ViewVC logotype

Contents of /code/trunk/ChangeLog

Parent Directory Parent Directory | Revision Log Revision Log

Revision 361 - (show annotations)
Thu Jul 10 16:03:28 2008 UTC (7 years, 4 months ago) by ph10
File size: 158751 byte(s)
Error occurred while calculating annotation data.
Fix off-end-of-buffer bug for patterns that match only at start of line.
1 ChangeLog for PCRE
2 ------------------
4 Version 8.0 02 Jul-08
5 ---------------------
7 1. Replaced UCP searching code with optimized version as implemented for Ad
8 Muncher (http://www.admuncher.com/) by Peter Kankowski. This uses a two-
9 stage table and inline lookup instead of a function, giving speed ups of 2
10 to 5 times on some simple patterns that I tested. Permission was given to
11 distribute the MultiStage2.py script that generates the tables (it's not in
12 the tarball, but is in the Subversion repository).
14 2. Updated the Unicode datatables to Unicode 5.1.0. This adds yet more
15 scripts.
17 3. Change 12 for 7.7 introduced a bug in pcre_study() when a pattern contained
18 a group with a zero qualifier. The result of the study could be incorrect,
19 or the function might crash, depending on the pattern.
21 4. Caseless matching was not working for non-ASCII characters in back
22 references. For example, /(\x{de})\1/8i was not matching \x{de}\x{fe}.
23 It now works when Unicode Property Support is available.
25 5. In pcretest, an escape such as \x{de} in the data was always generating
26 a UTF-8 string, even in non-UTF-8 mode. Now it generates a single byte in
27 non-UTF-8 mode. If the value is greater than 255, it gives a warning about
28 truncation.
30 6. Minor bugfix in pcrecpp.cc (change "" == ... to NULL == ...).
32 7. Added two (int) casts to pcregrep when printing the difference of two
33 pointers, in case they are 64-bit values.
35 8. Added comments about Mac OS X stack usage to the pcrestack man page and to
36 test 2 if it fails.
38 9. Added PCRE_CALL_CONVENTION just before the names of all exported functions,
39 and a #define of that name to empty if it is not externally set. This is to
40 allow users of MSVC to set it if necessary.
42 10. The PCRE_EXP_DEFN macro which precedes exported functions was missing from
43 the convenience functions in the pcre_get.c source file.
45 11. An option change at the start of a pattern that had top-level alternatives
46 could cause overwriting and/or a crash. This command provoked a crash in
47 some environments:
49 printf "/(?i)[\xc3\xa9\xc3\xbd]|[\xc3\xa9\xc3\xbdA]/8\n" | pcretest
51 This potential security problem was recorded as CVE-2008-2371.
53 12. For a pattern where the match had to start at the beginning or immediately
54 after a newline (e.g /.*anything/ without the DOTALL flag), pcre_exec() and
55 pcre_dfa_exec() could read past the end of the passed subject if there was
56 no match. To help with detecting such bugs (e.g. with valgrind), I modified
57 pcretest so that it places the subject at the end of its malloc-ed buffer.
60 Version 7.7 07-May-08
61 ---------------------
63 1. Applied Craig's patch to sort out a long long problem: "If we can't convert
64 a string to a long long, pretend we don't even have a long long." This is
65 done by checking for the strtoq, strtoll, and _strtoi64 functions.
67 2. Applied Craig's patch to pcrecpp.cc to restore ABI compatibility with
68 pre-7.6 versions, which defined a global no_arg variable instead of putting
69 it in the RE class. (See also #8 below.)
71 3. Remove a line of dead code, identified by coverity and reported by Nuno
72 Lopes.
74 4. Fixed two related pcregrep bugs involving -r with --include or --exclude:
76 (1) The include/exclude patterns were being applied to the whole pathnames
77 of files, instead of just to the final components.
79 (2) If there was more than one level of directory, the subdirectories were
80 skipped unless they satisfied the include/exclude conditions. This is
81 inconsistent with GNU grep (and could even be seen as contrary to the
82 pcregrep specification - which I improved to make it absolutely clear).
83 The action now is always to scan all levels of directory, and just
84 apply the include/exclude patterns to regular files.
86 5. Added the --include_dir and --exclude_dir patterns to pcregrep, and used
87 --exclude_dir in the tests to avoid scanning .svn directories.
89 6. Applied Craig's patch to the QuoteMeta function so that it escapes the
90 NUL character as backslash + 0 rather than backslash + NUL, because PCRE
91 doesn't support NULs in patterns.
93 7. Added some missing "const"s to declarations of static tables in
94 pcre_compile.c and pcre_dfa_exec.c.
96 8. Applied Craig's patch to pcrecpp.cc to fix a problem in OS X that was
97 caused by fix #2 above. (Subsequently also a second patch to fix the
98 first patch. And a third patch - this was a messy problem.)
100 9. Applied Craig's patch to remove the use of push_back().
102 10. Applied Alan Lehotsky's patch to add REG_STARTEND support to the POSIX
103 matching function regexec().
105 11. Added support for the Oniguruma syntax \g<name>, \g<n>, \g'name', \g'n',
106 which, however, unlike Perl's \g{...}, are subroutine calls, not back
107 references. PCRE supports relative numbers with this syntax (I don't think
108 Oniguruma does).
110 12. Previously, a group with a zero repeat such as (...){0} was completely
111 omitted from the compiled regex. However, this means that if the group
112 was called as a subroutine from elsewhere in the pattern, things went wrong
113 (an internal error was given). Such groups are now left in the compiled
114 pattern, with a new opcode that causes them to be skipped at execution
115 time.
117 13. Added the PCRE_JAVASCRIPT_COMPAT option. This makes the following changes
118 to the way PCRE behaves:
120 (a) A lone ] character is dis-allowed (Perl treats it as data).
122 (b) A back reference to an unmatched subpattern matches an empty string
123 (Perl fails the current match path).
125 (c) A data ] in a character class must be notated as \] because if the
126 first data character in a class is ], it defines an empty class. (In
127 Perl it is not possible to have an empty class.) The empty class []
128 never matches; it forces failure and is equivalent to (*FAIL) or (?!).
129 The negative empty class [^] matches any one character, independently
130 of the DOTALL setting.
132 14. A pattern such as /(?2)[]a()b](abc)/ which had a forward reference to a
133 non-existent subpattern following a character class starting with ']' and
134 containing () gave an internal compiling error instead of "reference to
135 non-existent subpattern". Fortunately, when the pattern did exist, the
136 compiled code was correct. (When scanning forwards to check for the
137 existencd of the subpattern, it was treating the data ']' as terminating
138 the class, so got the count wrong. When actually compiling, the reference
139 was subsequently set up correctly.)
141 15. The "always fail" assertion (?!) is optimzed to (*FAIL) by pcre_compile;
142 it was being rejected as not supported by pcre_dfa_exec(), even though
143 other assertions are supported. I have made pcre_dfa_exec() support
144 (*FAIL).
146 16. The implementation of 13c above involved the invention of a new opcode,
147 OP_ALLANY, which is like OP_ANY but doesn't check the /s flag. Since /s
148 cannot be changed at match time, I realized I could make a small
149 improvement to matching performance by compiling OP_ALLANY instead of
150 OP_ANY for "." when DOTALL was set, and then removing the runtime tests
151 on the OP_ANY path.
153 17. Compiling pcretest on Windows with readline support failed without the
154 following two fixes: (1) Make the unistd.h include conditional on
155 HAVE_UNISTD_H; (2) #define isatty and fileno as _isatty and _fileno.
157 18. Changed CMakeLists.txt and cmake/FindReadline.cmake to arrange for the
158 ncurses library to be included for pcretest when ReadLine support is
159 requested, but also to allow for it to be overridden. This patch came from
160 Daniel Bergström.
162 19. There was a typo in the file ucpinternal.h where f0_rangeflag was defined
163 as 0x00f00000 instead of 0x00800000. Luckily, this would not have caused
164 any errors with the current Unicode tables. Thanks to Peter Kankowski for
165 spotting this.
168 Version 7.6 28-Jan-08
169 ---------------------
171 1. A character class containing a very large number of characters with
172 codepoints greater than 255 (in UTF-8 mode, of course) caused a buffer
173 overflow.
175 2. Patch to cut out the "long long" test in pcrecpp_unittest when
176 HAVE_LONG_LONG is not defined.
178 3. Applied Christian Ehrlicher's patch to update the CMake build files to
179 bring them up to date and include new features. This patch includes:
181 - Fixed PH's badly added libz and libbz2 support.
182 - Fixed a problem with static linking.
183 - Added pcredemo. [But later removed - see 7 below.]
184 - Fixed dftables problem and added an option.
185 - Added a number of HAVE_XXX tests, including HAVE_WINDOWS_H and
187 - Added readline support for pcretest.
188 - Added an listing of the option settings after cmake has run.
190 4. A user submitted a patch to Makefile that makes it easy to create
191 "pcre.dll" under mingw when using Configure/Make. I added stuff to
192 Makefile.am that cause it to include this special target, without
193 affecting anything else. Note that the same mingw target plus all
194 the other distribution libraries and programs are now supported
195 when configuring with CMake (see 6 below) instead of with
196 Configure/Make.
198 5. Applied Craig's patch that moves no_arg into the RE class in the C++ code.
199 This is an attempt to solve the reported problem "pcrecpp::no_arg is not
200 exported in the Windows port". It has not yet been confirmed that the patch
201 solves the problem, but it does no harm.
203 6. Applied Sheri's patch to CMakeLists.txt to add NON_STANDARD_LIB_PREFIX and
204 NON_STANDARD_LIB_SUFFIX for dll names built with mingw when configured
205 with CMake, and also correct the comment about stack recursion.
207 7. Remove the automatic building of pcredemo from the ./configure system and
208 from CMakeLists.txt. The whole idea of pcredemo.c is that it is an example
209 of a program that users should build themselves after PCRE is installed, so
210 building it automatically is not really right. What is more, it gave
211 trouble in some build environments.
213 8. Further tidies to CMakeLists.txt from Sheri and Christian.
216 Version 7.5 10-Jan-08
217 ---------------------
219 1. Applied a patch from Craig: "This patch makes it possible to 'ignore'
220 values in parens when parsing an RE using the C++ wrapper."
222 2. Negative specials like \S did not work in character classes in UTF-8 mode.
223 Characters greater than 255 were excluded from the class instead of being
224 included.
226 3. The same bug as (2) above applied to negated POSIX classes such as
227 [:^space:].
229 4. PCRECPP_STATIC was referenced in pcrecpp_internal.h, but nowhere was it
230 defined or documented. It seems to have been a typo for PCRE_STATIC, so
231 I have changed it.
233 5. The construct (?&) was not diagnosed as a syntax error (it referenced the
234 first named subpattern) and a construct such as (?&a) would reference the
235 first named subpattern whose name started with "a" (in other words, the
236 length check was missing). Both these problems are fixed. "Subpattern name
237 expected" is now given for (?&) (a zero-length name), and this patch also
238 makes it give the same error for \k'' (previously it complained that that
239 was a reference to a non-existent subpattern).
241 6. The erroneous patterns (?+-a) and (?-+a) give different error messages;
242 this is right because (?- can be followed by option settings as well as by
243 digits. I have, however, made the messages clearer.
245 7. Patterns such as (?(1)a|b) (a pattern that contains fewer subpatterns
246 than the number used in the conditional) now cause a compile-time error.
247 This is actually not compatible with Perl, which accepts such patterns, but
248 treats the conditional as always being FALSE (as PCRE used to), but it
249 seems to me that giving a diagnostic is better.
251 8. Change "alphameric" to the more common word "alphanumeric" in comments
252 and messages.
254 9. Fix two occurrences of "backslash" in comments that should have been
255 "backspace".
257 10. Remove two redundant lines of code that can never be obeyed (their function
258 was moved elsewhere).
260 11. The program that makes PCRE's Unicode character property table had a bug
261 which caused it to generate incorrect table entries for sequences of
262 characters that have the same character type, but are in different scripts.
263 It amalgamated them into a single range, with the script of the first of
264 them. In other words, some characters were in the wrong script. There were
265 thirteen such cases, affecting characters in the following ranges:
267 U+002b0 - U+002c1
268 U+0060c - U+0060d
269 U+0061e - U+00612
270 U+0064b - U+0065e
271 U+0074d - U+0076d
272 U+01800 - U+01805
273 U+01d00 - U+01d77
274 U+01d9b - U+01dbf
275 U+0200b - U+0200f
276 U+030fc - U+030fe
277 U+03260 - U+0327f
278 U+0fb46 - U+0fbb1
279 U+10450 - U+1049d
281 12. The -o option (show only the matching part of a line) for pcregrep was not
282 compatible with GNU grep in that, if there was more than one match in a
283 line, it showed only the first of them. It now behaves in the same way as
284 GNU grep.
286 13. If the -o and -v options were combined for pcregrep, it printed a blank
287 line for every non-matching line. GNU grep prints nothing, and pcregrep now
288 does the same. The return code can be used to tell if there were any
289 non-matching lines.
291 14. Added --file-offsets and --line-offsets to pcregrep.
293 15. The pattern (?=something)(?R) was not being diagnosed as a potentially
294 infinitely looping recursion. The bug was that positive lookaheads were not
295 being skipped when checking for a possible empty match (negative lookaheads
296 and both kinds of lookbehind were skipped).
298 16. Fixed two typos in the Windows-only code in pcregrep.c, and moved the
299 inclusion of <windows.h> to before rather than after the definition of
300 INVALID_FILE_ATTRIBUTES (patch from David Byron).
302 17. Specifying a possessive quantifier with a specific limit for a Unicode
303 character property caused pcre_compile() to compile bad code, which led at
304 runtime to PCRE_ERROR_INTERNAL (-14). Examples of patterns that caused this
305 are: /\p{Zl}{2,3}+/8 and /\p{Cc}{2}+/8. It was the possessive "+" that
306 caused the error; without that there was no problem.
308 18. Added --enable-pcregrep-libz and --enable-pcregrep-libbz2.
310 19. Added --enable-pcretest-libreadline.
312 20. In pcrecpp.cc, the variable 'count' was incremented twice in
313 RE::GlobalReplace(). As a result, the number of replacements returned was
314 double what it should be. I removed one of the increments, but Craig sent a
315 later patch that removed the other one (the right fix) and added unit tests
316 that check the return values (which was not done before).
318 21. Several CMake things:
320 (1) Arranged that, when cmake is used on Unix, the libraries end up with
321 the names libpcre and libpcreposix, not just pcre and pcreposix.
323 (2) The above change means that pcretest and pcregrep are now correctly
324 linked with the newly-built libraries, not previously installed ones.
328 22. In UTF-8 mode, with newline set to "any", a pattern such as .*a.*=.b.*
329 crashed when matching a string such as a\x{2029}b (note that \x{2029} is a
330 UTF-8 newline character). The key issue is that the pattern starts .*;
331 this means that the match must be either at the beginning, or after a
332 newline. The bug was in the code for advancing after a failed match and
333 checking that the new position followed a newline. It was not taking
334 account of UTF-8 characters correctly.
336 23. PCRE was behaving differently from Perl in the way it recognized POSIX
337 character classes. PCRE was not treating the sequence [:...:] as a
338 character class unless the ... were all letters. Perl, however, seems to
339 allow any characters between [: and :], though of course it rejects as
340 unknown any "names" that contain non-letters, because all the known class
341 names consist only of letters. Thus, Perl gives an error for [[:1234:]],
342 for example, whereas PCRE did not - it did not recognize a POSIX character
343 class. This seemed a bit dangerous, so the code has been changed to be
344 closer to Perl. The behaviour is not identical to Perl, because PCRE will
345 diagnose an unknown class for, for example, [[:l\ower:]] where Perl will
346 treat it as [[:lower:]]. However, PCRE does now give "unknown" errors where
347 Perl does, and where it didn't before.
349 24. Rewrite so as to remove the single use of %n from pcregrep because in some
350 Windows environments %n is disabled by default.
353 Version 7.4 21-Sep-07
354 ---------------------
356 1. Change 7.3/28 was implemented for classes by looking at the bitmap. This
357 means that a class such as [\s] counted as "explicit reference to CR or
358 LF". That isn't really right - the whole point of the change was to try to
359 help when there was an actual mention of one of the two characters. So now
360 the change happens only if \r or \n (or a literal CR or LF) character is
361 encountered.
363 2. The 32-bit options word was also used for 6 internal flags, but the numbers
364 of both had grown to the point where there were only 3 bits left.
365 Fortunately, there was spare space in the data structure, and so I have
366 moved the internal flags into a new 16-bit field to free up more option
367 bits.
369 3. The appearance of (?J) at the start of a pattern set the DUPNAMES option,
370 but did not set the internal JCHANGED flag - either of these is enough to
371 control the way the "get" function works - but the PCRE_INFO_JCHANGED
372 facility is supposed to tell if (?J) was ever used, so now (?J) at the
373 start sets both bits.
375 4. Added options (at build time, compile time, exec time) to change \R from
376 matching any Unicode line ending sequence to just matching CR, LF, or CRLF.
378 5. doc/pcresyntax.html was missing from the distribution.
380 6. Put back the definition of PCRE_ERROR_NULLWSLIMIT, for backward
381 compatibility, even though it is no longer used.
383 7. Added macro for snprintf to pcrecpp_unittest.cc and also for strtoll and
384 strtoull to pcrecpp.cc to select the available functions in WIN32 when the
385 windows.h file is present (where different names are used). [This was
386 reversed later after testing - see 16 below.]
388 8. Changed all #include <config.h> to #include "config.h". There were also
389 some further <pcre.h> cases that I changed to "pcre.h".
391 9. When pcregrep was used with the --colour option, it missed the line ending
392 sequence off the lines that it output.
394 10. It was pointed out to me that arrays of string pointers cause lots of
395 relocations when a shared library is dynamically loaded. A technique of
396 using a single long string with a table of offsets can drastically reduce
397 these. I have refactored PCRE in four places to do this. The result is
398 dramatic:
400 Originally: 290
401 After changing UCP table: 187
402 After changing error message table: 43
403 After changing table of "verbs" 36
404 After changing table of Posix names 22
406 Thanks to the folks working on Gregex for glib for this insight.
408 11. --disable-stack-for-recursion caused compiling to fail unless -enable-
409 unicode-properties was also set.
411 12. Updated the tests so that they work when \R is defaulted to ANYCRLF.
413 13. Added checks for ANY and ANYCRLF to pcrecpp.cc where it previously
414 checked only for CRLF.
416 14. Added casts to pcretest.c to avoid compiler warnings.
418 15. Added Craig's patch to various pcrecpp modules to avoid compiler warnings.
420 16. Added Craig's patch to remove the WINDOWS_H tests, that were not working,
421 and instead check for _strtoi64 explicitly, and avoid the use of snprintf()
422 entirely. This removes changes made in 7 above.
424 17. The CMake files have been updated, and there is now more information about
425 building with CMake in the NON-UNIX-USE document.
428 Version 7.3 28-Aug-07
429 ---------------------
431 1. In the rejigging of the build system that eventually resulted in 7.1, the
432 line "#include <pcre.h>" was included in pcre_internal.h. The use of angle
433 brackets there is not right, since it causes compilers to look for an
434 installed pcre.h, not the version that is in the source that is being
435 compiled (which of course may be different). I have changed it back to:
437 #include "pcre.h"
439 I have a vague recollection that the change was concerned with compiling in
440 different directories, but in the new build system, that is taken care of
441 by the VPATH setting the Makefile.
443 2. The pattern .*$ when run in not-DOTALL UTF-8 mode with newline=any failed
444 when the subject happened to end in the byte 0x85 (e.g. if the last
445 character was \x{1ec5}). *Character* 0x85 is one of the "any" newline
446 characters but of course it shouldn't be taken as a newline when it is part
447 of another character. The bug was that, for an unlimited repeat of . in
448 not-DOTALL UTF-8 mode, PCRE was advancing by bytes rather than by
449 characters when looking for a newline.
451 3. A small performance improvement in the DOTALL UTF-8 mode .* case.
453 4. Debugging: adjusted the names of opcodes for different kinds of parentheses
454 in debug output.
456 5. Arrange to use "%I64d" instead of "%lld" and "%I64u" instead of "%llu" for
457 long printing in the pcrecpp unittest when running under MinGW.
459 6. ESC_K was left out of the EBCDIC table.
461 7. Change 7.0/38 introduced a new limit on the number of nested non-capturing
462 parentheses; I made it 1000, which seemed large enough. Unfortunately, the
463 limit also applies to "virtual nesting" when a pattern is recursive, and in
464 this case 1000 isn't so big. I have been able to remove this limit at the
465 expense of backing off one optimization in certain circumstances. Normally,
466 when pcre_exec() would call its internal match() function recursively and
467 immediately return the result unconditionally, it uses a "tail recursion"
468 feature to save stack. However, when a subpattern that can match an empty
469 string has an unlimited repetition quantifier, it no longer makes this
470 optimization. That gives it a stack frame in which to save the data for
471 checking that an empty string has been matched. Previously this was taken
472 from the 1000-entry workspace that had been reserved. So now there is no
473 explicit limit, but more stack is used.
475 8. Applied Daniel's patches to solve problems with the import/export magic
476 syntax that is required for Windows, and which was going wrong for the
477 pcreposix and pcrecpp parts of the library. These were overlooked when this
478 problem was solved for the main library.
480 9. There were some crude static tests to avoid integer overflow when computing
481 the size of patterns that contain repeated groups with explicit upper
482 limits. As the maximum quantifier is 65535, the maximum group length was
483 set at 30,000 so that the product of these two numbers did not overflow a
484 32-bit integer. However, it turns out that people want to use groups that
485 are longer than 30,000 bytes (though not repeat them that many times).
486 Change 7.0/17 (the refactoring of the way the pattern size is computed) has
487 made it possible to implement the integer overflow checks in a much more
488 dynamic way, which I have now done. The artificial limitation on group
489 length has been removed - we now have only the limit on the total length of
490 the compiled pattern, which depends on the LINK_SIZE setting.
492 10. Fixed a bug in the documentation for get/copy named substring when
493 duplicate names are permitted. If none of the named substrings are set, the
494 functions return PCRE_ERROR_NOSUBSTRING (7); the doc said they returned an
495 empty string.
497 11. Because Perl interprets \Q...\E at a high level, and ignores orphan \E
498 instances, patterns such as [\Q\E] or [\E] or even [^\E] cause an error,
499 because the ] is interpreted as the first data character and the
500 terminating ] is not found. PCRE has been made compatible with Perl in this
501 regard. Previously, it interpreted [\Q\E] as an empty class, and [\E] could
502 cause memory overwriting.
504 10. Like Perl, PCRE automatically breaks an unlimited repeat after an empty
505 string has been matched (to stop an infinite loop). It was not recognizing
506 a conditional subpattern that could match an empty string if that
507 subpattern was within another subpattern. For example, it looped when
508 trying to match (((?(1)X|))*) but it was OK with ((?(1)X|)*) where the
509 condition was not nested. This bug has been fixed.
511 12. A pattern like \X?\d or \P{L}?\d in non-UTF-8 mode could cause a backtrack
512 past the start of the subject in the presence of bytes with the top bit
513 set, for example "\x8aBCD".
515 13. Added Perl 5.10 experimental backtracking controls (*FAIL), (*F), (*PRUNE),
516 (*SKIP), (*THEN), (*COMMIT), and (*ACCEPT).
518 14. Optimized (?!) to (*FAIL).
520 15. Updated the test for a valid UTF-8 string to conform to the later RFC 3629.
521 This restricts code points to be within the range 0 to 0x10FFFF, excluding
522 the "low surrogate" sequence 0xD800 to 0xDFFF. Previously, PCRE allowed the
523 full range 0 to 0x7FFFFFFF, as defined by RFC 2279. Internally, it still
524 does: it's just the validity check that is more restrictive.
526 16. Inserted checks for integer overflows during escape sequence (backslash)
527 processing, and also fixed erroneous offset values for syntax errors during
528 backslash processing.
530 17. Fixed another case of looking too far back in non-UTF-8 mode (cf 12 above)
531 for patterns like [\PPP\x8a]{1,}\x80 with the subject "A\x80".
533 18. An unterminated class in a pattern like (?1)\c[ with a "forward reference"
534 caused an overrun.
536 19. A pattern like (?:[\PPa*]*){8,} which had an "extended class" (one with
537 something other than just ASCII characters) inside a group that had an
538 unlimited repeat caused a loop at compile time (while checking to see
539 whether the group could match an empty string).
541 20. Debugging a pattern containing \p or \P could cause a crash. For example,
542 [\P{Any}] did so. (Error in the code for printing property names.)
544 21. An orphan \E inside a character class could cause a crash.
546 22. A repeated capturing bracket such as (A)? could cause a wild memory
547 reference during compilation.
549 23. There are several functions in pcre_compile() that scan along a compiled
550 expression for various reasons (e.g. to see if it's fixed length for look
551 behind). There were bugs in these functions when a repeated \p or \P was
552 present in the pattern. These operators have additional parameters compared
553 with \d, etc, and these were not being taken into account when moving along
554 the compiled data. Specifically:
556 (a) A item such as \p{Yi}{3} in a lookbehind was not treated as fixed
557 length.
559 (b) An item such as \pL+ within a repeated group could cause crashes or
560 loops.
562 (c) A pattern such as \p{Yi}+(\P{Yi}+)(?1) could give an incorrect
563 "reference to non-existent subpattern" error.
565 (d) A pattern like (\P{Yi}{2}\277)? could loop at compile time.
567 24. A repeated \S or \W in UTF-8 mode could give wrong answers when multibyte
568 characters were involved (for example /\S{2}/8g with "A\x{a3}BC").
570 25. Using pcregrep in multiline, inverted mode (-Mv) caused it to loop.
572 26. Patterns such as [\P{Yi}A] which include \p or \P and just one other
573 character were causing crashes (broken optimization).
575 27. Patterns such as (\P{Yi}*\277)* (group with possible zero repeat containing
576 \p or \P) caused a compile-time loop.
578 28. More problems have arisen in unanchored patterns when CRLF is a valid line
579 break. For example, the unstudied pattern [\r\n]A does not match the string
580 "\r\nA" because change 7.0/46 below moves the current point on by two
581 characters after failing to match at the start. However, the pattern \nA
582 *does* match, because it doesn't start till \n, and if [\r\n]A is studied,
583 the same is true. There doesn't seem any very clean way out of this, but
584 what I have chosen to do makes the common cases work: PCRE now takes note
585 of whether there can be an explicit match for \r or \n anywhere in the
586 pattern, and if so, 7.0/46 no longer applies. As part of this change,
587 there's a new PCRE_INFO_HASCRORLF option for finding out whether a compiled
588 pattern has explicit CR or LF references.
590 29. Added (*CR) etc for changing newline setting at start of pattern.
593 Version 7.2 19-Jun-07
594 ---------------------
596 1. If the fr_FR locale cannot be found for test 3, try the "french" locale,
597 which is apparently normally available under Windows.
599 2. Re-jig the pcregrep tests with different newline settings in an attempt
600 to make them independent of the local environment's newline setting.
602 3. Add code to configure.ac to remove -g from the CFLAGS default settings.
604 4. Some of the "internals" tests were previously cut out when the link size
605 was not 2, because the output contained actual offsets. The recent new
606 "Z" feature of pcretest means that these can be cut out, making the tests
607 usable with all link sizes.
609 5. Implemented Stan Switzer's goto replacement for longjmp() when not using
610 stack recursion. This gives a massive performance boost under BSD, but just
611 a small improvement under Linux. However, it saves one field in the frame
612 in all cases.
614 6. Added more features from the forthcoming Perl 5.10:
616 (a) (?-n) (where n is a string of digits) is a relative subroutine or
617 recursion call. It refers to the nth most recently opened parentheses.
619 (b) (?+n) is also a relative subroutine call; it refers to the nth next
620 to be opened parentheses.
622 (c) Conditions that refer to capturing parentheses can be specified
623 relatively, for example, (?(-2)... or (?(+3)...
625 (d) \K resets the start of the current match so that everything before
626 is not part of it.
628 (e) \k{name} is synonymous with \k<name> and \k'name' (.NET compatible).
630 (f) \g{name} is another synonym - part of Perl 5.10's unification of
631 reference syntax.
633 (g) (?| introduces a group in which the numbering of parentheses in each
634 alternative starts with the same number.
636 (h) \h, \H, \v, and \V match horizontal and vertical whitespace.
638 7. Added two new calls to pcre_fullinfo(): PCRE_INFO_OKPARTIAL and
641 8. A pattern such as (.*(.)?)* caused pcre_exec() to fail by either not
642 terminating or by crashing. Diagnosed by Viktor Griph; it was in the code
643 for detecting groups that can match an empty string.
645 9. A pattern with a very large number of alternatives (more than several
646 hundred) was running out of internal workspace during the pre-compile
647 phase, where pcre_compile() figures out how much memory will be needed. A
648 bit of new cunning has reduced the workspace needed for groups with
649 alternatives. The 1000-alternative test pattern now uses 12 bytes of
650 workspace instead of running out of the 4096 that are available.
652 10. Inserted some missing (unsigned int) casts to get rid of compiler warnings.
654 11. Applied patch from Google to remove an optimization that didn't quite work.
655 The report of the bug said:
657 pcrecpp::RE("a*").FullMatch("aaa") matches, while
658 pcrecpp::RE("a*?").FullMatch("aaa") does not, and
659 pcrecpp::RE("a*?\\z").FullMatch("aaa") does again.
661 12. If \p or \P was used in non-UTF-8 mode on a character greater than 127
662 it matched the wrong number of bytes.
665 Version 7.1 24-Apr-07
666 ---------------------
668 1. Applied Bob Rossi and Daniel G's patches to convert the build system to one
669 that is more "standard", making use of automake and other Autotools. There
670 is some re-arrangement of the files and adjustment of comments consequent
671 on this.
673 2. Part of the patch fixed a problem with the pcregrep tests. The test of -r
674 for recursive directory scanning broke on some systems because the files
675 are not scanned in any specific order and on different systems the order
676 was different. A call to "sort" has been inserted into RunGrepTest for the
677 approprate test as a short-term fix. In the longer term there may be an
678 alternative.
680 3. I had an email from Eric Raymond about problems translating some of PCRE's
681 man pages to HTML (despite the fact that I distribute HTML pages, some
682 people do their own conversions for various reasons). The problems
683 concerned the use of low-level troff macros .br and .in. I have therefore
684 removed all such uses from the man pages (some were redundant, some could
685 be replaced by .nf/.fi pairs). The 132html script that I use to generate
686 HTML has been updated to handle .nf/.fi and to complain if it encounters
687 .br or .in.
689 4. Updated comments in configure.ac that get placed in config.h.in and also
690 arranged for config.h to be included in the distribution, with the name
691 config.h.generic, for the benefit of those who have to compile without
692 Autotools (compare pcre.h, which is now distributed as pcre.h.generic).
694 5. Updated the support (such as it is) for Virtual Pascal, thanks to Stefan
695 Weber: (1) pcre_internal.h was missing some function renames; (2) updated
696 makevp.bat for the current PCRE, using the additional files
697 makevp_c.txt, makevp_l.txt, and pcregexp.pas.
699 6. A Windows user reported a minor discrepancy with test 2, which turned out
700 to be caused by a trailing space on an input line that had got lost in his
701 copy. The trailing space was an accident, so I've just removed it.
703 7. Add -Wl,-R... flags in pcre-config.in for *BSD* systems, as I'm told
704 that is needed.
706 8. Mark ucp_table (in ucptable.h) and ucp_gentype (in pcre_ucp_searchfuncs.c)
707 as "const" (a) because they are and (b) because it helps the PHP
708 maintainers who have recently made a script to detect big data structures
709 in the php code that should be moved to the .rodata section. I remembered
710 to update Builducptable as well, so it won't revert if ucptable.h is ever
711 re-created.
713 9. Added some extra #ifdef SUPPORT_UTF8 conditionals into pcretest.c,
714 pcre_printint.src, pcre_compile.c, pcre_study.c, and pcre_tables.c, in
715 order to be able to cut out the UTF-8 tables in the latter when UTF-8
716 support is not required. This saves 1.5-2K of code, which is important in
717 some applications.
719 Later: more #ifdefs are needed in pcre_ord2utf8.c and pcre_valid_utf8.c
720 so as not to refer to the tables, even though these functions will never be
721 called when UTF-8 support is disabled. Otherwise there are problems with a
722 shared library.
724 10. Fixed two bugs in the emulated memmove() function in pcre_internal.h:
726 (a) It was defining its arguments as char * instead of void *.
728 (b) It was assuming that all moves were upwards in memory; this was true
729 a long time ago when I wrote it, but is no longer the case.
731 The emulated memove() is provided for those environments that have neither
732 memmove() nor bcopy(). I didn't think anyone used it these days, but that
733 is clearly not the case, as these two bugs were recently reported.
735 11. The script PrepareRelease is now distributed: it calls 132html, CleanTxt,
736 and Detrail to create the HTML documentation, the .txt form of the man
737 pages, and it removes trailing spaces from listed files. It also creates
738 pcre.h.generic and config.h.generic from pcre.h and config.h. In the latter
739 case, it wraps all the #defines with #ifndefs. This script should be run
740 before "make dist".
742 12. Fixed two fairly obscure bugs concerned with quantified caseless matching
743 with Unicode property support.
745 (a) For a maximizing quantifier, if the two different cases of the
746 character were of different lengths in their UTF-8 codings (there are
747 some cases like this - I found 11), and the matching function had to
748 back up over a mixture of the two cases, it incorrectly assumed they
749 were both the same length.
751 (b) When PCRE was configured to use the heap rather than the stack for
752 recursion during matching, it was not correctly preserving the data for
753 the other case of a UTF-8 character when checking ahead for a match
754 while processing a minimizing repeat. If the check also involved
755 matching a wide character, but failed, corruption could cause an
756 erroneous result when trying to check for a repeat of the original
757 character.
759 13. Some tidying changes to the testing mechanism:
761 (a) The RunTest script now detects the internal link size and whether there
762 is UTF-8 and UCP support by running ./pcretest -C instead of relying on
763 values substituted by "configure". (The RunGrepTest script already did
764 this for UTF-8.) The configure.ac script no longer substitutes the
765 relevant variables.
767 (b) The debugging options /B and /D in pcretest show the compiled bytecode
768 with length and offset values. This means that the output is different
769 for different internal link sizes. Test 2 is skipped for link sizes
770 other than 2 because of this, bypassing the problem. Unfortunately,
771 there was also a test in test 3 (the locale tests) that used /B and
772 failed for link sizes other than 2. Rather than cut the whole test out,
773 I have added a new /Z option to pcretest that replaces the length and
774 offset values with spaces. This is now used to make test 3 independent
775 of link size. (Test 2 will be tidied up later.)
777 14. If erroroffset was passed as NULL to pcre_compile, it provoked a
778 segmentation fault instead of returning the appropriate error message.
780 15. In multiline mode when the newline sequence was set to "any", the pattern
781 ^$ would give a match between the \r and \n of a subject such as "A\r\nB".
782 This doesn't seem right; it now treats the CRLF combination as the line
783 ending, and so does not match in that case. It's only a pattern such as ^$
784 that would hit this one: something like ^ABC$ would have failed after \r
785 and then tried again after \r\n.
787 16. Changed the comparison command for RunGrepTest from "diff -u" to "diff -ub"
788 in an attempt to make files that differ only in their line terminators
789 compare equal. This works on Linux.
791 17. Under certain error circumstances pcregrep might try to free random memory
792 as it exited. This is now fixed, thanks to valgrind.
794 19. In pcretest, if the pattern /(?m)^$/g<any> was matched against the string
795 "abc\r\n\r\n", it found an unwanted second match after the second \r. This
796 was because its rules for how to advance for /g after matching an empty
797 string at the end of a line did not allow for this case. They now check for
798 it specially.
800 20. pcretest is supposed to handle patterns and data of any length, by
801 extending its buffers when necessary. It was getting this wrong when the
802 buffer for a data line had to be extended.
804 21. Added PCRE_NEWLINE_ANYCRLF which is like ANY, but matches only CR, LF, or
805 CRLF as a newline sequence.
807 22. Code for handling Unicode properties in pcre_dfa_exec() wasn't being cut
808 out by #ifdef SUPPORT_UCP. This did no harm, as it could never be used, but
809 I have nevertheless tidied it up.
811 23. Added some casts to kill warnings from HP-UX ia64 compiler.
813 24. Added a man page for pcre-config.
816 Version 7.0 19-Dec-06
817 ---------------------
819 1. Fixed a signed/unsigned compiler warning in pcre_compile.c, shown up by
820 moving to gcc 4.1.1.
822 2. The -S option for pcretest uses setrlimit(); I had omitted to #include
823 sys/time.h, which is documented as needed for this function. It doesn't
824 seem to matter on Linux, but it showed up on some releases of OS X.
826 3. It seems that there are systems where bytes whose values are greater than
827 127 match isprint() in the "C" locale. The "C" locale should be the
828 default when a C program starts up. In most systems, only ASCII printing
829 characters match isprint(). This difference caused the output from pcretest
830 to vary, making some of the tests fail. I have changed pcretest so that:
832 (a) When it is outputting text in the compiled version of a pattern, bytes
833 other than 32-126 are always shown as hex escapes.
835 (b) When it is outputting text that is a matched part of a subject string,
836 it does the same, unless a different locale has been set for the match
837 (using the /L modifier). In this case, it uses isprint() to decide.
839 4. Fixed a major bug that caused incorrect computation of the amount of memory
840 required for a compiled pattern when options that changed within the
841 pattern affected the logic of the preliminary scan that determines the
842 length. The relevant options are -x, and -i in UTF-8 mode. The result was
843 that the computed length was too small. The symptoms of this bug were
844 either the PCRE error "internal error: code overflow" from pcre_compile(),
845 or a glibc crash with a message such as "pcretest: free(): invalid next
846 size (fast)". Examples of patterns that provoked this bug (shown in
847 pcretest format) are:
849 /(?-x: )/x
850 /(?x)(?-x: \s*#\s*)/
851 /((?i)[\x{c0}])/8
852 /(?i:[\x{c0}])/8
854 HOWEVER: Change 17 below makes this fix obsolete as the memory computation
855 is now done differently.
857 5. Applied patches from Google to: (a) add a QuoteMeta function to the C++
858 wrapper classes; (b) implement a new function in the C++ scanner that is
859 more efficient than the old way of doing things because it avoids levels of
860 recursion in the regex matching; (c) add a paragraph to the documentation
861 for the FullMatch() function.
863 6. The escape sequence \n was being treated as whatever was defined as
864 "newline". Not only was this contrary to the documentation, which states
865 that \n is character 10 (hex 0A), but it also went horribly wrong when
866 "newline" was defined as CRLF. This has been fixed.
868 7. In pcre_dfa_exec.c the value of an unsigned integer (the variable called c)
869 was being set to -1 for the "end of line" case (supposedly a value that no
870 character can have). Though this value is never used (the check for end of
871 line is "zero bytes in current character"), it caused compiler complaints.
872 I've changed it to 0xffffffff.
874 8. In pcre_version.c, the version string was being built by a sequence of
875 C macros that, in the event of PCRE_PRERELEASE being defined as an empty
876 string (as it is for production releases) called a macro with an empty
877 argument. The C standard says the result of this is undefined. The gcc
878 compiler treats it as an empty string (which was what was wanted) but it is
879 reported that Visual C gives an error. The source has been hacked around to
880 avoid this problem.
882 9. On the advice of a Windows user, included <io.h> and <fcntl.h> in Windows
883 builds of pcretest, and changed the call to _setmode() to use _O_BINARY
884 instead of 0x8000. Made all the #ifdefs test both _WIN32 and WIN32 (not all
885 of them did).
887 10. Originally, pcretest opened its input and output without "b"; then I was
888 told that "b" was needed in some environments, so it was added for release
889 5.0 to both the input and output. (It makes no difference on Unix-like
890 systems.) Later I was told that it is wrong for the input on Windows. I've
891 now abstracted the modes into two macros, to make it easier to fiddle with
892 them, and removed "b" from the input mode under Windows.
894 11. Added pkgconfig support for the C++ wrapper library, libpcrecpp.
896 12. Added -help and --help to pcretest as an official way of being reminded
897 of the options.
899 13. Removed some redundant semicolons after macro calls in pcrecpparg.h.in
900 and pcrecpp.cc because they annoy compilers at high warning levels.
902 14. A bit of tidying/refactoring in pcre_exec.c in the main bumpalong loop.
904 15. Fixed an occurrence of == in configure.ac that should have been = (shell
905 scripts are not C programs :-) and which was not noticed because it works
906 on Linux.
908 16. pcretest is supposed to handle any length of pattern and data line (as one
909 line or as a continued sequence of lines) by extending its input buffer if
910 necessary. This feature was broken for very long pattern lines, leading to
911 a string of junk being passed to pcre_compile() if the pattern was longer
912 than about 50K.
914 17. I have done a major re-factoring of the way pcre_compile() computes the
915 amount of memory needed for a compiled pattern. Previously, there was code
916 that made a preliminary scan of the pattern in order to do this. That was
917 OK when PCRE was new, but as the facilities have expanded, it has become
918 harder and harder to keep it in step with the real compile phase, and there
919 have been a number of bugs (see for example, 4 above). I have now found a
920 cunning way of running the real compile function in a "fake" mode that
921 enables it to compute how much memory it would need, while actually only
922 ever using a few hundred bytes of working memory and without too many
923 tests of the mode. This should make future maintenance and development
924 easier. A side effect of this work is that the limit of 200 on the nesting
925 depth of parentheses has been removed (though this was never a serious
926 limitation, I suspect). However, there is a downside: pcre_compile() now
927 runs more slowly than before (30% or more, depending on the pattern). I
928 hope this isn't a big issue. There is no effect on runtime performance.
930 18. Fixed a minor bug in pcretest: if a pattern line was not terminated by a
931 newline (only possible for the last line of a file) and it was a
932 pattern that set a locale (followed by /Lsomething), pcretest crashed.
934 19. Added additional timing features to pcretest. (1) The -tm option now times
935 matching only, not compiling. (2) Both -t and -tm can be followed, as a
936 separate command line item, by a number that specifies the number of
937 repeats to use when timing. The default is 50000; this gives better
938 precision, but takes uncomfortably long for very large patterns.
940 20. Extended pcre_study() to be more clever in cases where a branch of a
941 subpattern has no definite first character. For example, (a*|b*)[cd] would
942 previously give no result from pcre_study(). Now it recognizes that the
943 first character must be a, b, c, or d.
945 21. There was an incorrect error "recursive call could loop indefinitely" if
946 a subpattern (or the entire pattern) that was being tested for matching an
947 empty string contained only one non-empty item after a nested subpattern.
948 For example, the pattern (?>\x{100}*)\d(?R) provoked this error
949 incorrectly, because the \d was being skipped in the check.
951 22. The pcretest program now has a new pattern option /B and a command line
952 option -b, which is equivalent to adding /B to every pattern. This causes
953 it to show the compiled bytecode, without the additional information that
954 -d shows. The effect of -d is now the same as -b with -i (and similarly, /D
955 is the same as /B/I).
957 23. A new optimization is now able automatically to treat some sequences such
958 as a*b as a*+b. More specifically, if something simple (such as a character
959 or a simple class like \d) has an unlimited quantifier, and is followed by
960 something that cannot possibly match the quantified thing, the quantifier
961 is automatically "possessified".
963 24. A recursive reference to a subpattern whose number was greater than 39
964 went wrong under certain circumstances in UTF-8 mode. This bug could also
965 have affected the operation of pcre_study().
967 25. Realized that a little bit of performance could be had by replacing
968 (c & 0xc0) == 0xc0 with c >= 0xc0 when processing UTF-8 characters.
970 26. Timing data from pcretest is now shown to 4 decimal places instead of 3.
972 27. Possessive quantifiers such as a++ were previously implemented by turning
973 them into atomic groups such as ($>a+). Now they have their own opcodes,
974 which improves performance. This includes the automatically created ones
975 from 23 above.
977 28. A pattern such as (?=(\w+))\1: which simulates an atomic group using a
978 lookahead was broken if it was not anchored. PCRE was mistakenly expecting
979 the first matched character to be a colon. This applied both to named and
980 numbered groups.
982 29. The ucpinternal.h header file was missing its idempotency #ifdef.
984 30. I was sent a "project" file called libpcre.a.dev which I understand makes
985 building PCRE on Windows easier, so I have included it in the distribution.
987 31. There is now a check in pcretest against a ridiculously large number being
988 returned by pcre_exec() or pcre_dfa_exec(). If this happens in a /g or /G
989 loop, the loop is abandoned.
991 32. Forward references to subpatterns in conditions such as (?(2)...) where
992 subpattern 2 is defined later cause pcre_compile() to search forwards in
993 the pattern for the relevant set of parentheses. This search went wrong
994 when there were unescaped parentheses in a character class, parentheses
995 escaped with \Q...\E, or parentheses in a #-comment in /x mode.
997 33. "Subroutine" calls and backreferences were previously restricted to
998 referencing subpatterns earlier in the regex. This restriction has now
999 been removed.
1001 34. Added a number of extra features that are going to be in Perl 5.10. On the
1002 whole, these are just syntactic alternatives for features that PCRE had
1003 previously implemented using the Python syntax or my own invention. The
1004 other formats are all retained for compatibility.
1006 (a) Named groups can now be defined as (?<name>...) or (?'name'...) as well
1007 as (?P<name>...). The new forms, as well as being in Perl 5.10, are
1008 also .NET compatible.
1010 (b) A recursion or subroutine call to a named group can now be defined as
1011 (?&name) as well as (?P>name).
1013 (c) A backreference to a named group can now be defined as \k<name> or
1014 \k'name' as well as (?P=name). The new forms, as well as being in Perl
1015 5.10, are also .NET compatible.
1017 (d) A conditional reference to a named group can now use the syntax
1018 (?(<name>) or (?('name') as well as (?(name).
1020 (e) A "conditional group" of the form (?(DEFINE)...) can be used to define
1021 groups (named and numbered) that are never evaluated inline, but can be
1022 called as "subroutines" from elsewhere. In effect, the DEFINE condition
1023 is always false. There may be only one alternative in such a group.
1025 (f) A test for recursion can be given as (?(R1).. or (?(R&name)... as well
1026 as the simple (?(R). The condition is true only if the most recent
1027 recursion is that of the given number or name. It does not search out
1028 through the entire recursion stack.
1030 (g) The escape \gN or \g{N} has been added, where N is a positive or
1031 negative number, specifying an absolute or relative reference.
1033 35. Tidied to get rid of some further signed/unsigned compiler warnings and
1034 some "unreachable code" warnings.
1036 36. Updated the Unicode property tables to Unicode version 5.0.0. Amongst other
1037 things, this adds five new scripts.
1039 37. Perl ignores orphaned \E escapes completely. PCRE now does the same.
1040 There were also incompatibilities regarding the handling of \Q..\E inside
1041 character classes, for example with patterns like [\Qa\E-\Qz\E] where the
1042 hyphen was adjacent to \Q or \E. I hope I've cleared all this up now.
1044 38. Like Perl, PCRE detects when an indefinitely repeated parenthesized group
1045 matches an empty string, and forcibly breaks the loop. There were bugs in
1046 this code in non-simple cases. For a pattern such as ^(a()*)* matched
1047 against aaaa the result was just "a" rather than "aaaa", for example. Two
1048 separate and independent bugs (that affected different cases) have been
1049 fixed.
1051 39. Refactored the code to abolish the use of different opcodes for small
1052 capturing bracket numbers. This is a tidy that I avoided doing when I
1053 removed the limit on the number of capturing brackets for 3.5 back in 2001.
1054 The new approach is not only tidier, it makes it possible to reduce the
1055 memory needed to fix the previous bug (38).
1057 40. Implemented PCRE_NEWLINE_ANY to recognize any of the Unicode newline
1058 sequences (http://unicode.org/unicode/reports/tr18/) as "newline" when
1059 processing dot, circumflex, or dollar metacharacters, or #-comments in /x
1060 mode.
1062 41. Add \R to match any Unicode newline sequence, as suggested in the Unicode
1063 report.
1065 42. Applied patch, originally from Ari Pollak, modified by Google, to allow
1066 copy construction and assignment in the C++ wrapper.
1068 43. Updated pcregrep to support "--newline=any". In the process, I fixed a
1069 couple of bugs that could have given wrong results in the "--newline=crlf"
1070 case.
1072 44. Added a number of casts and did some reorganization of signed/unsigned int
1073 variables following suggestions from Dair Grant. Also renamed the variable
1074 "this" as "item" because it is a C++ keyword.
1076 45. Arranged for dftables to add
1078 #include "pcre_internal.h"
1080 to pcre_chartables.c because without it, gcc 4.x may remove the array
1081 definition from the final binary if PCRE is built into a static library and
1082 dead code stripping is activated.
1084 46. For an unanchored pattern, if a match attempt fails at the start of a
1085 newline sequence, and the newline setting is CRLF or ANY, and the next two
1086 characters are CRLF, advance by two characters instead of one.
1089 Version 6.7 04-Jul-06
1090 ---------------------
1092 1. In order to handle tests when input lines are enormously long, pcretest has
1093 been re-factored so that it automatically extends its buffers when
1094 necessary. The code is crude, but this _is_ just a test program. The
1095 default size has been increased from 32K to 50K.
1097 2. The code in pcre_study() was using the value of the re argument before
1098 testing it for NULL. (Of course, in any sensible call of the function, it
1099 won't be NULL.)
1101 3. The memmove() emulation function in pcre_internal.h, which is used on
1102 systems that lack both memmove() and bcopy() - that is, hardly ever -
1103 was missing a "static" storage class specifier.
1105 4. When UTF-8 mode was not set, PCRE looped when compiling certain patterns
1106 containing an extended class (one that cannot be represented by a bitmap
1107 because it contains high-valued characters or Unicode property items, e.g.
1108 [\pZ]). Almost always one would set UTF-8 mode when processing such a
1109 pattern, but PCRE should not loop if you do not (it no longer does).
1110 [Detail: two cases were found: (a) a repeated subpattern containing an
1111 extended class; (b) a recursive reference to a subpattern that followed a
1112 previous extended class. It wasn't skipping over the extended class
1113 correctly when UTF-8 mode was not set.]
1115 5. A negated single-character class was not being recognized as fixed-length
1116 in lookbehind assertions such as (?<=[^f]), leading to an incorrect
1117 compile error "lookbehind assertion is not fixed length".
1119 6. The RunPerlTest auxiliary script was showing an unexpected difference
1120 between PCRE and Perl for UTF-8 tests. It turns out that it is hard to
1121 write a Perl script that can interpret lines of an input file either as
1122 byte characters or as UTF-8, which is what "perltest" was being required to
1123 do for the non-UTF-8 and UTF-8 tests, respectively. Essentially what you
1124 can't do is switch easily at run time between having the "use utf8;" pragma
1125 or not. In the end, I fudged it by using the RunPerlTest script to insert
1126 "use utf8;" explicitly for the UTF-8 tests.
1128 7. In multiline (/m) mode, PCRE was matching ^ after a terminating newline at
1129 the end of the subject string, contrary to the documentation and to what
1130 Perl does. This was true of both matching functions. Now it matches only at
1131 the start of the subject and immediately after *internal* newlines.
1133 8. A call of pcre_fullinfo() from pcretest to get the option bits was passing
1134 a pointer to an int instead of a pointer to an unsigned long int. This
1135 caused problems on 64-bit systems.
1137 9. Applied a patch from the folks at Google to pcrecpp.cc, to fix "another
1138 instance of the 'standard' template library not being so standard".
1140 10. There was no check on the number of named subpatterns nor the maximum
1141 length of a subpattern name. The product of these values is used to compute
1142 the size of the memory block for a compiled pattern. By supplying a very
1143 long subpattern name and a large number of named subpatterns, the size
1144 computation could be caused to overflow. This is now prevented by limiting
1145 the length of names to 32 characters, and the number of named subpatterns
1146 to 10,000.
1148 11. Subpatterns that are repeated with specific counts have to be replicated in
1149 the compiled pattern. The size of memory for this was computed from the
1150 length of the subpattern and the repeat count. The latter is limited to
1151 65535, but there was no limit on the former, meaning that integer overflow
1152 could in principle occur. The compiled length of a repeated subpattern is
1153 now limited to 30,000 bytes in order to prevent this.
1155 12. Added the optional facility to have named substrings with the same name.
1157 13. Added the ability to use a named substring as a condition, using the
1158 Python syntax: (?(name)yes|no). This overloads (?(R)... and names that
1159 are numbers (not recommended). Forward references are permitted.
1161 14. Added forward references in named backreferences (if you see what I mean).
1163 15. In UTF-8 mode, with the PCRE_DOTALL option set, a quantified dot in the
1164 pattern could run off the end of the subject. For example, the pattern
1165 "(?s)(.{1,5})"8 did this with the subject "ab".
1167 16. If PCRE_DOTALL or PCRE_MULTILINE were set, pcre_dfa_exec() behaved as if
1168 PCRE_CASELESS was set when matching characters that were quantified with ?
1169 or *.
1171 17. A character class other than a single negated character that had a minimum
1172 but no maximum quantifier - for example [ab]{6,} - was not handled
1173 correctly by pce_dfa_exec(). It would match only one character.
1175 18. A valid (though odd) pattern that looked like a POSIX character
1176 class but used an invalid character after [ (for example [[,abc,]]) caused
1177 pcre_compile() to give the error "Failed: internal error: code overflow" or
1178 in some cases to crash with a glibc free() error. This could even happen if
1179 the pattern terminated after [[ but there just happened to be a sequence of
1180 letters, a binary zero, and a closing ] in the memory that followed.
1182 19. Perl's treatment of octal escapes in the range \400 to \777 has changed
1183 over the years. Originally (before any Unicode support), just the bottom 8
1184 bits were taken. Thus, for example, \500 really meant \100. Nowadays the
1185 output from "man perlunicode" includes this:
1187 The regular expression compiler produces polymorphic opcodes. That
1188 is, the pattern adapts to the data and automatically switches to
1189 the Unicode character scheme when presented with Unicode data--or
1190 instead uses a traditional byte scheme when presented with byte
1191 data.
1193 Sadly, a wide octal escape does not cause a switch, and in a string with
1194 no other multibyte characters, these octal escapes are treated as before.
1195 Thus, in Perl, the pattern /\500/ actually matches \100 but the pattern
1196 /\500|\x{1ff}/ matches \500 or \777 because the whole thing is treated as a
1197 Unicode string.
1199 I have not perpetrated such confusion in PCRE. Up till now, it took just
1200 the bottom 8 bits, as in old Perl. I have now made octal escapes with
1201 values greater than \377 illegal in non-UTF-8 mode. In UTF-8 mode they
1202 translate to the appropriate multibyte character.
1204 29. Applied some refactoring to reduce the number of warnings from Microsoft
1205 and Borland compilers. This has included removing the fudge introduced
1206 seven years ago for the OS/2 compiler (see 2.02/2 below) because it caused
1207 a warning about an unused variable.
1209 21. PCRE has not included VT (character 0x0b) in the set of whitespace
1210 characters since release 4.0, because Perl (from release 5.004) does not.
1211 [Or at least, is documented not to: some releases seem to be in conflict
1212 with the documentation.] However, when a pattern was studied with
1213 pcre_study() and all its branches started with \s, PCRE still included VT
1214 as a possible starting character. Of course, this did no harm; it just
1215 caused an unnecessary match attempt.
1217 22. Removed a now-redundant internal flag bit that recorded the fact that case
1218 dependency changed within the pattern. This was once needed for "required
1219 byte" processing, but is no longer used. This recovers a now-scarce options
1220 bit. Also moved the least significant internal flag bit to the most-
1221 significant bit of the word, which was not previously used (hangover from
1222 the days when it was an int rather than a uint) to free up another bit for
1223 the future.
1225 23. Added support for CRLF line endings as well as CR and LF. As well as the
1226 default being selectable at build time, it can now be changed at runtime
1227 via the PCRE_NEWLINE_xxx flags. There are now options for pcregrep to
1228 specify that it is scanning data with non-default line endings.
1230 24. Changed the definition of CXXLINK to make it agree with the definition of
1231 LINK in the Makefile, by replacing LDFLAGS to CXXFLAGS.
1233 25. Applied Ian Taylor's patches to avoid using another stack frame for tail
1234 recursions. This makes a big different to stack usage for some patterns.
1236 26. If a subpattern containing a named recursion or subroutine reference such
1237 as (?P>B) was quantified, for example (xxx(?P>B)){3}, the calculation of
1238 the space required for the compiled pattern went wrong and gave too small a
1239 value. Depending on the environment, this could lead to "Failed: internal
1240 error: code overflow at offset 49" or "glibc detected double free or
1241 corruption" errors.
1243 27. Applied patches from Google (a) to support the new newline modes and (b) to
1244 advance over multibyte UTF-8 characters in GlobalReplace.
1246 28. Change free() to pcre_free() in pcredemo.c. Apparently this makes a
1247 difference for some implementation of PCRE in some Windows version.
1249 29. Added some extra testing facilities to pcretest:
1251 \q<number> in a data line sets the "match limit" value
1252 \Q<number> in a data line sets the "match recursion limt" value
1253 -S <number> sets the stack size, where <number> is in megabytes
1255 The -S option isn't available for Windows.
1258 Version 6.6 06-Feb-06
1259 ---------------------
1261 1. Change 16(a) for 6.5 broke things, because PCRE_DATA_SCOPE was not defined
1262 in pcreposix.h. I have copied the definition from pcre.h.
1264 2. Change 25 for 6.5 broke compilation in a build directory out-of-tree
1265 because pcre.h is no longer a built file.
1267 3. Added Jeff Friedl's additional debugging patches to pcregrep. These are
1268 not normally included in the compiled code.
1271 Version 6.5 01-Feb-06
1272 ---------------------
1274 1. When using the partial match feature with pcre_dfa_exec(), it was not
1275 anchoring the second and subsequent partial matches at the new starting
1276 point. This could lead to incorrect results. For example, with the pattern
1277 /1234/, partially matching against "123" and then "a4" gave a match.
1279 2. Changes to pcregrep:
1281 (a) All non-match returns from pcre_exec() were being treated as failures
1282 to match the line. Now, unless the error is PCRE_ERROR_NOMATCH, an
1283 error message is output. Some extra information is given for the
1285 probably the only errors that are likely to be caused by users (by
1286 specifying a regex that has nested indefinite repeats, for instance).
1287 If there are more than 20 of these errors, pcregrep is abandoned.
1289 (b) A binary zero was treated as data while matching, but terminated the
1290 output line if it was written out. This has been fixed: binary zeroes
1291 are now no different to any other data bytes.
1293 (c) Whichever of the LC_ALL or LC_CTYPE environment variables is set is
1294 used to set a locale for matching. The --locale=xxxx long option has
1295 been added (no short equivalent) to specify a locale explicitly on the
1296 pcregrep command, overriding the environment variables.
1298 (d) When -B was used with -n, some line numbers in the output were one less
1299 than they should have been.
1301 (e) Added the -o (--only-matching) option.
1303 (f) If -A or -C was used with -c (count only), some lines of context were
1304 accidentally printed for the final match.
1306 (g) Added the -H (--with-filename) option.
1308 (h) The combination of options -rh failed to suppress file names for files
1309 that were found from directory arguments.
1311 (i) Added the -D (--devices) and -d (--directories) options.
1313 (j) Added the -F (--fixed-strings) option.
1315 (k) Allow "-" to be used as a file name for -f as well as for a data file.
1317 (l) Added the --colo(u)r option.
1319 (m) Added Jeffrey Friedl's -S testing option, but within #ifdefs so that it
1320 is not present by default.
1322 3. A nasty bug was discovered in the handling of recursive patterns, that is,
1323 items such as (?R) or (?1), when the recursion could match a number of
1324 alternatives. If it matched one of the alternatives, but subsequently,
1325 outside the recursion, there was a failure, the code tried to back up into
1326 the recursion. However, because of the way PCRE is implemented, this is not
1327 possible, and the result was an incorrect result from the match.
1329 In order to prevent this happening, the specification of recursion has
1330 been changed so that all such subpatterns are automatically treated as
1331 atomic groups. Thus, for example, (?R) is treated as if it were (?>(?R)).
1333 4. I had overlooked the fact that, in some locales, there are characters for
1334 which isalpha() is true but neither isupper() nor islower() are true. In
1335 the fr_FR locale, for instance, the \xAA and \xBA characters (ordmasculine
1336 and ordfeminine) are like this. This affected the treatment of \w and \W
1337 when they appeared in character classes, but not when they appeared outside
1338 a character class. The bit map for "word" characters is now created
1339 separately from the results of isalnum() instead of just taking it from the
1340 upper, lower, and digit maps. (Plus the underscore character, of course.)
1342 5. The above bug also affected the handling of POSIX character classes such as
1343 [[:alpha:]] and [[:alnum:]]. These do not have their own bit maps in PCRE's
1344 permanent tables. Instead, the bit maps for such a class were previously
1345 created as the appropriate unions of the upper, lower, and digit bitmaps.
1346 Now they are created by subtraction from the [[:word:]] class, which has
1347 its own bitmap.
1349 6. The [[:blank:]] character class matches horizontal, but not vertical space.
1350 It is created by subtracting the vertical space characters (\x09, \x0a,
1351 \x0b, \x0c) from the [[:space:]] bitmap. Previously, however, the
1352 subtraction was done in the overall bitmap for a character class, meaning
1353 that a class such as [\x0c[:blank:]] was incorrect because \x0c would not
1354 be recognized. This bug has been fixed.
1356 7. Patches from the folks at Google:
1358 (a) pcrecpp.cc: "to handle a corner case that may or may not happen in
1359 real life, but is still worth protecting against".
1361 (b) pcrecpp.cc: "corrects a bug when negative radixes are used with
1362 regular expressions".
1364 (c) pcre_scanner.cc: avoid use of std::count() because not all systems
1365 have it.
1367 (d) Split off pcrecpparg.h from pcrecpp.h and had the former built by
1368 "configure" and the latter not, in order to fix a problem somebody had
1369 with compiling the Arg class on HP-UX.
1371 (e) Improve the error-handling of the C++ wrapper a little bit.
1373 (f) New tests for checking recursion limiting.
1375 8. The pcre_memmove() function, which is used only if the environment does not
1376 have a standard memmove() function (and is therefore rarely compiled),
1377 contained two bugs: (a) use of int instead of size_t, and (b) it was not
1378 returning a result (though PCRE never actually uses the result).
1380 9. In the POSIX regexec() interface, if nmatch is specified as a ridiculously
1381 large number - greater than INT_MAX/(3*sizeof(int)) - REG_ESPACE is
1382 returned instead of calling malloc() with an overflowing number that would
1383 most likely cause subsequent chaos.
1385 10. The debugging option of pcretest was not showing the NO_AUTO_CAPTURE flag.
1387 11. The POSIX flag REG_NOSUB is now supported. When a pattern that was compiled
1388 with this option is matched, the nmatch and pmatch options of regexec() are
1389 ignored.
1391 12. Added REG_UTF8 to the POSIX interface. This is not defined by POSIX, but is
1392 provided in case anyone wants to the the POSIX interface with UTF-8
1393 strings.
1395 13. Added CXXLDFLAGS to the Makefile parameters to provide settings only on the
1396 C++ linking (needed for some HP-UX environments).
1398 14. Avoid compiler warnings in get_ucpname() when compiled without UCP support
1399 (unused parameter) and in the pcre_printint() function (omitted "default"
1400 switch label when the default is to do nothing).
1402 15. Added some code to make it possible, when PCRE is compiled as a C++
1403 library, to replace subject pointers for pcre_exec() with a smart pointer
1404 class, thus making it possible to process discontinuous strings.
1406 16. The two macros PCRE_EXPORT and PCRE_DATA_SCOPE are confusing, and perform
1407 much the same function. They were added by different people who were trying
1408 to make PCRE easy to compile on non-Unix systems. It has been suggested
1409 that PCRE_EXPORT be abolished now that there is more automatic apparatus
1410 for compiling on Windows systems. I have therefore replaced it with
1411 PCRE_DATA_SCOPE. This is set automatically for Windows; if not set it
1412 defaults to "extern" for C or "extern C" for C++, which works fine on
1413 Unix-like systems. It is now possible to override the value of PCRE_DATA_
1414 SCOPE with something explicit in config.h. In addition:
1416 (a) pcreposix.h still had just "extern" instead of either of these macros;
1417 I have replaced it with PCRE_DATA_SCOPE.
1419 (b) Functions such as _pcre_xclass(), which are internal to the library,
1420 but external in the C sense, all had PCRE_EXPORT in their definitions.
1421 This is apparently wrong for the Windows case, so I have removed it.
1422 (It makes no difference on Unix-like systems.)
1424 17. Added a new limit, MATCH_LIMIT_RECURSION, which limits the depth of nesting
1425 of recursive calls to match(). This is different to MATCH_LIMIT because
1426 that limits the total number of calls to match(), not all of which increase
1427 the depth of recursion. Limiting the recursion depth limits the amount of
1428 stack (or heap if NO_RECURSE is set) that is used. The default can be set
1429 when PCRE is compiled, and changed at run time. A patch from Google adds
1430 this functionality to the C++ interface.
1432 18. Changes to the handling of Unicode character properties:
1434 (a) Updated the table to Unicode 4.1.0.
1436 (b) Recognize characters that are not in the table as "Cn" (undefined).
1438 (c) I revised the way the table is implemented to a much improved format
1439 which includes recognition of ranges. It now supports the ranges that
1440 are defined in UnicodeData.txt, and it also amalgamates other
1441 characters into ranges. This has reduced the number of entries in the
1442 table from around 16,000 to around 3,000, thus reducing its size
1443 considerably. I realized I did not need to use a tree structure after
1444 all - a binary chop search is just as efficient. Having reduced the
1445 number of entries, I extended their size from 6 bytes to 8 bytes to
1446 allow for more data.
1448 (d) Added support for Unicode script names via properties such as \p{Han}.
1450 19. In UTF-8 mode, a backslash followed by a non-Ascii character was not
1451 matching that character.
1453 20. When matching a repeated Unicode property with a minimum greater than zero,
1454 (for example \pL{2,}), PCRE could look past the end of the subject if it
1455 reached it while seeking the minimum number of characters. This could
1456 happen only if some of the characters were more than one byte long, because
1457 there is a check for at least the minimum number of bytes.
1459 21. Refactored the implementation of \p and \P so as to be more general, to
1460 allow for more different types of property in future. This has changed the
1461 compiled form incompatibly. Anybody with saved compiled patterns that use
1462 \p or \P will have to recompile them.
1464 22. Added "Any" and "L&" to the supported property types.
1466 23. Recognize \x{...} as a code point specifier, even when not in UTF-8 mode,
1467 but give a compile time error if the value is greater than 0xff.
1469 24. The man pages for pcrepartial, pcreprecompile, and pcre_compile2 were
1470 accidentally not being installed or uninstalled.
1472 25. The pcre.h file was built from pcre.h.in, but the only changes that were
1473 made were to insert the current release number. This seemed silly, because
1474 it made things harder for people building PCRE on systems that don't run
1475 "configure". I have turned pcre.h into a distributed file, no longer built
1476 by "configure", with the version identification directly included. There is
1477 no longer a pcre.h.in file.
1479 However, this change necessitated a change to the pcre-config script as
1480 well. It is built from pcre-config.in, and one of the substitutions was the
1481 release number. I have updated configure.ac so that ./configure now finds
1482 the release number by grepping pcre.h.
1484 26. Added the ability to run the tests under valgrind.
1487 Version 6.4 05-Sep-05
1488 ---------------------
1490 1. Change 6.0/10/(l) to pcregrep introduced a bug that caused separator lines
1491 "--" to be printed when multiple files were scanned, even when none of the
1492 -A, -B, or -C options were used. This is not compatible with Gnu grep, so I
1493 consider it to be a bug, and have restored the previous behaviour.
1495 2. A couple of code tidies to get rid of compiler warnings.
1497 3. The pcretest program used to cheat by referring to symbols in the library
1498 whose names begin with _pcre_. These are internal symbols that are not
1499 really supposed to be visible externally, and in some environments it is
1500 possible to suppress them. The cheating is now confined to including
1501 certain files from the library's source, which is a bit cleaner.
1503 4. Renamed pcre.in as pcre.h.in to go with pcrecpp.h.in; it also makes the
1504 file's purpose clearer.
1506 5. Reorganized pcre_ucp_findchar().
1509 Version 6.3 15-Aug-05
1510 ---------------------
1512 1. The file libpcre.pc.in did not have general read permission in the tarball.
1514 2. There were some problems when building without C++ support:
1516 (a) If C++ support was not built, "make install" and "make test" still
1517 tried to test it.
1519 (b) There were problems when the value of CXX was explicitly set. Some
1520 changes have been made to try to fix these, and ...
1522 (c) --disable-cpp can now be used to explicitly disable C++ support.
1524 (d) The use of @CPP_OBJ@ directly caused a blank line preceded by a
1525 backslash in a target when C++ was disabled. This confuses some
1526 versions of "make", apparently. Using an intermediate variable solves
1527 this. (Same for CPP_LOBJ.)
1529 3. $(LINK_FOR_BUILD) now includes $(CFLAGS_FOR_BUILD) and $(LINK)
1530 (non-Windows) now includes $(CFLAGS) because these flags are sometimes
1531 necessary on certain architectures.
1533 4. Added a setting of -export-symbols-regex to the link command to remove
1534 those symbols that are exported in the C sense, but actually are local
1535 within the library, and not documented. Their names all begin with
1536 "_pcre_". This is not a perfect job, because (a) we have to except some
1537 symbols that pcretest ("illegally") uses, and (b) the facility isn't always
1538 available (and never for static libraries). I have made a note to try to
1539 find a way round (a) in the future.
1542 Version 6.2 01-Aug-05
1543 ---------------------
1545 1. There was no test for integer overflow of quantifier values. A construction
1546 such as {1111111111111111} would give undefined results. What is worse, if
1547 a minimum quantifier for a parenthesized subpattern overflowed and became
1548 negative, the calculation of the memory size went wrong. This could have
1549 led to memory overwriting.
1551 2. Building PCRE using VPATH was broken. Hopefully it is now fixed.
1553 3. Added "b" to the 2nd argument of fopen() in dftables.c, for non-Unix-like
1554 operating environments where this matters.
1556 4. Applied Giuseppe Maxia's patch to add additional features for controlling
1557 PCRE options from within the C++ wrapper.
1559 5. Named capturing subpatterns were not being correctly counted when a pattern
1560 was compiled. This caused two problems: (a) If there were more than 100
1561 such subpatterns, the calculation of the memory needed for the whole
1562 compiled pattern went wrong, leading to an overflow error. (b) Numerical
1563 back references of the form \12, where the number was greater than 9, were
1564 not recognized as back references, even though there were sufficient
1565 previous subpatterns.
1567 6. Two minor patches to pcrecpp.cc in order to allow it to compile on older
1568 versions of gcc, e.g. 2.95.4.
1571 Version 6.1 21-Jun-05
1572 ---------------------
1574 1. There was one reference to the variable "posix" in pcretest.c that was not
1575 surrounded by "#if !defined NOPOSIX".
1577 2. Make it possible to compile pcretest without DFA support, UTF8 support, or
1578 the cross-check on the old pcre_info() function, for the benefit of the
1579 cut-down version of PCRE that is currently imported into Exim.
1581 3. A (silly) pattern starting with (?i)(?-i) caused an internal space
1582 allocation error. I've done the easy fix, which wastes 2 bytes for sensible
1583 patterns that start (?i) but I don't think that matters. The use of (?i) is
1584 just an example; this all applies to the other options as well.
1586 4. Since libtool seems to echo the compile commands it is issuing, the output
1587 from "make" can be reduced a bit by putting "@" in front of each libtool
1588 compile command.
1590 5. Patch from the folks at Google for configure.in to be a bit more thorough
1591 in checking for a suitable C++ installation before trying to compile the
1592 C++ stuff. This should fix a reported problem when a compiler was present,
1593 but no suitable headers.
1595 6. The man pages all had just "PCRE" as their title. I have changed them to
1596 be the relevant file name. I have also arranged that these names are
1597 retained in the file doc/pcre.txt, which is a concatenation in text format
1598 of all the man pages except the little individual ones for each function.
1600 7. The NON-UNIX-USE file had not been updated for the different set of source
1601 files that come with release 6. I also added a few comments about the C++
1602 wrapper.
1605 Version 6.0 07-Jun-05
1606 ---------------------
1608 1. Some minor internal re-organization to help with my DFA experiments.
1610 2. Some missing #ifdef SUPPORT_UCP conditionals in pcretest and printint that
1611 didn't matter for the library itself when fully configured, but did matter
1612 when compiling without UCP support, or within Exim, where the ucp files are
1613 not imported.
1615 3. Refactoring of the library code to split up the various functions into
1616 different source modules. The addition of the new DFA matching code (see
1617 below) to a single monolithic source would have made it really too
1618 unwieldy, quite apart from causing all the code to be include in a
1619 statically linked application, when only some functions are used. This is
1620 relevant even without the DFA addition now that patterns can be compiled in
1621 one application and matched in another.
1623 The downside of splitting up is that there have to be some external
1624 functions and data tables that are used internally in different modules of
1625 the library but which are not part of the API. These have all had their
1626 names changed to start with "_pcre_" so that they are unlikely to clash
1627 with other external names.
1629 4. Added an alternate matching function, pcre_dfa_exec(), which matches using
1630 a different (DFA) algorithm. Although it is slower than the original
1631 function, it does have some advantages for certain types of matching
1632 problem.
1634 5. Upgrades to pcretest in order to test the features of pcre_dfa_exec(),
1635 including restarting after a partial match.
1637 6. A patch for pcregrep that defines INVALID_FILE_ATTRIBUTES if it is not
1638 defined when compiling for Windows was sent to me. I have put it into the
1639 code, though I have no means of testing or verifying it.
1641 7. Added the pcre_refcount() auxiliary function.
1643 8. Added the PCRE_FIRSTLINE option. This constrains an unanchored pattern to
1644 match before or at the first newline in the subject string. In pcretest,
1645 the /f option on a pattern can be used to set this.
1647 9. A repeated \w when used in UTF-8 mode with characters greater than 256
1648 would behave wrongly. This has been present in PCRE since release 4.0.
1650 10. A number of changes to the pcregrep command:
1652 (a) Refactored how -x works; insert ^(...)$ instead of setting
1653 PCRE_ANCHORED and checking the length, in preparation for adding
1654 something similar for -w.
1656 (b) Added the -w (match as a word) option.
1658 (c) Refactored the way lines are read and buffered so as to have more
1659 than one at a time available.
1661 (d) Implemented a pcregrep test script.
1663 (e) Added the -M (multiline match) option. This allows patterns to match
1664 over several lines of the subject. The buffering ensures that at least
1665 8K, or the rest of the document (whichever is the shorter) is available
1666 for matching (and similarly the previous 8K for lookbehind assertions).
1668 (f) Changed the --help output so that it now says
1670 -w, --word-regex(p)
1672 instead of two lines, one with "regex" and the other with "regexp"
1673 because that confused at least one person since the short forms are the
1674 same. (This required a bit of code, as the output is generated
1675 automatically from a table. It wasn't just a text change.)
1677 (g) -- can be used to terminate pcregrep options if the next thing isn't an
1678 option but starts with a hyphen. Could be a pattern or a path name
1679 starting with a hyphen, for instance.
1681 (h) "-" can be given as a file name to represent stdin.
1683 (i) When file names are being printed, "(standard input)" is used for
1684 the standard input, for compatibility with GNU grep. Previously
1685 "<stdin>" was used.
1687 (j) The option --label=xxx can be used to supply a name to be used for
1688 stdin when file names are being printed. There is no short form.
1690 (k) Re-factored the options decoding logic because we are going to add
1691 two more options that take data. Such options can now be given in four
1692 different ways, e.g. "-fname", "-f name", "--file=name", "--file name".
1694 (l) Added the -A, -B, and -C options for requesting that lines of context
1695 around matches be printed.
1697 (m) Added the -L option to print the names of files that do not contain
1698 any matching lines, that is, the complement of -l.
1700 (n) The return code is 2 if any file cannot be opened, but pcregrep does
1701 continue to scan other files.
1703 (o) The -s option was incorrectly implemented. For compatibility with other
1704 greps, it now suppresses the error message for a non-existent or non-
1705 accessible file (but not the return code). There is a new option called
1706 -q that suppresses the output of matching lines, which was what -s was
1707 previously doing.
1709 (p) Added --include and --exclude options to specify files for inclusion
1710 and exclusion when recursing.
1712 11. The Makefile was not using the Autoconf-supported LDFLAGS macro properly.
1713 Hopefully, it now does.
1715 12. Missing cast in pcre_study().
1717 13. Added an "uninstall" target to the makefile.
1719 14. Replaced "extern" in the function prototypes in Makefile.in with
1720 "PCRE_DATA_SCOPE", which defaults to 'extern' or 'extern "C"' in the Unix
1721 world, but is set differently for Windows.
1723 15. Added a second compiling function called pcre_compile2(). The only
1724 difference is that it has an extra argument, which is a pointer to an
1725 integer error code. When there is a compile-time failure, this is set
1726 non-zero, in addition to the error test pointer being set to point to an
1727 error message. The new argument may be NULL if no error number is required
1728 (but then you may as well call pcre_compile(), which is now just a
1729 wrapper). This facility is provided because some applications need a
1730 numeric error indication, but it has also enabled me to tidy up the way
1731 compile-time errors are handled in the POSIX wrapper.
1733 16. Added VPATH=.libs to the makefile; this should help when building with one
1734 prefix path and installing with another. (Or so I'm told by someone who
1735 knows more about this stuff than I do.)
1737 17. Added a new option, REG_DOTALL, to the POSIX function regcomp(). This
1738 passes PCRE_DOTALL to the pcre_compile() function, making the "." character
1739 match everything, including newlines. This is not POSIX-compatible, but
1740 somebody wanted the feature. From pcretest it can be activated by using
1741 both the P and the s flags.
1743 18. AC_PROG_LIBTOOL appeared twice in Makefile.in. Removed one.
1745 19. libpcre.pc was being incorrectly installed as executable.
1747 20. A couple of places in pcretest check for end-of-line by looking for '\n';
1748 it now also looks for '\r' so that it will work unmodified on Windows.
1750 21. Added Google's contributed C++ wrapper to the distribution.
1752 22. Added some untidy missing memory free() calls in pcretest, to keep
1753 Electric Fence happy when testing.
1757 Version 5.0 13-Sep-04
1758 ---------------------
1760 1. Internal change: literal characters are no longer packed up into items
1761 containing multiple characters in a single byte-string. Each character
1762 is now matched using a separate opcode. However, there may be more than one
1763 byte in the character in UTF-8 mode.
1765 2. The pcre_callout_block structure has two new fields: pattern_position and
1766 next_item_length. These contain the offset in the pattern to the next match
1767 item, and its length, respectively.
1769 3. The PCRE_AUTO_CALLOUT option for pcre_compile() requests the automatic
1770 insertion of callouts before each pattern item. Added the /C option to
1771 pcretest to make use of this.
1773 4. On the advice of a Windows user, the lines
1775 #if defined(_WIN32) || defined(WIN32)
1776 _setmode( _fileno( stdout ), 0x8000 );
1777 #endif /* defined(_WIN32) || defined(WIN32) */
1779 have been added to the source of pcretest. This apparently does useful
1780 magic in relation to line terminators.
1782 5. Changed "r" and "w" in the calls to fopen() in pcretest to "rb" and "wb"
1783 for the benefit of those environments where the "b" makes a difference.
1785 6. The icc compiler has the same options as gcc, but "configure" doesn't seem
1786 to know about it. I have put a hack into configure.in that adds in code
1787 to set GCC=yes if CC=icc. This seems to end up at a point in the
1788 generated configure script that is early enough to affect the setting of
1789 compiler options, which is what is needed, but I have no means of testing
1790 whether it really works. (The user who reported this had patched the
1791 generated configure script, which of course I cannot do.)
1793 LATER: After change 22 below (new libtool files), the configure script
1794 seems to know about icc (and also ecc). Therefore, I have commented out
1795 this hack in configure.in.
1797 7. Added support for pkg-config (2 patches were sent in).
1799 8. Negated POSIX character classes that used a combination of internal tables
1800 were completely broken. These were [[:^alpha:]], [[:^alnum:]], and
1801 [[:^ascii]]. Typically, they would match almost any characters. The other
1802 POSIX classes were not broken in this way.
1804 9. Matching the pattern "\b.*?" against "ab cd", starting at offset 1, failed
1805 to find the match, as PCRE was deluded into thinking that the match had to
1806 start at the start point or following a newline. The same bug applied to
1807 patterns with negative forward assertions or any backward assertions
1808 preceding ".*" at the start, unless the pattern required a fixed first
1809 character. This was a failing pattern: "(?!.bcd).*". The bug is now fixed.
1811 10. In UTF-8 mode, when moving forwards in the subject after a failed match
1812 starting at the last subject character, bytes beyond the end of the subject
1813 string were read.
1815 11. Renamed the variable "class" as "classbits" to make life easier for C++
1816 users. (Previously there was a macro definition, but it apparently wasn't
1817 enough.)
1819 12. Added the new field "tables" to the extra data so that tables can be passed
1820 in at exec time, or the internal tables can be re-selected. This allows
1821 a compiled regex to be saved and re-used at a later time by a different
1822 program that might have everything at different addresses.
1824 13. Modified the pcre-config script so that, when run on Solaris, it shows a
1825 -R library as well as a -L library.
1827 14. The debugging options of pcretest (-d on the command line or D on a
1828 pattern) showed incorrect output for anything following an extended class
1829 that contained multibyte characters and which was followed by a quantifier.
1831 15. Added optional support for general category Unicode character properties
1832 via the \p, \P, and \X escapes. Unicode property support implies UTF-8
1833 support. It adds about 90K to the size of the library. The meanings of the
1834 inbuilt class escapes such as \d and \s have NOT been changed.
1836 16. Updated pcredemo.c to include calls to free() to release the memory for the
1837 compiled pattern.
1839 17. The generated file chartables.c was being created in the source directory
1840 instead of in the building directory. This caused the build to fail if the
1841 source directory was different from the building directory, and was
1842 read-only.
1844 18. Added some sample Win commands from Mark Tetrode into the NON-UNIX-USE
1845 file. No doubt somebody will tell me if they don't make sense... Also added
1846 Dan Mooney's comments about building on OpenVMS.
1848 19. Added support for partial matching via the PCRE_PARTIAL option for
1849 pcre_exec() and the \P data escape in pcretest.
1851 20. Extended pcretest with 3 new pattern features:
1853 (i) A pattern option of the form ">rest-of-line" causes pcretest to
1854 write the compiled pattern to the file whose name is "rest-of-line".
1855 This is a straight binary dump of the data, with the saved pointer to
1856 the character tables forced to be NULL. The study data, if any, is
1857 written too. After writing, pcretest reads a new pattern.
1859 (ii) If, instead of a pattern, "<rest-of-line" is given, pcretest reads a
1860 compiled pattern from the given file. There must not be any
1861 occurrences of "<" in the file name (pretty unlikely); if there are,
1862 pcretest will instead treat the initial "<" as a pattern delimiter.
1863 After reading in the pattern, pcretest goes on to read data lines as
1864 usual.
1866 (iii) The F pattern option causes pcretest to flip the bytes in the 32-bit
1867 and 16-bit fields in a compiled pattern, to simulate a pattern that
1868 was compiled on a host of opposite endianness.
1870 21. The pcre-exec() function can now cope with patterns that were compiled on
1871 hosts of opposite endianness, with this restriction:
1873 As for any compiled expression that is saved and used later, the tables
1874 pointer field cannot be preserved; the extra_data field in the arguments
1875 to pcre_exec() should be used to pass in a tables address if a value
1876 other than the default internal tables were used at compile time.
1878 22. Calling pcre_exec() with a negative value of the "ovecsize" parameter is
1879 now diagnosed as an error. Previously, most of the time, a negative number
1880 would have been treated as zero, but if in addition "ovector" was passed as
1881 NULL, a crash could occur.
1883 23. Updated the files ltmain.sh, config.sub, config.guess, and aclocal.m4 with
1884 new versions from the libtool 1.5 distribution (the last one is a copy of
1885 a file called libtool.m4). This seems to have fixed the need to patch
1886 "configure" to support Darwin 1.3 (which I used to do). However, I still
1887 had to patch ltmain.sh to ensure that ${SED} is set (it isn't on my
1888 workstation).
1890 24. Changed the PCRE licence to be the more standard "BSD" licence.
1893 Version 4.5 01-Dec-03
1894 ---------------------
1896 1. There has been some re-arrangement of the code for the match() function so
1897 that it can be compiled in a version that does not call itself recursively.
1898 Instead, it keeps those local variables that need separate instances for
1899 each "recursion" in a frame on the heap, and gets/frees frames whenever it
1900 needs to "recurse". Keeping track of where control must go is done by means
1901 of setjmp/longjmp. The whole thing is implemented by a set of macros that
1902 hide most of the details from the main code, and operates only if
1903 NO_RECURSE is defined while compiling pcre.c. If PCRE is built using the
1904 "configure" mechanism, "--disable-stack-for-recursion" turns on this way of
1905 operating.
1907 To make it easier for callers to provide specially tailored get/free
1908 functions for this usage, two new functions, pcre_stack_malloc, and
1909 pcre_stack_free, are used. They are always called in strict stacking order,
1910 and the size of block requested is always the same.
1912 The PCRE_CONFIG_STACKRECURSE info parameter can be used to find out whether
1913 PCRE has been compiled to use the stack or the heap for recursion. The
1914 -C option of pcretest uses this to show which version is compiled.
1916 A new data escape \S, is added to pcretest; it causes the amounts of store
1917 obtained and freed by both kinds of malloc/free at match time to be added
1918 to the output.
1920 2. Changed the locale test to use "fr_FR" instead of "fr" because that's
1921 what's available on my current Linux desktop machine.
1923 3. When matching a UTF-8 string, the test for a valid string at the start has
1924 been extended. If start_offset is not zero, PCRE now checks that it points
1925 to a byte that is the start of a UTF-8 character. If not, it returns
1926 PCRE_ERROR_BADUTF8_OFFSET (-11). Note: the whole string is still checked;
1927 this is necessary because there may be backward assertions in the pattern.
1928 When matching the same subject several times, it may save resources to use
1929 PCRE_NO_UTF8_CHECK on all but the first call if the string is long.
1931 4. The code for checking the validity of UTF-8 strings has been tightened so
1932 that it rejects (a) strings containing 0xfe or 0xff bytes and (b) strings
1933 containing "overlong sequences".
1935 5. Fixed a bug (appearing twice) that I could not find any way of exploiting!
1936 I had written "if ((digitab[*p++] && chtab_digit) == 0)" where the "&&"
1937 should have been "&", but it just so happened that all the cases this let
1938 through by mistake were picked up later in the function.
1940 6. I had used a variable called "isblank" - this is a C99 function, causing
1941 some compilers to warn. To avoid this, I renamed it (as "blankclass").
1943 7. Cosmetic: (a) only output another newline at the end of pcretest if it is
1944 prompting; (b) run "./pcretest /dev/null" at the start of the test script
1945 so the version is shown; (c) stop "make test" echoing "./RunTest".
1947 8. Added patches from David Burgess to enable PCRE to run on EBCDIC systems.
1949 9. The prototype for memmove() for systems that don't have it was using
1950 size_t, but the inclusion of the header that defines size_t was later. I've
1951 moved the #includes for the C headers earlier to avoid this.
1953 10. Added some adjustments to the code to make it easier to compiler on certain
1954 special systems:
1956 (a) Some "const" qualifiers were missing.
1957 (b) Added the macro EXPORT before all exported functions; by default this
1958 is defined to be empty.
1959 (c) Changed the dftables auxiliary program (that builds chartables.c) so
1960 that it reads its output file name as an argument instead of writing
1961 to the standard output and assuming this can be redirected.
1963 11. In UTF-8 mode, if a recursive reference (e.g. (?1)) followed a character
1964 class containing characters with values greater than 255, PCRE compilation
1965 went into a loop.
1967 12. A recursive reference to a subpattern that was within another subpattern
1968 that had a minimum quantifier of zero caused PCRE to crash. For example,
1969 (x(y(?2))z)? provoked this bug with a subject that got as far as the
1970 recursion. If the recursively-called subpattern itself had a zero repeat,
1971 that was OK.
1973 13. In pcretest, the buffer for reading a data line was set at 30K, but the
1974 buffer into which it was copied (for escape processing) was still set at
1975 1024, so long lines caused crashes.
1977 14. A pattern such as /[ab]{1,3}+/ failed to compile, giving the error
1978 "internal error: code overflow...". This applied to any character class
1979 that was followed by a possessive quantifier.
1981 15. Modified the Makefile to add libpcre.la as a prerequisite for
1982 libpcreposix.la because I was told this is needed for a parallel build to
1983 work.
1985 16. If a pattern that contained .* following optional items at the start was
1986 studied, the wrong optimizing data was generated, leading to matching
1987 errors. For example, studying /[ab]*.*c/ concluded, erroneously, that any
1988 matching string must start with a or b or c. The correct conclusion for
1989 this pattern is that a match can start with any character.
1992 Version 4.4 13-Aug-03
1993 ---------------------
1995 1. In UTF-8 mode, a character class containing characters with values between
1996 127 and 255 was not handled correctly if the compiled pattern was studied.
1997 In fixing this, I have also improved the studying algorithm for such
1998 classes (slightly).
2000 2. Three internal functions had redundant arguments passed to them. Removal
2001 might give a very teeny performance improvement.
2003 3. Documentation bug: the value of the capture_top field in a callout is *one
2004 more than* the number of the hightest numbered captured substring.
2006 4. The Makefile linked pcretest and pcregrep with -lpcre, which could result
2007 in incorrectly linking with a previously installed version. They now link
2008 explicitly with libpcre.la.
2010 5. configure.in no longer needs to recognize Cygwin specially.
2012 6. A problem in pcre.in for Windows platforms is fixed.
2014 7. If a pattern was successfully studied, and the -d (or /D) flag was given to
2015 pcretest, it used to include the size of the study block as part of its
2016 output. Unfortunately, the structure contains a field that has a different
2017 size on different hardware architectures. This meant that the tests that
2018 showed this size failed. As the block is currently always of a fixed size,
2019 this information isn't actually particularly useful in pcretest output, so
2020 I have just removed it.
2022 8. Three pre-processor statements accidentally did not start in column 1.
2023 Sadly, there are *still* compilers around that complain, even though
2024 standard C has not required this for well over a decade. Sigh.
2026 9. In pcretest, the code for checking callouts passed small integers in the
2027 callout_data field, which is a void * field. However, some picky compilers
2028 complained about the casts involved for this on 64-bit systems. Now
2029 pcretest passes the address of the small integer instead, which should get
2030 rid of the warnings.
2032 10. By default, when in UTF-8 mode, PCRE now checks for valid UTF-8 strings at
2033 both compile and run time, and gives an error if an invalid UTF-8 sequence
2034 is found. There is a option for disabling this check in cases where the
2035 string is known to be correct and/or the maximum performance is wanted.
2037 11. In response to a bug report, I changed one line in Makefile.in from
2039 -Wl,--out-implib,.libs/lib@WIN_PREFIX@pcreposix.dll.a \
2040 to
2041 -Wl,--out-implib,.libs/@WIN_PREFIX@libpcreposix.dll.a \
2043 to look similar to other lines, but I have no way of telling whether this
2044 is the right thing to do, as I do not use Windows. No doubt I'll get told
2045 if it's wrong...
2048 Version 4.3 21-May-03
2049 ---------------------
2051 1. Two instances of @WIN_PREFIX@ omitted from the Windows targets in the
2052 Makefile.
2054 2. Some refactoring to improve the quality of the code:
2056 (i) The utf8_table... variables are now declared "const".
2058 (ii) The code for \cx, which used the "case flipping" table to upper case
2059 lower case letters, now just substracts 32. This is ASCII-specific,
2060 but the whole concept of \cx is ASCII-specific, so it seems
2061 reasonable.
2063 (iii) PCRE was using its character types table to recognize decimal and
2064 hexadecimal digits in the pattern. This is silly, because it handles
2065 only 0-9, a-f, and A-F, but the character types table is locale-
2066 specific, which means strange things might happen. A private
2067 table is now used for this - though it costs 256 bytes, a table is
2068 much faster than multiple explicit tests. Of course, the standard
2069 character types table is still used for matching digits in subject
2070 strings against \d.
2072 (iv) Strictly, the identifier ESC_t is reserved by POSIX (all identifiers
2073 ending in _t are). So I've renamed it as ESC_tee.
2075 3. The first argument for regexec() in the POSIX wrapper should have been
2076 defined as "const".
2078 4. Changed pcretest to use malloc() for its buffers so that they can be
2079 Electric Fenced for debugging.
2081 5. There were several places in the code where, in UTF-8 mode, PCRE would try
2082 to read one or more bytes before the start of the subject string. Often this
2083 had no effect on PCRE's behaviour, but in some circumstances it could
2084 provoke a segmentation fault.
2086 6. A lookbehind at the start of a pattern in UTF-8 mode could also cause PCRE
2087 to try to read one or more bytes before the start of the subject string.
2089 7. A lookbehind in a pattern matched in non-UTF-8 mode on a PCRE compiled with
2090 UTF-8 support could misbehave in various ways if the subject string
2091 contained bytes with the 0x80 bit set and the 0x40 bit unset in a lookbehind
2092 area. (PCRE was not checking for the UTF-8 mode flag, and trying to move
2093 back over UTF-8 characters.)
2096 Version 4.2 14-Apr-03
2097 ---------------------
2099 1. Typo "#if SUPPORT_UTF8" instead of "#ifdef SUPPORT_UTF8" fixed.
2101 2. Changes to the building process, supplied by Ronald Landheer-Cieslak
2102 [ON_WINDOWS]: new variable, "#" on non-Windows platforms
2103 [NOT_ON_WINDOWS]: new variable, "#" on Windows platforms
2104 [WIN_PREFIX]: new variable, "cyg" for Cygwin
2105 * Makefile.in: use autoconf substitution for OBJEXT, EXEEXT, BUILD_OBJEXT
2107 Note: automatic setting of the BUILD variables is not yet working
2108 set CPPFLAGS and BUILD_CPPFLAGS (but don't use yet) - should be used at
2109 compile-time but not at link-time
2110 [LINK]: use for linking executables only
2111 make different versions for Windows and non-Windows
2112 [LINKLIB]: new variable, copy of UNIX-style LINK, used for linking
2113 libraries
2114 [LINK_FOR_BUILD]: new variable
2115 [OBJEXT]: use throughout
2116 [EXEEXT]: use throughout
2117 <winshared>: new target
2118 <wininstall>: new target
2119 <dftables.o>: use native compiler
2120 <dftables>: use native linker
2121 <install>: handle Windows platform correctly
2122 <clean>: ditto
2123 <check>: ditto
2124 copy DLL to top builddir before testing
2126 As part of these changes, -no-undefined was removed again. This was reported
2127 to give trouble on HP-UX 11.0, so getting rid of it seems like a good idea
2128 in any case.
2130 3. Some tidies to get rid of compiler warnings:
2132 . In the match_data structure, match_limit was an unsigned long int, whereas
2133 match_call_count was an int. I've made them both unsigned long ints.
2135 . In pcretest the fact that a const uschar * doesn't automatically cast to
2136 a void * provoked a warning.
2138 . Turning on some more compiler warnings threw up some "shadow" variables
2139 and a few more missing casts.
2141 4. If PCRE was complied with UTF-8 support, but called without the PCRE_UTF8
2142 option, a class that contained a single character with a value between 128
2143 and 255 (e.g. /[\xFF]/) caused PCRE to crash.
2145 5. If PCRE was compiled with UTF-8 support, but called without the PCRE_UTF8
2146 option, a class that contained several characters, but with at least one
2147 whose value was between 128 and 255 caused PCRE to crash.
2150 Version 4.1 12-Mar-03
2151 ---------------------
2153 1. Compiling with gcc -pedantic found a couple of places where casts were
2154 needed, and a string in dftables.c that was longer than standard compilers are
2155 required to support.
2157 2. Compiling with Sun's compiler found a few more places where the code could
2158 be tidied up in order to avoid warnings.
2160 3. The variables for cross-compiling were called HOST_CC and HOST_CFLAGS; the
2161 first of these names is deprecated in the latest Autoconf in favour of the name
2162 CC_FOR_BUILD, because "host" is typically used to mean the system on which the
2163 compiled code will be run. I can't find a reference for HOST_CFLAGS, but by
2164 analogy I have changed it to CFLAGS_FOR_BUILD.
2166 4. Added -no-undefined to the linking command in the Makefile, because this is
2167 apparently helpful for Windows. To make it work, also added "-L. -lpcre" to the
2168 linking step for the pcreposix library.
2170 5. PCRE was failing to diagnose the case of two named groups with the same
2171 name.
2173 6. A problem with one of PCRE's optimizations was discovered. PCRE remembers a
2174 literal character that is needed in the subject for a match, and scans along to
2175 ensure that it is present before embarking on the full matching process. This
2176 saves time in cases of nested unlimited repeats that are never going to match.
2177 Problem: the scan can take a lot of time if the subject is very long (e.g.
2178 megabytes), thus penalizing straightforward matches. It is now done only if the
2179 amount of subject to be scanned is less than 1000 bytes.
2181 7. A lesser problem with the same optimization is that it was recording the
2182 first character of an anchored pattern as "needed", thus provoking a search
2183 right along the subject, even when the first match of the pattern was going to
2184 fail. The "needed" character is now not set for anchored patterns, unless it
2185 follows something in the pattern that is of non-fixed length. Thus, it still
2186 fulfils its original purpose of finding quick non-matches in cases of nested
2187 unlimited repeats, but isn't used for simple anchored patterns such as /^abc/.
2190 Version 4.0 17-Feb-03
2191 ---------------------
2193 1. If a comment in an extended regex that started immediately after a meta-item
2194 extended to the end of string, PCRE compiled incorrect data. This could lead to
2195 all kinds of weird effects. Example: /#/ was bad; /()#/ was bad; /a#/ was not.
2197 2. Moved to autoconf 2.53 and libtool 1.4.2.
2199 3. Perl 5.8 no longer needs "use utf8" for doing UTF-8 things. Consequently,
2200 the special perltest8 script is no longer needed - all the tests can be run
2201 from a single perltest script.
2203 4. From 5.004, Perl has not included the VT character (0x0b) in the set defined
2204 by \s. It has now been removed in PCRE. This means it isn't recognized as
2205 whitespace in /x regexes too, which is the same as Perl. Note that the POSIX
2206 class [:space:] *does* include VT, thereby creating a mess.
2208 5. Added the class [:blank:] (a GNU extension from Perl 5.8) to match only
2209 space and tab.
2211 6. Perl 5.005 was a long time ago. It's time to amalgamate the tests that use
2212 its new features into the main test script, reducing the number of scripts.
2214 7. Perl 5.8 has changed the meaning of patterns like /a(?i)b/. Earlier versions
2215 were backward compatible, and made the (?i) apply to the whole pattern, as if
2216 /i were given. Now it behaves more logically, and applies the option setting
2217 only to what follows. PCRE has been changed to follow suit. However, if it
2218 finds options settings right at the start of the pattern, it extracts them into
2219 the global options, as before. Thus, they show up in the info data.
2221 8. Added support for the \Q...\E escape sequence. Characters in between are
2222 treated as literals. This is slightly different from Perl in that $ and @ are
2223 also handled as literals inside the quotes. In Perl, they will cause variable
2224 interpolation. Note the following examples:
2226 Pattern PCRE matches Perl matches
2228 \Qabc$xyz\E abc$xyz abc followed by the contents of $xyz
2229 \Qabc\$xyz\E abc\$xyz abc\$xyz
2230 \Qabc\E\$\Qxyz\E abc$xyz abc$xyz
2232 For compatibility with Perl, \Q...\E sequences are recognized inside character
2233 classes as well as outside them.
2235 9. Re-organized 3 code statements in pcretest to avoid "overflow in
2236 floating-point constant arithmetic" warnings from a Microsoft compiler. Added a
2237 (size_t) cast to one statement in pcretest and one in pcreposix to avoid
2238 signed/unsigned warnings.
2240 10. SunOS4 doesn't have strtoul(). This was used only for unpicking the -o
2241 option for pcretest, so I've replaced it by a simple function that does just
2242 that job.
2244 11. pcregrep was ending with code 0 instead of 2 for the commands "pcregrep" or
2245 "pcregrep -".
2247 12. Added "possessive quantifiers" ?+, *+, ++, and {,}+ which come from Sun's
2248 Java package. This provides some syntactic sugar for simple cases of what my
2249 documentation calls "once-only subpatterns". A pattern such as x*+ is the same
2250 as (?>x*). In other words, if what is inside (?>...) is just a single repeated
2251 item, you can use this simplified notation. Note that only makes sense with
2252 greedy quantifiers. Consequently, the use of the possessive quantifier forces
2253 greediness, whatever the setting of the PCRE_UNGREEDY option.
2255 13. A change of greediness default within a pattern was not taking effect at
2256 the current level for patterns like /(b+(?U)a+)/. It did apply to parenthesized
2257 subpatterns that followed. Patterns like /b+(?U)a+/ worked because the option
2258 was abstracted outside.
2260 14. PCRE now supports the \G assertion. It is true when the current matching
2261 position is at the start point of the match. This differs from \A when the
2262 starting offset is non-zero. Used with the /g option of pcretest (or similar
2263 code), it works in the same way as it does for Perl's /g option. If all
2264 alternatives of a regex begin with \G, the expression is anchored to the start
2265 match position, and the "anchored" flag is set in the compiled expression.
2267 15. Some bugs concerning the handling of certain option changes within patterns
2268 have been fixed. These applied to options other than (?ims). For example,
2269 "a(?x: b c )d" did not match "XabcdY" but did match "Xa b c dY". It should have
2270 been the other way round. Some of this was related to change 7 above.
2272 16. PCRE now gives errors for /[.x.]/ and /[=x=]/ as unsupported POSIX
2273 features, as Perl does. Previously, PCRE gave the warnings only for /[[.x.]]/
2274 and /[[=x=]]/. PCRE now also gives an error for /[:name:]/ because it supports
2275 POSIX classes only within a class (e.g. /[[:alpha:]]/).
2277 17. Added support for Perl's \C escape. This matches one byte, even in UTF8
2278 mode. Unlike ".", it always matches newline, whatever the setting of
2279 PCRE_DOTALL. However, PCRE does not permit \C to appear in lookbehind
2280 assertions. Perl allows it, but it doesn't (in general) work because it can't
2281 calculate the length of the lookbehind. At least, that's the case for Perl
2282 5.8.0 - I've been told they are going to document that it doesn't work in
2283 future.
2285 18. Added an error diagnosis for escapes that PCRE does not support: these are
2286 \L, \l, \N, \P, \p, \U, \u, and \X.
2288 19. Although correctly diagnosing a missing ']' in a character class, PCRE was
2289 reading past the end of the pattern in cases such as /[abcd/.
2291 20. PCRE was getting more memory than necessary for patterns with classes that
2292 contained both POSIX named classes and other characters, e.g. /[[:space:]abc/.
2294 21. Added some code, conditional on #ifdef VPCOMPAT, to make life easier for
2295 compiling PCRE for use with Virtual Pascal.
2297 22. Small fix to the Makefile to make it work properly if the build is done
2298 outside the source tree.
2300 23. Added a new extension: a condition to go with recursion. If a conditional
2301 subpattern starts with (?(R) the "true" branch is used if recursion has
2302 happened, whereas the "false" branch is used only at the top level.
2304 24. When there was a very long string of literal characters (over 255 bytes
2305 without UTF support, over 250 bytes with UTF support), the computation of how
2306 much memory was required could be incorrect, leading to segfaults or other
2307 strange effects.
2309 25. PCRE was incorrectly assuming anchoring (either to start of subject or to
2310 start of line for a non-DOTALL pattern) when a pattern started with (.*) and
2311 there was a subsequent back reference to those brackets. This meant that, for
2312 example, /(.*)\d+\1/ failed to match "abc123bc". Unfortunately, it isn't
2313 possible to check for precisely this case. All we can do is abandon the
2314 optimization if .* occurs inside capturing brackets when there are any back
2315 references whatsoever. (See below for a better fix that came later.)
2317 26. The handling of the optimization for finding the first character of a
2318 non-anchored pattern, and for finding a character that is required later in the
2319 match were failing in some cases. This didn't break the matching; it just
2320 failed to optimize when it could. The way this is done has been re-implemented.
2322 27. Fixed typo in error message for invalid (?R item (it said "(?p").
2324 28. Added a new feature that provides some of the functionality that Perl
2325 provides with (?{...}). The facility is termed a "callout". The way it is done
2326 in PCRE is for the caller to provide an optional function, by setting
2327 pcre_callout to its entry point. Like pcre_malloc and pcre_free, this is a
2328 global variable. By default it is unset, which disables all calling out. To get
2329 the function called, the regex must include (?C) at appropriate points. This
2330 is, in fact, equivalent to (?C0), and any number <= 255 may be given with (?C).
2331 This provides a means of identifying different callout points. When PCRE
2332 reaches such a point in the regex, if pcre_callout has been set, the external
2333 function is called. It is provided with data in a structure called
2334 pcre_callout_block, which is defined in pcre.h. If the function returns 0,
2335 matching continues; if it returns a non-zero value, the match at the current
2336 point fails. However, backtracking will occur if possible. [This was changed
2337 later and other features added - see item 49 below.]
2339 29. pcretest is upgraded to test the callout functionality. It provides a
2340 callout function that displays information. By default, it shows the start of
2341 the match and the current position in the text. There are some new data escapes
2342 to vary what happens:
2344 \C+ in addition, show current contents of captured substrings
2345 \C- do not supply a callout function
2346 \C!n return 1 when callout number n is reached
2347 \C!n!m return 1 when callout number n is reached for the mth time
2349 30. If pcregrep was called with the -l option and just a single file name, it
2350 output "<stdin>" if a match was found, instead of the file name.
2352 31. Improve the efficiency of the POSIX API to PCRE. If the number of capturing
2353 slots is less than POSIX_MALLOC_THRESHOLD, use a block on the stack to pass to
2354 pcre_exec(). This saves a malloc/free per call. The default value of
2355 POSIX_MALLOC_THRESHOLD is 10; it can be changed by --with-posix-malloc-threshold
2356 when configuring.
2358 32. The default maximum size of a compiled pattern is 64K. There have been a
2359 few cases of people hitting this limit. The code now uses macros to handle the
2360 storing of links as offsets within the compiled pattern. It defaults to 2-byte
2361 links, but this can be changed to 3 or 4 bytes by --with-link-size when
2362 configuring. Tests 2 and 5 work only with 2-byte links because they output
2363 debugging information about compiled patterns.
2365 33. Internal code re-arrangements:
2367 (a) Moved the debugging function for printing out a compiled regex into
2368 its own source file (printint.c) and used #include to pull it into
2369 pcretest.c and, when DEBUG is defined, into pcre.c, instead of having two
2370 separate copies.
2372 (b) Defined the list of op-code names for debugging as a macro in
2373 internal.h so that it is next to the definition of the opcodes.
2375 (c) Defined a table of op-code lengths for simpler skipping along compiled
2376 code. This is again a macro in internal.h so that it is next to the
2377 definition of the opcodes.
2379 34. Added support for recursive calls to individual subpatterns, along the
2380 lines of Robin Houston's patch (but implemented somewhat differently).
2382 35. Further mods to the Makefile to help Win32. Also, added code to pcregrep to
2383 allow it to read and process whole directories in Win32. This code was
2384 contributed by Lionel Fourquaux; it has not been tested by me.
2386 36. Added support for named subpatterns. The Python syntax (?P<name>...) is
2387 used to name a group. Names consist of alphanumerics and underscores, and must
2388 be unique. Back references use the syntax (?P=name) and recursive calls use
2389 (?P>name) which is a PCRE extension to the Python extension. Groups still have
2390 numbers. The function pcre_fullinfo() can be used after compilation to extract
2391 a name/number map. There are three relevant calls:
2393 PCRE_INFO_NAMEENTRYSIZE yields the size of each entry in the map
2394 PCRE_INFO_NAMECOUNT yields the number of entries
2395 PCRE_INFO_NAMETABLE yields a pointer to the map.
2397 The map is a vector of fixed-size entries. The size of each entry depends on
2398 the length of the longest name used. The first two bytes of each entry are the
2399 group number, most significant byte first. There follows the corresponding
2400 name, zero terminated. The names are in alphabetical order.
2402 37. Make the maximum literal string in the compiled code 250 for the non-UTF-8
2403 case instead of 255. Making it the same both with and without UTF-8 support
2404 means that the same test output works with both.
2406 38. There was a case of malloc(0) in the POSIX testing code in pcretest. Avoid
2407 calling malloc() with a zero argument.
2409 39. Change 25 above had to resort to a heavy-handed test for the .* anchoring
2410 optimization. I've improved things by keeping a bitmap of backreferences with
2411 numbers 1-31 so that if .* occurs inside capturing brackets that are not in
2412 fact referenced, the optimization can be applied. It is unlikely that a
2413 relevant occurrence of .* (i.e. one which might indicate anchoring or forcing
2414 the match to follow \n) will appear inside brackets with a number greater than
2415 31, but if it does, any back reference > 31 suppresses the optimization.
2417 40. Added a new compile-time option PCRE_NO_AUTO_CAPTURE. This has the effect
2418 of disabling numbered capturing parentheses. Any opening parenthesis that is
2419 not followed by ? behaves as if it were followed by ?: but named parentheses
2420 can still be used for capturing (and they will acquire numbers in the usual
2421 way).
2423 41. Redesigned the return codes from the match() function into yes/no/error so
2424 that errors can be passed back from deep inside the nested calls. A malloc
2425 failure while inside a recursive subpattern call now causes the
2426 PCRE_ERROR_NOMEMORY return instead of quietly going wrong.
2428 42. It is now possible to set a limit on the number of times the match()
2429 function is called in a call to pcre_exec(). This facility makes it possible to
2430 limit the amount of recursion and backtracking, though not in a directly
2431 obvious way, because the match() function is used in a number of different
2432 circumstances. The count starts from zero for each position in the subject
2433 string (for non-anchored patterns). The default limit is, for compatibility, a
2434 large number, namely 10 000 000. You can change this in two ways:
2436 (a) When configuring PCRE before making, you can use --with-match-limit=n
2437 to set a default value for the compiled library.
2439 (b) For each call to pcre_exec(), you can pass a pcre_extra block in which
2440 a different value is set. See 45 below.
2442 If the limit is exceeded, pcre_exec() returns PCRE_ERROR_MATCHLIMIT.
2444 43. Added a new function pcre_config(int, void *) to enable run-time extraction
2445 of things that can be changed at compile time. The first argument specifies
2446 what is wanted and the second points to where the information is to be placed.
2447 The current list of available information is:
2451 The output is an integer that is set to one if UTF-8 support is available;
2452 otherwise it is set to zero.
2456 The output is an integer that it set to the value of the code that is used for
2457 newline. It is either LF (10) or CR (13).
2461 The output is an integer that contains the number of bytes used for internal
2462 linkage in compiled expressions. The value is 2, 3, or 4. See item 32 above.
2466 The output is an integer that contains the threshold above which the POSIX
2467 interface uses malloc() for output vectors. See item 31 above.
2471 The output is an unsigned integer that contains the default limit of the number
2472 of match() calls in a pcre_exec() execution. See 42 above.
2474 44. pcretest has been upgraded by the addition of the -C option. This causes it
2475 to extract all the available output from the new pcre_config() function, and to
2476 output it. The program then exits immediately.
2478 45. A need has arisen to pass over additional data with calls to pcre_exec() in
2479 order to support additional features. One way would have been to define
2480 pcre_exec2() (for example) with extra arguments, but this would not have been
2481 extensible, and would also have required all calls to the original function to
2482 be mapped to the new one. Instead, I have chosen to extend the mechanism that
2483 is used for passing in "extra" data from pcre_study().
2485 The pcre_extra structure is now exposed and defined in pcre.h. It currently
2486 contains the following fields:
2488 flags a bitmap indicating which of the following fields are set
2489 study_data opaque data from pcre_study()
2490 match_limit a way of specifying a limit on match() calls for a specific
2491 call to pcre_exec()
2492 callout_data data for callouts (see 49 below)
2494 The flag bits are also defined in pcre.h, and are
2500 The pcre_study() function now returns one of these new pcre_extra blocks, with
2501 the actual study data pointed to by the study_data field, and the
2502 PCRE_EXTRA_STUDY_DATA flag set. This can be passed directly to pcre_exec() as
2503 before. That is, this change is entirely upwards-compatible and requires no
2504 change to existing code.
2506 If you want to pass in additional data to pcre_exec(), you can either place it
2507 in a pcre_extra block provided by pcre_study(), or create your own pcre_extra
2508 block.
2510 46. pcretest has been extended to test the PCRE_EXTRA_MATCH_LIMIT feature. If a
2511 data string contains the escape sequence \M, pcretest calls pcre_exec() several
2512 times with different match limits, until it finds the minimum value needed for
2513 pcre_exec() to complete. The value is then output. This can be instructive; for
2514 most simple matches the number is quite small, but for pathological cases it
2515 gets very large very quickly.
2517 47. There's a new option for pcre_fullinfo() called PCRE_INFO_STUDYSIZE. It
2518 returns the size of the data block pointed to by the study_data field in a
2519 pcre_extra block, that is, the value that was passed as the argument to
2520 pcre_malloc() when PCRE was getting memory in which to place the information
2521 created by pcre_study(). The fourth argument should point to a size_t variable.
2522 pcretest has been extended so that this information is shown after a successful
2523 pcre_study() call when information about the compiled regex is being displayed.
2525 48. Cosmetic change to Makefile: there's no need to have / after $(DESTDIR)
2526 because what follows is always an absolute path. (Later: it turns out that this
2527 is more than cosmetic for MinGW, because it doesn't like empty path
2528 components.)
2530 49. Some changes have been made to the callout feature (see 28 above):
2532 (i) A callout function now has three choices for what it returns:
2534 0 => success, carry on matching
2535 > 0 => failure at this point, but backtrack if possible
2536 < 0 => serious error, return this value from pcre_exec()
2538 Negative values should normally be chosen from the set of PCRE_ERROR_xxx
2539 values. In particular, returning PCRE_ERROR_NOMATCH forces a standard
2540 "match failed" error. The error number PCRE_ERROR_CALLOUT is reserved for
2541 use by callout functions. It will never be used by PCRE itself.
2543 (ii) The pcre_extra structure (see 45 above) has a void * field called
2544 callout_data, with corresponding flag bit PCRE_EXTRA_CALLOUT_DATA. The
2545 pcre_callout_block structure has a field of the same name. The contents of
2546 the field passed in the pcre_extra structure are passed to the callout
2547 function in the corresponding field in the callout block. This makes it
2548 easier to use the same callout-containing regex from multiple threads. For
2549 testing, the pcretest program has a new data escape
2551 \C*n pass the number n (may be negative) as callout_data
2553 If the callout function in pcretest receives a non-zero value as
2554 callout_data, it returns that value.
2556 50. Makefile wasn't handling CFLAGS properly when compiling dftables. Also,
2557 there were some redundant $(CFLAGS) in commands that are now specified as
2558 $(LINK), which already includes $(CFLAGS).
2560 51. Extensions to UTF-8 support are listed below. These all apply when (a) PCRE
2561 has been compiled with UTF-8 support *and* pcre_compile() has been compiled
2562 with the PCRE_UTF8 flag. Patterns that are compiled without that flag assume
2563 one-byte characters throughout. Note that case-insensitive matching applies
2564 only to characters whose values are less than 256. PCRE doesn't support the
2565 notion of cases for higher-valued characters.
2567 (i) A character class whose characters are all within 0-255 is handled as
2568 a bit map, and the map is inverted for negative classes. Previously, a
2569 character > 255 always failed to match such a class; however it should
2570 match if the class was a negative one (e.g. [^ab]). This has been fixed.
2572 (ii) A negated character class with a single character < 255 is coded as
2573 "not this character" (OP_NOT). This wasn't working properly when the test
2574 character was multibyte, either singly or repeated.
2576 (iii) Repeats of multibyte characters are now handled correctly in UTF-8
2577 mode, for example: \x{100}{2,3}.
2579 (iv) The character escapes \b, \B, \d, \D, \s, \S, \w, and \W (either
2580 singly or repeated) now correctly test multibyte characters. However,
2581 PCRE doesn't recognize any characters with values greater than 255 as
2582 digits, spaces, or word characters. Such characters always match \D, \S,
2583 and \W, and never match \d, \s, or \w.
2585 (v) Classes may now contain characters and character ranges with values
2586 greater than 255. For example: [ab\x{100}-\x{400}].
2588 (vi) pcregrep now has a --utf-8 option (synonym -u) which makes it call
2589 PCRE in UTF-8 mode.
2591 52. The info request value PCRE_INFO_FIRSTCHAR has been renamed
2592 PCRE_INFO_FIRSTBYTE because it is a byte value. However, the old name is
2593 retained for backwards compatibility. (Note that LASTLITERAL is also a byte
2594 value.)
2596 53. The single man page has become too large. I have therefore split it up into
2597 a number of separate man pages. These also give rise to individual HTML pages;
2598 these are now put in a separate directory, and there is an index.html page that
2599 lists them all. Some hyperlinking between the pages has been installed.
2601 54. Added convenience functions for handling named capturing parentheses.
2603 55. Unknown escapes inside character classes (e.g. [\M]) and escapes that
2604 aren't interpreted therein (e.g. [\C]) are literals in Perl. This is now also
2605 true in PCRE, except when the PCRE_EXTENDED option is set, in which case they
2606 are faulted.
2608 56. Introduced HOST_CC and HOST_CFLAGS which can be set in the environment when
2609 calling configure. These values are used when compiling the dftables.c program
2610 which is run to generate the source of the default character tables. They
2611 default to the values of CC and CFLAGS. If you are cross-compiling PCRE,
2612 you will need to set these values.
2614 57. Updated the building process for Windows DLL, as provided by Fred Cox.
2617 Version 3.9 02-Jan-02
2618 ---------------------
2620 1. A bit of extraneous text had somehow crept into the pcregrep documentation.
2622 2. If --disable-static was given, the building process failed when trying to
2623 build pcretest and pcregrep. (For some reason it was using libtool to compile
2624 them, which is not right, as they aren't part of the library.)
2627 Version 3.8 18-Dec-01
2628 ---------------------
2630 1. The experimental UTF-8 code was completely screwed up. It was packing the
2631 bytes in the wrong order. How dumb can you get?
2634 Version 3.7 29-Oct-01
2635 ---------------------
2637 1. In updating pcretest to check change 1 of version 3.6, I screwed up.
2638 This caused pcretest, when used on the test data, to segfault. Unfortunately,
2639 this didn't happen under Solaris 8, where I normally test things.
2641 2. The Makefile had to be changed to make it work on BSD systems, where 'make'
2642 doesn't seem to recognize that ./xxx and xxx are the same file. (This entry
2643 isn't in ChangeLog distributed with 3.7 because I forgot when I hastily made
2644 this fix an hour or so after the initial 3.7 release.)
2647 Version 3.6 23-Oct-01
2648 ---------------------
2650 1. Crashed with /(sens|respons)e and \1ibility/ and "sense and sensibility" if
2651 offsets passed as NULL with zero offset count.
2653 2. The config.guess and config.sub files had not been updated when I moved to
2654 the latest autoconf.
2657 Version 3.5 15-Aug-01
2658 ---------------------
2660 1. Added some missing #if !defined NOPOSIX conditionals in pcretest.c that
2661 had been forgotten.
2663 2. By using declared but undefined structures, we can avoid using "void"
2664 definitions in pcre.h while keeping the internal definitions of the structures
2665 private.
2667 3. The distribution is now built using autoconf 2.50 and libtool 1.4. From a
2668 user point of view, this means that both static and shared libraries are built
2669 by default, but this can be individually controlled. More of the work of
2670 handling this static/shared cases is now inside libtool instead of PCRE's make
2671 file.
2673 4. The pcretest utility is now installed along with pcregrep because it is
2674 useful for users (to test regexs) and by doing this, it automatically gets
2675 relinked by libtool. The documentation has been turned into a man page, so
2676 there are now .1, .txt, and .html versions in /doc.
2678 5. Upgrades to pcregrep:
2679 (i) Added long-form option names like gnu grep.
2680 (ii) Added --help to list all options with an explanatory phrase.
2681 (iii) Added -r, --recursive to recurse into sub-directories.
2682 (iv) Added -f, --file to read patterns from a file.
2684 6. pcre_exec() was referring to its "code" argument before testing that
2685 argument for NULL (and giving an error if it was NULL).
2687 7. Upgraded Makefile.in to allow for compiling in a different directory from
2688 the source directory.
2690 8. Tiny buglet in pcretest: when pcre_fullinfo() was called to retrieve the
2691 options bits, the pointer it was passed was to an int instead of to an unsigned
2692 long int. This mattered only on 64-bit systems.
2694 9. Fixed typo (3.4/1) in pcre.h again. Sigh. I had changed pcre.h (which is
2695 generated) instead of pcre.in, which it its source. Also made the same change
2696 in several of the .c files.
2698 10. A new release of gcc defines printf() as a macro, which broke pcretest
2699 because it had an ifdef in the middle of a string argument for printf(). Fixed
2700 by using separate calls to printf().
2702 11. Added --enable-newline-is-cr and --enable-newline-is-lf to the configure
2703 script, to force use of CR or LF instead of \n in the source. On non-Unix
2704 systems, the value can be set in config.h.
2706 12. The limit of 200 on non-capturing parentheses is a _nesting_ limit, not an
2707 absolute limit. Changed the text of the error message to make this clear, and
2708 likewise updated the man page.
2710 13. The limit of 99 on the number of capturing subpatterns has been removed.
2711 The new limit is 65535, which I hope will not be a "real" limit.
2714 Version 3.4 22-Aug-00
2715 ---------------------
2717 1. Fixed typo in pcre.h: unsigned const char * changed to const unsigned char *.
2719 2. Diagnose condition (?(0) as an error instead of crashing on matching.
2722 Version 3.3 01-Aug-00
2723 ---------------------
2725 1. If an octal character was given, but the value was greater than \377, it
2726 was not getting masked to the least significant bits, as documented. This could
2727 lead to crashes in some systems.
2729 2. Perl 5.6 (if not earlier versions) accepts classes like [a-\d] and treats
2730 the hyphen as a literal. PCRE used to give an error; it now behaves like Perl.
2732 3. Added the functions pcre_free_substring() and pcre_free_substring_list().
2733 These just pass their arguments on to (pcre_free)(), but they are provided
2734 because some uses of PCRE bind it to non-C systems that can call its functions,
2735 but cannot call free() or pcre_free() directly.
2737 4. Add "make test" as a synonym for "make check". Corrected some comments in
2738 the Makefile.
2740 5. Add $(DESTDIR)/ in front of all the paths in the "install" target in the
2741 Makefile.
2743 6. Changed the name of pgrep to pcregrep, because Solaris has introduced a
2744 command called pgrep for grepping around the active processes.
2746 7. Added the beginnings of support for UTF-8 character strings.
2748 8. Arranged for the Makefile to pass over the settings of CC, CFLAGS, and
2749 RANLIB to ./ltconfig so that they are used by libtool. I think these are all
2750 the relevant ones. (AR is not passed because ./ltconfig does its own figuring
2751 out for the ar command.)
2754 Version 3.2 12-May-00
2755 ---------------------
2757 This is purely a bug fixing release.
2759 1. If the pattern /((Z)+|A)*/ was matched agained ZABCDEFG it matched Z instead
2760 of ZA. This was just one example of several cases that could provoke this bug,
2761 which was introduced by change 9 of version 2.00. The code for breaking
2762 infinite loops after an iteration that matches an empty string was't working
2763 correctly.
2765 2. The pcretest program was not imitating Perl correctly for the pattern /a*/g
2766 when matched against abbab (for example). After matching an empty string, it
2767 wasn't forcing anchoring when setting PCRE_NOTEMPTY for the next attempt; this
2768 caused it to match further down the string than it should.
2770 3. The code contained an inclusion of sys/types.h. It isn't clear why this
2771 was there because it doesn't seem to be needed, and it causes trouble on some
2772 systems, as it is not a Standard C header. It has been removed.
2774 4. Made 4 silly changes to the source to avoid stupid compiler warnings that
2775 were reported on the Macintosh. The changes were from
2777 while ((c = *(++ptr)) != 0 && c != '\n');
2778 to
2779 while ((c = *(++ptr)) != 0 && c != '\n') ;
2781 Totally extraordinary, but if that's what it takes...
2783 5. PCRE is being used in one environment where neither memmove() nor bcopy() is
2784 available. Added HAVE_BCOPY and an autoconf test for it; if neither
2785 HAVE_MEMMOVE nor HAVE_BCOPY is set, use a built-in emulation function which
2786 assumes the way PCRE uses memmove() (always moving upwards).
2788 6. PCRE is being used in one environment where strchr() is not available. There
2789 was only one use in pcre.c, and writing it out to avoid strchr() probably gives
2790 faster code anyway.
2793 Version 3.1 09-Feb-00
2794 ---------------------
2796 The only change in this release is the fixing of some bugs in Makefile.in for
2797 the "install" target:
2799 (1) It was failing to install pcreposix.h.
2801 (2) It was overwriting the pcre.3 man page with the pcreposix.3 man page.
2804 Version 3.0 01-Feb-00
2805 ---------------------
2807 1. Add support for the /+ modifier to perltest (to output $` like it does in
2808 pcretest).
2810 2. Add support for the /g modifier to perltest.
2812 3. Fix pcretest so that it behaves even more like Perl for /g when the pattern
2813 matches null strings.
2815 4. Fix perltest so that it doesn't do unwanted things when fed an empty
2816 pattern. Perl treats empty patterns specially - it reuses the most recent
2817 pattern, which is not what we want. Replace // by /(?#)/ in order to avoid this
2818 effect.
2820 5. The POSIX interface was broken in that it was just handing over the POSIX
2821 captured string vector to pcre_exec(), but (since release 2.00) PCRE has
2822 required a bigger vector, with some working space on the end. This means that
2823 the POSIX wrapper now has to get and free some memory, and copy the results.
2825 6. Added some simple autoconf support, placing the test data and the
2826 documentation in separate directories, re-organizing some of the
2827 information files, and making it build pcre-config (a GNU standard). Also added
2828 libtool support for building PCRE as a shared library, which is now the
2829 default.
2831 7. Got rid of the leading zero in the definition of PCRE_MINOR because 08 and
2832 09 are not valid octal constants. Single digits will be used for minor values
2833 less than 10.
2835 8. Defined REG_EXTENDED and REG_NOSUB as zero in the POSIX header, so that
2836 existing programs that set these in the POSIX interface can use PCRE without
2837 modification.
2839 9. Added a new function, pcre_fullinfo() with an extensible interface. It can
2840 return all that pcre_info() returns, plus additional data. The pcre_info()
2841 function is retained for compatibility, but is considered to be obsolete.
2843 10. Added experimental recursion feature (?R) to handle one common case that
2844 Perl 5.6 will be able to do with (?p{...}).
2846 11. Added support for POSIX character classes like [:alpha:], which Perl is
2847 adopting.
2850 Version 2.08 31-Aug-99
2851 ----------------------
2853 1. When startoffset was not zero and the pattern began with ".*", PCRE was not
2854 trying to match at the startoffset position, but instead was moving forward to
2855 the next newline as if a previous match had failed.
2857 2. pcretest was not making use of PCRE_NOTEMPTY when repeating for /g and /G,
2858 and could get into a loop if a null string was matched other than at the start
2859 of the subject.
2861 3. Added definitions of PCRE_MAJOR and PCRE_MINOR to pcre.h so the version can
2862 be distinguished at compile time, and for completeness also added PCRE_DATE.
2864 5. Added Paul Sokolovsky's minor changes to make it easy to compile a Win32 DLL
2865 in GnuWin32 environments.
2868 Version 2.07 29-Jul-99
2869 ----------------------
2871 1. The documentation is now supplied in plain text form and HTML as well as in
2872 the form of man page sources.
2874 2. C++ compilers don't like assigning (void *) values to other pointer types.
2875 In particular this affects malloc(). Although there is no problem in Standard
2876 C, I've put in casts to keep C++ compilers happy.
2878 3. Typo on pcretest.c; a cast of (unsigned char *) in the POSIX regexec() call
2879 should be (const char *).
2881 4. If NOPOSIX is defined, pcretest.c compiles without POSIX support. This may
2882 be useful for non-Unix systems who don't want to bother with the POSIX stuff.
2883 However, I haven't made this a standard facility. The documentation doesn't
2884 mention it, and the Makefile doesn't support it.
2886 5. The Makefile now contains an "install" target, with editable destinations at
2887 the top of the file. The pcretest program is not installed.
2889 6. pgrep -V now gives the PCRE version number and date.
2891 7. Fixed bug: a zero repetition after a literal string (e.g. /abcde{0}/) was
2892 causing the entire string to be ignored, instead of just the last character.
2894 8. If a pattern like /"([^\\"]+|\\.)*"/ is applied in the normal way to a
2895 non-matching string, it can take a very, very long time, even for strings of
2896 quite modest length, because of the nested recursion. PCRE now does better in
2897 some of these cases. It does this by remembering the last required literal
2898 character in the pattern, and pre-searching the subject to ensure it is present
2899 before running the real match. In other words, it applies a heuristic to detect
2900 some types of certain failure quickly, and in the above example, if presented
2901 with a string that has no trailing " it gives "no match" very quickly.
2903 9. A new runtime option PCRE_NOTEMPTY causes null string matches to be ignored;
2904 other alternatives are tried instead.
2907 Version 2.06 09-Jun-99
2908 ----------------------
2910 1. Change pcretest's output for amount of store used to show just the code
2911 space, because the remainder (the data block) varies in size between 32-bit and
2912 64-bit systems.
2914 2. Added an extra argument to pcre_exec() to supply an offset in the subject to
2915 start matching at. This allows lookbehinds to work when searching for multiple
2916 occurrences in a string.
2918 3. Added additional options to pcretest for testing multiple occurrences:
2920 /+ outputs the rest of the string that follows a match
2921 /g loops for multiple occurrences, using the new startoffset argument
2922 /G loops for multiple occurrences by passing an incremented pointer
2924 4. PCRE wasn't doing the "first character" optimization for patterns starting
2925 with \b or \B, though it was doing it for other lookbehind assertions. That is,
2926 it wasn't noticing that a match for a pattern such as /\bxyz/ has to start with
2927 the letter 'x'. On long subject strings, this gives a significant speed-up.
2930 Version 2.05 21-Apr-99
2931 ----------------------
2933 1. Changed the type of magic_number from int to long int so that it works
2934 properly on 16-bit systems.
2936 2. Fixed a bug which caused patterns starting with .* not to work correctly
2937 when the subject string contained newline characters. PCRE was assuming
2938 anchoring for such patterns in all cases, which is not correct because .* will
2939 not pass a newline unless PCRE_DOTALL is set. It now assumes anchoring only if
2940 DOTALL is set at top level; otherwise it knows that patterns starting with .*
2941 must be retried after every newline in the subject.
2944 Version 2.04 18-Feb-99
2945 ----------------------
2947 1. For parenthesized subpatterns with repeats whose minimum was zero, the
2948 computation of the store needed to hold the pattern was incorrect (too large).
2949 If such patterns were nested a few deep, this could multiply and become a real
2950 problem.
2952 2. Added /M option to pcretest to show the memory requirement of a specific
2953 pattern. Made -m a synonym of -s (which does this globally) for compatibility.
2955 3. Subpatterns of the form (regex){n,m} (i.e. limited maximum) were being
2956 compiled in such a way that the backtracking after subsequent failure was
2957 pessimal. Something like (a){0,3} was compiled as (a)?(a)?(a)? instead of
2958 ((a)((a)(a)?)?)? with disastrous performance if the maximum was of any size.
2961 Version 2.03 02-Feb-99
2962 ----------------------
2964 1. Fixed typo and small mistake in man page.
2966 2. Added 4th condition (GPL supersedes if conflict) and created separate
2967 LICENCE file containing the conditions.
2969 3. Updated pcretest so that patterns such as /abc\/def/ work like they do in
2970 Perl, that is the internal \ allows the delimiter to be included in the
2971 pattern. Locked out the use of \ as a delimiter. If \ immediately follows
2972 the final delimiter, add \ to the end of the pattern (to test the error).
2974 4. Added the convenience functions for extracting substrings after a successful
2975 match. Updated pcretest to make it able to test these functions.
2978 Version 2.02 14-Jan-99
2979 ----------------------
2981 1. Initialized the working variables associated with each extraction so that
2982 their saving and restoring doesn't refer to uninitialized store.
2984 2. Put dummy code into study.c in order to trick the optimizer of the IBM C
2985 compiler for OS/2 into generating correct code. Apparently IBM isn't going to
2986 fix the problem.
2988 3. Pcretest: the timing code wasn't using LOOPREPEAT for timing execution
2989 calls, and wasn't printing the correct value for compiling calls. Increased the
2990 default value of LOOPREPEAT, and the number of significant figures in the
2991 times.
2993 4. Changed "/bin/rm" in the Makefile to "-rm" so it works on Windows NT.
2995 5. Renamed "deftables" as "dftables" to get it down to 8 characters, to avoid
2996 a building problem on Windows NT with a FAT file system.
2999 Version 2.01 21-Oct-98
3000 ----------------------
3002 1. Changed the API for pcre_compile() to allow for the provision of a pointer
3003 to character tables built by pcre_maketables() in the current locale. If NULL
3004 is passed, the default tables are used.
3007 Version 2.00 24-Sep-98
3008 ----------------------
3010 1. Since the (>?) facility is in Perl 5.005, don't require PCRE_EXTRA to enable
3011 it any more.
3013 2. Allow quantification of (?>) groups, and make it work correctly.
3015 3. The first character computation wasn't working for (?>) groups.
3017 4. Correct the implementation of \Z (it is permitted to match on the \n at the
3018 end of the subject) and add 5.005's \z, which really does match only at the
3019 very end of the subject.
3021 5. Remove the \X "cut" facility; Perl doesn't have it, and (?> is neater.
3023 6. Remove the ability to specify CASELESS, MULTILINE, DOTALL, and
3024 DOLLAR_END_ONLY at runtime, to make it possible to implement the Perl 5.005
3025 localized options. All options to pcre_study() were also removed.
3027 7. Add other new features from 5.005:
3029 $(?<= positive lookbehind
3030 $(?<! negative lookbehind
3031 (?imsx-imsx) added the unsetting capability
3032 such a setting is global if at outer level; local otherwise
3033 (?imsx-imsx:) non-capturing groups with option setting
3034 (?(cond)re|re) conditional pattern matching
3036 A backreference to itself in a repeated group matches the previous
3037 captured string.
3039 8. General tidying up of studying (both automatic and via "study")
3040 consequential on the addition of new assertions.
3042 9. As in 5.005, unlimited repeated groups that could match an empty substring
3043 are no longer faulted at compile time. Instead, the loop is forcibly broken at
3044 runtime if any iteration does actually match an empty substring.
3046 10. Include the RunTest script in the distribution.
3048 11. Added tests from the Perl 5.005_02 distribution. This showed up a few
3049 discrepancies, some of which were old and were also with respect to 5.004. They
3050 have now been fixed.
3053 Version 1.09 28-Apr-98
3054 ----------------------
3056 1. A negated single character class followed by a quantifier with a minimum
3057 value of one (e.g. [^x]{1,6} ) was not compiled correctly. This could lead to
3058 program crashes, or just wrong answers. This did not apply to negated classes
3059 containing more than one character, or to minima other than one.
3062 Version 1.08 27-Mar-98
3063 ----------------------
3065 1. Add PCRE_UNGREEDY to invert the greediness of quantifiers.
3067 2. Add (?U) and (?X) to set PCRE_UNGREEDY and PCRE_EXTRA respectively. The
3068 latter must appear before anything that relies on it in the pattern.
3071 Version 1.07 16-Feb-98
3072 ----------------------
3074 1. A pattern such as /((a)*)*/ was not being diagnosed as in error (unlimited
3075 repeat of a potentially empty string).
3078 Version 1.06 23-Jan-98
3079 ----------------------
3081 1. Added Markus Oberhumer's little patches for C++.
3083 2. Literal strings longer than 255 characters were broken.
3086 Version 1.05 23-Dec-97
3087 ----------------------
3089 1. Negated character classes containing more than one character were failing if
3090 PCRE_CASELESS was set at run time.
3093 Version 1.04 19-Dec-97
3094 ----------------------
3096 1. Corrected the man page, where some "const" qualifiers had been omitted.
3098 2. Made debugging output print "{0,xxx}" instead of just "{,xxx}" to agree with
3099 input syntax.
3101 3. Fixed memory leak which occurred when a regex with back references was
3102 matched with an offsets vector that wasn't big enough. The temporary memory
3103 that is used in this case wasn't being freed if the match failed.
3105 4. Tidied pcretest to ensure it frees memory that it gets.
3107 5. Temporary memory was being obtained in the case where the passed offsets
3108 vector was exactly big enough.
3110 6. Corrected definition of offsetof() from change 5 below.
3112 7. I had screwed up change 6 below and broken the rules for the use of
3113 setjmp(). Now fixed.
3116 Version 1.03 18-Dec-97
3117 ----------------------
3119 1. A erroneous regex with a missing opening parenthesis was correctly
3120 diagnosed, but PCRE attempted to access brastack[-1], which could cause crashes
3121 on some systems.
3123 2. Replaced offsetof(real_pcre, code) by offsetof(real_pcre, code[0]) because
3124 it was reported that one broken compiler failed on the former because "code" is
3125 also an independent variable.
3127 3. The erroneous regex a[]b caused an array overrun reference.
3129 4. A regex ending with a one-character negative class (e.g. /[^k]$/) did not
3130 fail on data ending with that character. (It was going on too far, and checking
3131 the next character, typically a binary zero.) This was specific to the
3132 optimized code for single-character negative classes.
3134 5. Added a contributed patch from the TIN world which does the following:
3136 + Add an undef for memmove, in case the the system defines a macro for it.
3138 + Add a definition of offsetof(), in case there isn't one. (I don't know
3139 the reason behind this - offsetof() is part of the ANSI standard - but
3140 it does no harm).
3142 + Reduce the ifdef's in pcre.c using macro DPRINTF, thereby eliminating
3143 most of the places where whitespace preceded '#'. I have given up and
3144 allowed the remaining 2 cases to be at the margin.
3146 + Rename some variables in pcre to eliminate shadowing. This seems very
3147 pedantic, but does no harm, of course.
3149 6. Moved the call to setjmp() into its own function, to get rid of warnings
3150 from gcc -Wall, and avoided calling it at all unless PCRE_EXTRA is used.
3152 7. Constructs such as \d{8,} were compiling into the equivalent of
3153 \d{8}\d{0,65527} instead of \d{8}\d* which didn't make much difference to the
3154 outcome, but in this particular case used more store than had been allocated,
3155 which caused the bug to be discovered because it threw up an internal error.
3157 8. The debugging code in both pcre and pcretest for outputting the compiled
3158 form of a regex was going wrong in the case of back references followed by
3159 curly-bracketed repeats.
3162 Version 1.02 12-Dec-97
3163 ----------------------
3165 1. Typos in pcre.3 and comments in the source fixed.
3167 2. Applied a contributed patch to get rid of places where it used to remove
3168 'const' from variables, and fixed some signed/unsigned and uninitialized
3169 variable warnings.
3171 3. Added the "runtest" target to Makefile.
3173 4. Set default compiler flag to -O2 rather than just -O.
3176 Version 1.01 19-Nov-97
3177 ----------------------
3179 1. PCRE was failing to diagnose unlimited repeat of empty string for patterns
3180 like /([ab]*)*/, that is, for classes with more than one character in them.
3182 2. Likewise, it wasn't diagnosing patterns with "once-only" subpatterns, such
3183 as /((?>a*))*/ (a PCRE_EXTRA facility).
3186 Version 1.00 18-Nov-97
3187 ----------------------
3189 1. Added compile-time macros to support systems such as SunOS4 which don't have
3190 memmove() or strerror() but have other things that can be used instead.
3192 2. Arranged that "make clean" removes the executables.
3195 Version 0.99 27-Oct-97
3196 ----------------------
3198 1. Fixed bug in code for optimizing classes with only one character. It was
3199 initializing a 32-byte map regardless, which could cause it to run off the end
3200 of the memory it had got.
3202 2. Added, conditional on PCRE_EXTRA, the proposed (?>REGEX) construction.
3205 Version 0.98 22-Oct-97
3206 ----------------------
3208 1. Fixed bug in code for handling temporary memory usage when there are more
3209 back references than supplied space in the ovector. This could cause segfaults.
3212 Version 0.97 21-Oct-97
3213 ----------------------
3215 1. Added the \X "cut" facility, conditional on PCRE_EXTRA.
3217 2. Optimized negated single characters not to use a bit map.
3219 3. Brought error texts together as macro definitions; clarified some of them;
3220 fixed one that was wrong - it said "range out of order" when it meant "invalid
3221 escape sequence".
3223 4. Changed some char * arguments to const char *.
3225 5. Added PCRE_NOTBOL and PCRE_NOTEOL (from POSIX).
3227 6. Added the POSIX-style API wrapper in pcreposix.a and testing facilities in
3228 pcretest.
3231 Version 0.96 16-Oct-97
3232 ----------------------
3234 1. Added a simple "pgrep" utility to the distribution.
3236 2. Fixed an incompatibility with Perl: "{" is now treated as a normal character
3237 unless it appears in one of the precise forms "{ddd}", "{ddd,}", or "{ddd,ddd}"
3238 where "ddd" means "one or more decimal digits".
3240 3. Fixed serious bug. If a pattern had a back reference, but the call to
3241 pcre_exec() didn't supply a large enough ovector to record the related
3242 identifying subpattern, the match always failed. PCRE now remembers the number
3243 of the largest back reference, and gets some temporary memory in which to save
3244 the offsets during matching if necessary, in order to ensure that
3245 backreferences always work.
3247 4. Increased the compatibility with Perl in a number of ways:
3249 (a) . no longer matches \n by default; an option PCRE_DOTALL is provided
3250 to request this handling. The option can be set at compile or exec time.
3252 (b) $ matches before a terminating newline by default; an option
3253 PCRE_DOLLAR_ENDONLY is provided to override this (but not in multiline
3254 mode). The option can be set at compile or exec time.
3256 (c) The handling of \ followed by a digit other than 0 is now supposed to be
3257 the same as Perl's. If the decimal number it represents is less than 10
3258 or there aren't that many previous left capturing parentheses, an octal
3259 escape is read. Inside a character class, it's always an octal escape,
3260 even if it is a single digit.
3262 (d) An escaped but undefined alphabetic character is taken as a literal,
3263 unless PCRE_EXTRA is set. Currently this just reserves the remaining
3264 escapes.
3266 (e) {0} is now permitted. (The previous item is removed from the compiled
3267 pattern).
3269 5. Changed all the names of code files so that the basic parts are no longer
3270 than 10 characters, and abolished the teeny "globals.c" file.
3272 6. Changed the handling of character classes; they are now done with a 32-byte
3273 bit map always.
3275 7. Added the -d and /D options to pcretest to make it possible to look at the
3276 internals of compilation without having to recompile pcre.
3279 Version 0.95 23-Sep-97
3280 ----------------------
3282 1. Fixed bug in pre-pass concerning escaped "normal" characters such as \x5c or
3283 \x20 at the start of a run of normal characters. These were being treated as
3284 real characters, instead of the source characters being re-checked.
3287 Version 0.94 18-Sep-97
3288 ----------------------
3290 1. The functions are now thread-safe, with the caveat that the global variables
3291 containing pointers to malloc() and free() or alternative functions are the
3292 same for all threads.
3294 2. Get pcre_study() to generate a bitmap of initial characters for non-
3295 anchored patterns when this is possible, and use it if passed to pcre_exec().
3298 Version 0.93 15-Sep-97
3299 ----------------------
3301 1. /(b)|(:+)/ was computing an incorrect first character.
3303 2. Add pcre_study() to the API and the passing of pcre_extra to pcre_exec(),
3304 but not actually doing anything yet.
3306 3. Treat "-" characters in classes that cannot be part of ranges as literals,
3307 as Perl does (e.g. [-az] or [az-]).
3309 4. Set the anchored flag if a branch starts with .* or .*? because that tests
3310 all possible positions.
3312 5. Split up into different modules to avoid including unneeded functions in a
3313 compiled binary. However, compile and exec are still in one module. The "study"
3314 function is split off.
3316 6. The character tables are now in a separate module whose source is generated
3317 by an auxiliary program - but can then be edited by hand if required. There are
3318 now no calls to isalnum(), isspace(), isdigit(), isxdigit(), tolower() or
3319 toupper() in the code.
3321 7. Turn the malloc/free funtions variables into pcre_malloc and pcre_free and
3322 make them global. Abolish the function for setting them, as the caller can now
3323 set them directly.
3326 Version 0.92 11-Sep-97
3327 ----------------------
3329 1. A repeat with a fixed maximum and a minimum of 1 for an ordinary character
3330 (e.g. /a{1,3}/) was broken (I mis-optimized it).
3332 2. Caseless matching was not working in character classes if the characters in
3333 the pattern were in upper case.
3335 3. Make ranges like [W-c] work in the same way as Perl for caseless matching.
3337 4. Make PCRE_ANCHORED public and accept as a compile option.
3339 5. Add an options word to pcre_exec() and accept PCRE_ANCHORED and
3340 PCRE_CASELESS at run time. Add escapes \A and \I to pcretest to cause it to
3341 pass them.
3343 6. Give an error if bad option bits passed at compile or run time.
3345 7. Add PCRE_MULTILINE at compile and exec time, and (?m) as well. Add \M to
3346 pcretest to cause it to pass that flag.
3348 8. Add pcre_info(), to get the number of identifying subpatterns, the stored
3349 options, and the first character, if set.
3351 9. Recognize C+ or C{n,m} where n >= 1 as providing a fixed starting character.
3354 Version 0.91 10-Sep-97
3355 ----------------------
3357 1. PCRE was failing to diagnose unlimited repeats of subpatterns that could
3358 match the empty string as in /(a*)*/. It was looping and ultimately crashing.
3360 2. PCRE was looping on encountering an indefinitely repeated back reference to
3361 a subpattern that had matched an empty string, e.g. /(a|)\1*/. It now does what
3362 Perl does - treats the match as successful.
3364 ****


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

  ViewVC Help
Powered by ViewVC 1.1.5