. For example, the unstudied pattern [\r\n]A does not match the string
479 "\r\nA" because change 7.0/46 below moves the current point on by two
480 characters after failing to match at the start. However, the pattern \nA
481 *does* match, because it doesn't start till \n, and if [\r\n]A is studied,
482 the same is true. There doesn't seem any very clean way out of this, but
483 what I have chosen to do makes the common cases work: PCRE now takes note
484 of whether there can be an explicit match for \r or \n anywhere in the
485 pattern, and if so, 7.0/46 no longer applies. As part of this change,
486 there's a new PCRE_INFO_HASCRORLF option for finding out whether a compiled
487 pattern has explicit CR or LF references.
488
489 29. Added (*CR) etc for changing newline setting at start of pattern.
490
491
492 Version 7.2 19-Jun-07
493 ---------------------
494
495 1. If the fr_FR locale cannot be found for test 3, try the "french" locale,
496 which is apparently normally available under Windows.
497
498 2. Re-jig the pcregrep tests with different newline settings in an attempt
499 to make them independent of the local environment's newline setting.
500
501 3. Add code to configure.ac to remove -g from the CFLAGS default settings.
502
503 4. Some of the "internals" tests were previously cut out when the link size
504 was not 2, because the output contained actual offsets. The recent new
505 "Z" feature of pcretest means that these can be cut out, making the tests
506 usable with all link sizes.
507
508 5. Implemented Stan Switzer's goto replacement for longjmp() when not using
509 stack recursion. This gives a massive performance boost under BSD, but just
510 a small improvement under Linux. However, it saves one field in the frame
511 in all cases.
512
513 6. Added more features from the forthcoming Perl 5.10:
514
515 (a) (?-n) (where n is a string of digits) is a relative subroutine or
516 recursion call. It refers to the nth most recently opened parentheses.
517
518 (b) (?+n) is also a relative subroutine call; it refers to the nth next
519 to be opened parentheses.
520
521 (c) Conditions that refer to capturing parentheses can be specified
522 relatively, for example, (?(-2)... or (?(+3)...
523
524 (d) \K resets the start of the current match so that everything before
525 is not part of it.
526
527 (e) \k{name} is synonymous with \k<name> and \k'name' (.NET compatible).
528
529 (f) \g{name} is another synonym - part of Perl 5.10's unification of
530 reference syntax.
531
532 (g) (?| introduces a group in which the numbering of parentheses in each
533 alternative starts with the same number.
534
535 (h) \h, \H, \v, and \V match horizontal and vertical whitespace.
536
537 7. Added two new calls to pcre_fullinfo(): PCRE_INFO_OKPARTIAL and
538 PCRE_INFO_JCHANGED.
539
540 8. A pattern such as (.*(.)?)* caused pcre_exec() to fail by either not
541 terminating or by crashing. Diagnosed by Viktor Griph; it was in the code
542 for detecting groups that can match an empty string.
543
544 9. A pattern with a very large number of alternatives (more than several
545 hundred) was running out of internal workspace during the pre-compile
546 phase, where pcre_compile() figures out how much memory will be needed. A
547 bit of new cunning has reduced the workspace needed for groups with
548 alternatives. The 1000-alternative test pattern now uses 12 bytes of
549 workspace instead of running out of the 4096 that are available.
550
551 10. Inserted some missing (unsigned int) casts to get rid of compiler warnings.
552
553 11. Applied patch from Google to remove an optimization that didn't quite work.
554 The report of the bug said:
555
556 pcrecpp::RE("a*").FullMatch("aaa") matches, while
557 pcrecpp::RE("a*?").FullMatch("aaa") does not, and
558 pcrecpp::RE("a*?\\z").FullMatch("aaa") does again.
559
560 12. If \p or \P was used in non-UTF-8 mode on a character greater than 127
561 it matched the wrong number of bytes.
562
563
564 Version 7.1 24-Apr-07
565 ---------------------
566
567 1. Applied Bob Rossi and Daniel G's patches to convert the build system to one
568 that is more "standard", making use of automake and other Autotools. There
569 is some re-arrangement of the files and adjustment of comments consequent
570 on this.
571
572 2. Part of the patch fixed a problem with the pcregrep tests. The test of -r
573 for recursive directory scanning broke on some systems because the files
574 are not scanned in any specific order and on different systems the order
575 was different. A call to "sort" has been inserted into RunGrepTest for the
576 approprate test as a short-term fix. In the longer term there may be an
577 alternative.
578
579 3. I had an email from Eric Raymond about problems translating some of PCRE's
580 man pages to HTML (despite the fact that I distribute HTML pages, some
581 people do their own conversions for various reasons). The problems
582 concerned the use of low-level troff macros .br and .in. I have therefore
583 removed all such uses from the man pages (some were redundant, some could
584 be replaced by .nf/.fi pairs). The 132html script that I use to generate
585 HTML has been updated to handle .nf/.fi and to complain if it encounters
586 .br or .in.
587
588 4. Updated comments in configure.ac that get placed in config.h.in and also
589 arranged for config.h to be included in the distribution, with the name
590 config.h.generic, for the benefit of those who have to compile without
591 Autotools (compare pcre.h, which is now distributed as pcre.h.generic).
592
593 5. Updated the support (such as it is) for Virtual Pascal, thanks to Stefan
594 Weber: (1) pcre_internal.h was missing some function renames; (2) updated
595 makevp.bat for the current PCRE, using the additional files
596 makevp_c.txt, makevp_l.txt, and pcregexp.pas.
597
598 6. A Windows user reported a minor discrepancy with test 2, which turned out
599 to be caused by a trailing space on an input line that had got lost in his
600 copy. The trailing space was an accident, so I've just removed it.
601
602 7. Add -Wl,-R... flags in pcre-config.in for *BSD* systems, as I'm told
603 that is needed.
604
605 8. Mark ucp_table (in ucptable.h) and ucp_gentype (in pcre_ucp_searchfuncs.c)
606 as "const" (a) because they are and (b) because it helps the PHP
607 maintainers who have recently made a script to detect big data structures
608 in the php code that should be moved to the .rodata section. I remembered
609 to update Builducptable as well, so it won't revert if ucptable.h is ever
610 re-created.
611
612 9. Added some extra #ifdef SUPPORT_UTF8 conditionals into pcretest.c,
613 pcre_printint.src, pcre_compile.c, pcre_study.c, and pcre_tables.c, in
614 order to be able to cut out the UTF-8 tables in the latter when UTF-8
615 support is not required. This saves 1.5-2K of code, which is important in
616 some applications.
617
618 Later: more #ifdefs are needed in pcre_ord2utf8.c and pcre_valid_utf8.c
619 so as not to refer to the tables, even though these functions will never be
620 called when UTF-8 support is disabled. Otherwise there are problems with a
621 shared library.
622
623 10. Fixed two bugs in the emulated memmove() function in pcre_internal.h:
624
625 (a) It was defining its arguments as char * instead of void *.
626
627 (b) It was assuming that all moves were upwards in memory; this was true
628 a long time ago when I wrote it, but is no longer the case.
629
630 The emulated memove() is provided for those environments that have neither
631 memmove() nor bcopy(). I didn't think anyone used it these days, but that
632 is clearly not the case, as these two bugs were recently reported.
633
634 11. The script PrepareRelease is now distributed: it calls 132html, CleanTxt,
635 and Detrail to create the HTML documentation, the .txt form of the man
636 pages, and it removes trailing spaces from listed files. It also creates
637 pcre.h.generic and config.h.generic from pcre.h and config.h. In the latter
638 case, it wraps all the #defines with #ifndefs. This script should be run
639 before "make dist".
640
641 12. Fixed two fairly obscure bugs concerned with quantified caseless matching
642 with Unicode property support.
643
644 (a) For a maximizing quantifier, if the two different cases of the
645 character were of different lengths in their UTF-8 codings (there are
646 some cases like this - I found 11), and the matching function had to
647 back up over a mixture of the two cases, it incorrectly assumed they
648 were both the same length.
649
650 (b) When PCRE was configured to use the heap rather than the stack for
651 recursion during matching, it was not correctly preserving the data for
652 the other case of a UTF-8 character when checking ahead for a match
653 while processing a minimizing repeat. If the check also involved
654 matching a wide character, but failed, corruption could cause an
655 erroneous result when trying to check for a repeat of the original
656 character.
657
658 13. Some tidying changes to the testing mechanism:
659
660 (a) The RunTest script now detects the internal link size and whether there
661 is UTF-8 and UCP support by running ./pcretest -C instead of relying on
662 values substituted by "configure". (The RunGrepTest script already did
663 this for UTF-8.) The configure.ac script no longer substitutes the
664 relevant variables.
665
666 (b) The debugging options /B and /D in pcretest show the compiled bytecode
667 with length and offset values. This means that the output is different
668 for different internal link sizes. Test 2 is skipped for link sizes
669 other than 2 because of this, bypassing the problem. Unfortunately,
670 there was also a test in test 3 (the locale tests) that used /B and
671 failed for link sizes other than 2. Rather than cut the whole test out,
672 I have added a new /Z option to pcretest that replaces the length and
673 offset values with spaces. This is now used to make test 3 independent
674 of link size. (Test 2 will be tidied up later.)
675
676 14. If erroroffset was passed as NULL to pcre_compile, it provoked a
677 segmentation fault instead of returning the appropriate error message.
678
679 15. In multiline mode when the newline sequence was set to "any", the pattern
680 ^$ would give a match between the \r and \n of a subject such as "A\r\nB".
681 This doesn't seem right; it now treats the CRLF combination as the line
682 ending, and so does not match in that case. It's only a pattern such as ^$
683 that would hit this one: something like ^ABC$ would have failed after \r
684 and then tried again after \r\n.
685
686 16. Changed the comparison command for RunGrepTest from "diff -u" to "diff -ub"
687 in an attempt to make files that differ only in their line terminators
688 compare equal. This works on Linux.
689
690 17. Under certain error circumstances pcregrep might try to free random memory
691 as it exited. This is now fixed, thanks to valgrind.
692
693 19. In pcretest, if the pattern /(?m)^$/g<any> was matched against the string
694 "abc\r\n\r\n", it found an unwanted second match after the second \r. This
695 was because its rules for how to advance for /g after matching an empty
696 string at the end of a line did not allow for this case. They now check for
697 it specially.
698
699 20. pcretest is supposed to handle patterns and data of any length, by
700 extending its buffers when necessary. It was getting this wrong when the
701 buffer for a data line had to be extended.
702
703 21. Added PCRE_NEWLINE_ANYCRLF which is like ANY, but matches only CR, LF, or
704 CRLF as a newline sequence.
705
706 22. Code for handling Unicode properties in pcre_dfa_exec() wasn't being cut
707 out by #ifdef SUPPORT_UCP. This did no harm, as it could never be used, but
708 I have nevertheless tidied it up.
709
710 23. Added some casts to kill warnings from HP-UX ia64 compiler.
711
712 24. Added a man page for pcre-config.
713
714
715 Version 7.0 19-Dec-06
716 ---------------------
717
718 1. Fixed a signed/unsigned compiler warning in pcre_compile.c, shown up by
719 moving to gcc 4.1.1.
720
721 2. The -S option for pcretest uses setrlimit(); I had omitted to #include
722 sys/time.h, which is documented as needed for this function. It doesn't
723 seem to matter on Linux, but it showed up on some releases of OS X.
724
725 3. It seems that there are systems where bytes whose values are greater than
726 127 match isprint() in the "C" locale. The "C" locale should be the
727 default when a C program starts up. In most systems, only ASCII printing
728 characters match isprint(). This difference caused the output from pcretest
729 to vary, making some of the tests fail. I have changed pcretest so that:
730
731 (a) When it is outputting text in the compiled version of a pattern, bytes
732 other than 32-126 are always shown as hex escapes.
733
734 (b) When it is outputting text that is a matched part of a subject string,
735 it does the same, unless a different locale has been set for the match
736 (using the /L modifier). In this case, it uses isprint() to decide.
737
738 4. Fixed a major bug that caused incorrect computation of the amount of memory
739 required for a compiled pattern when options that changed within the
740 pattern affected the logic of the preliminary scan that determines the
741 length. The relevant options are -x, and -i in UTF-8 mode. The result was
742 that the computed length was too small. The symptoms of this bug were
743 either the PCRE error "internal error: code overflow" from pcre_compile(),
744 or a glibc crash with a message such as "pcretest: free(): invalid next
745 size (fast)". Examples of patterns that provoked this bug (shown in
746 pcretest format) are:
747
748 /(?-x: )/x
749 /(?x)(?-x: \s*#\s*)/
750 /((?i)[\x{c0}])/8
751 /(?i:[\x{c0}])/8
752
753 HOWEVER: Change 17 below makes this fix obsolete as the memory computation
754 is now done differently.
755
756 5. Applied patches from Google to: (a) add a QuoteMeta function to the C++
757 wrapper classes; (b) implement a new function in the C++ scanner that is
758 more efficient than the old way of doing things because it avoids levels of
759 recursion in the regex matching; (c) add a paragraph to the documentation
760 for the FullMatch() function.
761
762 6. The escape sequence \n was being treated as whatever was defined as
763 "newline". Not only was this contrary to the documentation, which states
764 that \n is character 10 (hex 0A), but it also went horribly wrong when
765 "newline" was defined as CRLF. This has been fixed.
766
767 7. In pcre_dfa_exec.c the value of an unsigned integer (the variable called c)
768 was being set to -1 for the "end of line" case (supposedly a value that no
769 character can have). Though this value is never used (the check for end of
770 line is "zero bytes in current character"), it caused compiler complaints.
771 I've changed it to 0xffffffff.
772
773 8. In pcre_version.c, the version string was being built by a sequence of
774 C macros that, in the event of PCRE_PRERELEASE being defined as an empty
775 string (as it is for production releases) called a macro with an empty
776 argument. The C standard says the result of this is undefined. The gcc
777 compiler treats it as an empty string (which was what was wanted) but it is
778 reported that Visual C gives an error. The source has been hacked around to
779 avoid this problem.
780
781 9. On the advice of a Windows user, included <io.h> and <fcntl.h> in Windows
782 builds of pcretest, and changed the call to _setmode() to use _O_BINARY
783 instead of 0x8000. Made all the #ifdefs test both _WIN32 and WIN32 (not all
784 of them did).
785
786 10. Originally, pcretest opened its input and output without "b"; then I was
787 told that "b" was needed in some environments, so it was added for release
788 5.0 to both the input and output. (It makes no difference on Unix-like
789 systems.) Later I was told that it is wrong for the input on Windows. I've
790 now abstracted the modes into two macros, to make it easier to fiddle with
791 them, and removed "b" from the input mode under Windows.
792
793 11. Added pkgconfig support for the C++ wrapper library, libpcrecpp.
794
795 12. Added -help and --help to pcretest as an official way of being reminded
796 of the options.
797
798 13. Removed some redundant semicolons after macro calls in pcrecpparg.h.in
799 and pcrecpp.cc because they annoy compilers at high warning levels.
800
801 14. A bit of tidying/refactoring in pcre_exec.c in the main bumpalong loop.
802
803 15. Fixed an occurrence of == in configure.ac that should have been = (shell
804 scripts are not C programs :-) and which was not noticed because it works
805 on Linux.
806
807 16. pcretest is supposed to handle any length of pattern and data line (as one
808 line or as a continued sequence of lines) by extending its input buffer if
809 necessary. This feature was broken for very long pattern lines, leading to
810 a string of junk being passed to pcre_compile() if the pattern was longer
811 than about 50K.
812
813 17. I have done a major re-factoring of the way pcre_compile() computes the
814 amount of memory needed for a compiled pattern. Previously, there was code
815 that made a preliminary scan of the pattern in order to do this. That was
816 OK when PCRE was new, but as the facilities have expanded, it has become
817 harder and harder to keep it in step with the real compile phase, and there
818 have been a number of bugs (see for example, 4 above). I have now found a
819 cunning way of running the real compile function in a "fake" mode that
820 enables it to compute how much memory it would need, while actually only
821 ever using a few hundred bytes of working memory and without too many
822 tests of the mode. This should make future maintenance and development
823 easier. A side effect of this work is that the limit of 200 on the nesting
824 depth of parentheses has been removed (though this was never a serious
825 limitation, I suspect). However, there is a downside: pcre_compile() now
826 runs more slowly than before (30% or more, depending on the pattern). I
827 hope this isn't a big issue. There is no effect on runtime performance.
828
829 18. Fixed a minor bug in pcretest: if a pattern line was not terminated by a
830 newline (only possible for the last line of a file) and it was a
831 pattern that set a locale (followed by /Lsomething), pcretest crashed.
832
833 19. Added additional timing features to pcretest. (1) The -tm option now times
834 matching only, not compiling. (2) Both -t and -tm can be followed, as a
835 separate command line item, by a number that specifies the number of
836 repeats to use when timing. The default is 50000; this gives better
837 precision, but takes uncomfortably long for very large patterns.
838
839 20. Extended pcre_study() to be more clever in cases where a branch of a
840 subpattern has no definite first character. For example, (a*|b*)[cd] would
841 previously give no result from pcre_study(). Now it recognizes that the
842 first character must be a, b, c, or d.
843
844 21. There was an incorrect error "recursive call could loop indefinitely" if
845 a subpattern (or the entire pattern) that was being tested for matching an
846 empty string contained only one non-empty item after a nested subpattern.
847 For example, the pattern (?>\x{100}*)\d(?R) provoked this error
848 incorrectly, because the \d was being skipped in the check.
849
850 22. The pcretest program now has a new pattern option /B and a command line
851 option -b, which is equivalent to adding /B to every pattern. This causes
852 it to show the compiled bytecode, without the additional information that
853 -d shows. The effect of -d is now the same as -b with -i (and similarly, /D
854 is the same as /B/I).
855
856 23. A new optimization is now able automatically to treat some sequences such
857 as a*b as a*+b. More specifically, if something simple (such as a character
858 or a simple class like \d) has an unlimited quantifier, and is followed by
859 something that cannot possibly match the quantified thing, the quantifier
860 is automatically "possessified".
861
862 24. A recursive reference to a subpattern whose number was greater than 39
863 went wrong under certain circumstances in UTF-8 mode. This bug could also
864 have affected the operation of pcre_study().
865
866 25. Realized that a little bit of performance could be had by replacing
867 (c & 0xc0) == 0xc0 with c >= 0xc0 when processing UTF-8 characters.
868
869 26. Timing data from pcretest is now shown to 4 decimal places instead of 3.
870
871 27. Possessive quantifiers such as a++ were previously implemented by turning
872 them into atomic groups such as ($>a+). Now they have their own opcodes,
873 which improves performance. This includes the automatically created ones
874 from 23 above.
875
876 28. A pattern such as (?=(\w+))\1: which simulates an atomic group using a
877 lookahead was broken if it was not anchored. PCRE was mistakenly expecting
878 the first matched character to be a colon. This applied both to named and
879 numbered groups.
880
881 29. The ucpinternal.h header file was missing its idempotency #ifdef.
882
883 30. I was sent a "project" file called libpcre.a.dev which I understand makes
884 building PCRE on Windows easier, so I have included it in the distribution.
885
886 31. There is now a check in pcretest against a ridiculously large number being
887 returned by pcre_exec() or pcre_dfa_exec(). If this happens in a /g or /G
888 loop, the loop is abandoned.
889
890 32. Forward references to subpatterns in conditions such as (?(2)...) where
891 subpattern 2 is defined later cause pcre_compile() to search forwards in
892 the pattern for the relevant set of parentheses. This search went wrong
893 when there were unescaped parentheses in a character class, parentheses
894 escaped with \Q...\E, or parentheses in a #-comment in /x mode.
895
896 33. "Subroutine" calls and backreferences were previously restricted to
897 referencing subpatterns earlier in the regex. This restriction has now
898 been removed.
899
900 34. Added a number of extra features that are going to be in Perl 5.10. On the
901 whole, these are just syntactic alternatives for features that PCRE had
902 previously implemented using the Python syntax or my own invention. The
903 other formats are all retained for compatibility.
904
905 (a) Named groups can now be defined as (?<name>...) or (?'name'...) as well
906 as (?P<name>...). The new forms, as well as being in Perl 5.10, are
907 also .NET compatible.
908
909 (b) A recursion or subroutine call to a named group can now be defined as
910 (?&name) as well as (?P>name).
911
912 (c) A backreference to a named group can now be defined as \k<name> or
913 \k'name' as well as (?P=name). The new forms, as well as being in Perl
914 5.10, are also .NET compatible.
915
916 (d) A conditional reference to a named group can now use the syntax
917 (?(<name>) or (?('name') as well as (?(name).
918
919 (e) A "conditional group" of the form (?(DEFINE)...) can be used to define
920 groups (named and numbered) that are never evaluated inline, but can be
921 called as "subroutines" from elsewhere. In effect, the DEFINE condition
922 is always false. There may be only one alternative in such a group.
923
924 (f) A test for recursion can be given as (?(R1).. or (?(R&name)... as well
925 as the simple (?(R). The condition is true only if the most recent
926 recursion is that of the given number or name. It does not search out
927 through the entire recursion stack.
928
929 (g) The escape \gN or \g{N} has been added, where N is a positive or
930 negative number, specifying an absolute or relative reference.
931
932 35. Tidied to get rid of some further signed/unsigned compiler warnings and
933 some "unreachable code" warnings.
934
935 36. Updated the Unicode property tables to Unicode version 5.0.0. Amongst other
936 things, this adds five new scripts.
937
938 37. Perl ignores orphaned \E escapes completely. PCRE now does the same.
939 There were also incompatibilities regarding the handling of \Q..\E inside
940 character classes, for example with patterns like [\Qa\E-\Qz\E] where the
941 hyphen was adjacent to \Q or \E. I hope I've cleared all this up now.
942
943 38. Like Perl, PCRE detects when an indefinitely repeated parenthesized group
944 matches an empty string, and forcibly breaks the loop. There were bugs in
945 this code in non-simple cases. For a pattern such as ^(a()*)* matched
946 against aaaa the result was just "a" rather than "aaaa", for example. Two
947 separate and independent bugs (that affected different cases) have been
948 fixed.
949
950 39. Refactored the code to abolish the use of different opcodes for small
951 capturing bracket numbers. This is a tidy that I avoided doing when I
952 removed the limit on the number of capturing brackets for 3.5 back in 2001.
953 The new approach is not only tidier, it makes it possible to reduce the
954 memory needed to fix the previous bug (38).
955
956 40. Implemented PCRE_NEWLINE_ANY to recognize any of the Unicode newline
957 sequences (http://unicode.org/unicode/reports/tr18/) as "newline" when
958 processing dot, circumflex, or dollar metacharacters, or #-comments in /x
959 mode.
960
961 41. Add \R to match any Unicode newline sequence, as suggested in the Unicode
962 report.
963
964 42. Applied patch, originally from Ari Pollak, modified by Google, to allow
965 copy construction and assignment in the C++ wrapper.
966
967 43. Updated pcregrep to support "--newline=any". In the process, I fixed a
968 couple of bugs that could have given wrong results in the "--newline=crlf"
969 case.
970
971 44. Added a number of casts and did some reorganization of signed/unsigned int
972 variables following suggestions from Dair Grant. Also renamed the variable
973 "this" as "item" because it is a C++ keyword.
974
975 45. Arranged for dftables to add
976
977 #include "pcre_internal.h"
978
979 to pcre_chartables.c because without it, gcc 4.x may remove the array
980 definition from the final binary if PCRE is built into a static library and
981 dead code stripping is activated.
982
983 46. For an unanchored pattern, if a match attempt fails at the start of a
984 newline sequence, and the newline setting is CRLF or ANY, and the next two
985 characters are CRLF, advance by two characters instead of one.
986
987
988 Version 6.7 04-Jul-06
989 ---------------------
990
991 1. In order to handle tests when input lines are enormously long, pcretest has
992 been re-factored so that it automatically extends its buffers when
993 necessary. The code is crude, but this _is_ just a test program. The
994 default size has been increased from 32K to 50K.
995
996 2. The code in pcre_study() was using the value of the re argument before
997 testing it for NULL. (Of course, in any sensible call of the function, it
998 won't be NULL.)
999
1000 3. The memmove() emulation function in pcre_internal.h, which is used on
1001 systems that lack both memmove() and bcopy() - that is, hardly ever -
1002 was missing a "static" storage class specifier.
1003
1004 4. When UTF-8 mode was not set, PCRE looped when compiling certain patterns
1005 containing an extended class (one that cannot be represented by a bitmap
1006 because it contains high-valued characters or Unicode property items, e.g.
1007 [\pZ]). Almost always one would set UTF-8 mode when processing such a
1008 pattern, but PCRE should not loop if you do not (it no longer does).
1009 [Detail: two cases were found: (a) a repeated subpattern containing an
1010 extended class; (b) a recursive reference to a subpattern that followed a
1011 previous extended class. It wasn't skipping over the extended class
1012 correctly when UTF-8 mode was not set.]
1013
1014 5. A negated single-character class was not being recognized as fixed-length
1015 in lookbehind assertions such as (?<=[^f]), leading to an incorrect
1016 compile error "lookbehind assertion is not fixed length".
1017
1018 6. The RunPerlTest auxiliary script was showing an unexpected difference
1019 between PCRE and Perl for UTF-8 tests. It turns out that it is hard to
1020 write a Perl script that can interpret lines of an input file either as
1021 byte characters or as UTF-8, which is what "perltest" was being required to
1022 do for the non-UTF-8 and UTF-8 tests, respectively. Essentially what you
1023 can't do is switch easily at run time between having the "use utf8;" pragma
1024 or not. In the end, I fudged it by using the RunPerlTest script to insert
1025 "use utf8;" explicitly for the UTF-8 tests.
1026
1027 7. In multiline (/m) mode, PCRE was matching ^ after a terminating newline at
1028 the end of the subject string, contrary to the documentation and to what
1029 Perl does. This was true of both matching functions. Now it matches only at
1030 the start of the subject and immediately after *internal* newlines.
1031
1032 8. A call of pcre_fullinfo() from pcretest to get the option bits was passing
1033 a pointer to an int instead of a pointer to an unsigned long int. This
1034 caused problems on 64-bit systems.
1035
1036 9. Applied a patch from the folks at Google to pcrecpp.cc, to fix "another
1037 instance of the 'standard' template library not being so standard".
1038
1039 10. There was no check on the number of named subpatterns nor the maximum
1040 length of a subpattern name. The product of these values is used to compute
1041 the size of the memory block for a compiled pattern. By supplying a very
1042 long subpattern name and a large number of named subpatterns, the size
1043 computation could be caused to overflow. This is now prevented by limiting
1044 the length of names to 32 characters, and the number of named subpatterns
1045 to 10,000.
1046
1047 11. Subpatterns that are repeated with specific counts have to be replicated in
1048 the compiled pattern. The size of memory for this was computed from the
1049 length of the subpattern and the repeat count. The latter is limited to
1050 65535, but there was no limit on the former, meaning that integer overflow
1051 could in principle occur. The compiled length of a repeated subpattern is
1052 now limited to 30,000 bytes in order to prevent this.
1053
1054 12. Added the optional facility to have named substrings with the same name.
1055
1056 13. Added the ability to use a named substring as a condition, using the
1057 Python syntax: (?(name)yes|no). This overloads (?(R)... and names that
1058 are numbers (not recommended). Forward references are permitted.
1059
1060 14. Added forward references in named backreferences (if you see what I mean).
1061
1062 15. In UTF-8 mode, with the PCRE_DOTALL option set, a quantified dot in the
1063 pattern could run off the end of the subject. For example, the pattern
1064 "(?s)(.{1,5})"8 did this with the subject "ab".
1065
1066 16. If PCRE_DOTALL or PCRE_MULTILINE were set, pcre_dfa_exec() behaved as if
1067 PCRE_CASELESS was set when matching characters that were quantified with ?
1068 or *.
1069
1070 17. A character class other than a single negated character that had a minimum
1071 but no maximum quantifier - for example [ab]{6,} - was not handled
1072 correctly by pce_dfa_exec(). It would match only one character.
1073
1074 18. A valid (though odd) pattern that looked like a POSIX character
1075 class but used an invalid character after [ (for example [[,abc,]]) caused
1076 pcre_compile() to give the error "Failed: internal error: code overflow" or
1077 in some cases to crash with a glibc free() error. This could even happen if
1078 the pattern terminated after [[ but there just happened to be a sequence of
1079 letters, a binary zero, and a closing ] in the memory that followed.
1080
1081 19. Perl's treatment of octal escapes in the range \400 to \777 has changed
1082 over the years. Originally (before any Unicode support), just the bottom 8
1083 bits were taken. Thus, for example, \500 really meant \100. Nowadays the
1084 output from "man perlunicode" includes this:
1085
1086 The regular expression compiler produces polymorphic opcodes. That
1087 is, the pattern adapts to the data and automatically switches to
1088 the Unicode character scheme when presented with Unicode data--or
1089 instead uses a traditional byte scheme when presented with byte
1090 data.
1091
1092 Sadly, a wide octal escape does not cause a switch, and in a string with
1093 no other multibyte characters, these octal escapes are treated as before.
1094 Thus, in Perl, the pattern /\500/ actually matches \100 but the pattern
1095 /\500|\x{1ff}/ matches \500 or \777 because the whole thing is treated as a
1096 Unicode string.
1097
1098 I have not perpetrated such confusion in PCRE. Up till now, it took just
1099 the bottom 8 bits, as in old Perl. I have now made octal escapes with
1100 values greater than \377 illegal in non-UTF-8 mode. In UTF-8 mode they
1101 translate to the appropriate multibyte character.
1102
1103 29. Applied some refactoring to reduce the number of warnings from Microsoft
1104 and Borland compilers. This has included removing the fudge introduced
1105 seven years ago for the OS/2 compiler (see 2.02/2 below) because it caused
1106 a warning about an unused variable.
1107
1108 21. PCRE has not included VT (character 0x0b) in the set of whitespace
1109 characters since release 4.0, because Perl (from release 5.004) does not.
1110 [Or at least, is documented not to: some releases seem to be in conflict
1111 with the documentation.] However, when a pattern was studied with
1112 pcre_study() and all its branches started with \s, PCRE still included VT
1113 as a possible starting character. Of course, this did no harm; it just
1114 caused an unnecessary match attempt.
1115
1116 22. Removed a now-redundant internal flag bit that recorded the fact that case
1117 dependency changed within the pattern. This was once needed for "required
1118 byte" processing, but is no longer used. This recovers a now-scarce options
1119 bit. Also moved the least significant internal flag bit to the most-
1120 significant bit of the word, which was not previously used (hangover from
1121 the days when it was an int rather than a uint) to free up another bit for
1122 the future.
1123
1124 23. Added support for CRLF line endings as well as CR and LF. As well as the
1125 default being selectable at build time, it can now be changed at runtime
1126 via the PCRE_NEWLINE_xxx flags. There are now options for pcregrep to
1127 specify that it is scanning data with non-default line endings.
1128
1129 24. Changed the definition of CXXLINK to make it agree with the definition of
1130 LINK in the Makefile, by replacing LDFLAGS to CXXFLAGS.
1131
1132 25. Applied Ian Taylor's patches to avoid using another stack frame for tail
1133 recursions. This makes a big different to stack usage for some patterns.
1134
1135 26. If a subpattern containing a named recursion or subroutine reference such
1136 as (?P>B) was quantified, for example (xxx(?P>B)){3}, the calculation of
1137 the space required for the compiled pattern went wrong and gave too small a
1138 value. Depending on the environment, this could lead to "Failed: internal
1139 error: code overflow at offset 49" or "glibc detected double free or
1140 corruption" errors.
1141
1142 27. Applied patches from Google (a) to support the new newline modes and (b) to
1143 advance over multibyte UTF-8 characters in GlobalReplace.
1144
1145 28. Change free() to pcre_free() in pcredemo.c. Apparently this makes a
1146 difference for some implementation of PCRE in some Windows version.
1147
1148 29. Added some extra testing facilities to pcretest:
1149
1150 \q<number> in a data line sets the "match limit" value
1151 \Q<number> in a data line sets the "match recursion limt" value
1152 -S <number> sets the stack size, where <number> is in megabytes
1153
1154 The -S option isn't available for Windows.
1155
1156
1157 Version 6.6 06-Feb-06
1158 ---------------------
1159
1160 1. Change 16(a) for 6.5 broke things, because PCRE_DATA_SCOPE was not defined
1161 in pcreposix.h. I have copied the definition from pcre.h.
1162
1163 2. Change 25 for 6.5 broke compilation in a build directory out-of-tree
1164 because pcre.h is no longer a built file.
1165
1166 3. Added Jeff Friedl's additional debugging patches to pcregrep. These are
1167 not normally included in the compiled code.
1168
1169
1170 Version 6.5 01-Feb-06
1171 ---------------------
1172
1173 1. When using the partial match feature with pcre_dfa_exec(), it was not
1174 anchoring the second and subsequent partial matches at the new starting
1175 point. This could lead to incorrect results. For example, with the pattern
1176 /1234/, partially matching against "123" and then "a4" gave a match.
1177
1178 2. Changes to pcregrep:
1179
1180 (a) All non-match returns from pcre_exec() were being treated as failures
1181 to match the line. Now, unless the error is PCRE_ERROR_NOMATCH, an
1182 error message is output. Some extra information is given for the
1183 PCRE_ERROR_MATCHLIMIT and PCRE_ERROR_RECURSIONLIMIT errors, which are
1184 probably the only errors that are likely to be caused by users (by
1185 specifying a regex that has nested indefinite repeats, for instance).
1186 If there are more than 20 of these errors, pcregrep is abandoned.
1187
1188 (b) A binary zero was treated as data while matching, but terminated the
1189 output line if it was written out. This has been fixed: binary zeroes
1190 are now no different to any other data bytes.
1191
1192 (c) Whichever of the LC_ALL or LC_CTYPE environment variables is set is
1193 used to set a locale for matching. The --locale=xxxx long option has
1194 been added (no short equivalent) to specify a locale explicitly on the
1195 pcregrep command, overriding the environment variables.
1196
1197 (d) When -B was used with -n, some line numbers in the output were one less
1198 than they should have been.
1199
1200 (e) Added the -o (--only-matching) option.
1201
1202 (f) If -A or -C was used with -c (count only), some lines of context were
1203 accidentally printed for the final match.
1204
1205 (g) Added the -H (--with-filename) option.
1206
1207 (h) The combination of options -rh failed to suppress file names for files
1208 that were found from directory arguments.
1209
1210 (i) Added the -D (--devices) and -d (--directories) options.
1211
1212 (j) Added the -F (--fixed-strings) option.
1213
1214 (k) Allow "-" to be used as a file name for -f as well as for a data file.
1215
1216 (l) Added the --colo(u)r option.
1217
1218 (m) Added Jeffrey Friedl's -S testing option, but within #ifdefs so that it
1219 is not present by default.
1220
1221 3. A nasty bug was discovered in the handling of recursive patterns, that is,
1222 items such as (?R) or (?1), when the recursion could match a number of
1223 alternatives. If it matched one of the alternatives, but subsequently,
1224 outside the recursion, there was a failure, the code tried to back up into
1225 the recursion. However, because of the way PCRE is implemented, this is not
1226 possible, and the result was an incorrect result from the match.
1227
1228 In order to prevent this happening, the specification of recursion has
1229 been changed so that all such subpatterns are automatically treated as
1230 atomic groups. Thus, for example, (?R) is treated as if it were (?>(?R)).
1231
1232 4. I had overlooked the fact that, in some locales, there are characters for
1233 which isalpha() is true but neither isupper() nor islower() are true. In
1234 the fr_FR locale, for instance, the \xAA and \xBA characters (ordmasculine
1235 and ordfeminine) are like this. This affected the treatment of \w and \W
1236 when they appeared in character classes, but not when they appeared outside
1237 a character class. The bit map for "word" characters is now created
1238 separately from the results of isalnum() instead of just taking it from the
1239 upper, lower, and digit maps. (Plus the underscore character, of course.)
1240
1241 5. The above bug also affected the handling of POSIX character classes such as
1242 [[:alpha:]] and [[:alnum:]]. These do not have their own bit maps in PCRE's
1243 permanent tables. Instead, the bit maps for such a class were previously
1244 created as the appropriate unions of the upper, lower, and digit bitmaps.
1245 Now they are created by subtraction from the [[:word:]] class, which has
1246 its own bitmap.
1247
1248 6. The [[:blank:]] character class matches horizontal, but not vertical space.
1249 It is created by subtracting the vertical space characters (\x09, \x0a,
1250 \x0b, \x0c) from the [[:space:]] bitmap. Previously, however, the
1251 subtraction was done in the overall bitmap for a character class, meaning
1252 that a class such as [\x0c[:blank:]] was incorrect because \x0c would not
1253 be recognized. This bug has been fixed.
1254
1255 7. Patches from the folks at Google:
1256
1257 (a) pcrecpp.cc: "to handle a corner case that may or may not happen in
1258 real life, but is still worth protecting against".
1259
1260 (b) pcrecpp.cc: "corrects a bug when negative radixes are used with
1261 regular expressions".
1262
1263 (c) pcre_scanner.cc: avoid use of std::count() because not all systems
1264 have it.
1265
1266 (d) Split off pcrecpparg.h from pcrecpp.h and had the former built by
1267 "configure" and the latter not, in order to fix a problem somebody had
1268 with compiling the Arg class on HP-UX.
1269
1270 (e) Improve the error-handling of the C++ wrapper a little bit.
1271
1272 (f) New tests for checking recursion limiting.
1273
1274 8. The pcre_memmove() function, which is used only if the environment does not
1275 have a standard memmove() function (and is therefore rarely compiled),
1276 contained two bugs: (a) use of int instead of size_t, and (b) it was not
1277 returning a result (though PCRE never actually uses the result).
1278
1279 9. In the POSIX regexec() interface, if nmatch is specified as a ridiculously
1280 large number - greater than INT_MAX/(3*sizeof(int)) - REG_ESPACE is
1281 returned instead of calling malloc() with an overflowing number that would
1282 most likely cause subsequent chaos.
1283
1284 10. The debugging option of pcretest was not showing the NO_AUTO_CAPTURE flag.
1285
1286 11. The POSIX flag REG_NOSUB is now supported. When a pattern that was compiled
1287 with this option is matched, the nmatch and pmatch options of regexec() are
1288 ignored.
1289
1290 12. Added REG_UTF8 to the POSIX interface. This is not defined by POSIX, but is
1291 provided in case anyone wants to the the POSIX interface with UTF-8
1292 strings.
1293
1294 13. Added CXXLDFLAGS to the Makefile parameters to provide settings only on the
1295 C++ linking (needed for some HP-UX environments).
1296
1297 14. Avoid compiler warnings in get_ucpname() when compiled without UCP support
1298 (unused parameter) and in the pcre_printint() function (omitted "default"
1299 switch label when the default is to do nothing).
1300
1301 15. Added some code to make it possible, when PCRE is compiled as a C++
1302 library, to replace subject pointers for pcre_exec() with a smart pointer
1303 class, thus making it possible to process discontinuous strings.
1304
1305 16. The two macros PCRE_EXPORT and PCRE_DATA_SCOPE are confusing, and perform
1306 much the same function. They were added by different people who were trying
1307 to make PCRE easy to compile on non-Unix systems. It has been suggested
1308 that PCRE_EXPORT be abolished now that there is more automatic apparatus
1309 for compiling on Windows systems. I have therefore replaced it with
1310 PCRE_DATA_SCOPE. This is set automatically for Windows; if not set it
1311 defaults to "extern" for C or "extern C" for C++, which works fine on
1312 Unix-like systems. It is now possible to override the value of PCRE_DATA_
1313 SCOPE with something explicit in config.h. In addition:
1314
1315 (a) pcreposix.h still had just "extern" instead of either of these macros;
1316 I have replaced it with PCRE_DATA_SCOPE.
1317
1318 (b) Functions such as _pcre_xclass(), which are internal to the library,
1319 but external in the C sense, all had PCRE_EXPORT in their definitions.
1320 This is apparently wrong for the Windows case, so I have removed it.
1321 (It makes no difference on Unix-like systems.)
1322
1323 17. Added a new limit, MATCH_LIMIT_RECURSION, which limits the depth of nesting
1324 of recursive calls to match(). This is different to MATCH_LIMIT because
1325 that limits the total number of calls to match(), not all of which increase
1326 the depth of recursion. Limiting the recursion depth limits the amount of
1327 stack (or heap if NO_RECURSE is set) that is used. The default can be set
1328 when PCRE is compiled, and changed at run time. A patch from Google adds
1329 this functionality to the C++ interface.
1330
1331 18. Changes to the handling of Unicode character properties:
1332
1333 (a) Updated the table to Unicode 4.1.0.
1334
1335 (b) Recognize characters that are not in the table as "Cn" (undefined).
1336
1337 (c) I revised the way the table is implemented to a much improved format
1338 which includes recognition of ranges. It now supports the ranges that
1339 are defined in UnicodeData.txt, and it also amalgamates other
1340 characters into ranges. This has reduced the number of entries in the
1341 table from around 16,000 to around 3,000, thus reducing its size
1342 considerably. I realized I did not need to use a tree structure after
1343 all - a binary chop search is just as efficient. Having reduced the
1344 number of entries, I extended their size from 6 bytes to 8 bytes to
1345 allow for more data.
1346
1347 (d) Added support for Unicode script names via properties such as \p{Han}.
1348
1349 19. In UTF-8 mode, a backslash followed by a non-Ascii character was not
1350 matching that character.
1351
1352 20. When matching a repeated Unicode property with a minimum greater than zero,
1353 (for example \pL{2,}), PCRE could look past the end of the subject if it
1354 reached it while seeking the minimum number of characters. This could
1355 happen only if some of the characters were more than one byte long, because
1356 there is a check for at least the minimum number of bytes.
1357
1358 21. Refactored the implementation of \p and \P so as to be more general, to
1359 allow for more different types of property in future. This has changed the
1360 compiled form incompatibly. Anybody with saved compiled patterns that use
1361 \p or \P will have to recompile them.
1362
1363 22. Added "Any" and "L&" to the supported property types.
1364
1365 23. Recognize \x{...} as a code point specifier, even when not in UTF-8 mode,
1366 but give a compile time error if the value is greater than 0xff.
1367
1368 24. The man pages for pcrepartial, pcreprecompile, and pcre_compile2 were
1369 accidentally not being installed or uninstalled.
1370
1371 25. The pcre.h file was built from pcre.h.in, but the only changes that were
1372 made were to insert the current release number. This seemed silly, because
1373 it made things harder for people building PCRE on systems that don't run
1374 "configure". I have turned pcre.h into a distributed file, no longer built
1375 by "configure", with the version identification directly included. There is
1376 no longer a pcre.h.in file.
1377
1378 However, this change necessitated a change to the pcre-config script as
1379 well. It is built from pcre-config.in, and one of the substitutions was the
1380 release number. I have updated configure.ac so that ./configure now finds
1381 the release number by grepping pcre.h.
1382
1383 26. Added the ability to run the tests under valgrind.
1384
1385
1386 Version 6.4 05-Sep-05
1387 ---------------------
1388
1389 1. Change 6.0/10/(l) to pcregrep introduced a bug that caused separator lines
1390 "--" to be printed when multiple files were scanned, even when none of the
1391 -A, -B, or -C options were used. This is not compatible with Gnu grep, so I
1392 consider it to be a bug, and have restored the previous behaviour.
1393
1394 2. A couple of code tidies to get rid of compiler warnings.
1395
1396 3. The pcretest program used to cheat by referring to symbols in the library
1397 whose names begin with _pcre_. These are internal symbols that are not
1398 really supposed to be visible externally, and in some environments it is
1399 possible to suppress them. The cheating is now confined to including
1400 certain files from the library's source, which is a bit cleaner.
1401
1402 4. Renamed pcre.in as pcre.h.in to go with pcrecpp.h.in; it also makes the
1403 file's purpose clearer.
1404
1405 5. Reorganized pcre_ucp_findchar().
1406
1407
1408 Version 6.3 15-Aug-05
1409 ---------------------
1410
1411 1. The file libpcre.pc.in did not have general read permission in the tarball.
1412
1413 2. There were some problems when building without C++ support:
1414
1415 (a) If C++ support was not built, "make install" and "make test" still
1416 tried to test it.
1417
1418 (b) There were problems when the value of CXX was explicitly set. Some
1419 changes have been made to try to fix these, and ...
1420
1421 (c) --disable-cpp can now be used to explicitly disable C++ support.
1422
1423 (d) The use of @CPP_OBJ@ directly caused a blank line preceded by a
1424 backslash in a target when C++ was disabled. This confuses some
1425 versions of "make", apparently. Using an intermediate variable solves
1426 this. (Same for CPP_LOBJ.)
1427
1428 3. $(LINK_FOR_BUILD) now includes $(CFLAGS_FOR_BUILD) and $(LINK)
1429 (non-Windows) now includes $(CFLAGS) because these flags are sometimes
1430 necessary on certain architectures.
1431
1432 4. Added a setting of -export-symbols-regex to the link command to remove
1433 those symbols that are exported in the C sense, but actually are local
1434 within the library, and not documented. Their names all begin with
1435 "_pcre_". This is not a perfect job, because (a) we have to except some
1436 symbols that pcretest ("illegally") uses, and (b) the facility isn't always
1437 available (and never for static libraries). I have made a note to try to
1438 find a way round (a) in the future.
1439
1440
1441 Version 6.2 01-Aug-05
1442 ---------------------
1443
1444 1. There was no test for integer overflow of quantifier values. A construction
1445 such as {1111111111111111} would give undefined results. What is worse, if
1446 a minimum quantifier for a parenthesized subpattern overflowed and became
1447 negative, the calculation of the memory size went wrong. This could have
1448 led to memory overwriting.
1449
1450 2. Building PCRE using VPATH was broken. Hopefully it is now fixed.
1451
1452 3. Added "b" to the 2nd argument of fopen() in dftables.c, for non-Unix-like
1453 operating environments where this matters.
1454
1455 4. Applied Giuseppe Maxia's patch to add additional features for controlling
1456 PCRE options from within the C++ wrapper.
1457
1458 5. Named capturing subpatterns were not being correctly counted when a pattern
1459 was compiled. This caused two problems: (a) If there were more than 100
1460 such subpatterns, the calculation of the memory needed for the whole
1461 compiled pattern went wrong, leading to an overflow error. (b) Numerical
1462 back references of the form \12, where the number was greater than 9, were
1463 not recognized as back references, even though there were sufficient
1464 previous subpatterns.
1465
1466 6. Two minor patches to pcrecpp.cc in order to allow it to compile on older
1467 versions of gcc, e.g. 2.95.4.
1468
1469
1470 Version 6.1 21-Jun-05
1471 ---------------------
1472
1473 1. There was one reference to the variable "posix" in pcretest.c that was not
1474 surrounded by "#if !defined NOPOSIX".
1475
1476 2. Make it possible to compile pcretest without DFA support, UTF8 support, or
1477 the cross-check on the old pcre_info() function, for the benefit of the
1478 cut-down version of PCRE that is currently imported into Exim.
1479
1480 3. A (silly) pattern starting with (?i)(?-i) caused an internal space
1481 allocation error. I've done the easy fix, which wastes 2 bytes for sensible
1482 patterns that start (?i) but I don't think that matters. The use of (?i) is
1483 just an example; this all applies to the other options as well.
1484
1485 4. Since libtool seems to echo the compile commands it is issuing, the output
1486 from "make" can be reduced a bit by putting "@" in front of each libtool
1487 compile command.
1488
1489 5. Patch from the folks at Google for configure.in to be a bit more thorough
1490 in checking for a suitable C++ installation before trying to compile the
1491 C++ stuff. This should fix a reported problem when a compiler was present,
1492 but no suitable headers.
1493
1494 6. The man pages all had just "PCRE" as their title. I have changed them to
1495 be the relevant file name. I have also arranged that these names are
1496 retained in the file doc/pcre.txt, which is a concatenation in text format
1497 of all the man pages except the little individual ones for each function.
1498
1499 7. The NON-UNIX-USE file had not been updated for the different set of source
1500 files that come with release 6. I also added a few comments about the C++
1501 wrapper.
1502
1503
1504 Version 6.0 07-Jun-05
1505 ---------------------
1506
1507 1. Some minor internal re-organization to help with my DFA experiments.
1508
1509 2. Some missing #ifdef SUPPORT_UCP conditionals in pcretest and printint that
1510 didn't matter for the library itself when fully configured, but did matter
1511 when compiling without UCP support, or within Exim, where the ucp files are
1512 not imported.
1513
1514 3. Refactoring of the library code to split up the various functions into
1515 different source modules. The addition of the new DFA matching code (see
1516 below) to a single monolithic source would have made it really too
1517 unwieldy, quite apart from causing all the code to be include in a
1518 statically linked application, when only some functions are used. This is
1519 relevant even without the DFA addition now that patterns can be compiled in
1520 one application and matched in another.
1521
1522 The downside of splitting up is that there have to be some external
1523 functions and data tables that are used internally in different modules of
1524 the library but which are not part of the API. These have all had their
1525 names changed to start with "_pcre_" so that they are unlikely to clash
1526 with other external names.
1527
1528 4. Added an alternate matching function, pcre_dfa_exec(), which matches using
1529 a different (DFA) algorithm. Although it is slower than the original
1530 function, it does have some advantages for certain types of matching
1531 problem.
1532
1533 5. Upgrades to pcretest in order to test the features of pcre_dfa_exec(),
1534 including restarting after a partial match.
1535
1536 6. A patch for pcregrep that defines INVALID_FILE_ATTRIBUTES if it is not
1537 defined when compiling for Windows was sent to me. I have put it into the
1538 code, though I have no means of testing or verifying it.
1539
1540 7. Added the pcre_refcount() auxiliary function.
1541
1542 8. Added the PCRE_FIRSTLINE option. This constrains an unanchored pattern to
1543 match before or at the first newline in the subject string. In pcretest,
1544 the /f option on a pattern can be used to set this.
1545
1546 9. A repeated \w when used in UTF-8 mode with characters greater than 256
1547 would behave wrongly. This has been present in PCRE since release 4.0.
1548
1549 10. A number of changes to the pcregrep command:
1550
1551 (a) Refactored how -x works; insert ^(...)$ instead of setting
1552 PCRE_ANCHORED and checking the length, in preparation for adding
1553 something similar for -w.
1554
1555 (b) Added the -w (match as a word) option.
1556
1557 (c) Refactored the way lines are read and buffered so as to have more
1558 than one at a time available.
1559
1560 (d) Implemented a pcregrep test script.
1561
1562 (e) Added the -M (multiline match) option. This allows patterns to match
1563 over several lines of the subject. The buffering ensures that at least
1564 8K, or the rest of the document (whichever is the shorter) is available
1565 for matching (and similarly the previous 8K for lookbehind assertions).
1566
1567 (f) Changed the --help output so that it now says
1568
1569 -w, --word-regex(p)
1570
1571 instead of two lines, one with "regex" and the other with "regexp"
1572 because that confused at least one person since the short forms are the
1573 same. (This required a bit of code, as the output is generated
1574 automatically from a table. It wasn't just a text change.)
1575
1576 (g) -- can be used to terminate pcregrep options if the next thing isn't an
1577 option but starts with a hyphen. Could be a pattern or a path name
1578 starting with a hyphen, for instance.
1579
1580 (h) "-" can be given as a file name to represent stdin.
1581
1582 (i) When file names are being printed, "(standard input)" is used for
1583 the standard input, for compatibility with GNU grep. Previously
1584 "<stdin>" was used.
1585
1586 (j) The option --label=xxx can be used to supply a name to be used for
1587 stdin when file names are being printed. There is no short form.
1588
1589 (k) Re-factored the options decoding logic because we are going to add
1590 two more options that take data. Such options can now be given in four
1591 different ways, e.g. "-fname", "-f name", "--file=name", "--file name".
1592
1593 (l) Added the -A, -B, and -C options for requesting that lines of context
1594 around matches be printed.
1595
1596 (m) Added the -L option to print the names of files that do not contain
1597 any matching lines, that is, the complement of -l.
1598
1599 (n) The return code is 2 if any file cannot be opened, but pcregrep does
1600 continue to scan other files.
1601
1602 (o) The -s option was incorrectly implemented. For compatibility with other
1603 greps, it now suppresses the error message for a non-existent or non-
1604 accessible file (but not the return code). There is a new option called
1605 -q that suppresses the output of matching lines, which was what -s was
1606 previously doing.
1607
1608 (p) Added --include and --exclude options to specify files for inclusion
1609 and exclusion when recursing.
1610
1611 11. The Makefile was not using the Autoconf-supported LDFLAGS macro properly.
1612 Hopefully, it now does.
1613
1614 12. Missing cast in pcre_study().
1615
1616 13. Added an "uninstall" target to the makefile.
1617
1618 14. Replaced "extern" in the function prototypes in Makefile.in with
1619 "PCRE_DATA_SCOPE", which defaults to 'extern' or 'extern "C"' in the Unix
1620 world, but is set differently for Windows.
1621
1622 15. Added a second compiling function called pcre_compile2(). The only
1623 difference is that it has an extra argument, which is a pointer to an
1624 integer error code. When there is a compile-time failure, this is set
1625 non-zero, in addition to the error test pointer being set to point to an
1626 error message. The new argument may be NULL if no error number is required
1627 (but then you may as well call pcre_compile(), which is now just a
1628 wrapper). This facility is provided because some applications need a
1629 numeric error indication, but it has also enabled me to tidy up the way
1630 compile-time errors are handled in the POSIX wrapper.
1631
1632 16. Added VPATH=.libs to the makefile; this should help when building with one
1633 prefix path and installing with another. (Or so I'm told by someone who
1634 knows more about this stuff than I do.)
1635
1636 17. Added a new option, REG_DOTALL, to the POSIX function regcomp(). This
1637 passes PCRE_DOTALL to the pcre_compile() function, making the "." character
1638 match everything, including newlines. This is not POSIX-compatible, but
1639 somebody wanted the feature. From pcretest it can be activated by using
1640 both the P and the s flags.
1641
1642 18. AC_PROG_LIBTOOL appeared twice in Makefile.in. Removed one.
1643
1644 19. libpcre.pc was being incorrectly installed as executable.
1645
1646 20. A couple of places in pcretest check for end-of-line by looking for '\n';
1647 it now also looks for '\r' so that it will work unmodified on Windows.
1648
1649 21. Added Google's contributed C++ wrapper to the distribution.
1650
1651 22. Added some untidy missing memory free() calls in pcretest, to keep
1652 Electric Fence happy when testing.
1653
1654
1655
1656 Version 5.0 13-Sep-04
1657 ---------------------
1658
1659 1. Internal change: literal characters are no longer packed up into items
1660 containing multiple characters in a single byte-string. Each character
1661 is now matched using a separate opcode. However, there may be more than one
1662 byte in the character in UTF-8 mode.
1663
1664 2. The pcre_callout_block structure has two new fields: pattern_position and
1665 next_item_length. These contain the offset in the pattern to the next match
1666 item, and its length, respectively.
1667
1668 3. The PCRE_AUTO_CALLOUT option for pcre_compile() requests the automatic
1669 insertion of callouts before each pattern item. Added the /C option to
1670 pcretest to make use of this.
1671
1672 4. On the advice of a Windows user, the lines
1673
1674 #if defined(_WIN32) || defined(WIN32)
1675 _setmode( _fileno( stdout ), 0x8000 );
1676 #endif /* defined(_WIN32) || defined(WIN32) */
1677
1678 have been added to the source of pcretest. This apparently does useful
1679 magic in relation to line terminators.
1680
1681 5. Changed "r" and "w" in the calls to fopen() in pcretest to "rb" and "wb"
1682 for the benefit of those environments where the "b" makes a difference.
1683
1684 6. The icc compiler has the same options as gcc, but "configure" doesn't seem
1685 to know about it. I have put a hack into configure.in that adds in code
1686 to set GCC=yes if CC=icc. This seems to end up at a point in the
1687 generated configure script that is early enough to affect the setting of
1688 compiler options, which is what is needed, but I have no means of testing
1689 whether it really works. (The user who reported this had patched the
1690 generated configure script, which of course I cannot do.)
1691
1692 LATER: After change 22 below (new libtool files), the configure script
1693 seems to know about icc (and also ecc). Therefore, I have commented out
1694 this hack in configure.in.
1695
1696 7. Added support for pkg-config (2 patches were sent in).
1697
1698 8. Negated POSIX character classes that used a combination of internal tables
1699 were completely broken. These were [[:^alpha:]], [[:^alnum:]], and
1700 [[:^ascii]]. Typically, they would match almost any characters. The other
1701 POSIX classes were not broken in this way.
1702
1703 9. Matching the pattern "\b.*?" against "ab cd", starting at offset 1, failed
1704 to find the match, as PCRE was deluded into thinking that the match had to
1705 start at the start point or following a newline. The same bug applied to
1706 patterns with negative forward assertions or any backward assertions
1707 preceding ".*" at the start, unless the pattern required a fixed first
1708 character. This was a failing pattern: "(?!.bcd).*". The bug is now fixed.
1709
1710 10. In UTF-8 mode, when moving forwards in the subject after a failed match
1711 starting at the last subject character, bytes beyond the end of the subject
1712 string were read.
1713
1714 11. Renamed the variable "class" as "classbits" to make life easier for C++
1715 users. (Previously there was a macro definition, but it apparently wasn't
1716 enough.)
1717
1718 12. Added the new field "tables" to the extra data so that tables can be passed
1719 in at exec time, or the internal tables can be re-selected. This allows
1720 a compiled regex to be saved and re-used at a later time by a different
1721 program that might have everything at different addresses.
1722
1723 13. Modified the pcre-config script so that, when run on Solaris, it shows a
1724 -R library as well as a -L library.
1725
1726 14. The debugging options of pcretest (-d on the command line or D on a
1727 pattern) showed incorrect output for anything following an extended class
1728 that contained multibyte characters and which was followed by a quantifier.
1729
1730 15. Added optional support for general category Unicode character properties
1731 via the \p, \P, and \X escapes. Unicode property support implies UTF-8
1732 support. It adds about 90K to the size of the library. The meanings of the
1733 inbuilt class escapes such as \d and \s have NOT been changed.
1734
1735 16. Updated pcredemo.c to include calls to free() to release the memory for the
1736 compiled pattern.
1737
1738 17. The generated file chartables.c was being created in the source directory
1739 instead of in the building directory. This caused the build to fail if the
1740 source directory was different from the building directory, and was
1741 read-only.
1742
1743 18. Added some sample Win commands from Mark Tetrode into the NON-UNIX-USE
1744 file. No doubt somebody will tell me if they don't make sense... Also added
1745 Dan Mooney's comments about building on OpenVMS.
1746
1747 19. Added support for partial matching via the PCRE_PARTIAL option for
1748 pcre_exec() and the \P data escape in pcretest.
1749
1750 20. Extended pcretest with 3 new pattern features:
1751
1752 (i) A pattern option of the form ">rest-of-line" causes pcretest to
1753 write the compiled pattern to the file whose name is "rest-of-line".
1754 This is a straight binary dump of the data, with the saved pointer to
1755 the character tables forced to be NULL. The study data, if any, is
1756 written too. After writing, pcretest reads a new pattern.
1757
1758 (ii) If, instead of a pattern, "<rest-of-line" is given, pcretest reads a
1759 compiled pattern from the given file. There must not be any
1760 occurrences of "<" in the file name (pretty unlikely); if there are,
1761 pcretest will instead treat the initial "<" as a pattern delimiter.
1762 After reading in the pattern, pcretest goes on to read data lines as
1763 usual.
1764
1765 (iii) The F pattern option causes pcretest to flip the bytes in the 32-bit
1766 and 16-bit fields in a compiled pattern, to simulate a pattern that
1767 was compiled on a host of opposite endianness.
1768
1769 21. The pcre-exec() function can now cope with patterns that were compiled on
1770 hosts of opposite endianness, with this restriction:
1771
1772 As for any compiled expression that is saved and used later, the tables
1773 pointer field cannot be preserved; the extra_data field in the arguments
1774 to pcre_exec() should be used to pass in a tables address if a value
1775 other than the default internal tables were used at compile time.
1776
1777 22. Calling pcre_exec() with a negative value of the "ovecsize" parameter is
1778 now diagnosed as an error. Previously, most of the time, a negative number
1779 would have been treated as zero, but if in addition "ovector" was passed as
1780 NULL, a crash could occur.
1781
1782 23. Updated the files ltmain.sh, config.sub, config.guess, and aclocal.m4 with
1783 new versions from the libtool 1.5 distribution (the last one is a copy of
1784 a file called libtool.m4). This seems to have fixed the need to patch
1785 "configure" to support Darwin 1.3 (which I used to do). However, I still
1786 had to patch ltmain.sh to ensure that ${SED} is set (it isn't on my
1787 workstation).
1788
1789 24. Changed the PCRE licence to be the more standard "BSD" licence.
1790
1791
1792 Version 4.5 01-Dec-03
1793 ---------------------
1794
1795 1. There has been some re-arrangement of the code for the match() function so
1796 that it can be compiled in a version that does not call itself recursively.
1797 Instead, it keeps those local variables that need separate instances for
1798 each "recursion" in a frame on the heap, and gets/frees frames whenever it
1799 needs to "recurse". Keeping track of where control must go is done by means
1800 of setjmp/longjmp. The whole thing is implemented by a set of macros that
1801 hide most of the details from the main code, and operates only if
1802 NO_RECURSE is defined while compiling pcre.c. If PCRE is built using the
1803 "configure" mechanism, "--disable-stack-for-recursion" turns on this way of
1804 operating.
1805
1806 To make it easier for callers to provide specially tailored get/free
1807 functions for this usage, two new functions, pcre_stack_malloc, and
1808 pcre_stack_free, are used. They are always called in strict stacking order,
1809 and the size of block requested is always the same.
1810
1811 The PCRE_CONFIG_STACKRECURSE info parameter can be used to find out whether
1812 PCRE has been compiled to use the stack or the heap for recursion. The
1813 -C option of pcretest uses this to show which version is compiled.
1814
1815 A new data escape \S, is added to pcretest; it causes the amounts of store
1816 obtained and freed by both kinds of malloc/free at match time to be added
1817 to the output.
1818
1819 2. Changed the locale test to use "fr_FR" instead of "fr" because that's
1820 what's available on my current Linux desktop machine.
1821
1822 3. When matching a UTF-8 string, the test for a valid string at the start has
1823 been extended. If start_offset is not zero, PCRE now checks that it points
1824 to a byte that is the start of a UTF-8 character. If not, it returns
1825 PCRE_ERROR_BADUTF8_OFFSET (-11). Note: the whole string is still checked;
1826 this is necessary because there may be backward assertions in the pattern.
1827 When matching the same subject several times, it may save resources to use
1828 PCRE_NO_UTF8_CHECK on all but the first call if the string is long.
1829
1830 4. The code for checking the validity of UTF-8 strings has been tightened so
1831 that it rejects (a) strings containing 0xfe or 0xff bytes and (b) strings
1832 containing "overlong sequences".
1833
1834 5. Fixed a bug (appearing twice) that I could not find any way of exploiting!
1835 I had written "if ((digitab[*p++] && chtab_digit) == 0)" where the "&&"
1836 should have been "&", but it just so happened that all the cases this let
1837 through by mistake were picked up later in the function.
1838
1839 6. I had used a variable called "isblank" - this is a C99 function, causing
1840 some compilers to warn. To avoid this, I renamed it (as "blankclass").
1841
1842 7. Cosmetic: (a) only output another newline at the end of pcretest if it is
1843 prompting; (b) run "./pcretest /dev/null" at the start of the test script
1844 so the version is shown; (c) stop "make test" echoing "./RunTest".
1845
1846 8. Added patches from David Burgess to enable PCRE to run on EBCDIC systems.
1847
1848 9. The prototype for memmove() for systems that don't have it was using
1849 size_t, but the inclusion of the header that defines size_t was later. I've
1850 moved the #includes for the C headers earlier to avoid this.
1851
1852 10. Added some adjustments to the code to make it easier to compiler on certain
1853 special systems:
1854
1855 (a) Some "const" qualifiers were missing.
1856 (b) Added the macro EXPORT before all exported functions; by default this
1857 is defined to be empty.
1858 (c) Changed the dftables auxiliary program (that builds chartables.c) so
1859 that it reads its output file name as an argument instead of writing
1860 to the standard output and assuming this can be redirected.
1861
1862 11. In UTF-8 mode, if a recursive reference (e.g. (?1)) followed a character
1863 class containing characters with values greater than 255, PCRE compilation
1864 went into a loop.
1865
1866 12. A recursive reference to a subpattern that was within another subpattern
1867 that had a minimum quantifier of zero caused PCRE to crash. For example,
1868 (x(y(?2))z)? provoked this bug with a subject that got as far as the
1869 recursion. If the recursively-called subpattern itself had a zero repeat,
1870 that was OK.
1871
1872 13. In pcretest, the buffer for reading a data line was set at 30K, but the
1873 buffer into which it was copied (for escape processing) was still set at
1874 1024, so long lines caused crashes.
1875
1876 14. A pattern such as /[ab]{1,3}+/ failed to compile, giving the error
1877 "internal error: code overflow...". This applied to any character class
1878 that was followed by a possessive quantifier.
1879
1880 15. Modified the Makefile to add libpcre.la as a prerequisite for
1881 libpcreposix.la because I was told this is needed for a parallel build to
1882 work.
1883
1884 16. If a pattern that contained .* following optional items at the start was
1885 studied, the wrong optimizing data was generated, leading to matching
1886 errors. For example, studying /[ab]*.*c/ concluded, erroneously, that any
1887 matching string must start with a or b or c. The correct conclusion for
1888 this pattern is that a match can start with any character.
1889
1890
1891 Version 4.4 13-Aug-03
1892 ---------------------
1893
1894 1. In UTF-8 mode, a character class containing characters with values between
1895 127 and 255 was not handled correctly if the compiled pattern was studied.
1896 In fixing this, I have also improved the studying algorithm for such
1897 classes (slightly).
1898
1899 2. Three internal functions had redundant arguments passed to them. Removal
1900 might give a very teeny performance improvement.
1901
1902 3. Documentation bug: the value of the capture_top field in a callout is *one
1903 more than* the number of the hightest numbered captured substring.
1904
1905 4. The Makefile linked pcretest and pcregrep with -lpcre, which could result
1906 in incorrectly linking with a previously installed version. They now link
1907 explicitly with libpcre.la.
1908
1909 5. configure.in no longer needs to recognize Cygwin specially.
1910
1911 6. A problem in pcre.in for Windows platforms is fixed.
1912
1913 7. If a pattern was successfully studied, and the -d (or /D) flag was given to
1914 pcretest, it used to include the size of the study block as part of its
1915 output. Unfortunately, the structure contains a field that has a different
1916 size on different hardware architectures. This meant that the tests that
1917 showed this size failed. As the block is currently always of a fixed size,
1918 this information isn't actually particularly useful in pcretest output, so
1919 I have just removed it.
1920
1921 8. Three pre-processor statements accidentally did not start in column 1.
1922 Sadly, there are *still* compilers around that complain, even though
1923 standard C has not required this for well over a decade. Sigh.
1924
1925 9. In pcretest, the code for checking callouts passed small integers in the
1926 callout_data field, which is a void * field. However, some picky compilers
1927 complained about the casts involved for this on 64-bit systems. Now
1928 pcretest passes the address of the small integer instead, which should get
1929 rid of the warnings.
1930
1931 10. By default, when in UTF-8 mode, PCRE now checks for valid UTF-8 strings at
1932 both compile and run time, and gives an error if an invalid UTF-8 sequence
1933 is found. There is a option for disabling this check in cases where the
1934 string is known to be correct and/or the maximum performance is wanted.
1935
1936 11. In response to a bug report, I changed one line in Makefile.in from
1937
1938 -Wl,--out-implib,.libs/lib@WIN_PREFIX@pcreposix.dll.a \
1939 to
1940 -Wl,--out-implib,.libs/@WIN_PREFIX@libpcreposix.dll.a \
1941
1942 to look similar to other lines, but I have no way of telling whether this
1943 is the right thing to do, as I do not use Windows. No doubt I'll get told
1944 if it's wrong...
1945
1946
1947 Version 4.3 21-May-03
1948 ---------------------
1949
1950 1. Two instances of @WIN_PREFIX@ omitted from the Windows targets in the
1951 Makefile.
1952
1953 2. Some refactoring to improve the quality of the code:
1954
1955 (i) The utf8_table... variables are now declared "const".
1956
1957 (ii) The code for \cx, which used the "case flipping" table to upper case
1958 lower case letters, now just substracts 32. This is ASCII-specific,
1959 but the whole concept of \cx is ASCII-specific, so it seems
1960 reasonable.
1961
1962 (iii) PCRE was using its character types table to recognize decimal and
1963 hexadecimal digits in the pattern. This is silly, because it handles
1964 only 0-9, a-f, and A-F, but the character types table is locale-
1965 specific, which means strange things might happen. A private
1966 table is now used for this - though it costs 256 bytes, a table is
1967 much faster than multiple explicit tests. Of course, the standard
1968 character types table is still used for matching digits in subject
1969 strings against \d.
1970
1971 (iv) Strictly, the identifier ESC_t is reserved by POSIX (all identifiers
1972 ending in _t are). So I've renamed it as ESC_tee.
1973
1974 3. The first argument for regexec() in the POSIX wrapper should have been
1975 defined as "const".
1976
1977 4. Changed pcretest to use malloc() for its buffers so that they can be
1978 Electric Fenced for debugging.
1979
1980 5. There were several places in the code where, in UTF-8 mode, PCRE would try
1981 to read one or more bytes before the start of the subject string. Often this
1982 had no effect on PCRE's behaviour, but in some circumstances it could
1983 provoke a segmentation fault.
1984
1985 6. A lookbehind at the start of a pattern in UTF-8 mode could also cause PCRE
1986 to try to read one or more bytes before the start of the subject string.
1987
1988 7. A lookbehind in a pattern matched in non-UTF-8 mode on a PCRE compiled with
1989 UTF-8 support could misbehave in various ways if the subject string
1990 contained bytes with the 0x80 bit set and the 0x40 bit unset in a lookbehind
1991 area. (PCRE was not checking for the UTF-8 mode flag, and trying to move
1992 back over UTF-8 characters.)
1993
1994
1995 Version 4.2 14-Apr-03
1996 ---------------------
1997
1998 1. Typo "#if SUPPORT_UTF8" instead of "#ifdef SUPPORT_UTF8" fixed.
1999
2000 2. Changes to the building process, supplied by Ronald Landheer-Cieslak
2001 [ON_WINDOWS]: new variable, "#" on non-Windows platforms
2002 [NOT_ON_WINDOWS]: new variable, "#" on Windows platforms
2003 [WIN_PREFIX]: new variable, "cyg" for Cygwin
2004 * Makefile.in: use autoconf substitution for OBJEXT, EXEEXT, BUILD_OBJEXT
2005 and BUILD_EXEEXT
2006 Note: automatic setting of the BUILD variables is not yet working
2007 set CPPFLAGS and BUILD_CPPFLAGS (but don't use yet) - should be used at
2008 compile-time but not at link-time
2009 [LINK]: use for linking executables only
2010 make different versions for Windows and non-Windows
2011 [LINKLIB]: new variable, copy of UNIX-style LINK, used for linking
2012 libraries
2013 [LINK_FOR_BUILD]: new variable
2014 [OBJEXT]: use throughout
2015 [EXEEXT]: use throughout
2016 <winshared>: new target
2017 <wininstall>: new target
2018 <dftables.o>: use native compiler
2019 <dftables>: use native linker
2020 <install>: handle Windows platform correctly
2021 <clean>: ditto
2022 <check>: ditto
2023 copy DLL to top builddir before testing
2024
2025 As part of these changes, -no-undefined was removed again. This was reported
2026 to give trouble on HP-UX 11.0, so getting rid of it seems like a good idea
2027 in any case.
2028
2029 3. Some tidies to get rid of compiler warnings:
2030
2031 . In the match_data structure, match_limit was an unsigned long int, whereas
2032 match_call_count was an int. I've made them both unsigned long ints.
2033
2034 . In pcretest the fact that a const uschar * doesn't automatically cast to
2035 a void * provoked a warning.
2036
2037 . Turning on some more compiler warnings threw up some "shadow" variables
2038 and a few more missing casts.
2039
2040 4. If PCRE was complied with UTF-8 support, but called without the PCRE_UTF8
2041 option, a class that contained a single character with a value between 128
2042 and 255 (e.g. /[\xFF]/) caused PCRE to crash.
2043
2044 5. If PCRE was compiled with UTF-8 support, but called without the PCRE_UTF8
2045 option, a class that contained several characters, but with at least one
2046 whose value was between 128 and 255 caused PCRE to crash.
2047
2048
2049 Version 4.1 12-Mar-03
2050 ---------------------
2051
2052 1. Compiling with gcc -pedantic found a couple of places where casts were
2053 needed, and a string in dftables.c that was longer than standard compilers are
2054 required to support.
2055
2056 2. Compiling with Sun's compiler found a few more places where the code could
2057 be tidied up in order to avoid warnings.
2058
2059 3. The variables for cross-compiling were called HOST_CC and HOST_CFLAGS; the
2060 first of these names is deprecated in the latest Autoconf in favour of the name
2061 CC_FOR_BUILD, because "host" is typically used to mean the system on which the
2062 compiled code will be run. I can't find a reference for HOST_CFLAGS, but by
2063 analogy I have changed it to CFLAGS_FOR_BUILD.
2064
2065 4. Added -no-undefined to the linking command in the Makefile, because this is
2066 apparently helpful for Windows. To make it work, also added "-L. -lpcre" to the
2067 linking step for the pcreposix library.
2068
2069 5. PCRE was failing to diagnose the case of two named groups with the same
2070 name.
2071
2072 6. A problem with one of PCRE's optimizations was discovered. PCRE remembers a
2073 literal character that is needed in the subject for a match, and scans along to
2074 ensure that it is present before embarking on the full matching process. This
2075 saves time in cases of nested unlimited repeats that are never going to match.
2076 Problem: the scan can take a lot of time if the subject is very long (e.g.
2077 megabytes), thus penalizing straightforward matches. It is now done only if the
2078 amount of subject to be scanned is less than 1000 bytes.
2079
2080 7. A lesser problem with the same optimization is that it was recording the
2081 first character of an anchored pattern as "needed", thus provoking a search
2082 right along the subject, even when the first match of the pattern was going to
2083 fail. The "needed" character is now not set for anchored patterns, unless it
2084 follows something in the pattern that is of non-fixed length. Thus, it still
2085 fulfils its original purpose of finding quick non-matches in cases of nested
2086 unlimited repeats, but isn't used for simple anchored patterns such as /^abc/.
2087
2088
2089 Version 4.0 17-Feb-03
2090 ---------------------
2091
2092 1. If a comment in an extended regex that started immediately after a meta-item
2093 extended to the end of string, PCRE compiled incorrect data. This could lead to
2094 all kinds of weird effects. Example: /#/ was bad; /()#/ was bad; /a#/ was not.
2095
2096 2. Moved to autoconf 2.53 and libtool 1.4.2.
2097
2098 3. Perl 5.8 no longer needs "use utf8" for doing UTF-8 things. Consequently,
2099 the special perltest8 script is no longer needed - all the tests can be run
2100 from a single perltest script.
2101
2102 4. From 5.004, Perl has not included the VT character (0x0b) in the set defined
2103 by \s. It has now been removed in PCRE. This means it isn't recognized as
2104 whitespace in /x regexes too, which is the same as Perl. Note that the POSIX
2105 class [:space:] *does* include VT, thereby creating a mess.
2106
2107 5. Added the class [:blank:] (a GNU extension from Perl 5.8) to match only
2108 space and tab.
2109
2110 6. Perl 5.005 was a long time ago. It's time to amalgamate the tests that use
2111 its new features into the main test script, reducing the number of scripts.
2112
2113 7. Perl 5.8 has changed the meaning of patterns like /a(?i)b/. Earlier versions
2114 were backward compatible, and made the (?i) apply to the whole pattern, as if
2115 /i were given. Now it behaves more logically, and applies the option setting
2116 only to what follows. PCRE has been changed to follow suit. However, if it
2117 finds options settings right at the start of the pattern, it extracts them into
2118 the global options, as before. Thus, they show up in the info data.
2119
2120 8. Added support for the \Q...\E escape sequence. Characters in between are
2121 treated as literals. This is slightly different from Perl in that $ and @ are
2122 also handled as literals inside the quotes. In Perl, they will cause variable
2123 interpolation. Note the following examples:
2124
2125 Pattern PCRE matches Perl matches
2126
2127 \Qabc$xyz\E abc$xyz abc followed by the contents of $xyz
2128 \Qabc\$xyz\E abc\$xyz abc\$xyz
2129 \Qabc\E\$\Qxyz\E abc$xyz abc$xyz
2130
2131 For compatibility with Perl, \Q...\E sequences are recognized inside character
2132 classes as well as outside them.
2133
2134 9. Re-organized 3 code statements in pcretest to avoid "overflow in
2135 floating-point constant arithmetic" warnings from a Microsoft compiler. Added a
2136 (size_t) cast to one statement in pcretest and one in pcreposix to avoid
2137 signed/unsigned warnings.
2138
2139 10. SunOS4 doesn't have strtoul(). This was used only for unpicking the -o
2140 option for pcretest, so I've replaced it by a simple function that does just
2141 that job.
2142
2143 11. pcregrep was ending with code 0 instead of 2 for the commands "pcregrep" or
2144 "pcregrep -".
2145
2146 12. Added "possessive quantifiers" ?+, *+, ++, and {,}+ which come from Sun's
2147 Java package. This provides some syntactic sugar for simple cases of what my
2148 documentation calls "once-only subpatterns". A pattern such as x*+ is the same
2149 as (?>x*). In other words, if what is inside (?>...) is just a single repeated
2150 item, you can use this simplified notation. Note that only makes sense with
2151 greedy quantifiers. Consequently, the use of the possessive quantifier forces
2152 greediness, whatever the setting of the PCRE_UNGREEDY option.
2153
2154 13. A change of greediness default within a pattern was not taking effect at
2155 the current level for patterns like /(b+(?U)a+)/. It did apply to parenthesized
2156 subpatterns that followed. Patterns like /b+(?U)a+/ worked because the option
2157 was abstracted outside.
2158
2159 14. PCRE now supports the \G assertion. It is true when the current matching
2160 position is at the start point of the match. This differs from \A when the
2161 starting offset is non-zero. Used with the /g option of pcretest (or similar
2162 code), it works in the same way as it does for Perl's /g option. If all
2163 alternatives of a regex begin with \G, the expression is anchored to the start
2164 match position, and the "anchored" flag is set in the compiled expression.
2165
2166 15. Some bugs concerning the handling of certain option changes within patterns
2167 have been fixed. These applied to options other than (?ims). For example,
2168 "a(?x: b c )d" did not match "XabcdY" but did match "Xa b c dY". It should have
2169 been the other way round. Some of this was related to change 7 above.
2170
2171 16. PCRE now gives errors for /[.x.]/ and /[=x=]/ as unsupported POSIX
2172 features, as Perl does. Previously, PCRE gave the warnings only for /[[.x.]]/
2173 and /[[=x=]]/. PCRE now also gives an error for /[:name:]/ because it supports
2174 POSIX classes only within a class (e.g. /[[:alpha:]]/).
2175
2176 17. Added support for Perl's \C escape. This matches one byte, even in UTF8
2177 mode. Unlike ".", it always matches newline, whatever the setting of
2178 PCRE_DOTALL. However, PCRE does not permit \C to appear in lookbehind
2179 assertions. Perl allows it, but it doesn't (in general) work because it can't
2180 calculate the length of the lookbehind. At least, that's the case for Perl
2181 5.8.0 - I've been told they are going to document that it doesn't work in
2182 future.
2183
2184 18. Added an error diagnosis for escapes that PCRE does not support: these are
2185 \L, \l, \N, \P, \p, \U, \u, and \X.
2186
2187 19. Although correctly diagnosing a missing ']' in a character class, PCRE was
2188 reading past the end of the pattern in cases such as /[abcd/.
2189
2190 20. PCRE was getting more memory than necessary for patterns with classes that
2191 contained both POSIX named classes and other characters, e.g. /[[:space:]abc/.
2192
2193 21. Added some code, conditional on #ifdef VPCOMPAT, to make life easier for
2194 compiling PCRE for use with Virtual Pascal.
2195
2196 22. Small fix to the Makefile to make it work properly if the build is done
2197 outside the source tree.
2198
2199 23. Added a new extension: a condition to go with recursion. If a conditional
2200 subpattern starts with (?(R) the "true" branch is used if recursion has
2201 happened, whereas the "false" branch is used only at the top level.
2202
2203 24. When there was a very long string of literal characters (over 255 bytes
2204 without UTF support, over 250 bytes with UTF support), the computation of how
2205 much memory was required could be incorrect, leading to segfaults or other
2206 strange effects.
2207
2208 25. PCRE was incorrectly assuming anchoring (either to start of subject or to
2209 start of line for a non-DOTALL pattern) when a pattern started with (.*) and
2210 there was a subsequent back reference to those brackets. This meant that, for
2211 example, /(.*)\d+\1/ failed to match "abc123bc". Unfortunately, it isn't
2212 possible to check for precisely this case. All we can do is abandon the
2213 optimization if .* occurs inside capturing brackets when there are any back
2214 references whatsoever. (See below for a better fix that came later.)
2215
2216 26. The handling of the optimization for finding the first character of a
2217 non-anchored pattern, and for finding a character that is required later in the
2218 match were failing in some cases. This didn't break the matching; it just
2219 failed to optimize when it could. The way this is done has been re-implemented.
2220
2221 27. Fixed typo in error message for invalid (?R item (it said "(?p").
2222
2223 28. Added a new feature that provides some of the functionality that Perl
2224 provides with (?{...}). The facility is termed a "callout". The way it is done
2225 in PCRE is for the caller to provide an optional function, by setting
2226 pcre_callout to its entry point. Like pcre_malloc and pcre_free, this is a
2227 global variable. By default it is unset, which disables all calling out. To get
2228 the function called, the regex must include (?C) at appropriate points. This
2229 is, in fact, equivalent to (?C0), and any number <= 255 may be given with (?C).
2230 This provides a means of identifying different callout points. When PCRE
2231 reaches such a point in the regex, if pcre_callout has been set, the external
2232 function is called. It is provided with data in a structure called
2233 pcre_callout_block, which is defined in pcre.h. If the function returns 0,
2234 matching continues; if it returns a non-zero value, the match at the current
2235 point fails. However, backtracking will occur if possible. [This was changed
2236 later and other features added - see item 49 below.]
2237
2238 29. pcretest is upgraded to test the callout functionality. It provides a
2239 callout function that displays information. By default, it shows the start of
2240 the match and the current position in the text. There are some new data escapes
2241 to vary what happens:
2242
2243 \C+ in addition, show current contents of captured substrings
2244 \C- do not supply a callout function
2245 \C!n return 1 when callout number n is reached
2246 \C!n!m return 1 when callout number n is reached for the mth time
2247
2248 30. If pcregrep was called with the -l option and just a single file name, it
2249 output "<stdin>" if a match was found, instead of the file name.
2250
2251 31. Improve the efficiency of the POSIX API to PCRE. If the number of capturing
2252 slots is less than POSIX_MALLOC_THRESHOLD, use a block on the stack to pass to
2253 pcre_exec(). This saves a malloc/free per call. The default value of
2254 POSIX_MALLOC_THRESHOLD is 10; it can be changed by --with-posix-malloc-threshold
2255 when configuring.
2256
2257 32. The default maximum size of a compiled pattern is 64K. There have been a
2258 few cases of people hitting this limit. The code now uses macros to handle the
2259 storing of links as offsets within the compiled pattern. It defaults to 2-byte
2260 links, but this can be changed to 3 or 4 bytes by --with-link-size when
2261 configuring. Tests 2 and 5 work only with 2-byte links because they output
2262 debugging information about compiled patterns.
2263
2264 33. Internal code re-arrangements:
2265
2266 (a) Moved the debugging function for printing out a compiled regex into
2267 its own source file (printint.c) and used #include to pull it into
2268 pcretest.c and, when DEBUG is defined, into pcre.c, instead of having two
2269 separate copies.
2270
2271 (b) Defined the list of op-code names for debugging as a macro in
2272 internal.h so that it is next to the definition of the opcodes.
2273
2274 (c) Defined a table of op-code lengths for simpler skipping along compiled
2275 code. This is again a macro in internal.h so that it is next to the
2276 definition of the opcodes.
2277
2278 34. Added support for recursive calls to individual subpatterns, along the
2279 lines of Robin Houston's patch (but implemented somewhat differently).
2280
2281 35. Further mods to the Makefile to help Win32. Also, added code to pcregrep to
2282 allow it to read and process whole directories in Win32. This code was
2283 contributed by Lionel Fourquaux; it has not been tested by me.
2284
2285 36. Added support for named subpatterns. The Python syntax (?P<name>...) is
2286 used to name a group. Names consist of alphanumerics and underscores, and must
2287 be unique. Back references use the syntax (?P=name) and recursive calls use
2288 (?P>name) which is a PCRE extension to the Python extension. Groups still have
2289 numbers. The function pcre_fullinfo() can be used after compilation to extract
2290 a name/number map. There are three relevant calls:
2291
2292 PCRE_INFO_NAMEENTRYSIZE yields the size of each entry in the map
2293 PCRE_INFO_NAMECOUNT yields the number of entries
2294 PCRE_INFO_NAMETABLE yields a pointer to the map.
2295
2296 The map is a vector of fixed-size entries. The size of each entry depends on
2297 the length of the longest name used. The first two bytes of each entry are the
2298 group number, most significant byte first. There follows the corresponding
2299 name, zero terminated. The names are in alphabetical order.
2300
2301 37. Make the maximum literal string in the compiled code 250 for the non-UTF-8
2302 case instead of 255. Making it the same both with and without UTF-8 support
2303 means that the same test output works with both.
2304
2305 38. There was a case of malloc(0) in the POSIX testing code in pcretest. Avoid
2306 calling malloc() with a zero argument.
2307
2308 39. Change 25 above had to resort to a heavy-handed test for the .* anchoring
2309 optimization. I've improved things by keeping a bitmap of backreferences with
2310 numbers 1-31 so that if .* occurs inside capturing brackets that are not in
2311 fact referenced, the optimization can be applied. It is unlikely that a
2312 relevant occurrence of .* (i.e. one which might indicate anchoring or forcing
2313 the match to follow \n) will appear inside brackets with a number greater than
2314 31, but if it does, any back reference > 31 suppresses the optimization.
2315
2316 40. Added a new compile-time option PCRE_NO_AUTO_CAPTURE. This has the effect
2317 of disabling numbered capturing parentheses. Any opening parenthesis that is
2318 not followed by ? behaves as if it were followed by ?: but named parentheses
2319 can still be used for capturing (and they will acquire numbers in the usual
2320 way).
2321
2322 41. Redesigned the return codes from the match() function into yes/no/error so
2323 that errors can be passed back from deep inside the nested calls. A malloc
2324 failure while inside a recursive subpattern call now causes the
2325 PCRE_ERROR_NOMEMORY return instead of quietly going wrong.
2326
2327 42. It is now possible to set a limit on the number of times the match()
2328 function is called in a call to pcre_exec(). This facility makes it possible to
2329 limit the amount of recursion and backtracking, though not in a directly
2330 obvious way, because the match() function is used in a number of different
2331 circumstances. The count starts from zero for each position in the subject
2332 string (for non-anchored patterns). The default limit is, for compatibility, a
2333 large number, namely 10 000 000. You can change this in two ways:
2334
2335 (a) When configuring PCRE before making, you can use --with-match-limit=n
2336 to set a default value for the compiled library.
2337
2338 (b) For each call to pcre_exec(), you can pass a pcre_extra block in which
2339 a different value is set. See 45 below.
2340
2341 If the limit is exceeded, pcre_exec() returns PCRE_ERROR_MATCHLIMIT.
2342
2343 43. Added a new function pcre_config(int, void *) to enable run-time extraction
2344 of things that can be changed at compile time. The first argument specifies
2345 what is wanted and the second points to where the information is to be placed.
2346 The current list of available information is:
2347
2348 PCRE_CONFIG_UTF8
2349
2350 The output is an integer that is set to one if UTF-8 support is available;
2351 otherwise it is set to zero.
2352
2353 PCRE_CONFIG_NEWLINE
2354
2355 The output is an integer that it set to the value of the code that is used for
2356 newline. It is either LF (10) or CR (13).
2357
2358 PCRE_CONFIG_LINK_SIZE
2359
2360 The output is an integer that contains the number of bytes used for internal
2361 linkage in compiled expressions. The value is 2, 3, or 4. See item 32 above.
2362
2363 PCRE_CONFIG_POSIX_MALLOC_THRESHOLD
2364
2365 The output is an integer that contains the threshold above which the POSIX
2366 interface uses malloc() for output vectors. See item 31 above.
2367
2368 PCRE_CONFIG_MATCH_LIMIT
2369
2370 The output is an unsigned integer that contains the default limit of the number
2371 of match() calls in a pcre_exec() execution. See 42 above.
2372
2373 44. pcretest has been upgraded by the addition of the -C option. This causes it
2374 to extract all the available output from the new pcre_config() function, and to
2375 output it. The program then exits immediately.
2376
2377 45. A need has arisen to pass over additional data with calls to pcre_exec() in
2378 order to support additional features. One way would have been to define
2379 pcre_exec2() (for example) with extra arguments, but this would not have been
2380 extensible, and would also have required all calls to the original function to
2381 be mapped to the new one. Instead, I have chosen to extend the mechanism that
2382 is used for passing in "extra" data from pcre_study().
2383
2384 The pcre_extra structure is now exposed and defined in pcre.h. It currently
2385 contains the following fields:
2386
2387 flags a bitmap indicating which of the following fields are set
2388 study_data opaque data from pcre_study()
2389 match_limit a way of specifying a limit on match() calls for a specific
2390 call to pcre_exec()
2391 callout_data data for callouts (see 49 below)
2392
2393 The flag bits are also defined in pcre.h, and are
2394
2395 PCRE_EXTRA_STUDY_DATA
2396 PCRE_EXTRA_MATCH_LIMIT
2397 PCRE_EXTRA_CALLOUT_DATA
2398
2399 The pcre_study() function now returns one of these new pcre_extra blocks, with
2400 the actual study data pointed to by the study_data field, and the
2401 PCRE_EXTRA_STUDY_DATA flag set. This can be passed directly to pcre_exec() as
2402 before. That is, this change is entirely upwards-compatible and requires no
2403 change to existing code.
2404
2405 If you want to pass in additional data to pcre_exec(), you can either place it
2406 in a pcre_extra block provided by pcre_study(), or create your own pcre_extra
2407 block.
2408
2409 46. pcretest has been extended to test the PCRE_EXTRA_MATCH_LIMIT feature. If a
2410 data string contains the escape sequence \M, pcretest calls pcre_exec() several
2411 times with different match limits, until it finds the minimum value needed for
2412 pcre_exec() to complete. The value is then output. This can be instructive; for
2413 most simple matches the number is quite small, but for pathological cases it
2414 gets very large very quickly.
2415
2416 47. There's a new option for pcre_fullinfo() called PCRE_INFO_STUDYSIZE. It
2417 returns the size of the data block pointed to by the study_data field in a
2418 pcre_extra block, that is, the value that was passed as the argument to
2419 pcre_malloc() when PCRE was getting memory in which to place the information
2420 created by pcre_study(). The fourth argument should point to a size_t variable.
2421 pcretest has been extended so that this information is shown after a successful
2422 pcre_study() call when information about the compiled regex is being displayed.
2423
2424 48. Cosmetic change to Makefile: there's no need to have / after $(DESTDIR)
2425 because what follows is always an absolute path. (Later: it turns out that this
2426 is more than cosmetic for MinGW, because it doesn't like empty path
2427 components.)
2428
2429 49. Some changes have been made to the callout feature (see 28 above):
2430
2431 (i) A callout function now has three choices for what it returns:
2432
2433 0 => success, carry on matching
2434 > 0 => failure at this point, but backtrack if possible
2435 < 0 => serious error, return this value from pcre_exec()
2436
2437 Negative values should normally be chosen from the set of PCRE_ERROR_xxx
2438 values. In particular, returning PCRE_ERROR_NOMATCH forces a standard
2439 "match failed" error. The error number PCRE_ERROR_CALLOUT is reserved for
2440 use by callout functions. It will never be used by PCRE itself.
2441
2442 (ii) The pcre_extra structure (see 45 above) has a void * field called
2443 callout_data, with corresponding flag bit PCRE_EXTRA_CALLOUT_DATA. The
2444 pcre_callout_block structure has a field of the same name. The contents of
2445 the field passed in the pcre_extra structure are passed to the callout
2446 function in the corresponding field in the callout block. This makes it
2447 easier to use the same callout-containing regex from multiple threads. For
2448 testing, the pcretest program has a new data escape
2449
2450 \C*n pass the number n (may be negative) as callout_data
2451
2452 If the callout function in pcretest receives a non-zero value as
2453 callout_data, it returns that value.
2454
2455 50. Makefile wasn't handling CFLAGS properly when compiling dftables. Also,
2456 there were some redundant $(CFLAGS) in commands that are now specified as
2457 $(LINK), which already includes $(CFLAGS).
2458
2459 51. Extensions to UTF-8 support are listed below. These all apply when (a) PCRE
2460 has been compiled with UTF-8 support *and* pcre_compile() has been compiled
2461 with the PCRE_UTF8 flag. Patterns that are compiled without that flag assume
2462 one-byte characters throughout. Note that case-insensitive matching applies
2463 only to characters whose values are less than 256. PCRE doesn't support the
2464 notion of cases for higher-valued characters.
2465
2466 (i) A character class whose characters are all within 0-255 is handled as
2467 a bit map, and the map is inverted for negative classes. Previously, a
2468 character > 255 always failed to match such a class; however it should
2469 match if the class was a negative one (e.g. [^ab]). This has been fixed.
2470
2471 (ii) A negated character class with a single character < 255 is coded as
2472 "not this character" (OP_NOT). This wasn't working properly when the test
2473 character was multibyte, either singly or repeated.
2474
2475 (iii) Repeats of multibyte characters are now handled correctly in UTF-8
2476 mode, for example: \x{100}{2,3}.
2477
2478 (iv) The character escapes \b, \B, \d, \D, \s, \S, \w, and \W (either
2479 singly or repeated) now correctly test multibyte characters. However,
2480 PCRE doesn't recognize any characters with values greater than 255 as
2481 digits, spaces, or word characters. Such characters always match \D, \S,
2482 and \W, and never match \d, \s, or \w.
2483
2484 (v) Classes may now contain characters and character ranges with values
2485 greater than 255. For example: [ab\x{100}-\x{400}].
2486
2487 (vi) pcregrep now has a --utf-8 option (synonym -u) which makes it call
2488 PCRE in UTF-8 mode.
2489
2490 52. The info request value PCRE_INFO_FIRSTCHAR has been renamed
2491 PCRE_INFO_FIRSTBYTE because it is a byte value. However, the old name is
2492 retained for backwards compatibility. (Note that LASTLITERAL is also a byte
2493 value.)
2494
2495 53. The single man page has become too large. I have therefore split it up into
2496 a number of separate man pages. These also give rise to individual HTML pages;
2497 these are now put in a separate directory, and there is an index.html page that
2498 lists them all. Some hyperlinking between the pages has been installed.
2499
2500 54. Added convenience functions for handling named capturing parentheses.
2501
2502 55. Unknown escapes inside character classes (e.g. [\M]) and escapes that
2503 aren't interpreted therein (e.g. [\C]) are literals in Perl. This is now also
2504 true in PCRE, except when the PCRE_EXTENDED option is set, in which case they
2505 are faulted.
2506
2507 56. Introduced HOST_CC and HOST_CFLAGS which can be set in the environment when
2508 calling configure. These values are used when compiling the dftables.c program
2509 which is run to generate the source of the default character tables. They
2510 default to the values of CC and CFLAGS. If you are cross-compiling PCRE,
2511 you will need to set these values.
2512
2513 57. Updated the building process for Windows DLL, as provided by Fred Cox.
2514
2515
2516 Version 3.9 02-Jan-02
2517 ---------------------
2518
2519 1. A bit of extraneous text had somehow crept into the pcregrep documentation.
2520
2521 2. If --disable-static was given, the building process failed when trying to
2522 build pcretest and pcregrep. (For some reason it was using libtool to compile
2523 them, which is not right, as they aren't part of the library.)
2524
2525
2526 Version 3.8 18-Dec-01
2527 ---------------------
2528
2529 1. The experimental UTF-8 code was completely screwed up. It was packing the
2530 bytes in the wrong order. How dumb can you get?
2531
2532
2533 Version 3.7 29-Oct-01
2534 ---------------------
2535
2536 1. In updating pcretest to check change 1 of version 3.6, I screwed up.
2537 This caused pcretest, when used on the test data, to segfault. Unfortunately,
2538 this didn't happen under Solaris 8, where I normally test things.
2539
2540 2. The Makefile had to be changed to make it work on BSD systems, where 'make'
2541 doesn't seem to recognize that ./xxx and xxx are the same file. (This entry
2542 isn't in ChangeLog distributed with 3.7 because I forgot when I hastily made
2543 this fix an hour or so after the initial 3.7 release.)
2544
2545
2546 Version 3.6 23-Oct-01
2547 ---------------------
2548
2549 1. Crashed with /(sens|respons)e and \1ibility/ and "sense and sensibility" if
2550 offsets passed as NULL with zero offset count.
2551
2552 2. The config.guess and config.sub files had not been updated when I moved to
2553 the latest autoconf.
2554
2555
2556 Version 3.5 15-Aug-01
2557 ---------------------
2558
2559 1. Added some missing #if !defined NOPOSIX conditionals in pcretest.c that
2560 had been forgotten.
2561
2562 2. By using declared but undefined structures, we can avoid using "void"
2563 definitions in pcre.h while keeping the internal definitions of the structures
2564 private.
2565
2566 3. The distribution is now built using autoconf 2.50 and libtool 1.4. From a
2567 user point of view, this means that both static and shared libraries are built
2568 by default, but this can be individually controlled. More of the work of
2569 handling this static/shared cases is now inside libtool instead of PCRE's make
2570 file.
2571
2572 4. The pcretest utility is now installed along with pcregrep because it is
2573 useful for users (to test regexs) and by doing this, it automatically gets
2574 relinked by libtool. The documentation has been turned into a man page, so
2575 there are now .1, .txt, and .html versions in /doc.
2576
2577 5. Upgrades to pcregrep:
2578 (i) Added long-form option names like gnu grep.
2579 (ii) Added --help to list all options with an explanatory phrase.
2580 (iii) Added -r, --recursive to recurse into sub-directories.
2581 (iv) Added -f, --file to read patterns from a file.
2582
2583 6. pcre_exec() was referring to its "code" argument before testing that
2584 argument for NULL (and giving an error if it was NULL).
2585
2586 7. Upgraded Makefile.in to allow for compiling in a different directory from
2587 the source directory.
2588
2589 8. Tiny buglet in pcretest: when pcre_fullinfo() was called to retrieve the
2590 options bits, the pointer it was passed was to an int instead of to an unsigned
2591 long int. This mattered only on 64-bit systems.
2592
2593 9. Fixed typo (3.4/1) in pcre.h again. Sigh. I had changed pcre.h (which is
2594 generated) instead of pcre.in, which it its source. Also made the same change
2595 in several of the .c files.
2596
2597 10. A new release of gcc defines printf() as a macro, which broke pcretest
2598 because it had an ifdef in the middle of a string argument for printf(). Fixed
2599 by using separate calls to printf().
2600
2601 11. Added --enable-newline-is-cr and --enable-newline-is-lf to the configure
2602 script, to force use of CR or LF instead of \n in the source. On non-Unix
2603 systems, the value can be set in config.h.
2604
2605 12. The limit of 200 on non-capturing parentheses is a _nesting_ limit, not an
2606 absolute limit. Changed the text of the error message to make this clear, and
2607 likewise updated the man page.
2608
2609 13. The limit of 99 on the number of capturing subpatterns has been removed.
2610 The new limit is 65535, which I hope will not be a "real" limit.
2611
2612
2613 Version 3.4 22-Aug-00
2614 ---------------------
2615
2616 1. Fixed typo in pcre.h: unsigned const char * changed to const unsigned char *.
2617
2618 2. Diagnose condition (?(0) as an error instead of crashing on matching.
2619
2620
2621 Version 3.3 01-Aug-00
2622 ---------------------
2623
2624 1. If an octal character was given, but the value was greater than \377, it
2625 was not getting masked to the least significant bits, as documented. This could
2626 lead to crashes in some systems.
2627
2628 2. Perl 5.6 (if not earlier versions) accepts classes like [a-\d] and treats
2629 the hyphen as a literal. PCRE used to give an error; it now behaves like Perl.
2630
2631 3. Added the functions pcre_free_substring() and pcre_free_substring_list().
2632 These just pass their arguments on to (pcre_free)(), but they are provided
2633 because some uses of PCRE bind it to non-C systems that can call its functions,
2634 but cannot call free() or pcre_free() directly.
2635
2636 4. Add "make test" as a synonym for "make check". Corrected some comments in
2637 the Makefile.
2638
2639 5. Add $(DESTDIR)/ in front of all the paths in the "install" target in the
2640 Makefile.
2641
2642 6. Changed the name of pgrep to pcregrep, because Solaris has introduced a
2643 command called pgrep for grepping around the active processes.
2644
2645 7. Added the beginnings of support for UTF-8 character strings.
2646
2647 8. Arranged for the Makefile to pass over the settings of CC, CFLAGS, and
2648 RANLIB to ./ltconfig so that they are used by libtool. I think these are all
2649 the relevant ones. (AR is not passed because ./ltconfig does its own figuring
2650 out for the ar command.)
2651
2652
2653 Version 3.2 12-May-00
2654 ---------------------
2655
2656 This is purely a bug fixing release.
2657
2658 1. If the pattern /((Z)+|A)*/ was matched agained ZABCDEFG it matched Z instead
2659 of ZA. This was just one example of several cases that could provoke this bug,
2660 which was introduced by change 9 of version 2.00. The code for breaking
2661 infinite loops after an iteration that matches an empty string was't working
2662 correctly.
2663
2664 2. The pcretest program was not imitating Perl correctly for the pattern /a*/g
2665 when matched against abbab (for example). After matching an empty string, it
2666 wasn't forcing anchoring when setting PCRE_NOTEMPTY for the next attempt; this
2667 caused it to match further down the string than it should.
2668
2669 3. The code contained an inclusion of sys/types.h. It isn't clear why this
2670 was there because it doesn't seem to be needed, and it causes trouble on some
2671 systems, as it is not a Standard C header. It has been removed.
2672
2673 4. Made 4 silly changes to the source to avoid stupid compiler warnings that
2674 were reported on the Macintosh. The changes were from
2675
2676 while ((c = *(++ptr)) != 0 && c != '\n');
2677 to
2678 while ((c = *(++ptr)) != 0 && c != '\n') ;
2679
2680 Totally extraordinary, but if that's what it takes...
2681
2682 5. PCRE is being used in one environment where neither memmove() nor bcopy() is
2683 available. Added HAVE_BCOPY and an autoconf test for it; if neither
2684 HAVE_MEMMOVE nor HAVE_BCOPY is set, use a built-in emulation function which
2685 assumes the way PCRE uses memmove() (always moving upwards).
2686
2687 6. PCRE is being used in one environment where strchr() is not available. There
2688 was only one use in pcre.c, and writing it out to avoid strchr() probably gives
2689 faster code anyway.
2690
2691
2692 Version 3.1 09-Feb-00
2693 ---------------------
2694
2695 The only change in this release is the fixing of some bugs in Makefile.in for
2696 the "install" target:
2697
2698 (1) It was failing to install pcreposix.h.
2699
2700 (2) It was overwriting the pcre.3 man page with the pcreposix.3 man page.
2701
2702
2703 Version 3.0 01-Feb-00
2704 ---------------------
2705
2706 1. Add support for the /+ modifier to perltest (to output $` like it does in
2707 pcretest).
2708
2709 2. Add support for the /g modifier to perltest.
2710
2711 3. Fix pcretest so that it behaves even more like Perl for /g when the pattern
2712 matches null strings.
2713
2714 4. Fix perltest so that it doesn't do unwanted things when fed an empty
2715 pattern. Perl treats empty patterns specially - it reuses the most recent
2716 pattern, which is not what we want. Replace // by /(?#)/ in order to avoid this
2717 effect.
2718
2719 5. The POSIX interface was broken in that it was just handing over the POSIX
2720 captured string vector to pcre_exec(), but (since release 2.00) PCRE has
2721 required a bigger vector, with some working space on the end. This means that
2722 the POSIX wrapper now has to get and free some memory, and copy the results.
2723
2724 6. Added some simple autoconf support, placing the test data and the
2725 documentation in separate directories, re-organizing some of the
2726 information files, and making it build pcre-config (a GNU standard). Also added
2727 libtool support for building PCRE as a shared library, which is now the
2728 default.
2729
2730 7. Got rid of the leading zero in the definition of PCRE_MINOR because 08 and
2731 09 are not valid octal constants. Single digits will be used for minor values
2732 less than 10.
2733
2734 8. Defined REG_EXTENDED and REG_NOSUB as zero in the POSIX header, so that
2735 existing programs that set these in the POSIX interface can use PCRE without
2736 modification.
2737
2738 9. Added a new function, pcre_fullinfo() with an extensible interface. It can
2739 return all that pcre_info() returns, plus additional data. The pcre_info()
2740 function is retained for compatibility, but is considered to be obsolete.
2741
2742 10. Added experimental recursion feature (?R) to handle one common case that
2743 Perl 5.6 will be able to do with (?p{...}).
2744
2745 11. Added support for POSIX character classes like [:alpha:], which Perl is
2746 adopting.
2747
2748
2749 Version 2.08 31-Aug-99
2750 ----------------------
2751
2752 1. When startoffset was not zero and the pattern began with ".*", PCRE was not
2753 trying to match at the startoffset position, but instead was moving forward to
2754 the next newline as if a previous match had failed.
2755
2756 2. pcretest was not making use of PCRE_NOTEMPTY when repeating for /g and /G,
2757 and could get into a loop if a null string was matched other than at the start
2758 of the subject.
2759
2760 3. Added definitions of PCRE_MAJOR and PCRE_MINOR to pcre.h so the version can
2761 be distinguished at compile time, and for completeness also added PCRE_DATE.
2762
2763 5. Added Paul Sokolovsky's minor changes to make it easy to compile a Win32 DLL
2764 in GnuWin32 environments.
2765
2766
2767 Version 2.07 29-Jul-99
2768 ----------------------
2769
2770 1. The documentation is now supplied in plain text form and HTML as well as in
2771 the form of man page sources.
2772
2773 2. C++ compilers don't like assigning (void *) values to other pointer types.
2774 In particular this affects malloc(). Although there is no problem in Standard
2775 C, I've put in casts to keep C++ compilers happy.
2776
2777 3. Typo on pcretest.c; a cast of (unsigned char *) in the POSIX regexec() call
2778 should be (const char *).
2779
2780 4. If NOPOSIX is defined, pcretest.c compiles without POSIX support. This may
2781 be useful for non-Unix systems who don't want to bother with the POSIX stuff.
2782 However, I haven't made this a standard facility. The documentation doesn't
2783 mention it, and the Makefile doesn't support it.
2784
2785 5. The Makefile now contains an "install" target, with editable destinations at
2786 the top of the file. The pcretest program is not installed.
2787
2788 6. pgrep -V now gives the PCRE version number and date.
2789
2790 7. Fixed bug: a zero repetition after a literal string (e.g. /abcde{0}/) was
2791 causing the entire string to be ignored, instead of just the last character.
2792
2793 8. If a pattern like /"([^\\"]+|\\.)*"/ is applied in the normal way to a
2794 non-matching string, it can take a very, very long time, even for strings of
2795 quite modest length, because of the nested recursion. PCRE now does better in
2796 some of these cases. It does this by remembering the last required literal
2797 character in the pattern, and pre-searching the subject to ensure it is present
2798 before running the real match. In other words, it applies a heuristic to detect
2799 some types of certain failure quickly, and in the above example, if presented
2800 with a string that has no trailing " it gives "no match" very quickly.
2801
2802 9. A new runtime option PCRE_NOTEMPTY causes null string matches to be ignored;
2803 other alternatives are tried instead.
2804
2805
2806 Version 2.06 09-Jun-99
2807 ----------------------
2808
2809 1. Change pcretest's output for amount of store used to show just the code
2810 space, because the remainder (the data block) varies in size between 32-bit and
2811 64-bit systems.
2812
2813 2. Added an extra argument to pcre_exec() to supply an offset in the subject to
2814 start matching at. This allows lookbehinds to work when searching for multiple
2815 occurrences in a string.
2816
2817 3. Added additional options to pcretest for testing multiple occurrences:
2818
2819 /+ outputs the rest of the string that follows a match
2820 /g loops for multiple occurrences, using the new startoffset argument
2821 /G loops for multiple occurrences by passing an incremented pointer
2822
2823 4. PCRE wasn't doing the "first character" optimization for patterns starting
2824 with \b or \B, though it was doing it for other lookbehind assertions. That is,
2825 it wasn't noticing that a match for a pattern such as /\bxyz/ has to start with
2826 the letter 'x'. On long subject strings, this gives a significant speed-up.
2827
2828
2829 Version 2.05 21-Apr-99
2830 ----------------------
2831
2832 1. Changed the type of magic_number from int to long int so that it works
2833 properly on 16-bit systems.
2834
2835 2. Fixed a bug which caused patterns starting with .* not to work correctly
2836 when the subject string contained newline characters. PCRE was assuming
2837 anchoring for such patterns in all cases, which is not correct because .* will
2838 not pass a newline unless PCRE_DOTALL is set. It now assumes anchoring only if
2839 DOTALL is set at top level; otherwise it knows that patterns starting with .*
2840 must be retried after every newline in the subject.
2841
2842
2843 Version 2.04 18-Feb-99
2844 ----------------------
2845
2846 1. For parenthesized subpatterns with repeats whose minimum was zero, the
2847 computation of the store needed to hold the pattern was incorrect (too large).
2848 If such patterns were nested a few deep, this could multiply and become a real
2849 problem.
2850
2851 2. Added /M option to pcretest to show the memory requirement of a specific
2852 pattern. Made -m a synonym of -s (which does this globally) for compatibility.
2853
2854 3. Subpatterns of the form (regex){n,m} (i.e. limited maximum) were being
2855 compiled in such a way that the backtracking after subsequent failure was
2856 pessimal. Something like (a){0,3} was compiled as (a)?(a)?(a)? instead of
2857 ((a)((a)(a)?)?)? with disastrous performance if the maximum was of any size.
2858
2859
2860 Version 2.03 02-Feb-99
2861 ----------------------
2862
2863 1. Fixed typo and small mistake in man page.
2864
2865 2. Added 4th condition (GPL supersedes if conflict) and created separate
2866 LICENCE file containing the conditions.
2867
2868 3. Updated pcretest so that patterns such as /abc\/def/ work like they do in
2869 Perl, that is the internal \ allows the delimiter to be included in the
2870 pattern. Locked out the use of \ as a delimiter. If \ immediately follows
2871 the final delimiter, add \ to the end of the pattern (to test the error).
2872
2873 4. Added the convenience functions for extracting substrings after a successful
2874 match. Updated pcretest to make it able to test these functions.
2875
2876
2877 Version 2.02 14-Jan-99
2878 ----------------------
2879
2880 1. Initialized the working variables associated with each extraction so that
2881 their saving and restoring doesn't refer to uninitialized store.
2882
2883 2. Put dummy code into study.c in order to trick the optimizer of the IBM C
2884 compiler for OS/2 into generating correct code. Apparently IBM isn't going to
2885 fix the problem.
2886
2887 3. Pcretest: the timing code wasn't using LOOPREPEAT for timing execution
2888 calls, and wasn't printing the correct value for compiling calls. Increased the
2889 default value of LOOPREPEAT, and the number of significant figures in the
2890 times.
2891
2892 4. Changed "/bin/rm" in the Makefile to "-rm" so it works on Windows NT.
2893
2894 5. Renamed "deftables" as "dftables" to get it down to 8 characters, to avoid
2895 a building problem on Windows NT with a FAT file system.
2896
2897
2898 Version 2.01 21-Oct-98
2899 ----------------------
2900
2901 1. Changed the API for pcre_compile() to allow for the provision of a pointer
2902 to character tables built by pcre_maketables() in the current locale. If NULL
2903 is passed, the default tables are used.
2904
2905
2906 Version 2.00 24-Sep-98
2907 ----------------------
2908
2909 1. Since the (>?) facility is in Perl 5.005, don't require PCRE_EXTRA to enable
2910 it any more.
2911
2912 2. Allow quantification of (?>) groups, and make it work correctly.
2913
2914 3. The first character computation wasn't working for (?>) groups.
2915
2916 4. Correct the implementation of \Z (it is permitted to match on the \n at the
2917 end of the subject) and add 5.005's \z, which really does match only at the
2918 very end of the subject.
2919
2920 5. Remove the \X "cut" facility; Perl doesn't have it, and (?> is neater.
2921
2922 6. Remove the ability to specify CASELESS, MULTILINE, DOTALL, and
2923 DOLLAR_END_ONLY at runtime, to make it possible to implement the Perl 5.005
2924 localized options. All options to pcre_study() were also removed.
2925
2926 7. Add other new features from 5.005:
2927
2928 $(?<= positive lookbehind
2929 $(?<! negative lookbehind
2930 (?imsx-imsx) added the unsetting capability
2931 such a setting is global if at outer level; local otherwise
2932 (?imsx-imsx:) non-capturing groups with option setting
2933 (?(cond)re|re) conditional pattern matching
2934
2935 A backreference to itself in a repeated group matches the previous
2936 captured string.
2937
2938 8. General tidying up of studying (both automatic and via "study")
2939 consequential on the addition of new assertions.
2940
2941 9. As in 5.005, unlimited repeated groups that could match an empty substring
2942 are no longer faulted at compile time. Instead, the loop is forcibly broken at
2943 runtime if any iteration does actually match an empty substring.
2944
2945 10. Include the RunTest script in the distribution.
2946
2947 11. Added tests from the Perl 5.005_02 distribution. This showed up a few
2948 discrepancies, some of which were old and were also with respect to 5.004. They
2949 have now been fixed.
2950
2951
2952 Version 1.09 28-Apr-98
2953 ----------------------
2954
2955 1. A negated single character class followed by a quantifier with a minimum
2956 value of one (e.g. [^x]{1,6} ) was not compiled correctly. This could lead to
2957 program crashes, or just wrong answers. This did not apply to negated classes
2958 containing more than one character, or to minima other than one.
2959
2960
2961 Version 1.08 27-Mar-98
2962 ----------------------
2963
2964 1. Add PCRE_UNGREEDY to invert the greediness of quantifiers.
2965
2966 2. Add (?U) and (?X) to set PCRE_UNGREEDY and PCRE_EXTRA respectively. The
2967 latter must appear before anything that relies on it in the pattern.
2968
2969
2970 Version 1.07 16-Feb-98
2971 ----------------------
2972
2973 1. A pattern such as /((a)*)*/ was not being diagnosed as in error (unlimited
2974 repeat of a potentially empty string).
2975
2976
2977 Version 1.06 23-Jan-98
2978 ----------------------
2979
2980 1. Added Markus Oberhumer's little patches for C++.
2981
2982 2. Literal strings longer than 255 characters were broken.
2983
2984
2985 Version 1.05 23-Dec-97
2986 ----------------------
2987
2988 1. Negated character classes containing more than one character were failing if
2989 PCRE_CASELESS was set at run time.
2990
2991
2992 Version 1.04 19-Dec-97
2993 ----------------------
2994
2995 1. Corrected the man page, where some "const" qualifiers had been omitted.
2996
2997 2. Made debugging output print "{0,xxx}" instead of just "{,xxx}" to agree with
2998 input syntax.
2999
3000 3. Fixed memory leak which occurred when a regex with back references was
3001 matched with an offsets vector that wasn't big enough. The temporary memory
3002 that is used in this case wasn't being freed if the match failed.
3003
3004 4. Tidied pcretest to ensure it frees memory that it gets.
3005
3006 5. Temporary memory was being obtained in the case where the passed offsets
3007 vector was exactly big enough.
3008
3009 6. Corrected definition of offsetof() from change 5 below.
3010
3011 7. I had screwed up change 6 below and broken the rules for the use of
3012 setjmp(). Now fixed.
3013
3014
3015 Version 1.03 18-Dec-97
3016 ----------------------
3017
3018 1. A erroneous regex with a missing opening parenthesis was correctly
3019 diagnosed, but PCRE attempted to access brastack[-1], which could cause crashes
3020 on some systems.
3021
3022 2. Replaced offsetof(real_pcre, code) by offsetof(real_pcre, code[0]) because
3023 it was reported that one broken compiler failed on the former because "code" is
3024 also an independent variable.
3025
3026 3. The erroneous regex a[]b caused an array overrun reference.
3027
3028 4. A regex ending with a one-character negative class (e.g. /[^k]$/) did not
3029 fail on data ending with that character. (It was going on too far, and checking
3030 the next character, typically a binary zero.) This was specific to the
3031 optimized code for single-character negative classes.
3032
3033 5. Added a contributed patch from the TIN world which does the following:
3034
3035 + Add an undef for memmove, in case the the system defines a macro for it.
3036
3037 + Add a definition of offsetof(), in case there isn't one. (I don't know
3038 the reason behind this - offsetof() is part of the ANSI standard - but
3039 it does no harm).
3040
3041 + Reduce the ifdef's in pcre.c using macro DPRINTF, thereby eliminating
3042 most of the places where whitespace preceded '#'. I have given up and
3043 allowed the remaining 2 cases to be at the margin.
3044
3045 + Rename some variables in pcre to eliminate shadowing. This seems very
3046 pedantic, but does no harm, of course.
3047
3048 6. Moved the call to setjmp() into its own function, to get rid of warnings
3049 from gcc -Wall, and avoided calling it at all unless PCRE_EXTRA is used.
3050
3051 7. Constructs such as \d{8,} were compiling into the equivalent of
3052 \d{8}\d{0,65527} instead of \d{8}\d* which didn't make much difference to the
3053 outcome, but in this particular case used more store than had been allocated,
3054 which caused the bug to be discovered because it threw up an internal error.
3055
3056 8. The debugging code in both pcre and pcretest for outputting the compiled
3057 form of a regex was going wrong in the case of back references followed by
3058 curly-bracketed repeats.
3059
3060
3061 Version 1.02 12-Dec-97
3062 ----------------------
3063
3064 1. Typos in pcre.3 and comments in the source fixed.
3065
3066 2. Applied a contributed patch to get rid of places where it used to remove
3067 'const' from variables, and fixed some signed/unsigned and uninitialized
3068 variable warnings.
3069
3070 3. Added the "runtest" target to Makefile.
3071
3072 4. Set default compiler flag to -O2 rather than just -O.
3073
3074
3075 Version 1.01 19-Nov-97
3076 ----------------------
3077
3078 1. PCRE was failing to diagnose unlimited repeat of empty string for patterns
3079 like /([ab]*)*/, that is, for classes with more than one character in them.
3080
3081 2. Likewise, it wasn't diagnosing patterns with "once-only" subpatterns, such
3082 as /((?>a*))*/ (a PCRE_EXTRA facility).
3083
3084
3085 Version 1.00 18-Nov-97
3086 ----------------------
3087
3088 1. Added compile-time macros to support systems such as SunOS4 which don't have
3089 memmove() or strerror() but have other things that can be used instead.
3090
3091 2. Arranged that "make clean" removes the executables.
3092
3093
3094 Version 0.99 27-Oct-97
3095 ----------------------
3096
3097 1. Fixed bug in code for optimizing classes with only one character. It was
3098 initializing a 32-byte map regardless, which could cause it to run off the end
3099 of the memory it had got.
3100
3101 2. Added, conditional on PCRE_EXTRA, the proposed (?>REGEX) construction.
3102
3103
3104 Version 0.98 22-Oct-97
3105 ----------------------
3106
3107 1. Fixed bug in code for handling temporary memory usage when there are more
3108 back references than supplied space in the ovector. This could cause segfaults.
3109
3110
3111 Version 0.97 21-Oct-97
3112 ----------------------
3113
3114 1. Added the \X "cut" facility, conditional on PCRE_EXTRA.
3115
3116 2. Optimized negated single characters not to use a bit map.
3117
3118 3. Brought error texts together as macro definitions; clarified some of them;
3119 fixed one that was wrong - it said "range out of order" when it meant "invalid
3120 escape sequence".
3121
3122 4. Changed some char * arguments to const char *.
3123
3124 5. Added PCRE_NOTBOL and PCRE_NOTEOL (from POSIX).
3125
3126 6. Added the POSIX-style API wrapper in pcreposix.a and testing facilities in
3127 pcretest.
3128
3129
3130 Version 0.96 16-Oct-97
3131 ----------------------
3132
3133 1. Added a simple "pgrep" utility to the distribution.
3134
3135 2. Fixed an incompatibility with Perl: "{" is now treated as a normal character
3136 unless it appears in one of the precise forms "{ddd}", "{ddd,}", or "{ddd,ddd}"
3137 where "ddd" means "one or more decimal digits".
3138
3139 3. Fixed serious bug. If a pattern had a back reference, but the call to
3140 pcre_exec() didn't supply a large enough ovector to record the related
3141 identifying subpattern, the match always failed. PCRE now remembers the number
3142 of the largest back reference, and gets some temporary memory in which to save
3143 the offsets during matching if necessary, in order to ensure that
3144 backreferences always work.
3145
3146 4. Increased the compatibility with Perl in a number of ways:
3147
3148 (a) . no longer matches \n by default; an option PCRE_DOTALL is provided
3149 to request this handling. The option can be set at compile or exec time.
3150
3151 (b) $ matches before a terminating newline by default; an option
3152 PCRE_DOLLAR_ENDONLY is provided to override this (but not in multiline
3153 mode). The option can be set at compile or exec time.
3154
3155 (c) The handling of \ followed by a digit other than 0 is now supposed to be
3156 the same as Perl's. If the decimal number it represents is less than 10
3157 or there aren't that many previous left capturing parentheses, an octal
3158 escape is read. Inside a character class, it's always an octal escape,
3159 even if it is a single digit.
3160
3161 (d) An escaped but undefined alphabetic character is taken as a literal,
3162 unless PCRE_EXTRA is set. Currently this just reserves the remaining
3163 escapes.
3164
3165 (e) {0} is now permitted. (The previous item is removed from the compiled
3166 pattern).
3167
3168 5. Changed all the names of code files so that the basic parts are no longer
3169 than 10 characters, and abolished the teeny "globals.c" file.
3170
3171 6. Changed the handling of character classes; they are now done with a 32-byte
3172 bit map always.
3173
3174 7. Added the -d and /D options to pcretest to make it possible to look at the
3175 internals of compilation without having to recompile pcre.
3176
3177
3178 Version 0.95 23-Sep-97
3179 ----------------------
3180
3181 1. Fixed bug in pre-pass concerning escaped "normal" characters such as \x5c or
3182 \x20 at the start of a run of normal characters. These were being treated as
3183 real characters, instead of the source characters being re-checked.
3184
3185
3186 Version 0.94 18-Sep-97
3187 ----------------------
3188
3189 1. The functions are now thread-safe, with the caveat that the global variables
3190 containing pointers to malloc() and free() or alternative functions are the
3191 same for all threads.
3192
3193 2. Get pcre_study() to generate a bitmap of initial characters for non-
3194 anchored patterns when this is possible, and use it if passed to pcre_exec().
3195
3196
3197 Version 0.93 15-Sep-97
3198 ----------------------
3199
3200 1. /(b)|(:+)/ was computing an incorrect first character.
3201
3202 2. Add pcre_study() to the API and the passing of pcre_extra to pcre_exec(),
3203 but not actually doing anything yet.
3204
3205 3. Treat "-" characters in classes that cannot be part of ranges as literals,
3206 as Perl does (e.g. [-az] or [az-]).
3207
3208 4. Set the anchored flag if a branch starts with .* or .*? because that tests
3209 all possible positions.
3210
3211 5. Split up into different modules to avoid including unneeded functions in a
3212 compiled binary. However, compile and exec are still in one module. The "study"
3213 function is split off.
3214
3215 6. The character tables are now in a separate module whose source is generated
3216 by an auxiliary program - but can then be edited by hand if required. There are
3217 now no calls to isalnum(), isspace(), isdigit(), isxdigit(), tolower() or
3218 toupper() in the code.
3219
3220 7. Turn the malloc/free funtions variables into pcre_malloc and pcre_free and
3221 make them global. Abolish the function for setting them, as the caller can now
3222 set them directly.
3223
3224
3225 Version 0.92 11-Sep-97
3226 ----------------------
3227
3228 1. A repeat with a fixed maximum and a minimum of 1 for an ordinary character
3229 (e.g. /a{1,3}/) was broken (I mis-optimized it).
3230
3231 2. Caseless matching was not working in character classes if the characters in
3232 the pattern were in upper case.
3233
3234 3. Make ranges like [W-c] work in the same way as Perl for caseless matching.
3235
3236 4. Make PCRE_ANCHORED public and accept as a compile option.
3237
3238 5. Add an options word to pcre_exec() and accept PCRE_ANCHORED and
3239 PCRE_CASELESS at run time. Add escapes \A and \I to pcretest to cause it to
3240 pass them.
3241
3242 6. Give an error if bad option bits passed at compile or run time.
3243
3244 7. Add PCRE_MULTILINE at compile and exec time, and (?m) as well. Add \M to
3245 pcretest to cause it to pass that flag.
3246
3247 8. Add pcre_info(), to get the number of identifying subpatterns, the stored
3248 options, and the first character, if set.
3249
3250 9. Recognize C+ or C{n,m} where n >= 1 as providing a fixed starting character.
3251
3252
3253 Version 0.91 10-Sep-97
3254 ----------------------
3255
3256 1. PCRE was failing to diagnose unlimited repeats of subpatterns that could
3257 match the empty string as in /(a*)*/. It was looping and ultimately crashing.
3258
3259 2. PCRE was looping on encountering an indefinitely repeated back reference to
3260 a subpattern that had matched an empty string, e.g. /(a|)\1*/. It now does what
3261 Perl does - treats the match as successful.
3262
3263 ****

Properties

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

  ViewVC Help
Powered by ViewVC 1.1.5