ViewVC logotype

Contents of /code/trunk/ChangeLog

Parent Directory Parent Directory | Revision Log Revision Log

Revision 340 - (show annotations)
Fri Apr 18 20:00:21 2008 UTC (12 years, 6 months ago) by ph10
File size: 154365 byte(s)
Fix incorrect error for patterns like /(?2)[]a()b](abc)/
1 ChangeLog for PCRE
2 ------------------
4 Version 7.7 05-Mar-08
5 ---------------------
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.
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.)
15 3. Remove a line of dead code, identified by coverity and reported by Nuno
16 Lopes.
18 4. Fixed two related pcregrep bugs involving -r with --include or --exclude:
20 (1) The include/exclude patterns were being applied to the whole pathnames
21 of files, instead of just to the final components.
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.
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.
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.
37 7. Added some missing "const"s to declarations of static tables in
38 pcre_compile.c and pcre_dfa_exec.c.
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.)
44 9. Applied Craig's patch to remove the use of push_back().
46 10. Applied Alan Lehotsky's patch to add REG_STARTEND support to the POSIX
47 matching function regexec().
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).
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.
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).
66 14. A pattern such as /(?2)[]a()b](abc)/ which had a forward reference to a
67 non-existent subpattern following a character class starting with ']' and
68 containing () gave an internal compiling error instead of "reference to
69 non-existent subpattern". Fortunately, when the pattern did exist, the
70 compiled code was correct. (When scanning forwards to check for the
71 existencd of the subpattern, it was treating the data ']' as terminating
72 the class, so got the count wrong. When actually compiling, the reference
73 was subsequently set up correctly.)
76 Version 7.6 28-Jan-08
77 ---------------------
79 1. A character class containing a very large number of characters with
80 codepoints greater than 255 (in UTF-8 mode, of course) caused a buffer
81 overflow.
83 2. Patch to cut out the "long long" test in pcrecpp_unittest when
84 HAVE_LONG_LONG is not defined.
86 3. Applied Christian Ehrlicher's patch to update the CMake build files to
87 bring them up to date and include new features. This patch includes:
89 - Fixed PH's badly added libz and libbz2 support.
90 - Fixed a problem with static linking.
91 - Added pcredemo. [But later removed - see 7 below.]
92 - Fixed dftables problem and added an option.
93 - Added a number of HAVE_XXX tests, including HAVE_WINDOWS_H and
95 - Added readline support for pcretest.
96 - Added an listing of the option settings after cmake has run.
98 4. A user submitted a patch to Makefile that makes it easy to create
99 "pcre.dll" under mingw when using Configure/Make. I added stuff to
100 Makefile.am that cause it to include this special target, without
101 affecting anything else. Note that the same mingw target plus all
102 the other distribution libraries and programs are now supported
103 when configuring with CMake (see 6 below) instead of with
104 Configure/Make.
106 5. Applied Craig's patch that moves no_arg into the RE class in the C++ code.
107 This is an attempt to solve the reported problem "pcrecpp::no_arg is not
108 exported in the Windows port". It has not yet been confirmed that the patch
109 solves the problem, but it does no harm.
111 6. Applied Sheri's patch to CMakeLists.txt to add NON_STANDARD_LIB_PREFIX and
112 NON_STANDARD_LIB_SUFFIX for dll names built with mingw when configured
113 with CMake, and also correct the comment about stack recursion.
115 7. Remove the automatic building of pcredemo from the ./configure system and
116 from CMakeLists.txt. The whole idea of pcredemo.c is that it is an example
117 of a program that users should build themselves after PCRE is installed, so
118 building it automatically is not really right. What is more, it gave
119 trouble in some build environments.
121 8. Further tidies to CMakeLists.txt from Sheri and Christian.
124 Version 7.5 10-Jan-08
125 ---------------------
127 1. Applied a patch from Craig: "This patch makes it possible to 'ignore'
128 values in parens when parsing an RE using the C++ wrapper."
130 2. Negative specials like \S did not work in character classes in UTF-8 mode.
131 Characters greater than 255 were excluded from the class instead of being
132 included.
134 3. The same bug as (2) above applied to negated POSIX classes such as
135 [:^space:].
137 4. PCRECPP_STATIC was referenced in pcrecpp_internal.h, but nowhere was it
138 defined or documented. It seems to have been a typo for PCRE_STATIC, so
139 I have changed it.
141 5. The construct (?&) was not diagnosed as a syntax error (it referenced the
142 first named subpattern) and a construct such as (?&a) would reference the
143 first named subpattern whose name started with "a" (in other words, the
144 length check was missing). Both these problems are fixed. "Subpattern name
145 expected" is now given for (?&) (a zero-length name), and this patch also
146 makes it give the same error for \k'' (previously it complained that that
147 was a reference to a non-existent subpattern).
149 6. The erroneous patterns (?+-a) and (?-+a) give different error messages;
150 this is right because (?- can be followed by option settings as well as by
151 digits. I have, however, made the messages clearer.
153 7. Patterns such as (?(1)a|b) (a pattern that contains fewer subpatterns
154 than the number used in the conditional) now cause a compile-time error.
155 This is actually not compatible with Perl, which accepts such patterns, but
156 treats the conditional as always being FALSE (as PCRE used to), but it
157 seems to me that giving a diagnostic is better.
159 8. Change "alphameric" to the more common word "alphanumeric" in comments
160 and messages.
162 9. Fix two occurrences of "backslash" in comments that should have been
163 "backspace".
165 10. Remove two redundant lines of code that can never be obeyed (their function
166 was moved elsewhere).
168 11. The program that makes PCRE's Unicode character property table had a bug
169 which caused it to generate incorrect table entries for sequences of
170 characters that have the same character type, but are in different scripts.
171 It amalgamated them into a single range, with the script of the first of
172 them. In other words, some characters were in the wrong script. There were
173 thirteen such cases, affecting characters in the following ranges:
175 U+002b0 - U+002c1
176 U+0060c - U+0060d
177 U+0061e - U+00612
178 U+0064b - U+0065e
179 U+0074d - U+0076d
180 U+01800 - U+01805
181 U+01d00 - U+01d77
182 U+01d9b - U+01dbf
183 U+0200b - U+0200f
184 U+030fc - U+030fe
185 U+03260 - U+0327f
186 U+0fb46 - U+0fbb1
187 U+10450 - U+1049d
189 12. The -o option (show only the matching part of a line) for pcregrep was not
190 compatible with GNU grep in that, if there was more than one match in a
191 line, it showed only the first of them. It now behaves in the same way as
192 GNU grep.
194 13. If the -o and -v options were combined for pcregrep, it printed a blank
195 line for every non-matching line. GNU grep prints nothing, and pcregrep now
196 does the same. The return code can be used to tell if there were any
197 non-matching lines.
199 14. Added --file-offsets and --line-offsets to pcregrep.
201 15. The pattern (?=something)(?R) was not being diagnosed as a potentially
202 infinitely looping recursion. The bug was that positive lookaheads were not
203 being skipped when checking for a possible empty match (negative lookaheads
204 and both kinds of lookbehind were skipped).
206 16. Fixed two typos in the Windows-only code in pcregrep.c, and moved the
207 inclusion of <windows.h> to before rather than after the definition of
208 INVALID_FILE_ATTRIBUTES (patch from David Byron).
210 17. Specifying a possessive quantifier with a specific limit for a Unicode
211 character property caused pcre_compile() to compile bad code, which led at
212 runtime to PCRE_ERROR_INTERNAL (-14). Examples of patterns that caused this
213 are: /\p{Zl}{2,3}+/8 and /\p{Cc}{2}+/8. It was the possessive "+" that
214 caused the error; without that there was no problem.
216 18. Added --enable-pcregrep-libz and --enable-pcregrep-libbz2.
218 19. Added --enable-pcretest-libreadline.
220 20. In pcrecpp.cc, the variable 'count' was incremented twice in
221 RE::GlobalReplace(). As a result, the number of replacements returned was
222 double what it should be. I removed one of the increments, but Craig sent a
223 later patch that removed the other one (the right fix) and added unit tests
224 that check the return values (which was not done before).
226 21. Several CMake things:
228 (1) Arranged that, when cmake is used on Unix, the libraries end up with
229 the names libpcre and libpcreposix, not just pcre and pcreposix.
231 (2) The above change means that pcretest and pcregrep are now correctly
232 linked with the newly-built libraries, not previously installed ones.
236 22. In UTF-8 mode, with newline set to "any", a pattern such as .*a.*=.b.*
237 crashed when matching a string such as a\x{2029}b (note that \x{2029} is a
238 UTF-8 newline character). The key issue is that the pattern starts .*;
239 this means that the match must be either at the beginning, or after a
240 newline. The bug was in the code for advancing after a failed match and
241 checking that the new position followed a newline. It was not taking
242 account of UTF-8 characters correctly.
244 23. PCRE was behaving differently from Perl in the way it recognized POSIX
245 character classes. PCRE was not treating the sequence [:...:] as a
246 character class unless the ... were all letters. Perl, however, seems to
247 allow any characters between [: and :], though of course it rejects as
248 unknown any "names" that contain non-letters, because all the known class
249 names consist only of letters. Thus, Perl gives an error for [[:1234:]],
250 for example, whereas PCRE did not - it did not recognize a POSIX character
251 class. This seemed a bit dangerous, so the code has been changed to be
252 closer to Perl. The behaviour is not identical to Perl, because PCRE will
253 diagnose an unknown class for, for example, [[:l\ower:]] where Perl will
254 treat it as [[:lower:]]. However, PCRE does now give "unknown" errors where
255 Perl does, and where it didn't before.
257 24. Rewrite so as to remove the single use of %n from pcregrep because in some
258 Windows environments %n is disabled by default.
261 Version 7.4 21-Sep-07
262 ---------------------
264 1. Change 7.3/28 was implemented for classes by looking at the bitmap. This
265 means that a class such as [\s] counted as "explicit reference to CR or
266 LF". That isn't really right - the whole point of the change was to try to
267 help when there was an actual mention of one of the two characters. So now
268 the change happens only if \r or \n (or a literal CR or LF) character is
269 encountered.
271 2. The 32-bit options word was also used for 6 internal flags, but the numbers
272 of both had grown to the point where there were only 3 bits left.
273 Fortunately, there was spare space in the data structure, and so I have
274 moved the internal flags into a new 16-bit field to free up more option
275 bits.
277 3. The appearance of (?J) at the start of a pattern set the DUPNAMES option,
278 but did not set the internal JCHANGED flag - either of these is enough to
279 control the way the "get" function works - but the PCRE_INFO_JCHANGED
280 facility is supposed to tell if (?J) was ever used, so now (?J) at the
281 start sets both bits.
283 4. Added options (at build time, compile time, exec time) to change \R from
284 matching any Unicode line ending sequence to just matching CR, LF, or CRLF.
286 5. doc/pcresyntax.html was missing from the distribution.
288 6. Put back the definition of PCRE_ERROR_NULLWSLIMIT, for backward
289 compatibility, even though it is no longer used.
291 7. Added macro for snprintf to pcrecpp_unittest.cc and also for strtoll and
292 strtoull to pcrecpp.cc to select the available functions in WIN32 when the
293 windows.h file is present (where different names are used). [This was
294 reversed later after testing - see 16 below.]
296 8. Changed all #include <config.h> to #include "config.h". There were also
297 some further <pcre.h> cases that I changed to "pcre.h".
299 9. When pcregrep was used with the --colour option, it missed the line ending
300 sequence off the lines that it output.
302 10. It was pointed out to me that arrays of string pointers cause lots of
303 relocations when a shared library is dynamically loaded. A technique of
304 using a single long string with a table of offsets can drastically reduce
305 these. I have refactored PCRE in four places to do this. The result is
306 dramatic:
308 Originally: 290
309 After changing UCP table: 187
310 After changing error message table: 43
311 After changing table of "verbs" 36
312 After changing table of Posix names 22
314 Thanks to the folks working on Gregex for glib for this insight.
316 11. --disable-stack-for-recursion caused compiling to fail unless -enable-
317 unicode-properties was also set.
319 12. Updated the tests so that they work when \R is defaulted to ANYCRLF.
321 13. Added checks for ANY and ANYCRLF to pcrecpp.cc where it previously
322 checked only for CRLF.
324 14. Added casts to pcretest.c to avoid compiler warnings.
326 15. Added Craig's patch to various pcrecpp modules to avoid compiler warnings.
328 16. Added Craig's patch to remove the WINDOWS_H tests, that were not working,
329 and instead check for _strtoi64 explicitly, and avoid the use of snprintf()
330 entirely. This removes changes made in 7 above.
332 17. The CMake files have been updated, and there is now more information about
333 building with CMake in the NON-UNIX-USE document.
336 Version 7.3 28-Aug-07
337 ---------------------
339 1. In the rejigging of the build system that eventually resulted in 7.1, the
340 line "#include <pcre.h>" was included in pcre_internal.h. The use of angle
341 brackets there is not right, since it causes compilers to look for an
342 installed pcre.h, not the version that is in the source that is being
343 compiled (which of course may be different). I have changed it back to:
345 #include "pcre.h"
347 I have a vague recollection that the change was concerned with compiling in
348 different directories, but in the new build system, that is taken care of
349 by the VPATH setting the Makefile.
351 2. The pattern .*$ when run in not-DOTALL UTF-8 mode with newline=any failed
352 when the subject happened to end in the byte 0x85 (e.g. if the last
353 character was \x{1ec5}). *Character* 0x85 is one of the "any" newline
354 characters but of course it shouldn't be taken as a newline when it is part
355 of another character. The bug was that, for an unlimited repeat of . in
356 not-DOTALL UTF-8 mode, PCRE was advancing by bytes rather than by
357 characters when looking for a newline.
359 3. A small performance improvement in the DOTALL UTF-8 mode .* case.
361 4. Debugging: adjusted the names of opcodes for different kinds of parentheses
362 in debug output.
364 5. Arrange to use "%I64d" instead of "%lld" and "%I64u" instead of "%llu" for
365 long printing in the pcrecpp unittest when running under MinGW.
367 6. ESC_K was left out of the EBCDIC table.
369 7. Change 7.0/38 introduced a new limit on the number of nested non-capturing
370 parentheses; I made it 1000, which seemed large enough. Unfortunately, the
371 limit also applies to "virtual nesting" when a pattern is recursive, and in
372 this case 1000 isn't so big. I have been able to remove this limit at the
373 expense of backing off one optimization in certain circumstances. Normally,
374 when pcre_exec() would call its internal match() function recursively and
375 immediately return the result unconditionally, it uses a "tail recursion"
376 feature to save stack. However, when a subpattern that can match an empty
377 string has an unlimited repetition quantifier, it no longer makes this
378 optimization. That gives it a stack frame in which to save the data for
379 checking that an empty string has been matched. Previously this was taken
380 from the 1000-entry workspace that had been reserved. So now there is no
381 explicit limit, but more stack is used.
383 8. Applied Daniel's patches to solve problems with the import/export magic
384 syntax that is required for Windows, and which was going wrong for the
385 pcreposix and pcrecpp parts of the library. These were overlooked when this
386 problem was solved for the main library.
388 9. There were some crude static tests to avoid integer overflow when computing
389 the size of patterns that contain repeated groups with explicit upper
390 limits. As the maximum quantifier is 65535, the maximum group length was
391 set at 30,000 so that the product of these two numbers did not overflow a
392 32-bit integer. However, it turns out that people want to use groups that
393 are longer than 30,000 bytes (though not repeat them that many times).
394 Change 7.0/17 (the refactoring of the way the pattern size is computed) has
395 made it possible to implement the integer overflow checks in a much more
396 dynamic way, which I have now done. The artificial limitation on group
397 length has been removed - we now have only the limit on the total length of
398 the compiled pattern, which depends on the LINK_SIZE setting.
400 10. Fixed a bug in the documentation for get/copy named substring when
401 duplicate names are permitted. If none of the named substrings are set, the
402 functions return PCRE_ERROR_NOSUBSTRING (7); the doc said they returned an
403 empty string.
405 11. Because Perl interprets \Q...\E at a high level, and ignores orphan \E
406 instances, patterns such as [\Q\E] or [\E] or even [^\E] cause an error,
407 because the ] is interpreted as the first data character and the
408 terminating ] is not found. PCRE has been made compatible with Perl in this
409 regard. Previously, it interpreted [\Q\E] as an empty class, and [\E] could
410 cause memory overwriting.
412 10. Like Perl, PCRE automatically breaks an unlimited repeat after an empty
413 string has been matched (to stop an infinite loop). It was not recognizing
414 a conditional subpattern that could match an empty string if that
415 subpattern was within another subpattern. For example, it looped when
416 trying to match (((?(1)X|))*) but it was OK with ((?(1)X|)*) where the
417 condition was not nested. This bug has been fixed.
419 12. A pattern like \X?\d or \P{L}?\d in non-UTF-8 mode could cause a backtrack
420 past the start of the subject in the presence of bytes with the top bit
421 set, for example "\x8aBCD".
423 13. Added Perl 5.10 experimental backtracking controls (*FAIL), (*F), (*PRUNE),
424 (*SKIP), (*THEN), (*COMMIT), and (*ACCEPT).
426 14. Optimized (?!) to (*FAIL).
428 15. Updated the test for a valid UTF-8 string to conform to the later RFC 3629.
429 This restricts code points to be within the range 0 to 0x10FFFF, excluding
430 the "low surrogate" sequence 0xD800 to 0xDFFF. Previously, PCRE allowed the
431 full range 0 to 0x7FFFFFFF, as defined by RFC 2279. Internally, it still
432 does: it's just the validity check that is more restrictive.
434 16. Inserted checks for integer overflows during escape sequence (backslash)
435 processing, and also fixed erroneous offset values for syntax errors during
436 backslash processing.
438 17. Fixed another case of looking too far back in non-UTF-8 mode (cf 12 above)
439 for patterns like [\PPP\x8a]{1,}\x80 with the subject "A\x80".
441 18. An unterminated class in a pattern like (?1)\c[ with a "forward reference"
442 caused an overrun.
444 19. A pattern like (?:[\PPa*]*){8,} which had an "extended class" (one with
445 something other than just ASCII characters) inside a group that had an
446 unlimited repeat caused a loop at compile time (while checking to see
447 whether the group could match an empty string).
449 20. Debugging a pattern containing \p or \P could cause a crash. For example,
450 [\P{Any}] did so. (Error in the code for printing property names.)
452 21. An orphan \E inside a character class could cause a crash.
454 22. A repeated capturing bracket such as (A)? could cause a wild memory
455 reference during compilation.
457 23. There are several functions in pcre_compile() that scan along a compiled
458 expression for various reasons (e.g. to see if it's fixed length for look
459 behind). There were bugs in these functions when a repeated \p or \P was
460 present in the pattern. These operators have additional parameters compared
461 with \d, etc, and these were not being taken into account when moving along
462 the compiled data. Specifically:
464 (a) A item such as \p{Yi}{3} in a lookbehind was not treated as fixed
465 length.
467 (b) An item such as \pL+ within a repeated group could cause crashes or
468 loops.
470 (c) A pattern such as \p{Yi}+(\P{Yi}+)(?1) could give an incorrect
471 "reference to non-existent subpattern" error.
473 (d) A pattern like (\P{Yi}{2}\277)? could loop at compile time.
475 24. A repeated \S or \W in UTF-8 mode could give wrong answers when multibyte
476 characters were involved (for example /\S{2}/8g with "A\x{a3}BC").
478 25. Using pcregrep in multiline, inverted mode (-Mv) caused it to loop.
480 26. Patterns such as [\P{Yi}A] which include \p or \P and just one other
481 character were causing crashes (broken optimization).
483 27. Patterns such as (\P{Yi}*\277)* (group with possible zero repeat containing
484 \p or \P) caused a compile-time loop.
486 28. More problems have arisen in unanchored patterns when CRLF is a valid line
487 break. For example, the unstudied pattern [\r\n]A does not match the string
488 "\r\nA" because change 7.0/46 below moves the current point on by two
489 characters after failing to match at the start. However, the pattern \nA
490 *does* match, because it doesn't start till \n, and if [\r\n]A is studied,
491 the same is true. There doesn't seem any very clean way out of this, but
492 what I have chosen to do makes the common cases work: PCRE now takes note
493 of whether there can be an explicit match for \r or \n anywhere in the
494 pattern, and if so, 7.0/46 no longer applies. As part of this change,
495 there's a new PCRE_INFO_HASCRORLF option for finding out whether a compiled
496 pattern has explicit CR or LF references.
498 29. Added (*CR) etc for changing newline setting at start of pattern.
501 Version 7.2 19-Jun-07
502 ---------------------
504 1. If the fr_FR locale cannot be found for test 3, try the "french" locale,
505 which is apparently normally available under Windows.
507 2. Re-jig the pcregrep tests with different newline settings in an attempt
508 to make them independent of the local environment's newline setting.
510 3. Add code to configure.ac to remove -g from the CFLAGS default settings.
512 4. Some of the "internals" tests were previously cut out when the link size
513 was not 2, because the output contained actual offsets. The recent new
514 "Z" feature of pcretest means that these can be cut out, making the tests
515 usable with all link sizes.
517 5. Implemented Stan Switzer's goto replacement for longjmp() when not using
518 stack recursion. This gives a massive performance boost under BSD, but just
519 a small improvement under Linux. However, it saves one field in the frame
520 in all cases.
522 6. Added more features from the forthcoming Perl 5.10:
524 (a) (?-n) (where n is a string of digits) is a relative subroutine or
525 recursion call. It refers to the nth most recently opened parentheses.
527 (b) (?+n) is also a relative subroutine call; it refers to the nth next
528 to be opened parentheses.
530 (c) Conditions that refer to capturing parentheses can be specified
531 relatively, for example, (?(-2)... or (?(+3)...
533 (d) \K resets the start of the current match so that everything before
534 is not part of it.
536 (e) \k{name} is synonymous with \k<name> and \k'name' (.NET compatible).
538 (f) \g{name} is another synonym - part of Perl 5.10's unification of
539 reference syntax.
541 (g) (?| introduces a group in which the numbering of parentheses in each
542 alternative starts with the same number.
544 (h) \h, \H, \v, and \V match horizontal and vertical whitespace.
546 7. Added two new calls to pcre_fullinfo(): PCRE_INFO_OKPARTIAL and
549 8. A pattern such as (.*(.)?)* caused pcre_exec() to fail by either not
550 terminating or by crashing. Diagnosed by Viktor Griph; it was in the code
551 for detecting groups that can match an empty string.
553 9. A pattern with a very large number of alternatives (more than several
554 hundred) was running out of internal workspace during the pre-compile
555 phase, where pcre_compile() figures out how much memory will be needed. A
556 bit of new cunning has reduced the workspace needed for groups with
557 alternatives. The 1000-alternative test pattern now uses 12 bytes of
558 workspace instead of running out of the 4096 that are available.
560 10. Inserted some missing (unsigned int) casts to get rid of compiler warnings.
562 11. Applied patch from Google to remove an optimization that didn't quite work.
563 The report of the bug said:
565 pcrecpp::RE("a*").FullMatch("aaa") matches, while
566 pcrecpp::RE("a*?").FullMatch("aaa") does not, and
567 pcrecpp::RE("a*?\\z").FullMatch("aaa") does again.
569 12. If \p or \P was used in non-UTF-8 mode on a character greater than 127
570 it matched the wrong number of bytes.
573 Version 7.1 24-Apr-07
574 ---------------------
576 1. Applied Bob Rossi and Daniel G's patches to convert the build system to one
577 that is more "standard", making use of automake and other Autotools. There
578 is some re-arrangement of the files and adjustment of comments consequent
579 on this.
581 2. Part of the patch fixed a problem with the pcregrep tests. The test of -r
582 for recursive directory scanning broke on some systems because the files
583 are not scanned in any specific order and on different systems the order
584 was different. A call to "sort" has been inserted into RunGrepTest for the
585 approprate test as a short-term fix. In the longer term there may be an
586 alternative.
588 3. I had an email from Eric Raymond about problems translating some of PCRE's
589 man pages to HTML (despite the fact that I distribute HTML pages, some
590 people do their own conversions for various reasons). The problems
591 concerned the use of low-level troff macros .br and .in. I have therefore
592 removed all such uses from the man pages (some were redundant, some could
593 be replaced by .nf/.fi pairs). The 132html script that I use to generate
594 HTML has been updated to handle .nf/.fi and to complain if it encounters
595 .br or .in.
597 4. Updated comments in configure.ac that get placed in config.h.in and also
598 arranged for config.h to be included in the distribution, with the name
599 config.h.generic, for the benefit of those who have to compile without
600 Autotools (compare pcre.h, which is now distributed as pcre.h.generic).
602 5. Updated the support (such as it is) for Virtual Pascal, thanks to Stefan
603 Weber: (1) pcre_internal.h was missing some function renames; (2) updated
604 makevp.bat for the current PCRE, using the additional files
605 makevp_c.txt, makevp_l.txt, and pcregexp.pas.
607 6. A Windows user reported a minor discrepancy with test 2, which turned out
608 to be caused by a trailing space on an input line that had got lost in his
609 copy. The trailing space was an accident, so I've just removed it.
611 7. Add -Wl,-R... flags in pcre-config.in for *BSD* systems, as I'm told
612 that is needed.
614 8. Mark ucp_table (in ucptable.h) and ucp_gentype (in pcre_ucp_searchfuncs.c)
615 as "const" (a) because they are and (b) because it helps the PHP
616 maintainers who have recently made a script to detect big data structures
617 in the php code that should be moved to the .rodata section. I remembered
618 to update Builducptable as well, so it won't revert if ucptable.h is ever
619 re-created.
621 9. Added some extra #ifdef SUPPORT_UTF8 conditionals into pcretest.c,
622 pcre_printint.src, pcre_compile.c, pcre_study.c, and pcre_tables.c, in
623 order to be able to cut out the UTF-8 tables in the latter when UTF-8
624 support is not required. This saves 1.5-2K of code, which is important in
625 some applications.
627 Later: more #ifdefs are needed in pcre_ord2utf8.c and pcre_valid_utf8.c
628 so as not to refer to the tables, even though these functions will never be
629 called when UTF-8 support is disabled. Otherwise there are problems with a
630 shared library.
632 10. Fixed two bugs in the emulated memmove() function in pcre_internal.h:
634 (a) It was defining its arguments as char * instead of void *.
636 (b) It was assuming that all moves were upwards in memory; this was true
637 a long time ago when I wrote it, but is no longer the case.
639 The emulated memove() is provided for those environments that have neither
640 memmove() nor bcopy(). I didn't think anyone used it these days, but that
641 is clearly not the case, as these two bugs were recently reported.
643 11. The script PrepareRelease is now distributed: it calls 132html, CleanTxt,
644 and Detrail to create the HTML documentation, the .txt form of the man
645 pages, and it removes trailing spaces from listed files. It also creates
646 pcre.h.generic and config.h.generic from pcre.h and config.h. In the latter
647 case, it wraps all the #defines with #ifndefs. This script should be run
648 before "make dist".
650 12. Fixed two fairly obscure bugs concerned with quantified caseless matching
651 with Unicode property support.
653 (a) For a maximizing quantifier, if the two different cases of the
654 character were of different lengths in their UTF-8 codings (there are
655 some cases like this - I found 11), and the matching function had to
656 back up over a mixture of the two cases, it incorrectly assumed they
657 were both the same length.
659 (b) When PCRE was configured to use the heap rather than the stack for
660 recursion during matching, it was not correctly preserving the data for
661 the other case of a UTF-8 character when checking ahead for a match
662 while processing a minimizing repeat. If the check also involved
663 matching a wide character, but failed, corruption could cause an
664 erroneous result when trying to check for a repeat of the original
665 character.
667 13. Some tidying changes to the testing mechanism:
669 (a) The RunTest script now detects the internal link size and whether there
670 is UTF-8 and UCP support by running ./pcretest -C instead of relying on
671 values substituted by "configure". (The RunGrepTest script already did
672 this for UTF-8.) The configure.ac script no longer substitutes the
673 relevant variables.
675 (b) The debugging options /B and /D in pcretest show the compiled bytecode
676 with length and offset values. This means that the output is different
677 for different internal link sizes. Test 2 is skipped for link sizes
678 other than 2 because of this, bypassing the problem. Unfortunately,
679 there was also a test in test 3 (the locale tests) that used /B and
680 failed for link sizes other than 2. Rather than cut the whole test out,
681 I have added a new /Z option to pcretest that replaces the length and
682 offset values with spaces. This is now used to make test 3 independent
683 of link size. (Test 2 will be tidied up later.)
685 14. If erroroffset was passed as NULL to pcre_compile, it provoked a
686 segmentation fault instead of returning the appropriate error message.
688 15. In multiline mode when the newline sequence was set to "any", the pattern
689 ^$ would give a match between the \r and \n of a subject such as "A\r\nB".
690 This doesn't seem right; it now treats the CRLF combination as the line
691 ending, and so does not match in that case. It's only a pattern such as ^$
692 that would hit this one: something like ^ABC$ would have failed after \r
693 and then tried again after \r\n.
695 16. Changed the comparison command for RunGrepTest from "diff -u" to "diff -ub"
696 in an attempt to make files that differ only in their line terminators
697 compare equal. This works on Linux.
699 17. Under certain error circumstances pcregrep might try to free random memory
700 as it exited. This is now fixed, thanks to valgrind.
702 19. In pcretest, if the pattern /(?m)^$/g<any> was matched against the string
703 "abc\r\n\r\n", it found an unwanted second match after the second \r. This
704 was because its rules for how to advance for /g after matching an empty
705 string at the end of a line did not allow for this case. They now check for
706 it specially.
708 20. pcretest is supposed to handle patterns and data of any length, by
709 extending its buffers when necessary. It was getting this wrong when the
710 buffer for a data line had to be extended.
712 21. Added PCRE_NEWLINE_ANYCRLF which is like ANY, but matches only CR, LF, or
713 CRLF as a newline sequence.
715 22. Code for handling Unicode properties in pcre_dfa_exec() wasn't being cut
716 out by #ifdef SUPPORT_UCP. This did no harm, as it could never be used, but
717 I have nevertheless tidied it up.
719 23. Added some casts to kill warnings from HP-UX ia64 compiler.
721 24. Added a man page for pcre-config.
724 Version 7.0 19-Dec-06
725 ---------------------
727 1. Fixed a signed/unsigned compiler warning in pcre_compile.c, shown up by
728 moving to gcc 4.1.1.
730 2. The -S option for pcretest uses setrlimit(); I had omitted to #include
731 sys/time.h, which is documented as needed for this function. It doesn't
732 seem to matter on Linux, but it showed up on some releases of OS X.
734 3. It seems that there are systems where bytes whose values are greater than
735 127 match isprint() in the "C" locale. The "C" locale should be the
736 default when a C program starts up. In most systems, only ASCII printing
737 characters match isprint(). This difference caused the output from pcretest
738 to vary, making some of the tests fail. I have changed pcretest so that:
740 (a) When it is outputting text in the compiled version of a pattern, bytes
741 other than 32-126 are always shown as hex escapes.
743 (b) When it is outputting text that is a matched part of a subject string,
744 it does the same, unless a different locale has been set for the match
745 (using the /L modifier). In this case, it uses isprint() to decide.
747 4. Fixed a major bug that caused incorrect computation of the amount of memory
748 required for a compiled pattern when options that changed within the
749 pattern affected the logic of the preliminary scan that determines the
750 length. The relevant options are -x, and -i in UTF-8 mode. The result was
751 that the computed length was too small. The symptoms of this bug were
752 either the PCRE error "internal error: code overflow" from pcre_compile(),
753 or a glibc crash with a message such as "pcretest: free(): invalid next
754 size (fast)". Examples of patterns that provoked this bug (shown in
755 pcretest format) are:
757 /(?-x: )/x
758 /(?x)(?-x: \s*#\s*)/
759 /((?i)[\x{c0}])/8
760 /(?i:[\x{c0}])/8
762 HOWEVER: Change 17 below makes this fix obsolete as the memory computation
763 is now done differently.
765 5. Applied patches from Google to: (a) add a QuoteMeta function to the C++
766 wrapper classes; (b) implement a new function in the C++ scanner that is
767 more efficient than the old way of doing things because it avoids levels of
768 recursion in the regex matching; (c) add a paragraph to the documentation
769 for the FullMatch() function.
771 6. The escape sequence \n was being treated as whatever was defined as
772 "newline". Not only was this contrary to the documentation, which states
773 that \n is character 10 (hex 0A), but it also went horribly wrong when
774 "newline" was defined as CRLF. This has been fixed.
776 7. In pcre_dfa_exec.c the value of an unsigned integer (the variable called c)
777 was being set to -1 for the "end of line" case (supposedly a value that no
778 character can have). Though this value is never used (the check for end of
779 line is "zero bytes in current character"), it caused compiler complaints.
780 I've changed it to 0xffffffff.
782 8. In pcre_version.c, the version string was being built by a sequence of
783 C macros that, in the event of PCRE_PRERELEASE being defined as an empty
784 string (as it is for production releases) called a macro with an empty
785 argument. The C standard says the result of this is undefined. The gcc
786 compiler treats it as an empty string (which was what was wanted) but it is
787 reported that Visual C gives an error. The source has been hacked around to
788 avoid this problem.
790 9. On the advice of a Windows user, included <io.h> and <fcntl.h> in Windows
791 builds of pcretest, and changed the call to _setmode() to use _O_BINARY
792 instead of 0x8000. Made all the #ifdefs test both _WIN32 and WIN32 (not all
793 of them did).
795 10. Originally, pcretest opened its input and output without "b"; then I was
796 told that "b" was needed in some environments, so it was added for release
797 5.0 to both the input and output. (It makes no difference on Unix-like
798 systems.) Later I was told that it is wrong for the input on Windows. I've
799 now abstracted the modes into two macros, to make it easier to fiddle with
800 them, and removed "b" from the input mode under Windows.
802 11. Added pkgconfig support for the C++ wrapper library, libpcrecpp.
804 12. Added -help and --help to pcretest as an official way of being reminded
805 of the options.
807 13. Removed some redundant semicolons after macro calls in pcrecpparg.h.in
808 and pcrecpp.cc because they annoy compilers at high warning levels.
810 14. A bit of tidying/refactoring in pcre_exec.c in the main bumpalong loop.
812 15. Fixed an occurrence of == in configure.ac that should have been = (shell
813 scripts are not C programs :-) and which was not noticed because it works
814 on Linux.
816 16. pcretest is supposed to handle any length of pattern and data line (as one
817 line or as a continued sequence of lines) by extending its input buffer if
818 necessary. This feature was broken for very long pattern lines, leading to
819 a string of junk being passed to pcre_compile() if the pattern was longer
820 than about 50K.
822 17. I have done a major re-factoring of the way pcre_compile() computes the
823 amount of memory needed for a compiled pattern. Previously, there was code
824 that made a preliminary scan of the pattern in order to do this. That was
825 OK when PCRE was new, but as the facilities have expanded, it has become
826 harder and harder to keep it in step with the real compile phase, and there
827 have been a number of bugs (see for example, 4 above). I have now found a
828 cunning way of running the real compile function in a "fake" mode that
829 enables it to compute how much memory it would need, while actually only
830 ever using a few hundred bytes of working memory and without too many
831 tests of the mode. This should make future maintenance and development
832 easier. A side effect of this work is that the limit of 200 on the nesting
833 depth of parentheses has been removed (though this was never a serious
834 limitation, I suspect). However, there is a downside: pcre_compile() now
835 runs more slowly than before (30% or more, depending on the pattern). I
836 hope this isn't a big issue. There is no effect on runtime performance.
838 18. Fixed a minor bug in pcretest: if a pattern line was not terminated by a
839 newline (only possible for the last line of a file) and it was a
840 pattern that set a locale (followed by /Lsomething), pcretest crashed.
842 19. Added additional timing features to pcretest. (1) The -tm option now times
843 matching only, not compiling. (2) Both -t and -tm can be followed, as a
844 separate command line item, by a number that specifies the number of
845 repeats to use when timing. The default is 50000; this gives better
846 precision, but takes uncomfortably long for very large patterns.
848 20. Extended pcre_study() to be more clever in cases where a branch of a
849 subpattern has no definite first character. For example, (a*|b*)[cd] would
850 previously give no result from pcre_study(). Now it recognizes that the
851 first character must be a, b, c, or d.
853 21. There was an incorrect error "recursive call could loop indefinitely" if
854 a subpattern (or the entire pattern) that was being tested for matching an
855 empty string contained only one non-empty item after a nested subpattern.
856 For example, the pattern (?>\x{100}*)\d(?R) provoked this error
857 incorrectly, because the \d was being skipped in the check.
859 22. The pcretest program now has a new pattern option /B and a command line
860 option -b, which is equivalent to adding /B to every pattern. This causes
861 it to show the compiled bytecode, without the additional information that
862 -d shows. The effect of -d is now the same as -b with -i (and similarly, /D
863 is the same as /B/I).
865 23. A new optimization is now able automatically to treat some sequences such
866 as a*b as a*+b. More specifically, if something simple (such as a character
867 or a simple class like \d) has an unlimited quantifier, and is followed by
868 something that cannot possibly match the quantified thing, the quantifier
869 is automatically "possessified".
871 24. A recursive reference to a subpattern whose number was greater than 39
872 went wrong under certain circumstances in UTF-8 mode. This bug could also
873 have affected the operation of pcre_study().
875 25. Realized that a little bit of performance could be had by replacing
876 (c & 0xc0) == 0xc0 with c >= 0xc0 when processing UTF-8 characters.
878 26. Timing data from pcretest is now shown to 4 decimal places instead of 3.
880 27. Possessive quantifiers such as a++ were previously implemented by turning
881 them into atomic groups such as ($>a+). Now they have their own opcodes,
882 which improves performance. This includes the automatically created ones
883 from 23 above.
885 28. A pattern such as (?=(\w+))\1: which simulates an atomic group using a
886 lookahead was broken if it was not anchored. PCRE was mistakenly expecting
887 the first matched character to be a colon. This applied both to named and
888 numbered groups.
890 29. The ucpinternal.h header file was missing its idempotency #ifdef.
892 30. I was sent a "project" file called libpcre.a.dev which I understand makes
893 building PCRE on Windows easier, so I have included it in the distribution.
895 31. There is now a check in pcretest against a ridiculously large number being
896 returned by pcre_exec() or pcre_dfa_exec(). If this happens in a /g or /G
897 loop, the loop is abandoned.
899 32. Forward references to subpatterns in conditions such as (?(2)...) where
900 subpattern 2 is defined later cause pcre_compile() to search forwards in
901 the pattern for the relevant set of parentheses. This search went wrong
902 when there were unescaped parentheses in a character class, parentheses
903 escaped with \Q...\E, or parentheses in a #-comment in /x mode.
905 33. "Subroutine" calls and backreferences were previously restricted to
906 referencing subpatterns earlier in the regex. This restriction has now
907 been removed.
909 34. Added a number of extra features that are going to be in Perl 5.10. On the
910 whole, these are just syntactic alternatives for features that PCRE had
911 previously implemented using the Python syntax or my own invention. The
912 other formats are all retained for compatibility.
914 (a) Named groups can now be defined as (?<name>...) or (?'name'...) as well
915 as (?P<name>...). The new forms, as well as being in Perl 5.10, are
916 also .NET compatible.
918 (b) A recursion or subroutine call to a named group can now be defined as
919 (?&name) as well as (?P>name).
921 (c) A backreference to a named group can now be defined as \k<name> or
922 \k'name' as well as (?P=name). The new forms, as well as being in Perl
923 5.10, are also .NET compatible.
925 (d) A conditional reference to a named group can now use the syntax
926 (?(<name>) or (?('name') as well as (?(name).
928 (e) A "conditional group" of the form (?(DEFINE)...) can be used to define
929 groups (named and numbered) that are never evaluated inline, but can be
930 called as "subroutines" from elsewhere. In effect, the DEFINE condition
931 is always false. There may be only one alternative in such a group.
933 (f) A test for recursion can be given as (?(R1).. or (?(R&name)... as well
934 as the simple (?(R). The condition is true only if the most recent
935 recursion is that of the given number or name. It does not search out
936 through the entire recursion stack.
938 (g) The escape \gN or \g{N} has been added, where N is a positive or
939 negative number, specifying an absolute or relative reference.
941 35. Tidied to get rid of some further signed/unsigned compiler warnings and
942 some "unreachable code" warnings.
944 36. Updated the Unicode property tables to Unicode version 5.0.0. Amongst other
945 things, this adds five new scripts.
947 37. Perl ignores orphaned \E escapes completely. PCRE now does the same.
948 There were also incompatibilities regarding the handling of \Q..\E inside
949 character classes, for example with patterns like [\Qa\E-\Qz\E] where the
950 hyphen was adjacent to \Q or \E. I hope I've cleared all this up now.
952 38. Like Perl, PCRE detects when an indefinitely repeated parenthesized group
953 matches an empty string, and forcibly breaks the loop. There were bugs in
954 this code in non-simple cases. For a pattern such as ^(a()*)* matched
955 against aaaa the result was just "a" rather than "aaaa", for example. Two
956 separate and independent bugs (that affected different cases) have been
957 fixed.
959 39. Refactored the code to abolish the use of different opcodes for small
960 capturing bracket numbers. This is a tidy that I avoided doing when I
961 removed the limit on the number of capturing brackets for 3.5 back in 2001.
962 The new approach is not only tidier, it makes it possible to reduce the
963 memory needed to fix the previous bug (38).
965 40. Implemented PCRE_NEWLINE_ANY to recognize any of the Unicode newline
966 sequences (http://unicode.org/unicode/reports/tr18/) as "newline" when
967 processing dot, circumflex, or dollar metacharacters, or #-comments in /x
968 mode.
970 41. Add \R to match any Unicode newline sequence, as suggested in the Unicode
971 report.
973 42. Applied patch, originally from Ari Pollak, modified by Google, to allow
974 copy construction and assignment in the C++ wrapper.
976 43. Updated pcregrep to support "--newline=any". In the process, I fixed a
977 couple of bugs that could have given wrong results in the "--newline=crlf"
978 case.
980 44. Added a number of casts and did some reorganization of signed/unsigned int
981 variables following suggestions from Dair Grant. Also renamed the variable
982 "this" as "item" because it is a C++ keyword.
984 45. Arranged for dftables to add
986 #include "pcre_internal.h"
988 to pcre_chartables.c because without it, gcc 4.x may remove the array
989 definition from the final binary if PCRE is built into a static library and
990 dead code stripping is activated.
992 46. For an unanchored pattern, if a match attempt fails at the start of a
993 newline sequence, and the newline setting is CRLF or ANY, and the next two
994 characters are CRLF, advance by two characters instead of one.
997 Version 6.7 04-Jul-06
998 ---------------------
1000 1. In order to handle tests when input lines are enormously long, pcretest has
1001 been re-factored so that it automatically extends its buffers when
1002 necessary. The code is crude, but this _is_ just a test program. The
1003 default size has been increased from 32K to 50K.
1005 2. The code in pcre_study() was using the value of the re argument before
1006 testing it for NULL. (Of course, in any sensible call of the function, it
1007 won't be NULL.)
1009 3. The memmove() emulation function in pcre_internal.h, which is used on
1010 systems that lack both memmove() and bcopy() - that is, hardly ever -
1011 was missing a "static" storage class specifier.
1013 4. When UTF-8 mode was not set, PCRE looped when compiling certain patterns
1014 containing an extended class (one that cannot be represented by a bitmap
1015 because it contains high-valued characters or Unicode property items, e.g.
1016 [\pZ]). Almost always one would set UTF-8 mode when processing such a
1017 pattern, but PCRE should not loop if you do not (it no longer does).
1018 [Detail: two cases were found: (a) a repeated subpattern containing an
1019 extended class; (b) a recursive reference to a subpattern that followed a
1020 previous extended class. It wasn't skipping over the extended class
1021 correctly when UTF-8 mode was not set.]
1023 5. A negated single-character class was not being recognized as fixed-length
1024 in lookbehind assertions such as (?<=[^f]), leading to an incorrect
1025 compile error "lookbehind assertion is not fixed length".
1027 6. The RunPerlTest auxiliary script was showing an unexpected difference
1028 between PCRE and Perl for UTF-8 tests. It turns out that it is hard to
1029 write a Perl script that can interpret lines of an input file either as
1030 byte characters or as UTF-8, which is what "perltest" was being required to
1031 do for the non-UTF-8 and UTF-8 tests, respectively. Essentially what you
1032 can't do is switch easily at run time between having the "use utf8;" pragma
1033 or not. In the end, I fudged it by using the RunPerlTest script to insert
1034 "use utf8;" explicitly for the UTF-8 tests.
1036 7. In multiline (/m) mode, PCRE was matching ^ after a terminating newline at
1037 the end of the subject string, contrary to the documentation and to what
1038 Perl does. This was true of both matching functions. Now it matches only at
1039 the start of the subject and immediately after *internal* newlines.
1041 8. A call of pcre_fullinfo() from pcretest to get the option bits was passing
1042 a pointer to an int instead of a pointer to an unsigned long int. This
1043 caused problems on 64-bit systems.
1045 9. Applied a patch from the folks at Google to pcrecpp.cc, to fix "another
1046 instance of the 'standard' template library not being so standard".
1048 10. There was no check on the number of named subpatterns nor the maximum
1049 length of a subpattern name. The product of these values is used to compute
1050 the size of the memory block for a compiled pattern. By supplying a very
1051 long subpattern name and a large number of named subpatterns, the size
1052 computation could be caused to overflow. This is now prevented by limiting
1053 the length of names to 32 characters, and the number of named subpatterns
1054 to 10,000.
1056 11. Subpatterns that are repeated with specific counts have to be replicated in
1057 the compiled pattern. The size of memory for this was computed from the
1058 length of the subpattern and the repeat count. The latter is limited to
1059 65535, but there was no limit on the former, meaning that integer overflow
1060 could in principle occur. The compiled length of a repeated subpattern is
1061 now limited to 30,000 bytes in order to prevent this.
1063 12. Added the optional facility to have named substrings with the same name.
1065 13. Added the ability to use a named substring as a condition, using the
1066 Python syntax: (?(name)yes|no). This overloads (?(R)... and names that
1067 are numbers (not recommended). Forward references are permitted.
1069 14. Added forward references in named backreferences (if you see what I mean).
1071 15. In UTF-8 mode, with the PCRE_DOTALL option set, a quantified dot in the
1072 pattern could run off the end of the subject. For example, the pattern
1073 "(?s)(.{1,5})"8 did this with the subject "ab".
1075 16. If PCRE_DOTALL or PCRE_MULTILINE were set, pcre_dfa_exec() behaved as if
1076 PCRE_CASELESS was set when matching characters that were quantified with ?
1077 or *.
1079 17. A character class other than a single negated character that had a minimum
1080 but no maximum quantifier - for example [ab]{6,} - was not handled
1081 correctly by pce_dfa_exec(). It would match only one character.
1083 18. A valid (though odd) pattern that looked like a POSIX character
1084 class but used an invalid character after [ (for example [[,abc,]]) caused
1085 pcre_compile() to give the error "Failed: internal error: code overflow" or
1086 in some cases to crash with a glibc free() error. This could even happen if
1087 the pattern terminated after [[ but there just happened to be a sequence of
1088 letters, a binary zero, and a closing ] in the memory that followed.
1090 19. Perl's treatment of octal escapes in the range \400 to \777 has changed
1091 over the years. Originally (before any Unicode support), just the bottom 8
1092 bits were taken. Thus, for example, \500 really meant \100. Nowadays the
1093 output from "man perlunicode" includes this:
1095 The regular expression compiler produces polymorphic opcodes. That
1096 is, the pattern adapts to the data and automatically switches to
1097 the Unicode character scheme when presented with Unicode data--or
1098 instead uses a traditional byte scheme when presented with byte
1099 data.
1101 Sadly, a wide octal escape does not cause a switch, and in a string with
1102 no other multibyte characters, these octal escapes are treated as before.
1103 Thus, in Perl, the pattern /\500/ actually matches \100 but the pattern
1104 /\500|\x{1ff}/ matches \500 or \777 because the whole thing is treated as a
1105 Unicode string.
1107 I have not perpetrated such confusion in PCRE. Up till now, it took just
1108 the bottom 8 bits, as in old Perl. I have now made octal escapes with
1109 values greater than \377 illegal in non-UTF-8 mode. In UTF-8 mode they
1110 translate to the appropriate multibyte character.
1112 29. Applied some refactoring to reduce the number of warnings from Microsoft
1113 and Borland compilers. This has included removing the fudge introduced
1114 seven years ago for the OS/2 compiler (see 2.02/2 below) because it caused
1115 a warning about an unused variable.
1117 21. PCRE has not included VT (character 0x0b) in the set of whitespace
1118 characters since release 4.0, because Perl (from release 5.004) does not.
1119 [Or at least, is documented not to: some releases seem to be in conflict
1120 with the documentation.] However, when a pattern was studied with
1121 pcre_study() and all its branches started with \s, PCRE still included VT
1122 as a possible starting character. Of course, this did no harm; it just
1123 caused an unnecessary match attempt.
1125 22. Removed a now-redundant internal flag bit that recorded the fact that case
1126 dependency changed within the pattern. This was once needed for "required
1127 byte" processing, but is no longer used. This recovers a now-scarce options
1128 bit. Also moved the least significant internal flag bit to the most-
1129 significant bit of the word, which was not previously used (hangover from
1130 the days when it was an int rather than a uint) to free up another bit for
1131 the future.
1133 23. Added support for CRLF line endings as well as CR and LF. As well as the
1134 default being selectable at build time, it can now be changed at runtime
1135 via the PCRE_NEWLINE_xxx flags. There are now options for pcregrep to
1136 specify that it is scanning data with non-default line endings.
1138 24. Changed the definition of CXXLINK to make it agree with the definition of
1139 LINK in the Makefile, by replacing LDFLAGS to CXXFLAGS.
1141 25. Applied Ian Taylor's patches to avoid using another stack frame for tail
1142 recursions. This makes a big different to stack usage for some patterns.
1144 26. If a subpattern containing a named recursion or subroutine reference such
1145 as (?P>B) was quantified, for example (xxx(?P>B)){3}, the calculation of
1146 the space required for the compiled pattern went wrong and gave too small a
1147 value. Depending on the environment, this could lead to "Failed: internal
1148 error: code overflow at offset 49" or "glibc detected double free or
1149 corruption" errors.
1151 27. Applied patches from Google (a) to support the new newline modes and (b) to
1152 advance over multibyte UTF-8 characters in GlobalReplace.
1154 28. Change free() to pcre_free() in pcredemo.c. Apparently this makes a
1155 difference for some implementation of PCRE in some Windows version.
1157 29. Added some extra testing facilities to pcretest:
1159 \q<number> in a data line sets the "match limit" value
1160 \Q<number> in a data line sets the "match recursion limt" value
1161 -S <number> sets the stack size, where <number> is in megabytes
1163 The -S option isn't available for Windows.
1166 Version 6.6 06-Feb-06
1167 ---------------------
1169 1. Change 16(a) for 6.5 broke things, because PCRE_DATA_SCOPE was not defined
1170 in pcreposix.h. I have copied the definition from pcre.h.
1172 2. Change 25 for 6.5 broke compilation in a build directory out-of-tree
1173 because pcre.h is no longer a built file.
1175 3. Added Jeff Friedl's additional debugging patches to pcregrep. These are
1176 not normally included in the compiled code.
1179 Version 6.5 01-Feb-06
1180 ---------------------
1182 1. When using the partial match feature with pcre_dfa_exec(), it was not
1183 anchoring the second and subsequent partial matches at the new starting
1184 point. This could lead to incorrect results. For example, with the pattern
1185 /1234/, partially matching against "123" and then "a4" gave a match.
1187 2. Changes to pcregrep:
1189 (a) All non-match returns from pcre_exec() were being treated as failures
1190 to match the line. Now, unless the error is PCRE_ERROR_NOMATCH, an
1191 error message is output. Some extra information is given for the
1193 probably the only errors that are likely to be caused by users (by
1194 specifying a regex that has nested indefinite repeats, for instance).
1195 If there are more than 20 of these errors, pcregrep is abandoned.
1197 (b) A binary zero was treated as data while matching, but terminated the
1198 output line if it was written out. This has been fixed: binary zeroes
1199 are now no different to any other data bytes.
1201 (c) Whichever of the LC_ALL or LC_CTYPE environment variables is set is
1202 used to set a locale for matching. The --locale=xxxx long option has
1203 been added (no short equivalent) to specify a locale explicitly on the
1204 pcregrep command, overriding the environment variables.
1206 (d) When -B was used with -n, some line numbers in the output were one less
1207 than they should have been.
1209 (e) Added the -o (--only-matching) option.
1211 (f) If -A or -C was used with -c (count only), some lines of context were
1212 accidentally printed for the final match.
1214 (g) Added the -H (--with-filename) option.
1216 (h) The combination of options -rh failed to suppress file names for files
1217 that were found from directory arguments.
1219 (i) Added the -D (--devices) and -d (--directories) options.
1221 (j) Added the -F (--fixed-strings) option.
1223 (k) Allow "-" to be used as a file name for -f as well as for a data file.
1225 (l) Added the --colo(u)r option.
1227 (m) Added Jeffrey Friedl's -S testing option, but within #ifdefs so that it
1228 is not present by default.
1230 3. A nasty bug was discovered in the handling of recursive patterns, that is,
1231 items such as (?R) or (?1), when the recursion could match a number of
1232 alternatives. If it matched one of the alternatives, but subsequently,
1233 outside the recursion, there was a failure, the code tried to back up into
1234 the recursion. However, because of the way PCRE is implemented, this is not
1235 possible, and the result was an incorrect result from the match.
1237 In order to prevent this happening, the specification of recursion has
1238 been changed so that all such subpatterns are automatically treated as
1239 atomic groups. Thus, for example, (?R) is treated as if it were (?>(?R)).
1241 4. I had overlooked the fact that, in some locales, there are characters for
1242 which isalpha() is true but neither isupper() nor islower() are true. In
1243 the fr_FR locale, for instance, the \xAA and \xBA characters (ordmasculine
1244 and ordfeminine) are like this. This affected the treatment of \w and \W
1245 when they appeared in character classes, but not when they appeared outside
1246 a character class. The bit map for "word" characters is now created
1247 separately from the results of isalnum() instead of just taking it from the
1248 upper, lower, and digit maps. (Plus the underscore character, of course.)
1250 5. The above bug also affected the handling of POSIX character classes such as
1251 [[:alpha:]] and [[:alnum:]]. These do not have their own bit maps in PCRE's
1252 permanent tables. Instead, the bit maps for such a class were previously
1253 created as the appropriate unions of the upper, lower, and digit bitmaps.
1254 Now they are created by subtraction from the [[:word:]] class, which has
1255 its own bitmap.
1257 6. The [[:blank:]] character class matches horizontal, but not vertical space.
1258 It is created by subtracting the vertical space characters (\x09, \x0a,
1259 \x0b, \x0c) from the [[:space:]] bitmap. Previously, however, the
1260 subtraction was done in the overall bitmap for a character class, meaning
1261 that a class such as [\x0c[:blank:]] was incorrect because \x0c would not
1262 be recognized. This bug has been fixed.
1264 7. Patches from the folks at Google:
1266 (a) pcrecpp.cc: "to handle a corner case that may or may not happen in
1267 real life, but is still worth protecting against".
1269 (b) pcrecpp.cc: "corrects a bug when negative radixes are used with
1270 regular expressions".
1272 (c) pcre_scanner.cc: avoid use of std::count() because not all systems
1273 have it.
1275 (d) Split off pcrecpparg.h from pcrecpp.h and had the former built by
1276 "configure" and the latter not, in order to fix a problem somebody had
1277 with compiling the Arg class on HP-UX.
1279 (e) Improve the error-handling of the C++ wrapper a little bit.
1281 (f) New tests for checking recursion limiting.
1283 8. The pcre_memmove() function, which is used only if the environment does not
1284 have a standard memmove() function (and is therefore rarely compiled),
1285 contained two bugs: (a) use of int instead of size_t, and (b) it was not
1286 returning a result (though PCRE never actually uses the result).
1288 9. In the POSIX regexec() interface, if nmatch is specified as a ridiculously
1289 large number - greater than INT_MAX/(3*sizeof(int)) - REG_ESPACE is
1290 returned instead of calling malloc() with an overflowing number that would
1291 most likely cause subsequent chaos.
1293 10. The debugging option of pcretest was not showing the NO_AUTO_CAPTURE flag.
1295 11. The POSIX flag REG_NOSUB is now supported. When a pattern that was compiled
1296 with this option is matched, the nmatch and pmatch options of regexec() are
1297 ignored.
1299 12. Added REG_UTF8 to the POSIX interface. This is not defined by POSIX, but is
1300 provided in case anyone wants to the the POSIX interface with UTF-8
1301 strings.
1303 13. Added CXXLDFLAGS to the Makefile parameters to provide settings only on the
1304 C++ linking (needed for some HP-UX environments).
1306 14. Avoid compiler warnings in get_ucpname() when compiled without UCP support
1307 (unused parameter) and in the pcre_printint() function (omitted "default"
1308 switch label when the default is to do nothing).
1310 15. Added some code to make it possible, when PCRE is compiled as a C++
1311 library, to replace subject pointers for pcre_exec() with a smart pointer
1312 class, thus making it possible to process discontinuous strings.
1314 16. The two macros PCRE_EXPORT and PCRE_DATA_SCOPE are confusing, and perform
1315 much the same function. They were added by different people who were trying
1316 to make PCRE easy to compile on non-Unix systems. It has been suggested
1317 that PCRE_EXPORT be abolished now that there is more automatic apparatus
1318 for compiling on Windows systems. I have therefore replaced it with
1319 PCRE_DATA_SCOPE. This is set automatically for Windows; if not set it
1320 defaults to "extern" for C or "extern C" for C++, which works fine on
1321 Unix-like systems. It is now possible to override the value of PCRE_DATA_
1322 SCOPE with something explicit in config.h. In addition:
1324 (a) pcreposix.h still had just "extern" instead of either of these macros;
1325 I have replaced it with PCRE_DATA_SCOPE.
1327 (b) Functions such as _pcre_xclass(), which are internal to the library,
1328 but external in the C sense, all had PCRE_EXPORT in their definitions.
1329 This is apparently wrong for the Windows case, so I have removed it.
1330 (It makes no difference on Unix-like systems.)
1332 17. Added a new limit, MATCH_LIMIT_RECURSION, which limits the depth of nesting
1333 of recursive calls to match(). This is different to MATCH_LIMIT because
1334 that limits the total number of calls to match(), not all of which increase
1335 the depth of recursion. Limiting the recursion depth limits the amount of
1336 stack (or heap if NO_RECURSE is set) that is used. The default can be set
1337 when PCRE is compiled, and changed at run time. A patch from Google adds
1338 this functionality to the C++ interface.
1340 18. Changes to the handling of Unicode character properties:
1342 (a) Updated the table to Unicode 4.1.0.
1344 (b) Recognize characters that are not in the table as "Cn" (undefined).
1346 (c) I revised the way the table is implemented to a much improved format
1347 which includes recognition of ranges. It now supports the ranges that
1348 are defined in UnicodeData.txt, and it also amalgamates other
1349 characters into ranges. This has reduced the number of entries in the
1350 table from around 16,000 to around 3,000, thus reducing its size
1351 considerably. I realized I did not need to use a tree structure after
1352 all - a binary chop search is just as efficient. Having reduced the
1353 number of entries, I extended their size from 6 bytes to 8 bytes to
1354 allow for more data.
1356 (d) Added support for Unicode script names via properties such as \p{Han}.
1358 19. In UTF-8 mode, a backslash followed by a non-Ascii character was not
1359 matching that character.
1361 20. When matching a repeated Unicode property with a minimum greater than zero,
1362 (for example \pL{2,}), PCRE could look past the end of the subject if it
1363 reached it while seeking the minimum number of characters. This could
1364 happen only if some of the characters were more than one byte long, because
1365 there is a check for at least the minimum number of bytes.
1367 21. Refactored the implementation of \p and \P so as to be more general, to
1368 allow for more different types of property in future. This has changed the
1369 compiled form incompatibly. Anybody with saved compiled patterns that use
1370 \p or \P will have to recompile them.
1372 22. Added "Any" and "L&" to the supported property types.
1374 23. Recognize \x{...} as a code point specifier, even when not in UTF-8 mode,
1375 but give a compile time error if the value is greater than 0xff.
1377 24. The man pages for pcrepartial, pcreprecompile, and pcre_compile2 were
1378 accidentally not being installed or uninstalled.
1380 25. The pcre.h file was built from pcre.h.in, but the only changes that were
1381 made were to insert the current release number. This seemed silly, because
1382 it made things harder for people building PCRE on systems that don't run
1383 "configure". I have turned pcre.h into a distributed file, no longer built
1384 by "configure", with the version identification directly included. There is
1385 no longer a pcre.h.in file.
1387 However, this change necessitated a change to the pcre-config script as
1388 well. It is built from pcre-config.in, and one of the substitutions was the
1389 release number. I have updated configure.ac so that ./configure now finds
1390 the release number by grepping pcre.h.
1392 26. Added the ability to run the tests under valgrind.
1395 Version 6.4 05-Sep-05
1396 ---------------------
1398 1. Change 6.0/10/(l) to pcregrep introduced a bug that caused separator lines
1399 "--" to be printed when multiple files were scanned, even when none of the
1400 -A, -B, or -C options were used. This is not compatible with Gnu grep, so I
1401 consider it to be a bug, and have restored the previous behaviour.
1403 2. A couple of code tidies to get rid of compiler warnings.
1405 3. The pcretest program used to cheat by referring to symbols in the library
1406 whose names begin with _pcre_. These are internal symbols that are not
1407 really supposed to be visible externally, and in some environments it is
1408 possible to suppress them. The cheating is now confined to including
1409 certain files from the library's source, which is a bit cleaner.
1411 4. Renamed pcre.in as pcre.h.in to go with pcrecpp.h.in; it also makes the
1412 file's purpose clearer.
1414 5. Reorganized pcre_ucp_findchar().
1417 Version 6.3 15-Aug-05
1418 ---------------------
1420 1. The file libpcre.pc.in did not have general read permission in the tarball.
1422 2. There were some problems when building without C++ support:
1424 (a) If C++ support was not built, "make install" and "make test" still
1425 tried to test it.
1427 (b) There were problems when the value of CXX was explicitly set. Some
1428 changes have been made to try to fix these, and ...
1430 (c) --disable-cpp can now be used to explicitly disable C++ support.
1432 (d) The use of @CPP_OBJ@ directly caused a blank line preceded by a
1433 backslash in a target when C++ was disabled. This confuses some
1434 versions of "make", apparently. Using an intermediate variable solves
1435 this. (Same for CPP_LOBJ.)
1437 3. $(LINK_FOR_BUILD) now includes $(CFLAGS_FOR_BUILD) and $(LINK)
1438 (non-Windows) now includes $(CFLAGS) because these flags are sometimes
1439 necessary on certain architectures.
1441 4. Added a setting of -export-symbols-regex to the link command to remove
1442 those symbols that are exported in the C sense, but actually are local
1443 within the library, and not documented. Their names all begin with
1444 "_pcre_". This is not a perfect job, because (a) we have to except some
1445 symbols that pcretest ("illegally") uses, and (b) the facility isn't always
1446 available (and never for static libraries). I have made a note to try to
1447 find a way round (a) in the future.
1450 Version 6.2 01-Aug-05
1451 ---------------------
1453 1. There was no test for integer overflow of quantifier values. A construction
1454 such as {1111111111111111} would give undefined results. What is worse, if
1455 a minimum quantifier for a parenthesized subpattern overflowed and became
1456 negative, the calculation of the memory size went wrong. This could have
1457 led to memory overwriting.
1459 2. Building PCRE using VPATH was broken. Hopefully it is now fixed.
1461 3. Added "b" to the 2nd argument of fopen() in dftables.c, for non-Unix-like
1462 operating environments where this matters.
1464 4. Applied Giuseppe Maxia's patch to add additional features for controlling
1465 PCRE options from within the C++ wrapper.
1467 5. Named capturing subpatterns were not being correctly counted when a pattern
1468 was compiled. This caused two problems: (a) If there were more than 100
1469 such subpatterns, the calculation of the memory needed for the whole
1470 compiled pattern went wrong, leading to an overflow error. (b) Numerical
1471 back references of the form \12, where the number was greater than 9, were
1472 not recognized as back references, even though there were sufficient
1473 previous subpatterns.
1475 6. Two minor patches to pcrecpp.cc in order to allow it to compile on older
1476 versions of gcc, e.g. 2.95.4.
1479 Version 6.1 21-Jun-05
1480 ---------------------
1482 1. There was one reference to the variable "posix" in pcretest.c that was not
1483 surrounded by "#if !defined NOPOSIX".
1485 2. Make it possible to compile pcretest without DFA support, UTF8 support, or
1486 the cross-check on the old pcre_info() function, for the benefit of the
1487 cut-down version of PCRE that is currently imported into Exim.
1489 3. A (silly) pattern starting with (?i)(?-i) caused an internal space
1490 allocation error. I've done the easy fix, which wastes 2 bytes for sensible
1491 patterns that start (?i) but I don't think that matters. The use of (?i) is
1492 just an example; this all applies to the other options as well.
1494 4. Since libtool seems to echo the compile commands it is issuing, the output
1495 from "make" can be reduced a bit by putting "@" in front of each libtool
1496 compile command.
1498 5. Patch from the folks at Google for configure.in to be a bit more thorough
1499 in checking for a suitable C++ installation before trying to compile the
1500 C++ stuff. This should fix a reported problem when a compiler was present,
1501 but no suitable headers.
1503 6. The man pages all had just "PCRE" as their title. I have changed them to
1504 be the relevant file name. I have also arranged that these names are
1505 retained in the file doc/pcre.txt, which is a concatenation in text format
1506 of all the man pages except the little individual ones for each function.
1508 7. The NON-UNIX-USE file had not been updated for the different set of source
1509 files that come with release 6. I also added a few comments about the C++
1510 wrapper.
1513 Version 6.0 07-Jun-05
1514 ---------------------
1516 1. Some minor internal re-organization to help with my DFA experiments.
1518 2. Some missing #ifdef SUPPORT_UCP conditionals in pcretest and printint that
1519 didn't matter for the library itself when fully configured, but did matter
1520 when compiling without UCP support, or within Exim, where the ucp files are
1521 not imported.
1523 3. Refactoring of the library code to split up the various functions into
1524 different source modules. The addition of the new DFA matching code (see
1525 below) to a single monolithic source would have made it really too
1526 unwieldy, quite apart from causing all the code to be include in a
1527 statically linked application, when only some functions are used. This is
1528 relevant even without the DFA addition now that patterns can be compiled in
1529 one application and matched in another.
1531 The downside of splitting up is that there have to be some external
1532 functions and data tables that are used internally in different modules of
1533 the library but which are not part of the API. These have all had their
1534 names changed to start with "_pcre_" so that they are unlikely to clash
1535 with other external names.
1537 4. Added an alternate matching function, pcre_dfa_exec(), which matches using
1538 a different (DFA) algorithm. Although it is slower than the original
1539 function, it does have some advantages for certain types of matching
1540 problem.
1542 5. Upgrades to pcretest in order to test the features of pcre_dfa_exec(),
1543 including restarting after a partial match.
1545 6. A patch for pcregrep that defines INVALID_FILE_ATTRIBUTES if it is not
1546 defined when compiling for Windows was sent to me. I have put it into the
1547 code, though I have no means of testing or verifying it.
1549 7. Added the pcre_refcount() auxiliary function.
1551 8. Added the PCRE_FIRSTLINE option. This constrains an unanchored pattern to
1552 match before or at the first newline in the subject string. In pcretest,
1553 the /f option on a pattern can be used to set this.
1555 9. A repeated \w when used in UTF-8 mode with characters greater than 256
1556 would behave wrongly. This has been present in PCRE since release 4.0.
1558 10. A number of changes to the pcregrep command:
1560 (a) Refactored how -x works; insert ^(...)$ instead of setting
1561 PCRE_ANCHORED and checking the length, in preparation for adding
1562 something similar for -w.
1564 (b) Added the -w (match as a word) option.
1566 (c) Refactored the way lines are read and buffered so as to have more
1567 than one at a time available.
1569 (d) Implemented a pcregrep test script.
1571 (e) Added the -M (multiline match) option. This allows patterns to match
1572 over several lines of the subject. The buffering ensures that at least
1573 8K, or the rest of the document (whichever is the shorter) is available
1574 for matching (and similarly the previous 8K for lookbehind assertions).
1576 (f) Changed the --help output so that it now says
1578 -w, --word-regex(p)
1580 instead of two lines, one with "regex" and the other with "regexp"
1581 because that confused at least one person since the short forms are the
1582 same. (This required a bit of code, as the output is generated
1583 automatically from a table. It wasn't just a text change.)
1585 (g) -- can be used to terminate pcregrep options if the next thing isn't an
1586 option but starts with a hyphen. Could be a pattern or a path name
1587 starting with a hyphen, for instance.
1589 (h) "-" can be given as a file name to represent stdin.
1591 (i) When file names are being printed, "(standard input)" is used for
1592 the standard input, for compatibility with GNU grep. Previously
1593 "<stdin>" was used.
1595 (j) The option --label=xxx can be used to supply a name to be used for
1596 stdin when file names are being printed. There is no short form.
1598 (k) Re-factored the options decoding logic because we are going to add
1599 two more options that take data. Such options can now be given in four
1600 different ways, e.g. "-fname", "-f name", "--file=name", "--file name".
1602 (l) Added the -A, -B, and -C options for requesting that lines of context
1603 around matches be printed.
1605 (m) Added the -L option to print the names of files that do not contain
1606 any matching lines, that is, the complement of -l.
1608 (n) The return code is 2 if any file cannot be opened, but pcregrep does
1609 continue to scan other files.
1611 (o) The -s option was incorrectly implemented. For compatibility with other
1612 greps, it now suppresses the error message for a non-existent or non-
1613 accessible file (but not the return code). There is a new option called
1614 -q that suppresses the output of matching lines, which was what -s was
1615 previously doing.
1617 (p) Added --include and --exclude options to specify files for inclusion
1618 and exclusion when recursing.
1620 11. The Makefile was not using the Autoconf-supported LDFLAGS macro properly.
1621 Hopefully, it now does.
1623 12. Missing cast in pcre_study().
1625 13. Added an "uninstall" target to the makefile.
1627 14. Replaced "extern" in the function prototypes in Makefile.in with
1628 "PCRE_DATA_SCOPE", which defaults to 'extern' or 'extern "C"' in the Unix
1629 world, but is set differently for Windows.
1631 15. Added a second compiling function called pcre_compile2(). The only
1632 difference is that it has an extra argument, which is a pointer to an
1633 integer error code. When there is a compile-time failure, this is set
1634 non-zero, in addition to the error test pointer being set to point to an
1635 error message. The new argument may be NULL if no error number is required
1636 (but then you may as well call pcre_compile(), which is now just a
1637 wrapper). This facility is provided because some applications need a
1638 numeric error indication, but it has also enabled me to tidy up the way
1639 compile-time errors are handled in the POSIX wrapper.
1641 16. Added VPATH=.libs to the makefile; this should help when building with one
1642 prefix path and installing with another. (Or so I'm told by someone who
1643 knows more about this stuff than I do.)
1645 17. Added a new option, REG_DOTALL, to the POSIX function regcomp(). This
1646 passes PCRE_DOTALL to the pcre_compile() function, making the "." character
1647 match everything, including newlines. This is not POSIX-compatible, but
1648 somebody wanted the feature. From pcretest it can be activated by using
1649 both the P and the s flags.
1651 18. AC_PROG_LIBTOOL appeared twice in Makefile.in. Removed one.
1653 19. libpcre.pc was being incorrectly installed as executable.
1655 20. A couple of places in pcretest check for end-of-line by looking for '\n';
1656 it now also looks for '\r' so that it will work unmodified on Windows.
1658 21. Added Google's contributed C++ wrapper to the distribution.
1660 22. Added some untidy missing memory free() calls in pcretest, to keep
1661 Electric Fence happy when testing.
1665 Version 5.0 13-Sep-04
1666 ---------------------
1668 1. Internal change: literal characters are no longer packed up into items
1669 containing multiple characters in a single byte-string. Each character
1670 is now matched using a separate opcode. However, there may be more than one
1671 byte in the character in UTF-8 mode.
1673 2. The pcre_callout_block structure has two new fields: pattern_position and
1674 next_item_length. These contain the offset in the pattern to the next match
1675 item, and its length, respectively.
1677 3. The PCRE_AUTO_CALLOUT option for pcre_compile() requests the automatic
1678 insertion of callouts before each pattern item. Added the /C option to
1679 pcretest to make use of this.
1681 4. On the advice of a Windows user, the lines
1683 #if defined(_WIN32) || defined(WIN32)
1684 _setmode( _fileno( stdout ), 0x8000 );
1685 #endif /* defined(_WIN32) || defined(WIN32) */
1687 have been added to the source of pcretest. This apparently does useful
1688 magic in relation to line terminators.
1690 5. Changed "r" and "w" in the calls to fopen() in pcretest to "rb" and "wb"
1691 for the benefit of those environments where the "b" makes a difference.
1693 6. The icc compiler has the same options as gcc, but "configure" doesn't seem
1694 to know about it. I have put a hack into configure.in that adds in code
1695 to set GCC=yes if CC=icc. This seems to end up at a point in the
1696 generated configure script that is early enough to affect the setting of
1697 compiler options, which is what is needed, but I have no means of testing
1698 whether it really works. (The user who reported this had patched the
1699 generated configure script, which of course I cannot do.)
1701 LATER: After change 22 below (new libtool files), the configure script
1702 seems to know about icc (and also ecc). Therefore, I have commented out
1703 this hack in configure.in.
1705 7. Added support for pkg-config (2 patches were sent in).
1707 8. Negated POSIX character classes that used a combination of internal tables
1708 were completely broken. These were [[:^alpha:]], [[:^alnum:]], and
1709 [[:^ascii]]. Typically, they would match almost any characters. The other
1710 POSIX classes were not broken in this way.
1712 9. Matching the pattern "\b.*?" against "ab cd", starting at offset 1, failed
1713 to find the match, as PCRE was deluded into thinking that the match had to
1714 start at the start point or following a newline. The same bug applied to
1715 patterns with negative forward assertions or any backward assertions
1716 preceding ".*" at the start, unless the pattern required a fixed first
1717 character. This was a failing pattern: "(?!.bcd).*". The bug is now fixed.
1719 10. In UTF-8 mode, when moving forwards in the subject after a failed match
1720 starting at the last subject character, bytes beyond the end of the subject
1721 string were read.
1723 11. Renamed the variable "class" as "classbits" to make life easier for C++
1724 users. (Previously there was a macro definition, but it apparently wasn't
1725 enough.)
1727 12. Added the new field "tables" to the extra data so that tables can be passed
1728 in at exec time, or the internal tables can be re-selected. This allows
1729 a compiled regex to be saved and re-used at a later time by a different
1730 program that might have everything at different addresses.
1732 13. Modified the pcre-config script so that, when run on Solaris, it shows a
1733 -R library as well as a -L library.
1735 14. The debugging options of pcretest (-d on the command line or D on a
1736 pattern) showed incorrect output for anything following an extended class
1737 that contained multibyte characters and which was followed by a quantifier.
1739 15. Added optional support for general category Unicode character properties
1740 via the \p, \P, and \X escapes. Unicode property support implies UTF-8
1741 support. It adds about 90K to the size of the library. The meanings of the
1742 inbuilt class escapes such as \d and \s have NOT been changed.
1744 16. Updated pcredemo.c to include calls to free() to release the memory for the
1745 compiled pattern.
1747 17. The generated file chartables.c was being created in the source directory
1748 instead of in the building directory. This caused the build to fail if the
1749 source directory was different from the building directory, and was
1750 read-only.
1752 18. Added some sample Win commands from Mark Tetrode into the NON-UNIX-USE
1753 file. No doubt somebody will tell me if they don't make sense... Also added
1754 Dan Mooney's comments about building on OpenVMS.
1756 19. Added support for partial matching via the PCRE_PARTIAL option for
1757 pcre_exec() and the \P data escape in pcretest.
1759 20. Extended pcretest with 3 new pattern features:
1761 (i) A pattern option of the form ">rest-of-line" causes pcretest to
1762 write the compiled pattern to the file whose name is "rest-of-line".
1763 This is a straight binary dump of the data, with the saved pointer to
1764 the character tables forced to be NULL. The study data, if any, is
1765 written too. After writing, pcretest reads a new pattern.
1767 (ii) If, instead of a pattern, "<rest-of-line" is given, pcretest reads a
1768 compiled pattern from the given file. There must not be any
1769 occurrences of "<" in the file name (pretty unlikely); if there are,
1770 pcretest will instead treat the initial "<" as a pattern delimiter.
1771 After reading in the pattern, pcretest goes on to read data lines as
1772 usual.
1774 (iii) The F pattern option causes pcretest to flip the bytes in the 32-bit
1775 and 16-bit fields in a compiled pattern, to simulate a pattern that
1776 was compiled on a host of opposite endianness.
1778 21. The pcre-exec() function can now cope with patterns that were compiled on
1779 hosts of opposite endianness, with this restriction:
1781 As for any compiled expression that is saved and used later, the tables
1782 pointer field cannot be preserved; the extra_data field in the arguments
1783 to pcre_exec() should be used to pass in a tables address if a value
1784 other than the default internal tables were used at compile time.
1786 22. Calling pcre_exec() with a negative value of the "ovecsize" parameter is
1787 now diagnosed as an error. Previously, most of the time, a negative number
1788 would have been treated as zero, but if in addition "ovector" was passed as
1789 NULL, a crash could occur.
1791 23. Updated the files ltmain.sh, config.sub, config.guess, and aclocal.m4 with
1792 new versions from the libtool 1.5 distribution (the last one is a copy of
1793 a file called libtool.m4). This seems to have fixed the need to patch
1794 "configure" to support Darwin 1.3 (which I used to do). However, I still
1795 had to patch ltmain.sh to ensure that ${SED} is set (it isn't on my
1796 workstation).
1798 24. Changed the PCRE licence to be the more standard "BSD" licence.
1801 Version 4.5 01-Dec-03
1802 ---------------------
1804 1. There has been some re-arrangement of the code for the match() function so
1805 that it can be compiled in a version that does not call itself recursively.
1806 Instead, it keeps those local variables that need separate instances for
1807 each "recursion" in a frame on the heap, and gets/frees frames whenever it
1808 needs to "recurse". Keeping track of where control must go is done by means
1809 of setjmp/longjmp. The whole thing is implemented by a set of macros that
1810 hide most of the details from the main code, and operates only if
1811 NO_RECURSE is defined while compiling pcre.c. If PCRE is built using the
1812 "configure" mechanism, "--disable-stack-for-recursion" turns on this way of
1813 operating.
1815 To make it easier for callers to provide specially tailored get/free
1816 functions for this usage, two new functions, pcre_stack_malloc, and
1817 pcre_stack_free, are used. They are always called in strict stacking order,
1818 and the size of block requested is always the same.
1820 The PCRE_CONFIG_STACKRECURSE info parameter can be used to find out whether
1821 PCRE has been compiled to use the stack or the heap for recursion. The
1822 -C option of pcretest uses this to show which version is compiled.
1824 A new data escape \S, is added to pcretest; it causes the amounts of store
1825 obtained and freed by both kinds of malloc/free at match time to be added
1826 to the output.
1828 2. Changed the locale test to use "fr_FR" instead of "fr" because that's
1829 what's available on my current Linux desktop machine.
1831 3. When matching a UTF-8 string, the test for a valid string at the start has
1832 been extended. If start_offset is not zero, PCRE now checks that it points
1833 to a byte that is the start of a UTF-8 character. If not, it returns
1834 PCRE_ERROR_BADUTF8_OFFSET (-11). Note: the whole string is still checked;
1835 this is necessary because there may be backward assertions in the pattern.
1836 When matching the same subject several times, it may save resources to use
1837 PCRE_NO_UTF8_CHECK on all but the first call if the string is long.
1839 4. The code for checking the validity of UTF-8 strings has been tightened so
1840 that it rejects (a) strings containing 0xfe or 0xff bytes and (b) strings
1841 containing "overlong sequences".
1843 5. Fixed a bug (appearing twice) that I could not find any way of exploiting!
1844 I had written "if ((digitab[*p++] && chtab_digit) == 0)" where the "&&"
1845 should have been "&", but it just so happened that all the cases this let
1846 through by mistake were picked up later in the function.
1848 6. I had used a variable called "isblank" - this is a C99 function, causing
1849 some compilers to warn. To avoid this, I renamed it (as "blankclass").
1851 7. Cosmetic: (a) only output another newline at the end of pcretest if it is
1852 prompting; (b) run "./pcretest /dev/null" at the start of the test script
1853 so the version is shown; (c) stop "make test" echoing "./RunTest".
1855 8. Added patches from David Burgess to enable PCRE to run on EBCDIC systems.
1857 9. The prototype for memmove() for systems that don't have it was using
1858 size_t, but the inclusion of the header that defines size_t was later. I've
1859 moved the #includes for the C headers earlier to avoid this.
1861 10. Added some adjustments to the code to make it easier to compiler on certain
1862 special systems:
1864 (a) Some "const" qualifiers were missing.
1865 (b) Added the macro EXPORT before all exported functions; by default this
1866 is defined to be empty.
1867 (c) Changed the dftables auxiliary program (that builds chartables.c) so
1868 that it reads its output file name as an argument instead of writing
1869 to the standard output and assuming this can be redirected.
1871 11. In UTF-8 mode, if a recursive reference (e.g. (?1)) followed a character
1872 class containing characters with values greater than 255, PCRE compilation
1873 went into a loop.
1875 12. A recursive reference to a subpattern that was within another subpattern
1876 that had a minimum quantifier of zero caused PCRE to crash. For example,
1877 (x(y(?2))z)? provoked this bug with a subject that got as far as the
1878 recursion. If the recursively-called subpattern itself had a zero repeat,
1879 that was OK.
1881 13. In pcretest, the buffer for reading a data line was set at 30K, but the
1882 buffer into which it was copied (for escape processing) was still set at
1883 1024, so long lines caused crashes.
1885 14. A pattern such as /[ab]{1,3}+/ failed to compile, giving the error
1886 "internal error: code overflow...". This applied to any character class
1887 that was followed by a possessive quantifier.
1889 15. Modified the Makefile to add libpcre.la as a prerequisite for
1890 libpcreposix.la because I was told this is needed for a parallel build to
1891 work.
1893 16. If a pattern that contained .* following optional items at the start was
1894 studied, the wrong optimizing data was generated, leading to matching
1895 errors. For example, studying /[ab]*.*c/ concluded, erroneously, that any
1896 matching string must start with a or b or c. The correct conclusion for
1897 this pattern is that a match can start with any character.
1900 Version 4.4 13-Aug-03
1901 ---------------------
1903 1. In UTF-8 mode, a character class containing characters with values between
1904 127 and 255 was not handled correctly if the compiled pattern was studied.
1905 In fixing this, I have also improved the studying algorithm for such
1906 classes (slightly).
1908 2. Three internal functions had redundant arguments passed to them. Removal
1909 might give a very teeny performance improvement.
1911 3. Documentation bug: the value of the capture_top field in a callout is *one
1912 more than* the number of the hightest numbered captured substring.
1914 4. The Makefile linked pcretest and pcregrep with -lpcre, which could result
1915 in incorrectly linking with a previously installed version. They now link
1916 explicitly with libpcre.la.
1918 5. configure.in no longer needs to recognize Cygwin specially.
1920 6. A problem in pcre.in for Windows platforms is fixed.
1922 7. If a pattern was successfully studied, and the -d (or /D) flag was given to
1923 pcretest, it used to include the size of the study block as part of its
1924 output. Unfortunately, the structure contains a field that has a different
1925 size on different hardware architectures. This meant that the tests that
1926 showed this size failed. As the block is currently always of a fixed size,
1927 this information isn't actually particularly useful in pcretest output, so
1928 I have just removed it.
1930 8. Three pre-processor statements accidentally did not start in column 1.
1931 Sadly, there are *still* compilers around that complain, even though
1932 standard C has not required this for well over a decade. Sigh.
1934 9. In pcretest, the code for checking callouts passed small integers in the
1935 callout_data field, which is a void * field. However, some picky compilers
1936 complained about the casts involved for this on 64-bit systems. Now
1937 pcretest passes the address of the small integer instead, which should get
1938 rid of the warnings.
1940 10. By default, when in UTF-8 mode, PCRE now checks for valid UTF-8 strings at
1941 both compile and run time, and gives an error if an invalid UTF-8 sequence
1942 is found. There is a option for disabling this check in cases where the
1943 string is known to be correct and/or the maximum performance is wanted.
1945 11. In response to a bug report, I changed one line in Makefile.in from
1947 -Wl,--out-implib,.libs/lib@WIN_PREFIX@pcreposix.dll.a \
1948 to
1949 -Wl,--out-implib,.libs/@WIN_PREFIX@libpcreposix.dll.a \
1951 to look similar to other lines, but I have no way of telling whether this
1952 is the right thing to do, as I do not use Windows. No doubt I'll get told
1953 if it's wrong...
1956 Version 4.3 21-May-03
1957 ---------------------
1959 1. Two instances of @WIN_PREFIX@ omitted from the Windows targets in the
1960 Makefile.
1962 2. Some refactoring to improve the quality of the code:
1964 (i) The utf8_table... variables are now declared "const".
1966 (ii) The code for \cx, which used the "case flipping" table to upper case
1967 lower case letters, now just substracts 32. This is ASCII-specific,
1968 but the whole concept of \cx is ASCII-specific, so it seems
1969 reasonable.
1971 (iii) PCRE was using its character types table to recognize decimal and
1972 hexadecimal digits in the pattern. This is silly, because it handles
1973 only 0-9, a-f, and A-F, but the character types table is locale-
1974 specific, which means strange things might happen. A private
1975 table is now used for this - though it costs 256 bytes, a table is
1976 much faster than multiple explicit tests. Of course, the standard
1977 character types table is still used for matching digits in subject
1978 strings against \d.
1980 (iv) Strictly, the identifier ESC_t is reserved by POSIX (all identifiers
1981 ending in _t are). So I've renamed it as ESC_tee.
1983 3. The first argument for regexec() in the POSIX wrapper should have been
1984 defined as "const".
1986 4. Changed pcretest to use malloc() for its buffers so that they can be
1987 Electric Fenced for debugging.
1989 5. There were several places in the code where, in UTF-8 mode, PCRE would try
1990 to read one or more bytes before the start of the subject string. Often this
1991 had no effect on PCRE's behaviour, but in some circumstances it could
1992 provoke a segmentation fault.
1994 6. A lookbehind at the start of a pattern in UTF-8 mode could also cause PCRE
1995 to try to read one or more bytes before the start of the subject string.
1997 7. A lookbehind in a pattern matched in non-UTF-8 mode on a PCRE compiled with
1998 UTF-8 support could misbehave in various ways if the subject string
1999 contained bytes with the 0x80 bit set and the 0x40 bit unset in a lookbehind
2000 area. (PCRE was not checking for the UTF-8 mode flag, and trying to move
2001 back over UTF-8 characters.)
2004 Version 4.2 14-Apr-03
2005 ---------------------
2007 1. Typo "#if SUPPORT_UTF8" instead of "#ifdef SUPPORT_UTF8" fixed.
2009 2. Changes to the building process, supplied by Ronald Landheer-Cieslak
2010 [ON_WINDOWS]: new variable, "#" on non-Windows platforms
2011 [NOT_ON_WINDOWS]: new variable, "#" on Windows platforms
2012 [WIN_PREFIX]: new variable, "cyg" for Cygwin
2013 * Makefile.in: use autoconf substitution for OBJEXT, EXEEXT, BUILD_OBJEXT
2015 Note: automatic setting of the BUILD variables is not yet working
2016 set CPPFLAGS and BUILD_CPPFLAGS (but don't use yet) - should be used at
2017 compile-time but not at link-time
2018 [LINK]: use for linking executables only
2019 make different versions for Windows and non-Windows
2020 [LINKLIB]: new variable, copy of UNIX-style LINK, used for linking
2021 libraries
2022 [LINK_FOR_BUILD]: new variable
2023 [OBJEXT]: use throughout
2024 [EXEEXT]: use throughout
2025 <winshared>: new target
2026 <wininstall>: new target
2027 <dftables.o>: use native compiler
2028 <dftables>: use native linker
2029 <install>: handle Windows platform correctly
2030 <clean>: ditto
2031 <check>: ditto
2032 copy DLL to top builddir before testing
2034 As part of these changes, -no-undefined was removed again. This was reported
2035 to give trouble on HP-UX 11.0, so getting rid of it seems like a good idea
2036 in any case.
2038 3. Some tidies to get rid of compiler warnings:
2040 . In the match_data structure, match_limit was an unsigned long int, whereas
2041 match_call_count was an int. I've made them both unsigned long ints.
2043 . In pcretest the fact that a const uschar * doesn't automatically cast to
2044 a void * provoked a warning.
2046 . Turning on some more compiler warnings threw up some "shadow" variables
2047 and a few more missing casts.
2049 4. If PCRE was complied with UTF-8 support, but called without the PCRE_UTF8
2050 option, a class that contained a single character with a value between 128
2051 and 255 (e.g. /[\xFF]/) caused PCRE to crash.
2053 5. If PCRE was compiled with UTF-8 support, but called without the PCRE_UTF8
2054 option, a class that contained several characters, but with at least one
2055 whose value was between 128 and 255 caused PCRE to crash.
2058 Version 4.1 12-Mar-03
2059 ---------------------
2061 1. Compiling with gcc -pedantic found a couple of places where casts were
2062 needed, and a string in dftables.c that was longer than standard compilers are
2063 required to support.
2065 2. Compiling with Sun's compiler found a few more places where the code could
2066 be tidied up in order to avoid warnings.
2068 3. The variables for cross-compiling were called HOST_CC and HOST_CFLAGS; the
2069 first of these names is deprecated in the latest Autoconf in favour of the name
2070 CC_FOR_BUILD, because "host" is typically used to mean the system on which the
2071 compiled code will be run. I can't find a reference for HOST_CFLAGS, but by
2072 analogy I have changed it to CFLAGS_FOR_BUILD.
2074 4. Added -no-undefined to the linking command in the Makefile, because this is
2075 apparently helpful for Windows. To make it work, also added "-L. -lpcre" to the
2076 linking step for the pcreposix library.
2078 5. PCRE was failing to diagnose the case of two named groups with the same
2079 name.
2081 6. A problem with one of PCRE's optimizations was discovered. PCRE remembers a
2082 literal character that is needed in the subject for a match, and scans along to
2083 ensure that it is present before embarking on the full matching process. This
2084 saves time in cases of nested unlimited repeats that are never going to match.
2085 Problem: the scan can take a lot of time if the subject is very long (e.g.
2086 megabytes), thus penalizing straightforward matches. It is now done only if the
2087 amount of subject to be scanned is less than 1000 bytes.
2089 7. A lesser problem with the same optimization is that it was recording the
2090 first character of an anchored pattern as "needed", thus provoking a search
2091 right along the subject, even when the first match of the pattern was going to
2092 fail. The "needed" character is now not set for anchored patterns, unless it
2093 follows something in the pattern that is of non-fixed length. Thus, it still
2094 fulfils its original purpose of finding quick non-matches in cases of nested
2095 unlimited repeats, but isn't used for simple anchored patterns such as /^abc/.
2098 Version 4.0 17-Feb-03
2099 ---------------------
2101 1. If a comment in an extended regex that started immediately after a meta-item
2102 extended to the end of string, PCRE compiled incorrect data. This could lead to
2103 all kinds of weird effects. Example: /#/ was bad; /()#/ was bad; /a#/ was not.
2105 2. Moved to autoconf 2.53 and libtool 1.4.2.
2107 3. Perl 5.8 no longer needs "use utf8" for doing UTF-8 things. Consequently,
2108 the special perltest8 script is no longer needed - all the tests can be run
2109 from a single perltest script.
2111 4. From 5.004, Perl has not included the VT character (0x0b) in the set defined
2112 by \s. It has now been removed in PCRE. This means it isn't recognized as
2113 whitespace in /x regexes too, which is the same as Perl. Note that the POSIX
2114 class [:space:] *does* include VT, thereby creating a mess.
2116 5. Added the class [:blank:] (a GNU extension from Perl 5.8) to match only
2117 space and tab.
2119 6. Perl 5.005 was a long time ago. It's time to amalgamate the tests that use
2120 its new features into the main test script, reducing the number of scripts.
2122 7. Perl 5.8 has changed the meaning of patterns like /a(?i)b/. Earlier versions
2123 were backward compatible, and made the (?i) apply to the whole pattern, as if
2124 /i were given. Now it behaves more logically, and applies the option setting
2125 only to what follows. PCRE has been changed to follow suit. However, if it
2126 finds options settings right at the start of the pattern, it extracts them into
2127 the global options, as before. Thus, they show up in the info data.
2129 8. Added support for the \Q...\E escape sequence. Characters in between are
2130 treated as literals. This is slightly different from Perl in that $ and @ are
2131 also handled as literals inside the quotes. In Perl, they will cause variable
2132 interpolation. Note the following examples:
2134 Pattern PCRE matches Perl matches
2136 \Qabc$xyz\E abc$xyz abc followed by the contents of $xyz
2137 \Qabc\$xyz\E abc\$xyz abc\$xyz
2138 \Qabc\E\$\Qxyz\E abc$xyz abc$xyz
2140 For compatibility with Perl, \Q...\E sequences are recognized inside character
2141 classes as well as outside them.
2143 9. Re-organized 3 code statements in pcretest to avoid "overflow in
2144 floating-point constant arithmetic" warnings from a Microsoft compiler. Added a
2145 (size_t) cast to one statement in pcretest and one in pcreposix to avoid
2146 signed/unsigned warnings.
2148 10. SunOS4 doesn't have strtoul(). This was used only for unpicking the -o
2149 option for pcretest, so I've replaced it by a simple function that does just
2150 that job.
2152 11. pcregrep was ending with code 0 instead of 2 for the commands "pcregrep" or
2153 "pcregrep -".
2155 12. Added "possessive quantifiers" ?+, *+, ++, and {,}+ which come from Sun's
2156 Java package. This provides some syntactic sugar for simple cases of what my
2157 documentation calls "once-only subpatterns". A pattern such as x*+ is the same
2158 as (?>x*). In other words, if what is inside (?>...) is just a single repeated
2159 item, you can use this simplified notation. Note that only makes sense with
2160 greedy quantifiers. Consequently, the use of the possessive quantifier forces
2161 greediness, whatever the setting of the PCRE_UNGREEDY option.
2163 13. A change of greediness default within a pattern was not taking effect at
2164 the current level for patterns like /(b+(?U)a+)/. It did apply to parenthesized
2165 subpatterns that followed. Patterns like /b+(?U)a+/ worked because the option
2166 was abstracted outside.
2168 14. PCRE now supports the \G assertion. It is true when the current matching
2169 position is at the start point of the match. This differs from \A when the
2170 starting offset is non-zero. Used with the /g option of pcretest (or similar
2171 code), it works in the same way as it does for Perl's /g option. If all
2172 alternatives of a regex begin with \G, the expression is anchored to the start
2173 match position, and the "anchored" flag is set in the compiled expression.
2175 15. Some bugs concerning the handling of certain option changes within patterns
2176 have been fixed. These applied to options other than (?ims). For example,
2177 "a(?x: b c )d" did not match "XabcdY" but did match "Xa b c dY". It should have
2178 been the other way round. Some of this was related to change 7 above.
2180 16. PCRE now gives errors for /[.x.]/ and /[=x=]/ as unsupported POSIX
2181 features, as Perl does. Previously, PCRE gave the warnings only for /[[.x.]]/
2182 and /[[=x=]]/. PCRE now also gives an error for /[:name:]/ because it supports
2183 POSIX classes only within a class (e.g. /[[:alpha:]]/).
2185 17. Added support for Perl's \C escape. This matches one byte, even in UTF8
2186 mode. Unlike ".", it always matches newline, whatever the setting of
2187 PCRE_DOTALL. However, PCRE does not permit \C to appear in lookbehind
2188 assertions. Perl allows it, but it doesn't (in general) work because it can't
2189 calculate the length of the lookbehind. At least, that's the case for Perl
2190 5.8.0 - I've been told they are going to document that it doesn't work in
2191 future.
2193 18. Added an error diagnosis for escapes that PCRE does not support: these are
2194 \L, \l, \N, \P, \p, \U, \u, and \X.
2196 19. Although correctly diagnosing a missing ']' in a character class, PCRE was
2197 reading past the end of the pattern in cases such as /[abcd/.
2199 20. PCRE was getting more memory than necessary for patterns with classes that
2200 contained both POSIX named classes and other characters, e.g. /[[:space:]abc/.
2202 21. Added some code, conditional on #ifdef VPCOMPAT, to make life easier for
2203 compiling PCRE for use with Virtual Pascal.
2205 22. Small fix to the Makefile to make it work properly if the build is done
2206 outside the source tree.
2208 23. Added a new extension: a condition to go with recursion. If a conditional
2209 subpattern starts with (?(R) the "true" branch is used if recursion has
2210 happened, whereas the "false" branch is used only at the top level.
2212 24. When there was a very long string of literal characters (over 255 bytes
2213 without UTF support, over 250 bytes with UTF support), the computation of how
2214 much memory was required could be incorrect, leading to segfaults or other
2215 strange effects.
2217 25. PCRE was incorrectly assuming anchoring (either to start of subject or to
2218 start of line for a non-DOTALL pattern) when a pattern started with (.*) and
2219 there was a subsequent back reference to those brackets. This meant that, for
2220 example, /(.*)\d+\1/ failed to match "abc123bc". Unfortunately, it isn't
2221 possible to check for precisely this case. All we can do is abandon the
2222 optimization if .* occurs inside capturing brackets when there are any back
2223 references whatsoever. (See below for a better fix that came later.)
2225 26. The handling of the optimization for finding the first character of a
2226 non-anchored pattern, and for finding a character that is required later in the
2227 match were failing in some cases. This didn't break the matching; it just
2228 failed to optimize when it could. The way this is done has been re-implemented.
2230 27. Fixed typo in error message for invalid (?R item (it said "(?p").
2232 28. Added a new feature that provides some of the functionality that Perl
2233 provides with (?{...}). The facility is termed a "callout". The way it is done
2234 in PCRE is for the caller to provide an optional function, by setting
2235 pcre_callout to its entry point. Like pcre_malloc and pcre_free, this is a
2236 global variable. By default it is unset, which disables all calling out. To get
2237 the function called, the regex must include (?C) at appropriate points. This
2238 is, in fact, equivalent to (?C0), and any number <= 255 may be given with (?C).
2239 This provides a means of identifying different callout points. When PCRE
2240 reaches such a point in the regex, if pcre_callout has been set, the external
2241 function is called. It is provided with data in a structure called
2242 pcre_callout_block, which is defined in pcre.h. If the function returns 0,
2243 matching continues; if it returns a non-zero value, the match at the current
2244 point fails. However, backtracking will occur if possible. [This was changed
2245 later and other features added - see item 49 below.]
2247 29. pcretest is upgraded to test the callout functionality. It provides a
2248 callout function that displays information. By default, it shows the start of
2249 the match and the current position in the text. There are some new data escapes
2250 to vary what happens:
2252 \C+ in addition, show current contents of captured substrings
2253 \C- do not supply a callout function
2254 \C!n return 1 when callout number n is reached
2255 \C!n!m return 1 when callout number n is reached for the mth time
2257 30. If pcregrep was called with the -l option and just a single file name, it
2258 output "<stdin>" if a match was found, instead of the file name.
2260 31. Improve the efficiency of the POSIX API to PCRE. If the number of capturing
2261 slots is less than POSIX_MALLOC_THRESHOLD, use a block on the stack to pass to
2262 pcre_exec(). This saves a malloc/free per call. The default value of
2263 POSIX_MALLOC_THRESHOLD is 10; it can be changed by --with-posix-malloc-threshold
2264 when configuring.
2266 32. The default maximum size of a compiled pattern is 64K. There have been a
2267 few cases of people hitting this limit. The code now uses macros to handle the
2268 storing of links as offsets within the compiled pattern. It defaults to 2-byte
2269 links, but this can be changed to 3 or 4 bytes by --with-link-size when
2270 configuring. Tests 2 and 5 work only with 2-byte links because they output
2271 debugging information about compiled patterns.
2273 33. Internal code re-arrangements:
2275 (a) Moved the debugging function for printing out a compiled regex into
2276 its own source file (printint.c) and used #include to pull it into
2277 pcretest.c and, when DEBUG is defined, into pcre.c, instead of having two
2278 separate copies.
2280 (b) Defined the list of op-code names for debugging as a macro in
2281 internal.h so that it is next to the definition of the opcodes.
2283 (c) Defined a table of op-code lengths for simpler skipping along compiled
2284 code. This is again a macro in internal.h so that it is next to the
2285 definition of the opcodes.
2287 34. Added support for recursive calls to individual subpatterns, along the
2288 lines of Robin Houston's patch (but implemented somewhat differently).
2290 35. Further mods to the Makefile to help Win32. Also, added code to pcregrep to
2291 allow it to read and process whole directories in Win32. This code was
2292 contributed by Lionel Fourquaux; it has not been tested by me.
2294 36. Added support for named subpatterns. The Python syntax (?P<name>...) is
2295 used to name a group. Names consist of alphanumerics and underscores, and must
2296 be unique. Back references use the syntax (?P=name) and recursive calls use
2297 (?P>name) which is a PCRE extension to the Python extension. Groups still have
2298 numbers. The function pcre_fullinfo() can be used after compilation to extract
2299 a name/number map. There are three relevant calls:
2301 PCRE_INFO_NAMEENTRYSIZE yields the size of each entry in the map
2302 PCRE_INFO_NAMECOUNT yields the number of entries
2303 PCRE_INFO_NAMETABLE yields a pointer to the map.
2305 The map is a vector of fixed-size entries. The size of each entry depends on
2306 the length of the longest name used. The first two bytes of each entry are the
2307 group number, most significant byte first. There follows the corresponding
2308 name, zero terminated. The names are in alphabetical order.
2310 37. Make the maximum literal string in the compiled code 250 for the non-UTF-8
2311 case instead of 255. Making it the same both with and without UTF-8 support
2312 means that the same test output works with both.
2314 38. There was a case of malloc(0) in the POSIX testing code in pcretest. Avoid
2315 calling malloc() with a zero argument.
2317 39. Change 25 above had to resort to a heavy-handed test for the .* anchoring
2318 optimization. I've improved things by keeping a bitmap of backreferences with
2319 numbers 1-31 so that if .* occurs inside capturing brackets that are not in
2320 fact referenced, the optimization can be applied. It is unlikely that a
2321 relevant occurrence of .* (i.e. one which might indicate anchoring or forcing
2322 the match to follow \n) will appear inside brackets with a number greater than
2323 31, but if it does, any back reference > 31 suppresses the optimization.
2325 40. Added a new compile-time option PCRE_NO_AUTO_CAPTURE. This has the effect
2326 of disabling numbered capturing parentheses. Any opening parenthesis that is
2327 not followed by ? behaves as if it were followed by ?: but named parentheses
2328 can still be used for capturing (and they will acquire numbers in the usual
2329 way).
2331 41. Redesigned the return codes from the match() function into yes/no/error so
2332 that errors can be passed back from deep inside the nested calls. A malloc
2333 failure while inside a recursive subpattern call now causes the
2334 PCRE_ERROR_NOMEMORY return instead of quietly going wrong.
2336 42. It is now possible to set a limit on the number of times the match()
2337 function is called in a call to pcre_exec(). This facility makes it possible to
2338 limit the amount of recursion and backtracking, though not in a directly
2339 obvious way, because the match() function is used in a number of different
2340 circumstances. The count starts from zero for each position in the subject
2341 string (for non-anchored patterns). The default limit is, for compatibility, a
2342 large number, namely 10 000 000. You can change this in two ways:
2344 (a) When configuring PCRE before making, you can use --with-match-limit=n
2345 to set a default value for the compiled library.
2347 (b) For each call to pcre_exec(), you can pass a pcre_extra block in which
2348 a different value is set. See 45 below.
2350 If the limit is exceeded, pcre_exec() returns PCRE_ERROR_MATCHLIMIT.
2352 43. Added a new function pcre_config(int, void *) to enable run-time extraction
2353 of things that can be changed at compile time. The first argument specifies
2354 what is wanted and the second points to where the information is to be placed.
2355 The current list of available information is:
2359 The output is an integer that is set to one if UTF-8 support is available;
2360 otherwise it is set to zero.
2364 The output is an integer that it set to the value of the code that is used for
2365 newline. It is either LF (10) or CR (13).
2369 The output is an integer that contains the number of bytes used for internal
2370 linkage in compiled expressions. The value is 2, 3, or 4. See item 32 above.
2374 The output is an integer that contains the threshold above which the POSIX
2375 interface uses malloc() for output vectors. See item 31 above.
2379 The output is an unsigned integer that contains the default limit of the number
2380 of match() calls in a pcre_exec() execution. See 42 above.
2382 44. pcretest has been upgraded by the addition of the -C option. This causes it
2383 to extract all the available output from the new pcre_config() function, and to
2384 output it. The program then exits immediately.
2386 45. A need has arisen to pass over additional data with calls to pcre_exec() in
2387 order to support additional features. One way would have been to define
2388 pcre_exec2() (for example) with extra arguments, but this would not have been
2389 extensible, and would also have required all calls to the original function to
2390 be mapped to the new one. Instead, I have chosen to extend the mechanism that
2391 is used for passing in "extra" data from pcre_study().
2393 The pcre_extra structure is now exposed and defined in pcre.h. It currently
2394 contains the following fields:
2396 flags a bitmap indicating which of the following fields are set
2397 study_data opaque data from pcre_study()
2398 match_limit a way of specifying a limit on match() calls for a specific
2399 call to pcre_exec()
2400 callout_data data for callouts (see 49 below)
2402 The flag bits are also defined in pcre.h, and are
2408 The pcre_study() function now returns one of these new pcre_extra blocks, with
2409 the actual study data pointed to by the study_data field, and the
2410 PCRE_EXTRA_STUDY_DATA flag set. This can be passed directly to pcre_exec() as
2411 before. That is, this change is entirely upwards-compatible and requires no
2412 change to existing code.
2414 If you want to pass in additional data to pcre_exec(), you can either place it
2415 in a pcre_extra block provided by pcre_study(), or create your own pcre_extra
2416 block.
2418 46. pcretest has been extended to test the PCRE_EXTRA_MATCH_LIMIT feature. If a
2419 data string contains the escape sequence \M, pcretest calls pcre_exec() several
2420 times with different match limits, until it finds the minimum value needed for
2421 pcre_exec() to complete. The value is then output. This can be instructive; for
2422 most simple matches the number is quite small, but for pathological cases it
2423 gets very large very quickly.
2425 47. There's a new option for pcre_fullinfo() called PCRE_INFO_STUDYSIZE. It
2426 returns the size of the data block pointed to by the study_data field in a
2427 pcre_extra block, that is, the value that was passed as the argument to
2428 pcre_malloc() when PCRE was getting memory in which to place the information
2429 created by pcre_study(). The fourth argument should point to a size_t variable.
2430 pcretest has been extended so that this information is shown after a successful
2431 pcre_study() call when information about the compiled regex is being displayed.
2433 48. Cosmetic change to Makefile: there's no need to have / after $(DESTDIR)
2434 because what follows is always an absolute path. (Later: it turns out that this
2435 is more than cosmetic for MinGW, because it doesn't like empty path
2436 components.)
2438 49. Some changes have been made to the callout feature (see 28 above):
2440 (i) A callout function now has three choices for what it returns:
2442 0 => success, carry on matching
2443 > 0 => failure at this point, but backtrack if possible
2444 < 0 => serious error, return this value from pcre_exec()
2446 Negative values should normally be chosen from the set of PCRE_ERROR_xxx
2447 values. In particular, returning PCRE_ERROR_NOMATCH forces a standard
2448 "match failed" error. The error number PCRE_ERROR_CALLOUT is reserved for
2449 use by callout functions. It will never be used by PCRE itself.
2451 (ii) The pcre_extra structure (see 45 above) has a void * field called
2452 callout_data, with corresponding flag bit PCRE_EXTRA_CALLOUT_DATA. The
2453 pcre_callout_block structure has a field of the same name. The contents of
2454 the field passed in the pcre_extra structure are passed to the callout
2455 function in the corresponding field in the callout block. This makes it
2456 easier to use the same callout-containing regex from multiple threads. For
2457 testing, the pcretest program has a new data escape
2459 \C*n pass the number n (may be negative) as callout_data
2461 If the callout function in pcretest receives a non-zero value as
2462 callout_data, it returns that value.
2464 50. Makefile wasn't handling CFLAGS properly when compiling dftables. Also,
2465 there were some redundant $(CFLAGS) in commands that are now specified as
2466 $(LINK), which already includes $(CFLAGS).
2468 51. Extensions to UTF-8 support are listed below. These all apply when (a) PCRE
2469 has been compiled with UTF-8 support *and* pcre_compile() has been compiled
2470 with the PCRE_UTF8 flag. Patterns that are compiled without that flag assume
2471 one-byte characters throughout. Note that case-insensitive matching applies
2472 only to characters whose values are less than 256. PCRE doesn't support the
2473 notion of cases for higher-valued characters.
2475 (i) A character class whose characters are all within 0-255 is handled as
2476 a bit map, and the map is inverted for negative classes. Previously, a
2477 character > 255 always failed to match such a class; however it should
2478 match if the class was a negative one (e.g. [^ab]). This has been fixed.
2480 (ii) A negated character class with a single character < 255 is coded as
2481 "not this character" (OP_NOT). This wasn't working properly when the test
2482 character was multibyte, either singly or repeated.
2484 (iii) Repeats of multibyte characters are now handled correctly in UTF-8
2485 mode, for example: \x{100}{2,3}.
2487 (iv) The character escapes \b, \B, \d, \D, \s, \S, \w, and \W (either
2488 singly or repeated) now correctly test multibyte characters. However,
2489 PCRE doesn't recognize any characters with values greater than 255 as
2490 digits, spaces, or word characters. Such characters always match \D, \S,
2491 and \W, and never match \d, \s, or \w.
2493 (v) Classes may now contain characters and character ranges with values
2494 greater than 255. For example: [ab\x{100}-\x{400}].
2496 (vi) pcregrep now has a --utf-8 option (synonym -u) which makes it call
2497 PCRE in UTF-8 mode.
2499 52. The info request value PCRE_INFO_FIRSTCHAR has been renamed
2500 PCRE_INFO_FIRSTBYTE because it is a byte value. However, the old name is
2501 retained for backwards compatibility. (Note that LASTLITERAL is also a byte
2502 value.)
2504 53. The single man page has become too large. I have therefore split it up into
2505 a number of separate man pages. These also give rise to individual HTML pages;
2506 these are now put in a separate directory, and there is an index.html page that
2507 lists them all. Some hyperlinking between the pages has been installed.
2509 54. Added convenience functions for handling named capturing parentheses.
2511 55. Unknown escapes inside character classes (e.g. [\M]) and escapes that
2512 aren't interpreted therein (e.g. [\C]) are literals in Perl. This is now also
2513 true in PCRE, except when the PCRE_EXTENDED option is set, in which case they
2514 are faulted.
2516 56. Introduced HOST_CC and HOST_CFLAGS which can be set in the environment when
2517 calling configure. These values are used when compiling the dftables.c program
2518 which is run to generate the source of the default character tables. They
2519 default to the values of CC and CFLAGS. If you are cross-compiling PCRE,
2520 you will need to set these values.
2522 57. Updated the building process for Windows DLL, as provided by Fred Cox.
2525 Version 3.9 02-Jan-02
2526 ---------------------
2528 1. A bit of extraneous text had somehow crept into the pcregrep documentation.
2530 2. If --disable-static was given, the building process failed when trying to
2531 build pcretest and pcregrep. (For some reason it was using libtool to compile
2532 them, which is not right, as they aren't part of the library.)
2535 Version 3.8 18-Dec-01
2536 ---------------------
2538 1. The experimental UTF-8 code was completely screwed up. It was packing the
2539 bytes in the wrong order. How dumb can you get?
2542 Version 3.7 29-Oct-01
2543 ---------------------
2545 1. In updating pcretest to check change 1 of version 3.6, I screwed up.
2546 This caused pcretest, when used on the test data, to segfault. Unfortunately,
2547 this didn't happen under Solaris 8, where I normally test things.
2549 2. The Makefile had to be changed to make it work on BSD systems, where 'make'
2550 doesn't seem to recognize that ./xxx and xxx are the same file. (This entry
2551 isn't in ChangeLog distributed with 3.7 because I forgot when I hastily made
2552 this fix an hour or so after the initial 3.7 release.)
2555 Version 3.6 23-Oct-01
2556 ---------------------
2558 1. Crashed with /(sens|respons)e and \1ibility/ and "sense and sensibility" if
2559 offsets passed as NULL with zero offset count.
2561 2. The config.guess and config.sub files had not been updated when I moved to
2562 the latest autoconf.
2565 Version 3.5 15-Aug-01
2566 ---------------------
2568 1. Added some missing #if !defined NOPOSIX conditionals in pcretest.c that
2569 had been forgotten.
2571 2. By using declared but undefined structures, we can avoid using "void"
2572 definitions in pcre.h while keeping the internal definitions of the structures
2573 private.
2575 3. The distribution is now built using autoconf 2.50 and libtool 1.4. From a
2576 user point of view, this means that both static and shared libraries are built
2577 by default, but this can be individually controlled. More of the work of
2578 handling this static/shared cases is now inside libtool instead of PCRE's make
2579 file.
2581 4. The pcretest utility is now installed along with pcregrep because it is
2582 useful for users (to test regexs) and by doing this, it automatically gets
2583 relinked by libtool. The documentation has been turned into a man page, so
2584 there are now .1, .txt, and .html versions in /doc.
2586 5. Upgrades to pcregrep:
2587 (i) Added long-form option names like gnu grep.
2588 (ii) Added --help to list all options with an explanatory phrase.
2589 (iii) Added -r, --recursive to recurse into sub-directories.
2590 (iv) Added -f, --file to read patterns from a file.
2592 6. pcre_exec() was referring to its "code" argument before testing that
2593 argument for NULL (and giving an error if it was NULL).
2595 7. Upgraded Makefile.in to allow for compiling in a different directory from
2596 the source directory.
2598 8. Tiny buglet in pcretest: when pcre_fullinfo() was called to retrieve the
2599 options bits, the pointer it was passed was to an int instead of to an unsigned
2600 long int. This mattered only on 64-bit systems.
2602 9. Fixed typo (3.4/1) in pcre.h again. Sigh. I had changed pcre.h (which is
2603 generated) instead of pcre.in, which it its source. Also made the same change
2604 in several of the .c files.
2606 10. A new release of gcc defines printf() as a macro, which broke pcretest
2607 because it had an ifdef in the middle of a string argument for printf(). Fixed
2608 by using separate calls to printf().
2610 11. Added --enable-newline-is-cr and --enable-newline-is-lf to the configure
2611 script, to force use of CR or LF instead of \n in the source. On non-Unix
2612 systems, the value can be set in config.h.
2614 12. The limit of 200 on non-capturing parentheses is a _nesting_ limit, not an
2615 absolute limit. Changed the text of the error message to make this clear, and
2616 likewise updated the man page.
2618 13. The limit of 99 on the number of capturing subpatterns has been removed.
2619 The new limit is 65535, which I hope will not be a "real" limit.
2622 Version 3.4 22-Aug-00
2623 ---------------------
2625 1. Fixed typo in pcre.h: unsigned const char * changed to const unsigned char *.
2627 2. Diagnose condition (?(0) as an error instead of crashing on matching.
2630 Version 3.3 01-Aug-00
2631 ---------------------
2633 1. If an octal character was given, but the value was greater than \377, it
2634 was not getting masked to the least significant bits, as documented. This could
2635 lead to crashes in some systems.
2637 2. Perl 5.6 (if not earlier versions) accepts classes like [a-\d] and treats
2638 the hyphen as a literal. PCRE used to give an error; it now behaves like Perl.
2640 3. Added the functions pcre_free_substring() and pcre_free_substring_list().
2641 These just pass their arguments on to (pcre_free)(), but they are provided
2642 because some uses of PCRE bind it to non-C systems that can call its functions,
2643 but cannot call free() or pcre_free() directly.
2645 4. Add "make test" as a synonym for "make check". Corrected some comments in
2646 the Makefile.
2648 5. Add $(DESTDIR)/ in front of all the paths in the "install" target in the
2649 Makefile.
2651 6. Changed the name of pgrep to pcregrep, because Solaris has introduced a
2652 command called pgrep for grepping around the active processes.
2654 7. Added the beginnings of support for UTF-8 character strings.
2656 8. Arranged for the Makefile to pass over the settings of CC, CFLAGS, and
2657 RANLIB to ./ltconfig so that they are used by libtool. I think these are all
2658 the relevant ones. (AR is not passed because ./ltconfig does its own figuring
2659 out for the ar command.)
2662 Version 3.2 12-May-00
2663 ---------------------
2665 This is purely a bug fixing release.
2667 1. If the pattern /((Z)+|A)*/ was matched agained ZABCDEFG it matched Z instead
2668 of ZA. This was just one example of several cases that could provoke this bug,
2669 which was introduced by change 9 of version 2.00. The code for breaking
2670 infinite loops after an iteration that matches an empty string was't working
2671 correctly.
2673 2. The pcretest program was not imitating Perl correctly for the pattern /a*/g
2674 when matched against abbab (for example). After matching an empty string, it
2675 wasn't forcing anchoring when setting PCRE_NOTEMPTY for the next attempt; this
2676 caused it to match further down the string than it should.
2678 3. The code contained an inclusion of sys/types.h. It isn't clear why this
2679 was there because it doesn't seem to be needed, and it causes trouble on some
2680 systems, as it is not a Standard C header. It has been removed.
2682 4. Made 4 silly changes to the source to avoid stupid compiler warnings that
2683 were reported on the Macintosh. The changes were from
2685 while ((c = *(++ptr)) != 0 && c != '\n');
2686 to
2687 while ((c = *(++ptr)) != 0 && c != '\n') ;
2689 Totally extraordinary, but if that's what it takes...
2691 5. PCRE is being used in one environment where neither memmove() nor bcopy() is
2692 available. Added HAVE_BCOPY and an autoconf test for it; if neither
2693 HAVE_MEMMOVE nor HAVE_BCOPY is set, use a built-in emulation function which
2694 assumes the way PCRE uses memmove() (always moving upwards).
2696 6. PCRE is being used in one environment where strchr() is not available. There
2697 was only one use in pcre.c, and writing it out to avoid strchr() probably gives
2698 faster code anyway.
2701 Version 3.1 09-Feb-00
2702 ---------------------
2704 The only change in this release is the fixing of some bugs in Makefile.in for
2705 the "install" target:
2707 (1) It was failing to install pcreposix.h.
2709 (2) It was overwriting the pcre.3 man page with the pcreposix.3 man page.
2712 Version 3.0 01-Feb-00
2713 ---------------------
2715 1. Add support for the /+ modifier to perltest (to output $` like it does in
2716 pcretest).
2718 2. Add support for the /g modifier to perltest.
2720 3. Fix pcretest so that it behaves even more like Perl for /g when the pattern
2721 matches null strings.
2723 4. Fix perltest so that it doesn't do unwanted things when fed an empty
2724 pattern. Perl treats empty patterns specially - it reuses the most recent
2725 pattern, which is not what we want. Replace // by /(?#)/ in order to avoid this
2726 effect.
2728 5. The POSIX interface was broken in that it was just handing over the POSIX
2729 captured string vector to pcre_exec(), but (since release 2.00) PCRE has
2730 required a bigger vector, with some working space on the end. This means that
2731 the POSIX wrapper now has to get and free some memory, and copy the results.
2733 6. Added some simple autoconf support, placing the test data and the
2734 documentation in separate directories, re-organizing some of the
2735 information files, and making it build pcre-config (a GNU standard). Also added
2736 libtool support for building PCRE as a shared library, which is now the
2737 default.
2739 7. Got rid of the leading zero in the definition of PCRE_MINOR because 08 and
2740 09 are not valid octal constants. Single digits will be used for minor values
2741 less than 10.
2743 8. Defined REG_EXTENDED and REG_NOSUB as zero in the POSIX header, so that
2744 existing programs that set these in the POSIX interface can use PCRE without
2745 modification.
2747 9. Added a new function, pcre_fullinfo() with an extensible interface. It can
2748 return all that pcre_info() returns, plus additional data. The pcre_info()
2749 function is retained for compatibility, but is considered to be obsolete.
2751 10. Added experimental recursion feature (?R) to handle one common case that
2752 Perl 5.6 will be able to do with (?p{...}).
2754 11. Added support for POSIX character classes like [:alpha:], which Perl is
2755 adopting.
2758 Version 2.08 31-Aug-99
2759 ----------------------
2761 1. When startoffset was not zero and the pattern began with ".*", PCRE was not
2762 trying to match at the startoffset position, but instead was moving forward to
2763 the next newline as if a previous match had failed.
2765 2. pcretest was not making use of PCRE_NOTEMPTY when repeating for /g and /G,
2766 and could get into a loop if a null string was matched other than at the start
2767 of the subject.
2769 3. Added definitions of PCRE_MAJOR and PCRE_MINOR to pcre.h so the version can
2770 be distinguished at compile time, and for completeness also added PCRE_DATE.
2772 5. Added Paul Sokolovsky's minor changes to make it easy to compile a Win32 DLL
2773 in GnuWin32 environments.
2776 Version 2.07 29-Jul-99
2777 ----------------------
2779 1. The documentation is now supplied in plain text form and HTML as well as in
2780 the form of man page sources.
2782 2. C++ compilers don't like assigning (void *) values to other pointer types.
2783 In particular this affects malloc(). Although there is no problem in Standard
2784 C, I've put in casts to keep C++ compilers happy.
2786 3. Typo on pcretest.c; a cast of (unsigned char *) in the POSIX regexec() call
2787 should be (const char *).
2789 4. If NOPOSIX is defined, pcretest.c compiles without POSIX support. This may
2790 be useful for non-Unix systems who don't want to bother with the POSIX stuff.
2791 However, I haven't made this a standard facility. The documentation doesn't
2792 mention it, and the Makefile doesn't support it.
2794 5. The Makefile now contains an "install" target, with editable destinations at
2795 the top of the file. The pcretest program is not installed.
2797 6. pgrep -V now gives the PCRE version number and date.
2799 7. Fixed bug: a zero repetition after a literal string (e.g. /abcde{0}/) was
2800 causing the entire string to be ignored, instead of just the last character.
2802 8. If a pattern like /"([^\\"]+|\\.)*"/ is applied in the normal way to a
2803 non-matching string, it can take a very, very long time, even for strings of
2804 quite modest length, because of the nested recursion. PCRE now does better in
2805 some of these cases. It does this by remembering the last required literal
2806 character in the pattern, and pre-searching the subject to ensure it is present
2807 before running the real match. In other words, it applies a heuristic to detect
2808 some types of certain failure quickly, and in the above example, if presented
2809 with a string that has no trailing " it gives "no match" very quickly.
2811 9. A new runtime option PCRE_NOTEMPTY causes null string matches to be ignored;
2812 other alternatives are tried instead.
2815 Version 2.06 09-Jun-99
2816 ----------------------
2818 1. Change pcretest's output for amount of store used to show just the code
2819 space, because the remainder (the data block) varies in size between 32-bit and
2820 64-bit systems.
2822 2. Added an extra argument to pcre_exec() to supply an offset in the subject to
2823 start matching at. This allows lookbehinds to work when searching for multiple
2824 occurrences in a string.
2826 3. Added additional options to pcretest for testing multiple occurrences:
2828 /+ outputs the rest of the string that follows a match
2829 /g loops for multiple occurrences, using the new startoffset argument
2830 /G loops for multiple occurrences by passing an incremented pointer
2832 4. PCRE wasn't doing the "first character" optimization for patterns starting
2833 with \b or \B, though it was doing it for other lookbehind assertions. That is,
2834 it wasn't noticing that a match for a pattern such as /\bxyz/ has to start with
2835 the letter 'x'. On long subject strings, this gives a significant speed-up.
2838 Version 2.05 21-Apr-99
2839 ----------------------
2841 1. Changed the type of magic_number from int to long int so that it works
2842 properly on 16-bit systems.
2844 2. Fixed a bug which caused patterns starting with .* not to work correctly
2845 when the subject string contained newline characters. PCRE was assuming
2846 anchoring for such patterns in all cases, which is not correct because .* will
2847 not pass a newline unless PCRE_DOTALL is set. It now assumes anchoring only if
2848 DOTALL is set at top level; otherwise it knows that patterns starting with .*
2849 must be retried after every newline in the subject.
2852 Version 2.04 18-Feb-99
2853 ----------------------
2855 1. For parenthesized subpatterns with repeats whose minimum was zero, the
2856 computation of the store needed to hold the pattern was incorrect (too large).
2857 If such patterns were nested a few deep, this could multiply and become a real
2858 problem.
2860 2. Added /M option to pcretest to show the memory requirement of a specific
2861 pattern. Made -m a synonym of -s (which does this globally) for compatibility.
2863 3. Subpatterns of the form (regex){n,m} (i.e. limited maximum) were being
2864 compiled in such a way that the backtracking after subsequent failure was
2865 pessimal. Something like (a){0,3} was compiled as (a)?(a)?(a)? instead of
2866 ((a)((a)(a)?)?)? with disastrous performance if the maximum was of any size.
2869 Version 2.03 02-Feb-99
2870 ----------------------
2872 1. Fixed typo and small mistake in man page.
2874 2. Added 4th condition (GPL supersedes if conflict) and created separate
2875 LICENCE file containing the conditions.
2877 3. Updated pcretest so that patterns such as /abc\/def/ work like they do in
2878 Perl, that is the internal \ allows the delimiter to be included in the
2879 pattern. Locked out the use of \ as a delimiter. If \ immediately follows
2880 the final delimiter, add \ to the end of the pattern (to test the error).
2882 4. Added the convenience functions for extracting substrings after a successful
2883 match. Updated pcretest to make it able to test these functions.
2886 Version 2.02 14-Jan-99
2887 ----------------------
2889 1. Initialized the working variables associated with each extraction so that
2890 their saving and restoring doesn't refer to uninitialized store.
2892 2. Put dummy code into study.c in order to trick the optimizer of the IBM C
2893 compiler for OS/2 into generating correct code. Apparently IBM isn't going to
2894 fix the problem.
2896 3. Pcretest: the timing code wasn't using LOOPREPEAT for timing execution
2897 calls, and wasn't printing the correct value for compiling calls. Increased the
2898 default value of LOOPREPEAT, and the number of significant figures in the
2899 times.
2901 4. Changed "/bin/rm" in the Makefile to "-rm" so it works on Windows NT.
2903 5. Renamed "deftables" as "dftables" to get it down to 8 characters, to avoid
2904 a building problem on Windows NT with a FAT file system.
2907 Version 2.01 21-Oct-98
2908 ----------------------
2910 1. Changed the API for pcre_compile() to allow for the provision of a pointer
2911 to character tables built by pcre_maketables() in the current locale. If NULL
2912 is passed, the default tables are used.
2915 Version 2.00 24-Sep-98
2916 ----------------------
2918 1. Since the (>?) facility is in Perl 5.005, don't require PCRE_EXTRA to enable
2919 it any more.
2921 2. Allow quantification of (?>) groups, and make it work correctly.
2923 3. The first character computation wasn't working for (?>) groups.
2925 4. Correct the implementation of \Z (it is permitted to match on the \n at the
2926 end of the subject) and add 5.005's \z, which really does match only at the
2927 very end of the subject.
2929 5. Remove the \X "cut" facility; Perl doesn't have it, and (?> is neater.
2931 6. Remove the ability to specify CASELESS, MULTILINE, DOTALL, and
2932 DOLLAR_END_ONLY at runtime, to make it possible to implement the Perl 5.005
2933 localized options. All options to pcre_study() were also removed.
2935 7. Add other new features from 5.005:
2937 $(?<= positive lookbehind
2938 $(?<! negative lookbehind
2939 (?imsx-imsx) added the unsetting capability
2940 such a setting is global if at outer level; local otherwise
2941 (?imsx-imsx:) non-capturing groups with option setting
2942 (?(cond)re|re) conditional pattern matching
2944 A backreference to itself in a repeated group matches the previous
2945 captured string.
2947 8. General tidying up of studying (both automatic and via "study")
2948 consequential on the addition of new assertions.
2950 9. As in 5.005, unlimited repeated groups that could match an empty substring
2951 are no longer faulted at compile time. Instead, the loop is forcibly broken at
2952 runtime if any iteration does actually match an empty substring.
2954 10. Include the RunTest script in the distribution.
2956 11. Added tests from the Perl 5.005_02 distribution. This showed up a few
2957 discrepancies, some of which were old and were also with respect to 5.004. They
2958 have now been fixed.
2961 Version 1.09 28-Apr-98
2962 ----------------------
2964 1. A negated single character class followed by a quantifier with a minimum
2965 value of one (e.g. [^x]{1,6} ) was not compiled correctly. This could lead to
2966 program crashes, or just wrong answers. This did not apply to negated classes
2967 containing more than one character, or to minima other than one.
2970 Version 1.08 27-Mar-98
2971 ----------------------
2973 1. Add PCRE_UNGREEDY to invert the greediness of quantifiers.
2975 2. Add (?U) and (?X) to set PCRE_UNGREEDY and PCRE_EXTRA respectively. The
2976 latter must appear before anything that relies on it in the pattern.
2979 Version 1.07 16-Feb-98
2980 ----------------------
2982 1. A pattern such as /((a)*)*/ was not being diagnosed as in error (unlimited
2983 repeat of a potentially empty string).
2986 Version 1.06 23-Jan-98
2987 ----------------------
2989 1. Added Markus Oberhumer's little patches for C++.
2991 2. Literal strings longer than 255 characters were broken.
2994 Version 1.05 23-Dec-97
2995 ----------------------
2997 1. Negated character classes containing more than one character were failing if
2998 PCRE_CASELESS was set at run time.
3001 Version 1.04 19-Dec-97
3002 ----------------------
3004 1. Corrected the man page, where some "const" qualifiers had been omitted.
3006 2. Made debugging output print "{0,xxx}" instead of just "{,xxx}" to agree with
3007 input syntax.
3009 3. Fixed memory leak which occurred when a regex with back references was
3010 matched with an offsets vector that wasn't big enough. The temporary memory
3011 that is used in this case wasn't being freed if the match failed.
3013 4. Tidied pcretest to ensure it frees memory that it gets.
3015 5. Temporary memory was being obtained in the case where the passed offsets
3016 vector was exactly big enough.
3018 6. Corrected definition of offsetof() from change 5 below.
3020 7. I had screwed up change 6 below and broken the rules for the use of
3021 setjmp(). Now fixed.
3024 Version 1.03 18-Dec-97
3025 ----------------------
3027 1. A erroneous regex with a missing opening parenthesis was correctly
3028 diagnosed, but PCRE attempted to access brastack[-1], which could cause crashes
3029 on some systems.
3031 2. Replaced offsetof(real_pcre, code) by offsetof(real_pcre, code[0]) because
3032 it was reported that one broken compiler failed on the former because "code" is
3033 also an independent variable.
3035 3. The erroneous regex a[]b caused an array overrun reference.
3037 4. A regex ending with a one-character negative class (e.g. /[^k]$/) did not
3038 fail on data ending with that character. (It was going on too far, and checking
3039 the next character, typically a binary zero.) This was specific to the
3040 optimized code for single-character negative classes.
3042 5. Added a contributed patch from the TIN world which does the following:
3044 + Add an undef for memmove, in case the the system defines a macro for it.
3046 + Add a definition of offsetof(), in case there isn't one. (I don't know
3047 the reason behind this - offsetof() is part of the ANSI standard - but
3048 it does no harm).
3050 + Reduce the ifdef's in pcre.c using macro DPRINTF, thereby eliminating
3051 most of the places where whitespace preceded '#'. I have given up and
3052 allowed the remaining 2 cases to be at the margin.
3054 + Rename some variables in pcre to eliminate shadowing. This seems very
3055 pedantic, but does no harm, of course.
3057 6. Moved the call to setjmp() into its own function, to get rid of warnings
3058 from gcc -Wall, and avoided calling it at all unless PCRE_EXTRA is used.
3060 7. Constructs such as \d{8,} were compiling into the equivalent of
3061 \d{8}\d{0,65527} instead of \d{8}\d* which didn't make much difference to the
3062 outcome, but in this particular case used more store than had been allocated,
3063 which caused the bug to be discovered because it threw up an internal error.
3065 8. The debugging code in both pcre and pcretest for outputting the compiled
3066 form of a regex was going wrong in the case of back references followed by
3067 curly-bracketed repeats.
3070 Version 1.02 12-Dec-97
3071 ----------------------
3073 1. Typos in pcre.3 and comments in the source fixed.
3075 2. Applied a contributed patch to get rid of places where it used to remove
3076 'const' from variables, and fixed some signed/unsigned and uninitialized
3077 variable warnings.
3079 3. Added the "runtest" target to Makefile.
3081 4. Set default compiler flag to -O2 rather than just -O.
3084 Version 1.01 19-Nov-97
3085 ----------------------
3087 1. PCRE was failing to diagnose unlimited repeat of empty string for patterns
3088 like /([ab]*)*/, that is, for classes with more than one character in them.
3090 2. Likewise, it wasn't diagnosing patterns with "once-only" subpatterns, such
3091 as /((?>a*))*/ (a PCRE_EXTRA facility).
3094 Version 1.00 18-Nov-97
3095 ----------------------
3097 1. Added compile-time macros to support systems such as SunOS4 which don't have
3098 memmove() or strerror() but have other things that can be used instead.
3100 2. Arranged that "make clean" removes the executables.
3103 Version 0.99 27-Oct-97
3104 ----------------------
3106 1. Fixed bug in code for optimizing classes with only one character. It was
3107 initializing a 32-byte map regardless, which could cause it to run off the end
3108 of the memory it had got.
3110 2. Added, conditional on PCRE_EXTRA, the proposed (?>REGEX) construction.
3113 Version 0.98 22-Oct-97
3114 ----------------------
3116 1. Fixed bug in code for handling temporary memory usage when there are more
3117 back references than supplied space in the ovector. This could cause segfaults.
3120 Version 0.97 21-Oct-97
3121 ----------------------
3123 1. Added the \X "cut" facility, conditional on PCRE_EXTRA.
3125 2. Optimized negated single characters not to use a bit map.
3127 3. Brought error texts together as macro definitions; clarified some of them;
3128 fixed one that was wrong - it said "range out of order" when it meant "invalid
3129 escape sequence".
3131 4. Changed some char * arguments to const char *.
3133 5. Added PCRE_NOTBOL and PCRE_NOTEOL (from POSIX).
3135 6. Added the POSIX-style API wrapper in pcreposix.a and testing facilities in
3136 pcretest.
3139 Version 0.96 16-Oct-97
3140 ----------------------
3142 1. Added a simple "pgrep" utility to the distribution.
3144 2. Fixed an incompatibility with Perl: "{" is now treated as a normal character
3145 unless it appears in one of the precise forms "{ddd}", "{ddd,}", or "{ddd,ddd}"
3146 where "ddd" means "one or more decimal digits".
3148 3. Fixed serious bug. If a pattern had a back reference, but the call to
3149 pcre_exec() didn't supply a large enough ovector to record the related
3150 identifying subpattern, the match always failed. PCRE now remembers the number
3151 of the largest back reference, and gets some temporary memory in which to save
3152 the offsets during matching if necessary, in order to ensure that
3153 backreferences always work.
3155 4. Increased the compatibility with Perl in a number of ways:
3157 (a) . no longer matches \n by default; an option PCRE_DOTALL is provided
3158 to request this handling. The option can be set at compile or exec time.
3160 (b) $ matches before a terminating newline by default; an option
3161 PCRE_DOLLAR_ENDONLY is provided to override this (but not in multiline
3162 mode). The option can be set at compile or exec time.
3164 (c) The handling of \ followed by a digit other than 0 is now supposed to be
3165 the same as Perl's. If the decimal number it represents is less than 10
3166 or there aren't that many previous left capturing parentheses, an octal
3167 escape is read. Inside a character class, it's always an octal escape,
3168 even if it is a single digit.
3170 (d) An escaped but undefined alphabetic character is taken as a literal,
3171 unless PCRE_EXTRA is set. Currently this just reserves the remaining
3172 escapes.
3174 (e) {0} is now permitted. (The previous item is removed from the compiled
3175 pattern).
3177 5. Changed all the names of code files so that the basic parts are no longer
3178 than 10 characters, and abolished the teeny "globals.c" file.
3180 6. Changed the handling of character classes; they are now done with a 32-byte
3181 bit map always.
3183 7. Added the -d and /D options to pcretest to make it possible to look at the
3184 internals of compilation without having to recompile pcre.
3187 Version 0.95 23-Sep-97
3188 ----------------------
3190 1. Fixed bug in pre-pass concerning escaped "normal" characters such as \x5c or
3191 \x20 at the start of a run of normal characters. These were being treated as
3192 real characters, instead of the source characters being re-checked.
3195 Version 0.94 18-Sep-97
3196 ----------------------
3198 1. The functions are now thread-safe, with the caveat that the global variables
3199 containing pointers to malloc() and free() or alternative functions are the
3200 same for all threads.
3202 2. Get pcre_study() to generate a bitmap of initial characters for non-
3203 anchored patterns when this is possible, and use it if passed to pcre_exec().
3206 Version 0.93 15-Sep-97
3207 ----------------------
3209 1. /(b)|(:+)/ was computing an incorrect first character.
3211 2. Add pcre_study() to the API and the passing of pcre_extra to pcre_exec(),
3212 but not actually doing anything yet.
3214 3. Treat "-" characters in classes that cannot be part of ranges as literals,
3215 as Perl does (e.g. [-az] or [az-]).
3217 4. Set the anchored flag if a branch starts with .* or .*? because that tests
3218 all possible positions.
3220 5. Split up into different modules to avoid including unneeded functions in a
3221 compiled binary. However, compile and exec are still in one module. The "study"
3222 function is split off.
3224 6. The character tables are now in a separate module whose source is generated
3225 by an auxiliary program - but can then be edited by hand if required. There are
3226 now no calls to isalnum(), isspace(), isdigit(), isxdigit(), tolower() or
3227 toupper() in the code.
3229 7. Turn the malloc/free funtions variables into pcre_malloc and pcre_free and
3230 make them global. Abolish the function for setting them, as the caller can now
3231 set them directly.
3234 Version 0.92 11-Sep-97
3235 ----------------------
3237 1. A repeat with a fixed maximum and a minimum of 1 for an ordinary character
3238 (e.g. /a{1,3}/) was broken (I mis-optimized it).
3240 2. Caseless matching was not working in character classes if the characters in
3241 the pattern were in upper case.
3243 3. Make ranges like [W-c] work in the same way as Perl for caseless matching.
3245 4. Make PCRE_ANCHORED public and accept as a compile option.
3247 5. Add an options word to pcre_exec() and accept PCRE_ANCHORED and
3248 PCRE_CASELESS at run time. Add escapes \A and \I to pcretest to cause it to
3249 pass them.
3251 6. Give an error if bad option bits passed at compile or run time.
3253 7. Add PCRE_MULTILINE at compile and exec time, and (?m) as well. Add \M to
3254 pcretest to cause it to pass that flag.
3256 8. Add pcre_info(), to get the number of identifying subpatterns, the stored
3257 options, and the first character, if set.
3259 9. Recognize C+ or C{n,m} where n >= 1 as providing a fixed starting character.
3262 Version 0.91 10-Sep-97
3263 ----------------------
3265 1. PCRE was failing to diagnose unlimited repeats of subpatterns that could
3266 match the empty string as in /(a*)*/. It was looping and ultimately crashing.
3268 2. PCRE was looping on encountering an indefinitely repeated back reference to
3269 a subpattern that had matched an empty string, e.g. /(a|)\1*/. It now does what
3270 Perl does - treats the match as successful.
3272 ****


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

  ViewVC Help
Powered by ViewVC 1.1.5