ViewVC logotype

Contents of /code/trunk/ChangeLog

Parent Directory Parent Directory | Revision Log Revision Log

Revision 354 - (show annotations)
Mon Jul 7 16:30:33 2008 UTC (13 years ago) by ph10
File size: 157108 byte(s)
Fix caseless backreferences for non-ASCII characters.
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.
26 Version 7.7 07-May-08
27 ---------------------
29 1. Applied Craig's patch to sort out a long long problem: "If we can't convert
30 a string to a long long, pretend we don't even have a long long." This is
31 done by checking for the strtoq, strtoll, and _strtoi64 functions.
33 2. Applied Craig's patch to pcrecpp.cc to restore ABI compatibility with
34 pre-7.6 versions, which defined a global no_arg variable instead of putting
35 it in the RE class. (See also #8 below.)
37 3. Remove a line of dead code, identified by coverity and reported by Nuno
38 Lopes.
40 4. Fixed two related pcregrep bugs involving -r with --include or --exclude:
42 (1) The include/exclude patterns were being applied to the whole pathnames
43 of files, instead of just to the final components.
45 (2) If there was more than one level of directory, the subdirectories were
46 skipped unless they satisfied the include/exclude conditions. This is
47 inconsistent with GNU grep (and could even be seen as contrary to the
48 pcregrep specification - which I improved to make it absolutely clear).
49 The action now is always to scan all levels of directory, and just
50 apply the include/exclude patterns to regular files.
52 5. Added the --include_dir and --exclude_dir patterns to pcregrep, and used
53 --exclude_dir in the tests to avoid scanning .svn directories.
55 6. Applied Craig's patch to the QuoteMeta function so that it escapes the
56 NUL character as backslash + 0 rather than backslash + NUL, because PCRE
57 doesn't support NULs in patterns.
59 7. Added some missing "const"s to declarations of static tables in
60 pcre_compile.c and pcre_dfa_exec.c.
62 8. Applied Craig's patch to pcrecpp.cc to fix a problem in OS X that was
63 caused by fix #2 above. (Subsequently also a second patch to fix the
64 first patch. And a third patch - this was a messy problem.)
66 9. Applied Craig's patch to remove the use of push_back().
68 10. Applied Alan Lehotsky's patch to add REG_STARTEND support to the POSIX
69 matching function regexec().
71 11. Added support for the Oniguruma syntax \g<name>, \g<n>, \g'name', \g'n',
72 which, however, unlike Perl's \g{...}, are subroutine calls, not back
73 references. PCRE supports relative numbers with this syntax (I don't think
74 Oniguruma does).
76 12. Previously, a group with a zero repeat such as (...){0} was completely
77 omitted from the compiled regex. However, this means that if the group
78 was called as a subroutine from elsewhere in the pattern, things went wrong
79 (an internal error was given). Such groups are now left in the compiled
80 pattern, with a new opcode that causes them to be skipped at execution
81 time.
83 13. Added the PCRE_JAVASCRIPT_COMPAT option. This makes the following changes
84 to the way PCRE behaves:
86 (a) A lone ] character is dis-allowed (Perl treats it as data).
88 (b) A back reference to an unmatched subpattern matches an empty string
89 (Perl fails the current match path).
91 (c) A data ] in a character class must be notated as \] because if the
92 first data character in a class is ], it defines an empty class. (In
93 Perl it is not possible to have an empty class.) The empty class []
94 never matches; it forces failure and is equivalent to (*FAIL) or (?!).
95 The negative empty class [^] matches any one character, independently
96 of the DOTALL setting.
98 14. A pattern such as /(?2)[]a()b](abc)/ which had a forward reference to a
99 non-existent subpattern following a character class starting with ']' and
100 containing () gave an internal compiling error instead of "reference to
101 non-existent subpattern". Fortunately, when the pattern did exist, the
102 compiled code was correct. (When scanning forwards to check for the
103 existencd of the subpattern, it was treating the data ']' as terminating
104 the class, so got the count wrong. When actually compiling, the reference
105 was subsequently set up correctly.)
107 15. The "always fail" assertion (?!) is optimzed to (*FAIL) by pcre_compile;
108 it was being rejected as not supported by pcre_dfa_exec(), even though
109 other assertions are supported. I have made pcre_dfa_exec() support
110 (*FAIL).
112 16. The implementation of 13c above involved the invention of a new opcode,
113 OP_ALLANY, which is like OP_ANY but doesn't check the /s flag. Since /s
114 cannot be changed at match time, I realized I could make a small
115 improvement to matching performance by compiling OP_ALLANY instead of
116 OP_ANY for "." when DOTALL was set, and then removing the runtime tests
117 on the OP_ANY path.
119 17. Compiling pcretest on Windows with readline support failed without the
120 following two fixes: (1) Make the unistd.h include conditional on
121 HAVE_UNISTD_H; (2) #define isatty and fileno as _isatty and _fileno.
123 18. Changed CMakeLists.txt and cmake/FindReadline.cmake to arrange for the
124 ncurses library to be included for pcretest when ReadLine support is
125 requested, but also to allow for it to be overridden. This patch came from
126 Daniel Bergström.
128 19. There was a typo in the file ucpinternal.h where f0_rangeflag was defined
129 as 0x00f00000 instead of 0x00800000. Luckily, this would not have caused
130 any errors with the current Unicode tables. Thanks to Peter Kankowski for
131 spotting this.
134 Version 7.6 28-Jan-08
135 ---------------------
137 1. A character class containing a very large number of characters with
138 codepoints greater than 255 (in UTF-8 mode, of course) caused a buffer
139 overflow.
141 2. Patch to cut out the "long long" test in pcrecpp_unittest when
142 HAVE_LONG_LONG is not defined.
144 3. Applied Christian Ehrlicher's patch to update the CMake build files to
145 bring them up to date and include new features. This patch includes:
147 - Fixed PH's badly added libz and libbz2 support.
148 - Fixed a problem with static linking.
149 - Added pcredemo. [But later removed - see 7 below.]
150 - Fixed dftables problem and added an option.
151 - Added a number of HAVE_XXX tests, including HAVE_WINDOWS_H and
153 - Added readline support for pcretest.
154 - Added an listing of the option settings after cmake has run.
156 4. A user submitted a patch to Makefile that makes it easy to create
157 "pcre.dll" under mingw when using Configure/Make. I added stuff to
158 Makefile.am that cause it to include this special target, without
159 affecting anything else. Note that the same mingw target plus all
160 the other distribution libraries and programs are now supported
161 when configuring with CMake (see 6 below) instead of with
162 Configure/Make.
164 5. Applied Craig's patch that moves no_arg into the RE class in the C++ code.
165 This is an attempt to solve the reported problem "pcrecpp::no_arg is not
166 exported in the Windows port". It has not yet been confirmed that the patch
167 solves the problem, but it does no harm.
169 6. Applied Sheri's patch to CMakeLists.txt to add NON_STANDARD_LIB_PREFIX and
170 NON_STANDARD_LIB_SUFFIX for dll names built with mingw when configured
171 with CMake, and also correct the comment about stack recursion.
173 7. Remove the automatic building of pcredemo from the ./configure system and
174 from CMakeLists.txt. The whole idea of pcredemo.c is that it is an example
175 of a program that users should build themselves after PCRE is installed, so
176 building it automatically is not really right. What is more, it gave
177 trouble in some build environments.
179 8. Further tidies to CMakeLists.txt from Sheri and Christian.
182 Version 7.5 10-Jan-08
183 ---------------------
185 1. Applied a patch from Craig: "This patch makes it possible to 'ignore'
186 values in parens when parsing an RE using the C++ wrapper."
188 2. Negative specials like \S did not work in character classes in UTF-8 mode.
189 Characters greater than 255 were excluded from the class instead of being
190 included.
192 3. The same bug as (2) above applied to negated POSIX classes such as
193 [:^space:].
195 4. PCRECPP_STATIC was referenced in pcrecpp_internal.h, but nowhere was it
196 defined or documented. It seems to have been a typo for PCRE_STATIC, so
197 I have changed it.
199 5. The construct (?&) was not diagnosed as a syntax error (it referenced the
200 first named subpattern) and a construct such as (?&a) would reference the
201 first named subpattern whose name started with "a" (in other words, the
202 length check was missing). Both these problems are fixed. "Subpattern name
203 expected" is now given for (?&) (a zero-length name), and this patch also
204 makes it give the same error for \k'' (previously it complained that that
205 was a reference to a non-existent subpattern).
207 6. The erroneous patterns (?+-a) and (?-+a) give different error messages;
208 this is right because (?- can be followed by option settings as well as by
209 digits. I have, however, made the messages clearer.
211 7. Patterns such as (?(1)a|b) (a pattern that contains fewer subpatterns
212 than the number used in the conditional) now cause a compile-time error.
213 This is actually not compatible with Perl, which accepts such patterns, but
214 treats the conditional as always being FALSE (as PCRE used to), but it
215 seems to me that giving a diagnostic is better.
217 8. Change "alphameric" to the more common word "alphanumeric" in comments
218 and messages.
220 9. Fix two occurrences of "backslash" in comments that should have been
221 "backspace".
223 10. Remove two redundant lines of code that can never be obeyed (their function
224 was moved elsewhere).
226 11. The program that makes PCRE's Unicode character property table had a bug
227 which caused it to generate incorrect table entries for sequences of
228 characters that have the same character type, but are in different scripts.
229 It amalgamated them into a single range, with the script of the first of
230 them. In other words, some characters were in the wrong script. There were
231 thirteen such cases, affecting characters in the following ranges:
233 U+002b0 - U+002c1
234 U+0060c - U+0060d
235 U+0061e - U+00612
236 U+0064b - U+0065e
237 U+0074d - U+0076d
238 U+01800 - U+01805
239 U+01d00 - U+01d77
240 U+01d9b - U+01dbf
241 U+0200b - U+0200f
242 U+030fc - U+030fe
243 U+03260 - U+0327f
244 U+0fb46 - U+0fbb1
245 U+10450 - U+1049d
247 12. The -o option (show only the matching part of a line) for pcregrep was not
248 compatible with GNU grep in that, if there was more than one match in a
249 line, it showed only the first of them. It now behaves in the same way as
250 GNU grep.
252 13. If the -o and -v options were combined for pcregrep, it printed a blank
253 line for every non-matching line. GNU grep prints nothing, and pcregrep now
254 does the same. The return code can be used to tell if there were any
255 non-matching lines.
257 14. Added --file-offsets and --line-offsets to pcregrep.
259 15. The pattern (?=something)(?R) was not being diagnosed as a potentially
260 infinitely looping recursion. The bug was that positive lookaheads were not
261 being skipped when checking for a possible empty match (negative lookaheads
262 and both kinds of lookbehind were skipped).
264 16. Fixed two typos in the Windows-only code in pcregrep.c, and moved the
265 inclusion of <windows.h> to before rather than after the definition of
266 INVALID_FILE_ATTRIBUTES (patch from David Byron).
268 17. Specifying a possessive quantifier with a specific limit for a Unicode
269 character property caused pcre_compile() to compile bad code, which led at
270 runtime to PCRE_ERROR_INTERNAL (-14). Examples of patterns that caused this
271 are: /\p{Zl}{2,3}+/8 and /\p{Cc}{2}+/8. It was the possessive "+" that
272 caused the error; without that there was no problem.
274 18. Added --enable-pcregrep-libz and --enable-pcregrep-libbz2.
276 19. Added --enable-pcretest-libreadline.
278 20. In pcrecpp.cc, the variable 'count' was incremented twice in
279 RE::GlobalReplace(). As a result, the number of replacements returned was
280 double what it should be. I removed one of the increments, but Craig sent a
281 later patch that removed the other one (the right fix) and added unit tests
282 that check the return values (which was not done before).
284 21. Several CMake things:
286 (1) Arranged that, when cmake is used on Unix, the libraries end up with
287 the names libpcre and libpcreposix, not just pcre and pcreposix.
289 (2) The above change means that pcretest and pcregrep are now correctly
290 linked with the newly-built libraries, not previously installed ones.
294 22. In UTF-8 mode, with newline set to "any", a pattern such as .*a.*=.b.*
295 crashed when matching a string such as a\x{2029}b (note that \x{2029} is a
296 UTF-8 newline character). The key issue is that the pattern starts .*;
297 this means that the match must be either at the beginning, or after a
298 newline. The bug was in the code for advancing after a failed match and
299 checking that the new position followed a newline. It was not taking
300 account of UTF-8 characters correctly.
302 23. PCRE was behaving differently from Perl in the way it recognized POSIX
303 character classes. PCRE was not treating the sequence [:...:] as a
304 character class unless the ... were all letters. Perl, however, seems to
305 allow any characters between [: and :], though of course it rejects as
306 unknown any "names" that contain non-letters, because all the known class
307 names consist only of letters. Thus, Perl gives an error for [[:1234:]],
308 for example, whereas PCRE did not - it did not recognize a POSIX character
309 class. This seemed a bit dangerous, so the code has been changed to be
310 closer to Perl. The behaviour is not identical to Perl, because PCRE will
311 diagnose an unknown class for, for example, [[:l\ower:]] where Perl will
312 treat it as [[:lower:]]. However, PCRE does now give "unknown" errors where
313 Perl does, and where it didn't before.
315 24. Rewrite so as to remove the single use of %n from pcregrep because in some
316 Windows environments %n is disabled by default.
319 Version 7.4 21-Sep-07
320 ---------------------
322 1. Change 7.3/28 was implemented for classes by looking at the bitmap. This
323 means that a class such as [\s] counted as "explicit reference to CR or
324 LF". That isn't really right - the whole point of the change was to try to
325 help when there was an actual mention of one of the two characters. So now
326 the change happens only if \r or \n (or a literal CR or LF) character is
327 encountered.
329 2. The 32-bit options word was also used for 6 internal flags, but the numbers
330 of both had grown to the point where there were only 3 bits left.
331 Fortunately, there was spare space in the data structure, and so I have
332 moved the internal flags into a new 16-bit field to free up more option
333 bits.
335 3. The appearance of (?J) at the start of a pattern set the DUPNAMES option,
336 but did not set the internal JCHANGED flag - either of these is enough to
337 control the way the "get" function works - but the PCRE_INFO_JCHANGED
338 facility is supposed to tell if (?J) was ever used, so now (?J) at the
339 start sets both bits.
341 4. Added options (at build time, compile time, exec time) to change \R from
342 matching any Unicode line ending sequence to just matching CR, LF, or CRLF.
344 5. doc/pcresyntax.html was missing from the distribution.
346 6. Put back the definition of PCRE_ERROR_NULLWSLIMIT, for backward
347 compatibility, even though it is no longer used.
349 7. Added macro for snprintf to pcrecpp_unittest.cc and also for strtoll and
350 strtoull to pcrecpp.cc to select the available functions in WIN32 when the
351 windows.h file is present (where different names are used). [This was
352 reversed later after testing - see 16 below.]
354 8. Changed all #include <config.h> to #include "config.h". There were also
355 some further <pcre.h> cases that I changed to "pcre.h".
357 9. When pcregrep was used with the --colour option, it missed the line ending
358 sequence off the lines that it output.
360 10. It was pointed out to me that arrays of string pointers cause lots of
361 relocations when a shared library is dynamically loaded. A technique of
362 using a single long string with a table of offsets can drastically reduce
363 these. I have refactored PCRE in four places to do this. The result is
364 dramatic:
366 Originally: 290
367 After changing UCP table: 187
368 After changing error message table: 43
369 After changing table of "verbs" 36
370 After changing table of Posix names 22
372 Thanks to the folks working on Gregex for glib for this insight.
374 11. --disable-stack-for-recursion caused compiling to fail unless -enable-
375 unicode-properties was also set.
377 12. Updated the tests so that they work when \R is defaulted to ANYCRLF.
379 13. Added checks for ANY and ANYCRLF to pcrecpp.cc where it previously
380 checked only for CRLF.
382 14. Added casts to pcretest.c to avoid compiler warnings.
384 15. Added Craig's patch to various pcrecpp modules to avoid compiler warnings.
386 16. Added Craig's patch to remove the WINDOWS_H tests, that were not working,
387 and instead check for _strtoi64 explicitly, and avoid the use of snprintf()
388 entirely. This removes changes made in 7 above.
390 17. The CMake files have been updated, and there is now more information about
391 building with CMake in the NON-UNIX-USE document.
394 Version 7.3 28-Aug-07
395 ---------------------
397 1. In the rejigging of the build system that eventually resulted in 7.1, the
398 line "#include <pcre.h>" was included in pcre_internal.h. The use of angle
399 brackets there is not right, since it causes compilers to look for an
400 installed pcre.h, not the version that is in the source that is being
401 compiled (which of course may be different). I have changed it back to:
403 #include "pcre.h"
405 I have a vague recollection that the change was concerned with compiling in
406 different directories, but in the new build system, that is taken care of
407 by the VPATH setting the Makefile.
409 2. The pattern .*$ when run in not-DOTALL UTF-8 mode with newline=any failed
410 when the subject happened to end in the byte 0x85 (e.g. if the last
411 character was \x{1ec5}). *Character* 0x85 is one of the "any" newline
412 characters but of course it shouldn't be taken as a newline when it is part
413 of another character. The bug was that, for an unlimited repeat of . in
414 not-DOTALL UTF-8 mode, PCRE was advancing by bytes rather than by
415 characters when looking for a newline.
417 3. A small performance improvement in the DOTALL UTF-8 mode .* case.
419 4. Debugging: adjusted the names of opcodes for different kinds of parentheses
420 in debug output.
422 5. Arrange to use "%I64d" instead of "%lld" and "%I64u" instead of "%llu" for
423 long printing in the pcrecpp unittest when running under MinGW.
425 6. ESC_K was left out of the EBCDIC table.
427 7. Change 7.0/38 introduced a new limit on the number of nested non-capturing
428 parentheses; I made it 1000, which seemed large enough. Unfortunately, the
429 limit also applies to "virtual nesting" when a pattern is recursive, and in
430 this case 1000 isn't so big. I have been able to remove this limit at the
431 expense of backing off one optimization in certain circumstances. Normally,
432 when pcre_exec() would call its internal match() function recursively and
433 immediately return the result unconditionally, it uses a "tail recursion"
434 feature to save stack. However, when a subpattern that can match an empty
435 string has an unlimited repetition quantifier, it no longer makes this
436 optimization. That gives it a stack frame in which to save the data for
437 checking that an empty string has been matched. Previously this was taken
438 from the 1000-entry workspace that had been reserved. So now there is no
439 explicit limit, but more stack is used.
441 8. Applied Daniel's patches to solve problems with the import/export magic
442 syntax that is required for Windows, and which was going wrong for the
443 pcreposix and pcrecpp parts of the library. These were overlooked when this
444 problem was solved for the main library.
446 9. There were some crude static tests to avoid integer overflow when computing
447 the size of patterns that contain repeated groups with explicit upper
448 limits. As the maximum quantifier is 65535, the maximum group length was
449 set at 30,000 so that the product of these two numbers did not overflow a
450 32-bit integer. However, it turns out that people want to use groups that
451 are longer than 30,000 bytes (though not repeat them that many times).
452 Change 7.0/17 (the refactoring of the way the pattern size is computed) has
453 made it possible to implement the integer overflow checks in a much more
454 dynamic way, which I have now done. The artificial limitation on group
455 length has been removed - we now have only the limit on the total length of
456 the compiled pattern, which depends on the LINK_SIZE setting.
458 10. Fixed a bug in the documentation for get/copy named substring when
459 duplicate names are permitted. If none of the named substrings are set, the
460 functions return PCRE_ERROR_NOSUBSTRING (7); the doc said they returned an
461 empty string.
463 11. Because Perl interprets \Q...\E at a high level, and ignores orphan \E
464 instances, patterns such as [\Q\E] or [\E] or even [^\E] cause an error,
465 because the ] is interpreted as the first data character and the
466 terminating ] is not found. PCRE has been made compatible with Perl in this
467 regard. Previously, it interpreted [\Q\E] as an empty class, and [\E] could
468 cause memory overwriting.
470 10. Like Perl, PCRE automatically breaks an unlimited repeat after an empty
471 string has been matched (to stop an infinite loop). It was not recognizing
472 a conditional subpattern that could match an empty string if that
473 subpattern was within another subpattern. For example, it looped when
474 trying to match (((?(1)X|))*) but it was OK with ((?(1)X|)*) where the
475 condition was not nested. This bug has been fixed.
477 12. A pattern like \X?\d or \P{L}?\d in non-UTF-8 mode could cause a backtrack
478 past the start of the subject in the presence of bytes with the top bit
479 set, for example "\x8aBCD".
481 13. Added Perl 5.10 experimental backtracking controls (*FAIL), (*F), (*PRUNE),
482 (*SKIP), (*THEN), (*COMMIT), and (*ACCEPT).
484 14. Optimized (?!) to (*FAIL).
486 15. Updated the test for a valid UTF-8 string to conform to the later RFC 3629.
487 This restricts code points to be within the range 0 to 0x10FFFF, excluding
488 the "low surrogate" sequence 0xD800 to 0xDFFF. Previously, PCRE allowed the
489 full range 0 to 0x7FFFFFFF, as defined by RFC 2279. Internally, it still
490 does: it's just the validity check that is more restrictive.
492 16. Inserted checks for integer overflows during escape sequence (backslash)
493 processing, and also fixed erroneous offset values for syntax errors during
494 backslash processing.
496 17. Fixed another case of looking too far back in non-UTF-8 mode (cf 12 above)
497 for patterns like [\PPP\x8a]{1,}\x80 with the subject "A\x80".
499 18. An unterminated class in a pattern like (?1)\c[ with a "forward reference"
500 caused an overrun.
502 19. A pattern like (?:[\PPa*]*){8,} which had an "extended class" (one with
503 something other than just ASCII characters) inside a group that had an
504 unlimited repeat caused a loop at compile time (while checking to see
505 whether the group could match an empty string).
507 20. Debugging a pattern containing \p or \P could cause a crash. For example,
508 [\P{Any}] did so. (Error in the code for printing property names.)
510 21. An orphan \E inside a character class could cause a crash.
512 22. A repeated capturing bracket such as (A)? could cause a wild memory
513 reference during compilation.
515 23. There are several functions in pcre_compile() that scan along a compiled
516 expression for various reasons (e.g. to see if it's fixed length for look
517 behind). There were bugs in these functions when a repeated \p or \P was
518 present in the pattern. These operators have additional parameters compared
519 with \d, etc, and these were not being taken into account when moving along
520 the compiled data. Specifically:
522 (a) A item such as \p{Yi}{3} in a lookbehind was not treated as fixed
523 length.
525 (b) An item such as \pL+ within a repeated group could cause crashes or
526 loops.
528 (c) A pattern such as \p{Yi}+(\P{Yi}+)(?1) could give an incorrect
529 "reference to non-existent subpattern" error.
531 (d) A pattern like (\P{Yi}{2}\277)? could loop at compile time.
533 24. A repeated \S or \W in UTF-8 mode could give wrong answers when multibyte
534 characters were involved (for example /\S{2}/8g with "A\x{a3}BC").
536 25. Using pcregrep in multiline, inverted mode (-Mv) caused it to loop.
538 26. Patterns such as [\P{Yi}A] which include \p or \P and just one other
539 character were causing crashes (broken optimization).
541 27. Patterns such as (\P{Yi}*\277)* (group with possible zero repeat containing
542 \p or \P) caused a compile-time loop.
544 28. More problems have arisen in unanchored patterns when CRLF is a valid line
545 break. For example, the unstudied pattern [\r\n]A does not match the string
546 "\r\nA" because change 7.0/46 below moves the current point on by two
547 characters after failing to match at the start. However, the pattern \nA
548 *does* match, because it doesn't start till \n, and if [\r\n]A is studied,
549 the same is true. There doesn't seem any very clean way out of this, but
550 what I have chosen to do makes the common cases work: PCRE now takes note
551 of whether there can be an explicit match for \r or \n anywhere in the
552 pattern, and if so, 7.0/46 no longer applies. As part of this change,
553 there's a new PCRE_INFO_HASCRORLF option for finding out whether a compiled
554 pattern has explicit CR or LF references.
556 29. Added (*CR) etc for changing newline setting at start of pattern.
559 Version 7.2 19-Jun-07
560 ---------------------
562 1. If the fr_FR locale cannot be found for test 3, try the "french" locale,
563 which is apparently normally available under Windows.
565 2. Re-jig the pcregrep tests with different newline settings in an attempt
566 to make them independent of the local environment's newline setting.
568 3. Add code to configure.ac to remove -g from the CFLAGS default settings.
570 4. Some of the "internals" tests were previously cut out when the link size
571 was not 2, because the output contained actual offsets. The recent new
572 "Z" feature of pcretest means that these can be cut out, making the tests
573 usable with all link sizes.
575 5. Implemented Stan Switzer's goto replacement for longjmp() when not using
576 stack recursion. This gives a massive performance boost under BSD, but just
577 a small improvement under Linux. However, it saves one field in the frame
578 in all cases.
580 6. Added more features from the forthcoming Perl 5.10:
582 (a) (?-n) (where n is a string of digits) is a relative subroutine or
583 recursion call. It refers to the nth most recently opened parentheses.
585 (b) (?+n) is also a relative subroutine call; it refers to the nth next
586 to be opened parentheses.
588 (c) Conditions that refer to capturing parentheses can be specified
589 relatively, for example, (?(-2)... or (?(+3)...
591 (d) \K resets the start of the current match so that everything before
592 is not part of it.
594 (e) \k{name} is synonymous with \k<name> and \k'name' (.NET compatible).
596 (f) \g{name} is another synonym - part of Perl 5.10's unification of
597 reference syntax.
599 (g) (?| introduces a group in which the numbering of parentheses in each
600 alternative starts with the same number.
602 (h) \h, \H, \v, and \V match horizontal and vertical whitespace.
604 7. Added two new calls to pcre_fullinfo(): PCRE_INFO_OKPARTIAL and
607 8. A pattern such as (.*(.)?)* caused pcre_exec() to fail by either not
608 terminating or by crashing. Diagnosed by Viktor Griph; it was in the code
609 for detecting groups that can match an empty string.
611 9. A pattern with a very large number of alternatives (more than several
612 hundred) was running out of internal workspace during the pre-compile
613 phase, where pcre_compile() figures out how much memory will be needed. A
614 bit of new cunning has reduced the workspace needed for groups with
615 alternatives. The 1000-alternative test pattern now uses 12 bytes of
616 workspace instead of running out of the 4096 that are available.
618 10. Inserted some missing (unsigned int) casts to get rid of compiler warnings.
620 11. Applied patch from Google to remove an optimization that didn't quite work.
621 The report of the bug said:
623 pcrecpp::RE("a*").FullMatch("aaa") matches, while
624 pcrecpp::RE("a*?").FullMatch("aaa") does not, and
625 pcrecpp::RE("a*?\\z").FullMatch("aaa") does again.
627 12. If \p or \P was used in non-UTF-8 mode on a character greater than 127
628 it matched the wrong number of bytes.
631 Version 7.1 24-Apr-07
632 ---------------------
634 1. Applied Bob Rossi and Daniel G's patches to convert the build system to one
635 that is more "standard", making use of automake and other Autotools. There
636 is some re-arrangement of the files and adjustment of comments consequent
637 on this.
639 2. Part of the patch fixed a problem with the pcregrep tests. The test of -r
640 for recursive directory scanning broke on some systems because the files
641 are not scanned in any specific order and on different systems the order
642 was different. A call to "sort" has been inserted into RunGrepTest for the
643 approprate test as a short-term fix. In the longer term there may be an
644 alternative.
646 3. I had an email from Eric Raymond about problems translating some of PCRE's
647 man pages to HTML (despite the fact that I distribute HTML pages, some
648 people do their own conversions for various reasons). The problems
649 concerned the use of low-level troff macros .br and .in. I have therefore
650 removed all such uses from the man pages (some were redundant, some could
651 be replaced by .nf/.fi pairs). The 132html script that I use to generate
652 HTML has been updated to handle .nf/.fi and to complain if it encounters
653 .br or .in.
655 4. Updated comments in configure.ac that get placed in config.h.in and also
656 arranged for config.h to be included in the distribution, with the name
657 config.h.generic, for the benefit of those who have to compile without
658 Autotools (compare pcre.h, which is now distributed as pcre.h.generic).
660 5. Updated the support (such as it is) for Virtual Pascal, thanks to Stefan
661 Weber: (1) pcre_internal.h was missing some function renames; (2) updated
662 makevp.bat for the current PCRE, using the additional files
663 makevp_c.txt, makevp_l.txt, and pcregexp.pas.
665 6. A Windows user reported a minor discrepancy with test 2, which turned out
666 to be caused by a trailing space on an input line that had got lost in his
667 copy. The trailing space was an accident, so I've just removed it.
669 7. Add -Wl,-R... flags in pcre-config.in for *BSD* systems, as I'm told
670 that is needed.
672 8. Mark ucp_table (in ucptable.h) and ucp_gentype (in pcre_ucp_searchfuncs.c)
673 as "const" (a) because they are and (b) because it helps the PHP
674 maintainers who have recently made a script to detect big data structures
675 in the php code that should be moved to the .rodata section. I remembered
676 to update Builducptable as well, so it won't revert if ucptable.h is ever
677 re-created.
679 9. Added some extra #ifdef SUPPORT_UTF8 conditionals into pcretest.c,
680 pcre_printint.src, pcre_compile.c, pcre_study.c, and pcre_tables.c, in
681 order to be able to cut out the UTF-8 tables in the latter when UTF-8
682 support is not required. This saves 1.5-2K of code, which is important in
683 some applications.
685 Later: more #ifdefs are needed in pcre_ord2utf8.c and pcre_valid_utf8.c
686 so as not to refer to the tables, even though these functions will never be
687 called when UTF-8 support is disabled. Otherwise there are problems with a
688 shared library.
690 10. Fixed two bugs in the emulated memmove() function in pcre_internal.h:
692 (a) It was defining its arguments as char * instead of void *.
694 (b) It was assuming that all moves were upwards in memory; this was true
695 a long time ago when I wrote it, but is no longer the case.
697 The emulated memove() is provided for those environments that have neither
698 memmove() nor bcopy(). I didn't think anyone used it these days, but that
699 is clearly not the case, as these two bugs were recently reported.
701 11. The script PrepareRelease is now distributed: it calls 132html, CleanTxt,
702 and Detrail to create the HTML documentation, the .txt form of the man
703 pages, and it removes trailing spaces from listed files. It also creates
704 pcre.h.generic and config.h.generic from pcre.h and config.h. In the latter
705 case, it wraps all the #defines with #ifndefs. This script should be run
706 before "make dist".
708 12. Fixed two fairly obscure bugs concerned with quantified caseless matching
709 with Unicode property support.
711 (a) For a maximizing quantifier, if the two different cases of the
712 character were of different lengths in their UTF-8 codings (there are
713 some cases like this - I found 11), and the matching function had to
714 back up over a mixture of the two cases, it incorrectly assumed they
715 were both the same length.
717 (b) When PCRE was configured to use the heap rather than the stack for
718 recursion during matching, it was not correctly preserving the data for
719 the other case of a UTF-8 character when checking ahead for a match
720 while processing a minimizing repeat. If the check also involved
721 matching a wide character, but failed, corruption could cause an
722 erroneous result when trying to check for a repeat of the original
723 character.
725 13. Some tidying changes to the testing mechanism:
727 (a) The RunTest script now detects the internal link size and whether there
728 is UTF-8 and UCP support by running ./pcretest -C instead of relying on
729 values substituted by "configure". (The RunGrepTest script already did
730 this for UTF-8.) The configure.ac script no longer substitutes the
731 relevant variables.
733 (b) The debugging options /B and /D in pcretest show the compiled bytecode
734 with length and offset values. This means that the output is different
735 for different internal link sizes. Test 2 is skipped for link sizes
736 other than 2 because of this, bypassing the problem. Unfortunately,
737 there was also a test in test 3 (the locale tests) that used /B and
738 failed for link sizes other than 2. Rather than cut the whole test out,
739 I have added a new /Z option to pcretest that replaces the length and
740 offset values with spaces. This is now used to make test 3 independent
741 of link size. (Test 2 will be tidied up later.)
743 14. If erroroffset was passed as NULL to pcre_compile, it provoked a
744 segmentation fault instead of returning the appropriate error message.
746 15. In multiline mode when the newline sequence was set to "any", the pattern
747 ^$ would give a match between the \r and \n of a subject such as "A\r\nB".
748 This doesn't seem right; it now treats the CRLF combination as the line
749 ending, and so does not match in that case. It's only a pattern such as ^$
750 that would hit this one: something like ^ABC$ would have failed after \r
751 and then tried again after \r\n.
753 16. Changed the comparison command for RunGrepTest from "diff -u" to "diff -ub"
754 in an attempt to make files that differ only in their line terminators
755 compare equal. This works on Linux.
757 17. Under certain error circumstances pcregrep might try to free random memory
758 as it exited. This is now fixed, thanks to valgrind.
760 19. In pcretest, if the pattern /(?m)^$/g<any> was matched against the string
761 "abc\r\n\r\n", it found an unwanted second match after the second \r. This
762 was because its rules for how to advance for /g after matching an empty
763 string at the end of a line did not allow for this case. They now check for
764 it specially.
766 20. pcretest is supposed to handle patterns and data of any length, by
767 extending its buffers when necessary. It was getting this wrong when the
768 buffer for a data line had to be extended.
770 21. Added PCRE_NEWLINE_ANYCRLF which is like ANY, but matches only CR, LF, or
771 CRLF as a newline sequence.
773 22. Code for handling Unicode properties in pcre_dfa_exec() wasn't being cut
774 out by #ifdef SUPPORT_UCP. This did no harm, as it could never be used, but
775 I have nevertheless tidied it up.
777 23. Added some casts to kill warnings from HP-UX ia64 compiler.
779 24. Added a man page for pcre-config.
782 Version 7.0 19-Dec-06
783 ---------------------
785 1. Fixed a signed/unsigned compiler warning in pcre_compile.c, shown up by
786 moving to gcc 4.1.1.
788 2. The -S option for pcretest uses setrlimit(); I had omitted to #include
789 sys/time.h, which is documented as needed for this function. It doesn't
790 seem to matter on Linux, but it showed up on some releases of OS X.
792 3. It seems that there are systems where bytes whose values are greater than
793 127 match isprint() in the "C" locale. The "C" locale should be the
794 default when a C program starts up. In most systems, only ASCII printing
795 characters match isprint(). This difference caused the output from pcretest
796 to vary, making some of the tests fail. I have changed pcretest so that:
798 (a) When it is outputting text in the compiled version of a pattern, bytes
799 other than 32-126 are always shown as hex escapes.
801 (b) When it is outputting text that is a matched part of a subject string,
802 it does the same, unless a different locale has been set for the match
803 (using the /L modifier). In this case, it uses isprint() to decide.
805 4. Fixed a major bug that caused incorrect computation of the amount of memory
806 required for a compiled pattern when options that changed within the
807 pattern affected the logic of the preliminary scan that determines the
808 length. The relevant options are -x, and -i in UTF-8 mode. The result was
809 that the computed length was too small. The symptoms of this bug were
810 either the PCRE error "internal error: code overflow" from pcre_compile(),
811 or a glibc crash with a message such as "pcretest: free(): invalid next
812 size (fast)". Examples of patterns that provoked this bug (shown in
813 pcretest format) are:
815 /(?-x: )/x
816 /(?x)(?-x: \s*#\s*)/
817 /((?i)[\x{c0}])/8
818 /(?i:[\x{c0}])/8
820 HOWEVER: Change 17 below makes this fix obsolete as the memory computation
821 is now done differently.
823 5. Applied patches from Google to: (a) add a QuoteMeta function to the C++
824 wrapper classes; (b) implement a new function in the C++ scanner that is
825 more efficient than the old way of doing things because it avoids levels of
826 recursion in the regex matching; (c) add a paragraph to the documentation
827 for the FullMatch() function.
829 6. The escape sequence \n was being treated as whatever was defined as
830 "newline". Not only was this contrary to the documentation, which states
831 that \n is character 10 (hex 0A), but it also went horribly wrong when
832 "newline" was defined as CRLF. This has been fixed.
834 7. In pcre_dfa_exec.c the value of an unsigned integer (the variable called c)
835 was being set to -1 for the "end of line" case (supposedly a value that no
836 character can have). Though this value is never used (the check for end of
837 line is "zero bytes in current character"), it caused compiler complaints.
838 I've changed it to 0xffffffff.
840 8. In pcre_version.c, the version string was being built by a sequence of
841 C macros that, in the event of PCRE_PRERELEASE being defined as an empty
842 string (as it is for production releases) called a macro with an empty
843 argument. The C standard says the result of this is undefined. The gcc
844 compiler treats it as an empty string (which was what was wanted) but it is
845 reported that Visual C gives an error. The source has been hacked around to
846 avoid this problem.
848 9. On the advice of a Windows user, included <io.h> and <fcntl.h> in Windows
849 builds of pcretest, and changed the call to _setmode() to use _O_BINARY
850 instead of 0x8000. Made all the #ifdefs test both _WIN32 and WIN32 (not all
851 of them did).
853 10. Originally, pcretest opened its input and output without "b"; then I was
854 told that "b" was needed in some environments, so it was added for release
855 5.0 to both the input and output. (It makes no difference on Unix-like
856 systems.) Later I was told that it is wrong for the input on Windows. I've
857 now abstracted the modes into two macros, to make it easier to fiddle with
858 them, and removed "b" from the input mode under Windows.
860 11. Added pkgconfig support for the C++ wrapper library, libpcrecpp.
862 12. Added -help and --help to pcretest as an official way of being reminded
863 of the options.
865 13. Removed some redundant semicolons after macro calls in pcrecpparg.h.in
866 and pcrecpp.cc because they annoy compilers at high warning levels.
868 14. A bit of tidying/refactoring in pcre_exec.c in the main bumpalong loop.
870 15. Fixed an occurrence of == in configure.ac that should have been = (shell
871 scripts are not C programs :-) and which was not noticed because it works
872 on Linux.
874 16. pcretest is supposed to handle any length of pattern and data line (as one
875 line or as a continued sequence of lines) by extending its input buffer if
876 necessary. This feature was broken for very long pattern lines, leading to
877 a string of junk being passed to pcre_compile() if the pattern was longer
878 than about 50K.
880 17. I have done a major re-factoring of the way pcre_compile() computes the
881 amount of memory needed for a compiled pattern. Previously, there was code
882 that made a preliminary scan of the pattern in order to do this. That was
883 OK when PCRE was new, but as the facilities have expanded, it has become
884 harder and harder to keep it in step with the real compile phase, and there
885 have been a number of bugs (see for example, 4 above). I have now found a
886 cunning way of running the real compile function in a "fake" mode that
887 enables it to compute how much memory it would need, while actually only
888 ever using a few hundred bytes of working memory and without too many
889 tests of the mode. This should make future maintenance and development
890 easier. A side effect of this work is that the limit of 200 on the nesting
891 depth of parentheses has been removed (though this was never a serious
892 limitation, I suspect). However, there is a downside: pcre_compile() now
893 runs more slowly than before (30% or more, depending on the pattern). I
894 hope this isn't a big issue. There is no effect on runtime performance.
896 18. Fixed a minor bug in pcretest: if a pattern line was not terminated by a
897 newline (only possible for the last line of a file) and it was a
898 pattern that set a locale (followed by /Lsomething), pcretest crashed.
900 19. Added additional timing features to pcretest. (1) The -tm option now times
901 matching only, not compiling. (2) Both -t and -tm can be followed, as a
902 separate command line item, by a number that specifies the number of
903 repeats to use when timing. The default is 50000; this gives better
904 precision, but takes uncomfortably long for very large patterns.
906 20. Extended pcre_study() to be more clever in cases where a branch of a
907 subpattern has no definite first character. For example, (a*|b*)[cd] would
908 previously give no result from pcre_study(). Now it recognizes that the
909 first character must be a, b, c, or d.
911 21. There was an incorrect error "recursive call could loop indefinitely" if
912 a subpattern (or the entire pattern) that was being tested for matching an
913 empty string contained only one non-empty item after a nested subpattern.
914 For example, the pattern (?>\x{100}*)\d(?R) provoked this error
915 incorrectly, because the \d was being skipped in the check.
917 22. The pcretest program now has a new pattern option /B and a command line
918 option -b, which is equivalent to adding /B to every pattern. This causes
919 it to show the compiled bytecode, without the additional information that
920 -d shows. The effect of -d is now the same as -b with -i (and similarly, /D
921 is the same as /B/I).
923 23. A new optimization is now able automatically to treat some sequences such
924 as a*b as a*+b. More specifically, if something simple (such as a character
925 or a simple class like \d) has an unlimited quantifier, and is followed by
926 something that cannot possibly match the quantified thing, the quantifier
927 is automatically "possessified".
929 24. A recursive reference to a subpattern whose number was greater than 39
930 went wrong under certain circumstances in UTF-8 mode. This bug could also
931 have affected the operation of pcre_study().
933 25. Realized that a little bit of performance could be had by replacing
934 (c & 0xc0) == 0xc0 with c >= 0xc0 when processing UTF-8 characters.
936 26. Timing data from pcretest is now shown to 4 decimal places instead of 3.
938 27. Possessive quantifiers such as a++ were previously implemented by turning
939 them into atomic groups such as ($>a+). Now they have their own opcodes,
940 which improves performance. This includes the automatically created ones
941 from 23 above.
943 28. A pattern such as (?=(\w+))\1: which simulates an atomic group using a
944 lookahead was broken if it was not anchored. PCRE was mistakenly expecting
945 the first matched character to be a colon. This applied both to named and
946 numbered groups.
948 29. The ucpinternal.h header file was missing its idempotency #ifdef.
950 30. I was sent a "project" file called libpcre.a.dev which I understand makes
951 building PCRE on Windows easier, so I have included it in the distribution.
953 31. There is now a check in pcretest against a ridiculously large number being
954 returned by pcre_exec() or pcre_dfa_exec(). If this happens in a /g or /G
955 loop, the loop is abandoned.
957 32. Forward references to subpatterns in conditions such as (?(2)...) where
958 subpattern 2 is defined later cause pcre_compile() to search forwards in
959 the pattern for the relevant set of parentheses. This search went wrong
960 when there were unescaped parentheses in a character class, parentheses
961 escaped with \Q...\E, or parentheses in a #-comment in /x mode.
963 33. "Subroutine" calls and backreferences were previously restricted to
964 referencing subpatterns earlier in the regex. This restriction has now
965 been removed.
967 34. Added a number of extra features that are going to be in Perl 5.10. On the
968 whole, these are just syntactic alternatives for features that PCRE had
969 previously implemented using the Python syntax or my own invention. The
970 other formats are all retained for compatibility.
972 (a) Named groups can now be defined as (?<name>...) or (?'name'...) as well
973 as (?P<name>...). The new forms, as well as being in Perl 5.10, are
974 also .NET compatible.
976 (b) A recursion or subroutine call to a named group can now be defined as
977 (?&name) as well as (?P>name).
979 (c) A backreference to a named group can now be defined as \k<name> or
980 \k'name' as well as (?P=name). The new forms, as well as being in Perl
981 5.10, are also .NET compatible.
983 (d) A conditional reference to a named group can now use the syntax
984 (?(<name>) or (?('name') as well as (?(name).
986 (e) A "conditional group" of the form (?(DEFINE)...) can be used to define
987 groups (named and numbered) that are never evaluated inline, but can be
988 called as "subroutines" from elsewhere. In effect, the DEFINE condition
989 is always false. There may be only one alternative in such a group.
991 (f) A test for recursion can be given as (?(R1).. or (?(R&name)... as well
992 as the simple (?(R). The condition is true only if the most recent
993 recursion is that of the given number or name. It does not search out
994 through the entire recursion stack.
996 (g) The escape \gN or \g{N} has been added, where N is a positive or
997 negative number, specifying an absolute or relative reference.
999 35. Tidied to get rid of some further signed/unsigned compiler warnings and
1000 some "unreachable code" warnings.
1002 36. Updated the Unicode property tables to Unicode version 5.0.0. Amongst other
1003 things, this adds five new scripts.
1005 37. Perl ignores orphaned \E escapes completely. PCRE now does the same.
1006 There were also incompatibilities regarding the handling of \Q..\E inside
1007 character classes, for example with patterns like [\Qa\E-\Qz\E] where the
1008 hyphen was adjacent to \Q or \E. I hope I've cleared all this up now.
1010 38. Like Perl, PCRE detects when an indefinitely repeated parenthesized group
1011 matches an empty string, and forcibly breaks the loop. There were bugs in
1012 this code in non-simple cases. For a pattern such as ^(a()*)* matched
1013 against aaaa the result was just "a" rather than "aaaa", for example. Two
1014 separate and independent bugs (that affected different cases) have been
1015 fixed.
1017 39. Refactored the code to abolish the use of different opcodes for small
1018 capturing bracket numbers. This is a tidy that I avoided doing when I
1019 removed the limit on the number of capturing brackets for 3.5 back in 2001.
1020 The new approach is not only tidier, it makes it possible to reduce the
1021 memory needed to fix the previous bug (38).
1023 40. Implemented PCRE_NEWLINE_ANY to recognize any of the Unicode newline
1024 sequences (http://unicode.org/unicode/reports/tr18/) as "newline" when
1025 processing dot, circumflex, or dollar metacharacters, or #-comments in /x
1026 mode.
1028 41. Add \R to match any Unicode newline sequence, as suggested in the Unicode
1029 report.
1031 42. Applied patch, originally from Ari Pollak, modified by Google, to allow
1032 copy construction and assignment in the C++ wrapper.
1034 43. Updated pcregrep to support "--newline=any". In the process, I fixed a
1035 couple of bugs that could have given wrong results in the "--newline=crlf"
1036 case.
1038 44. Added a number of casts and did some reorganization of signed/unsigned int
1039 variables following suggestions from Dair Grant. Also renamed the variable
1040 "this" as "item" because it is a C++ keyword.
1042 45. Arranged for dftables to add
1044 #include "pcre_internal.h"
1046 to pcre_chartables.c because without it, gcc 4.x may remove the array
1047 definition from the final binary if PCRE is built into a static library and
1048 dead code stripping is activated.
1050 46. For an unanchored pattern, if a match attempt fails at the start of a
1051 newline sequence, and the newline setting is CRLF or ANY, and the next two
1052 characters are CRLF, advance by two characters instead of one.
1055 Version 6.7 04-Jul-06
1056 ---------------------
1058 1. In order to handle tests when input lines are enormously long, pcretest has
1059 been re-factored so that it automatically extends its buffers when
1060 necessary. The code is crude, but this _is_ just a test program. The
1061 default size has been increased from 32K to 50K.
1063 2. The code in pcre_study() was using the value of the re argument before
1064 testing it for NULL. (Of course, in any sensible call of the function, it
1065 won't be NULL.)
1067 3. The memmove() emulation function in pcre_internal.h, which is used on
1068 systems that lack both memmove() and bcopy() - that is, hardly ever -
1069 was missing a "static" storage class specifier.
1071 4. When UTF-8 mode was not set, PCRE looped when compiling certain patterns
1072 containing an extended class (one that cannot be represented by a bitmap
1073 because it contains high-valued characters or Unicode property items, e.g.
1074 [\pZ]). Almost always one would set UTF-8 mode when processing such a
1075 pattern, but PCRE should not loop if you do not (it no longer does).
1076 [Detail: two cases were found: (a) a repeated subpattern containing an
1077 extended class; (b) a recursive reference to a subpattern that followed a
1078 previous extended class. It wasn't skipping over the extended class
1079 correctly when UTF-8 mode was not set.]
1081 5. A negated single-character class was not being recognized as fixed-length
1082 in lookbehind assertions such as (?<=[^f]), leading to an incorrect
1083 compile error "lookbehind assertion is not fixed length".
1085 6. The RunPerlTest auxiliary script was showing an unexpected difference
1086 between PCRE and Perl for UTF-8 tests. It turns out that it is hard to
1087 write a Perl script that can interpret lines of an input file either as
1088 byte characters or as UTF-8, which is what "perltest" was being required to
1089 do for the non-UTF-8 and UTF-8 tests, respectively. Essentially what you
1090 can't do is switch easily at run time between having the "use utf8;" pragma
1091 or not. In the end, I fudged it by using the RunPerlTest script to insert
1092 "use utf8;" explicitly for the UTF-8 tests.
1094 7. In multiline (/m) mode, PCRE was matching ^ after a terminating newline at
1095 the end of the subject string, contrary to the documentation and to what
1096 Perl does. This was true of both matching functions. Now it matches only at
1097 the start of the subject and immediately after *internal* newlines.
1099 8. A call of pcre_fullinfo() from pcretest to get the option bits was passing
1100 a pointer to an int instead of a pointer to an unsigned long int. This
1101 caused problems on 64-bit systems.
1103 9. Applied a patch from the folks at Google to pcrecpp.cc, to fix "another
1104 instance of the 'standard' template library not being so standard".
1106 10. There was no check on the number of named subpatterns nor the maximum
1107 length of a subpattern name. The product of these values is used to compute
1108 the size of the memory block for a compiled pattern. By supplying a very
1109 long subpattern name and a large number of named subpatterns, the size
1110 computation could be caused to overflow. This is now prevented by limiting
1111 the length of names to 32 characters, and the number of named subpatterns
1112 to 10,000.
1114 11. Subpatterns that are repeated with specific counts have to be replicated in
1115 the compiled pattern. The size of memory for this was computed from the
1116 length of the subpattern and the repeat count. The latter is limited to
1117 65535, but there was no limit on the former, meaning that integer overflow
1118 could in principle occur. The compiled length of a repeated subpattern is
1119 now limited to 30,000 bytes in order to prevent this.
1121 12. Added the optional facility to have named substrings with the same name.
1123 13. Added the ability to use a named substring as a condition, using the
1124 Python syntax: (?(name)yes|no). This overloads (?(R)... and names that
1125 are numbers (not recommended). Forward references are permitted.
1127 14. Added forward references in named backreferences (if you see what I mean).
1129 15. In UTF-8 mode, with the PCRE_DOTALL option set, a quantified dot in the
1130 pattern could run off the end of the subject. For example, the pattern
1131 "(?s)(.{1,5})"8 did this with the subject "ab".
1133 16. If PCRE_DOTALL or PCRE_MULTILINE were set, pcre_dfa_exec() behaved as if
1134 PCRE_CASELESS was set when matching characters that were quantified with ?
1135 or *.
1137 17. A character class other than a single negated character that had a minimum
1138 but no maximum quantifier - for example [ab]{6,} - was not handled
1139 correctly by pce_dfa_exec(). It would match only one character.
1141 18. A valid (though odd) pattern that looked like a POSIX character
1142 class but used an invalid character after [ (for example [[,abc,]]) caused
1143 pcre_compile() to give the error "Failed: internal error: code overflow" or
1144 in some cases to crash with a glibc free() error. This could even happen if
1145 the pattern terminated after [[ but there just happened to be a sequence of
1146 letters, a binary zero, and a closing ] in the memory that followed.
1148 19. Perl's treatment of octal escapes in the range \400 to \777 has changed
1149 over the years. Originally (before any Unicode support), just the bottom 8
1150 bits were taken. Thus, for example, \500 really meant \100. Nowadays the
1151 output from "man perlunicode" includes this:
1153 The regular expression compiler produces polymorphic opcodes. That
1154 is, the pattern adapts to the data and automatically switches to
1155 the Unicode character scheme when presented with Unicode data--or
1156 instead uses a traditional byte scheme when presented with byte
1157 data.
1159 Sadly, a wide octal escape does not cause a switch, and in a string with
1160 no other multibyte characters, these octal escapes are treated as before.
1161 Thus, in Perl, the pattern /\500/ actually matches \100 but the pattern
1162 /\500|\x{1ff}/ matches \500 or \777 because the whole thing is treated as a
1163 Unicode string.
1165 I have not perpetrated such confusion in PCRE. Up till now, it took just
1166 the bottom 8 bits, as in old Perl. I have now made octal escapes with
1167 values greater than \377 illegal in non-UTF-8 mode. In UTF-8 mode they
1168 translate to the appropriate multibyte character.
1170 29. Applied some refactoring to reduce the number of warnings from Microsoft
1171 and Borland compilers. This has included removing the fudge introduced
1172 seven years ago for the OS/2 compiler (see 2.02/2 below) because it caused
1173 a warning about an unused variable.
1175 21. PCRE has not included VT (character 0x0b) in the set of whitespace
1176 characters since release 4.0, because Perl (from release 5.004) does not.
1177 [Or at least, is documented not to: some releases seem to be in conflict
1178 with the documentation.] However, when a pattern was studied with
1179 pcre_study() and all its branches started with \s, PCRE still included VT
1180 as a possible starting character. Of course, this did no harm; it just
1181 caused an unnecessary match attempt.
1183 22. Removed a now-redundant internal flag bit that recorded the fact that case
1184 dependency changed within the pattern. This was once needed for "required
1185 byte" processing, but is no longer used. This recovers a now-scarce options
1186 bit. Also moved the least significant internal flag bit to the most-
1187 significant bit of the word, which was not previously used (hangover from
1188 the days when it was an int rather than a uint) to free up another bit for
1189 the future.
1191 23. Added support for CRLF line endings as well as CR and LF. As well as the
1192 default being selectable at build time, it can now be changed at runtime
1193 via the PCRE_NEWLINE_xxx flags. There are now options for pcregrep to
1194 specify that it is scanning data with non-default line endings.
1196 24. Changed the definition of CXXLINK to make it agree with the definition of
1197 LINK in the Makefile, by replacing LDFLAGS to CXXFLAGS.
1199 25. Applied Ian Taylor's patches to avoid using another stack frame for tail
1200 recursions. This makes a big different to stack usage for some patterns.
1202 26. If a subpattern containing a named recursion or subroutine reference such
1203 as (?P>B) was quantified, for example (xxx(?P>B)){3}, the calculation of
1204 the space required for the compiled pattern went wrong and gave too small a
1205 value. Depending on the environment, this could lead to "Failed: internal
1206 error: code overflow at offset 49" or "glibc detected double free or
1207 corruption" errors.
1209 27. Applied patches from Google (a) to support the new newline modes and (b) to
1210 advance over multibyte UTF-8 characters in GlobalReplace.
1212 28. Change free() to pcre_free() in pcredemo.c. Apparently this makes a
1213 difference for some implementation of PCRE in some Windows version.
1215 29. Added some extra testing facilities to pcretest:
1217 \q<number> in a data line sets the "match limit" value
1218 \Q<number> in a data line sets the "match recursion limt" value
1219 -S <number> sets the stack size, where <number> is in megabytes
1221 The -S option isn't available for Windows.
1224 Version 6.6 06-Feb-06
1225 ---------------------
1227 1. Change 16(a) for 6.5 broke things, because PCRE_DATA_SCOPE was not defined
1228 in pcreposix.h. I have copied the definition from pcre.h.
1230 2. Change 25 for 6.5 broke compilation in a build directory out-of-tree
1231 because pcre.h is no longer a built file.
1233 3. Added Jeff Friedl's additional debugging patches to pcregrep. These are
1234 not normally included in the compiled code.
1237 Version 6.5 01-Feb-06
1238 ---------------------
1240 1. When using the partial match feature with pcre_dfa_exec(), it was not
1241 anchoring the second and subsequent partial matches at the new starting
1242 point. This could lead to incorrect results. For example, with the pattern
1243 /1234/, partially matching against "123" and then "a4" gave a match.
1245 2. Changes to pcregrep:
1247 (a) All non-match returns from pcre_exec() were being treated as failures
1248 to match the line. Now, unless the error is PCRE_ERROR_NOMATCH, an
1249 error message is output. Some extra information is given for the
1251 probably the only errors that are likely to be caused by users (by
1252 specifying a regex that has nested indefinite repeats, for instance).
1253 If there are more than 20 of these errors, pcregrep is abandoned.
1255 (b) A binary zero was treated as data while matching, but terminated the
1256 output line if it was written out. This has been fixed: binary zeroes
1257 are now no different to any other data bytes.
1259 (c) Whichever of the LC_ALL or LC_CTYPE environment variables is set is
1260 used to set a locale for matching. The --locale=xxxx long option has
1261 been added (no short equivalent) to specify a locale explicitly on the
1262 pcregrep command, overriding the environment variables.
1264 (d) When -B was used with -n, some line numbers in the output were one less
1265 than they should have been.
1267 (e) Added the -o (--only-matching) option.
1269 (f) If -A or -C was used with -c (count only), some lines of context were
1270 accidentally printed for the final match.
1272 (g) Added the -H (--with-filename) option.
1274 (h) The combination of options -rh failed to suppress file names for files
1275 that were found from directory arguments.
1277 (i) Added the -D (--devices) and -d (--directories) options.
1279 (j) Added the -F (--fixed-strings) option.
1281 (k) Allow "-" to be used as a file name for -f as well as for a data file.
1283 (l) Added the --colo(u)r option.
1285 (m) Added Jeffrey Friedl's -S testing option, but within #ifdefs so that it
1286 is not present by default.
1288 3. A nasty bug was discovered in the handling of recursive patterns, that is,
1289 items such as (?R) or (?1), when the recursion could match a number of
1290 alternatives. If it matched one of the alternatives, but subsequently,
1291 outside the recursion, there was a failure, the code tried to back up into
1292 the recursion. However, because of the way PCRE is implemented, this is not
1293 possible, and the result was an incorrect result from the match.
1295 In order to prevent this happening, the specification of recursion has
1296 been changed so that all such subpatterns are automatically treated as
1297 atomic groups. Thus, for example, (?R) is treated as if it were (?>(?R)).
1299 4. I had overlooked the fact that, in some locales, there are characters for
1300 which isalpha() is true but neither isupper() nor islower() are true. In
1301 the fr_FR locale, for instance, the \xAA and \xBA characters (ordmasculine
1302 and ordfeminine) are like this. This affected the treatment of \w and \W
1303 when they appeared in character classes, but not when they appeared outside
1304 a character class. The bit map for "word" characters is now created
1305 separately from the results of isalnum() instead of just taking it from the
1306 upper, lower, and digit maps. (Plus the underscore character, of course.)
1308 5. The above bug also affected the handling of POSIX character classes such as
1309 [[:alpha:]] and [[:alnum:]]. These do not have their own bit maps in PCRE's
1310 permanent tables. Instead, the bit maps for such a class were previously
1311 created as the appropriate unions of the upper, lower, and digit bitmaps.
1312 Now they are created by subtraction from the [[:word:]] class, which has
1313 its own bitmap.
1315 6. The [[:blank:]] character class matches horizontal, but not vertical space.
1316 It is created by subtracting the vertical space characters (\x09, \x0a,
1317 \x0b, \x0c) from the [[:space:]] bitmap. Previously, however, the
1318 subtraction was done in the overall bitmap for a character class, meaning
1319 that a class such as [\x0c[:blank:]] was incorrect because \x0c would not
1320 be recognized. This bug has been fixed.
1322 7. Patches from the folks at Google:
1324 (a) pcrecpp.cc: "to handle a corner case that may or may not happen in
1325 real life, but is still worth protecting against".
1327 (b) pcrecpp.cc: "corrects a bug when negative radixes are used with
1328 regular expressions".
1330 (c) pcre_scanner.cc: avoid use of std::count() because not all systems
1331 have it.
1333 (d) Split off pcrecpparg.h from pcrecpp.h and had the former built by
1334 "configure" and the latter not, in order to fix a problem somebody had
1335 with compiling the Arg class on HP-UX.
1337 (e) Improve the error-handling of the C++ wrapper a little bit.
1339 (f) New tests for checking recursion limiting.
1341 8. The pcre_memmove() function, which is used only if the environment does not
1342 have a standard memmove() function (and is therefore rarely compiled),
1343 contained two bugs: (a) use of int instead of size_t, and (b) it was not
1344 returning a result (though PCRE never actually uses the result).
1346 9. In the POSIX regexec() interface, if nmatch is specified as a ridiculously
1347 large number - greater than INT_MAX/(3*sizeof(int)) - REG_ESPACE is
1348 returned instead of calling malloc() with an overflowing number that would
1349 most likely cause subsequent chaos.
1351 10. The debugging option of pcretest was not showing the NO_AUTO_CAPTURE flag.
1353 11. The POSIX flag REG_NOSUB is now supported. When a pattern that was compiled
1354 with this option is matched, the nmatch and pmatch options of regexec() are
1355 ignored.
1357 12. Added REG_UTF8 to the POSIX interface. This is not defined by POSIX, but is
1358 provided in case anyone wants to the the POSIX interface with UTF-8
1359 strings.
1361 13. Added CXXLDFLAGS to the Makefile parameters to provide settings only on the
1362 C++ linking (needed for some HP-UX environments).
1364 14. Avoid compiler warnings in get_ucpname() when compiled without UCP support
1365 (unused parameter) and in the pcre_printint() function (omitted "default"
1366 switch label when the default is to do nothing).
1368 15. Added some code to make it possible, when PCRE is compiled as a C++
1369 library, to replace subject pointers for pcre_exec() with a smart pointer
1370 class, thus making it possible to process discontinuous strings.
1372 16. The two macros PCRE_EXPORT and PCRE_DATA_SCOPE are confusing, and perform
1373 much the same function. They were added by different people who were trying
1374 to make PCRE easy to compile on non-Unix systems. It has been suggested
1375 that PCRE_EXPORT be abolished now that there is more automatic apparatus
1376 for compiling on Windows systems. I have therefore replaced it with
1377 PCRE_DATA_SCOPE. This is set automatically for Windows; if not set it
1378 defaults to "extern" for C or "extern C" for C++, which works fine on
1379 Unix-like systems. It is now possible to override the value of PCRE_DATA_
1380 SCOPE with something explicit in config.h. In addition:
1382 (a) pcreposix.h still had just "extern" instead of either of these macros;
1383 I have replaced it with PCRE_DATA_SCOPE.
1385 (b) Functions such as _pcre_xclass(), which are internal to the library,
1386 but external in the C sense, all had PCRE_EXPORT in their definitions.
1387 This is apparently wrong for the Windows case, so I have removed it.
1388 (It makes no difference on Unix-like systems.)
1390 17. Added a new limit, MATCH_LIMIT_RECURSION, which limits the depth of nesting
1391 of recursive calls to match(). This is different to MATCH_LIMIT because
1392 that limits the total number of calls to match(), not all of which increase
1393 the depth of recursion. Limiting the recursion depth limits the amount of
1394 stack (or heap if NO_RECURSE is set) that is used. The default can be set
1395 when PCRE is compiled, and changed at run time. A patch from Google adds
1396 this functionality to the C++ interface.
1398 18. Changes to the handling of Unicode character properties:
1400 (a) Updated the table to Unicode 4.1.0.
1402 (b) Recognize characters that are not in the table as "Cn" (undefined).
1404 (c) I revised the way the table is implemented to a much improved format
1405 which includes recognition of ranges. It now supports the ranges that
1406 are defined in UnicodeData.txt, and it also amalgamates other
1407 characters into ranges. This has reduced the number of entries in the
1408 table from around 16,000 to around 3,000, thus reducing its size
1409 considerably. I realized I did not need to use a tree structure after
1410 all - a binary chop search is just as efficient. Having reduced the
1411 number of entries, I extended their size from 6 bytes to 8 bytes to
1412 allow for more data.
1414 (d) Added support for Unicode script names via properties such as \p{Han}.
1416 19. In UTF-8 mode, a backslash followed by a non-Ascii character was not
1417 matching that character.
1419 20. When matching a repeated Unicode property with a minimum greater than zero,
1420 (for example \pL{2,}), PCRE could look past the end of the subject if it
1421 reached it while seeking the minimum number of characters. This could
1422 happen only if some of the characters were more than one byte long, because
1423 there is a check for at least the minimum number of bytes.
1425 21. Refactored the implementation of \p and \P so as to be more general, to
1426 allow for more different types of property in future. This has changed the
1427 compiled form incompatibly. Anybody with saved compiled patterns that use
1428 \p or \P will have to recompile them.
1430 22. Added "Any" and "L&" to the supported property types.
1432 23. Recognize \x{...} as a code point specifier, even when not in UTF-8 mode,
1433 but give a compile time error if the value is greater than 0xff.
1435 24. The man pages for pcrepartial, pcreprecompile, and pcre_compile2 were
1436 accidentally not being installed or uninstalled.
1438 25. The pcre.h file was built from pcre.h.in, but the only changes that were
1439 made were to insert the current release number. This seemed silly, because
1440 it made things harder for people building PCRE on systems that don't run
1441 "configure". I have turned pcre.h into a distributed file, no longer built
1442 by "configure", with the version identification directly included. There is
1443 no longer a pcre.h.in file.
1445 However, this change necessitated a change to the pcre-config script as
1446 well. It is built from pcre-config.in, and one of the substitutions was the
1447 release number. I have updated configure.ac so that ./configure now finds
1448 the release number by grepping pcre.h.
1450 26. Added the ability to run the tests under valgrind.
1453 Version 6.4 05-Sep-05
1454 ---------------------
1456 1. Change 6.0/10/(l) to pcregrep introduced a bug that caused separator lines
1457 "--" to be printed when multiple files were scanned, even when none of the
1458 -A, -B, or -C options were used. This is not compatible with Gnu grep, so I
1459 consider it to be a bug, and have restored the previous behaviour.
1461 2. A couple of code tidies to get rid of compiler warnings.
1463 3. The pcretest program used to cheat by referring to symbols in the library
1464 whose names begin with _pcre_. These are internal symbols that are not
1465 really supposed to be visible externally, and in some environments it is
1466 possible to suppress them. The cheating is now confined to including
1467 certain files from the library's source, which is a bit cleaner.
1469 4. Renamed pcre.in as pcre.h.in to go with pcrecpp.h.in; it also makes the
1470 file's purpose clearer.
1472 5. Reorganized pcre_ucp_findchar().
1475 Version 6.3 15-Aug-05
1476 ---------------------
1478 1. The file libpcre.pc.in did not have general read permission in the tarball.
1480 2. There were some problems when building without C++ support:
1482 (a) If C++ support was not built, "make install" and "make test" still
1483 tried to test it.
1485 (b) There were problems when the value of CXX was explicitly set. Some
1486 changes have been made to try to fix these, and ...
1488 (c) --disable-cpp can now be used to explicitly disable C++ support.
1490 (d) The use of @CPP_OBJ@ directly caused a blank line preceded by a
1491 backslash in a target when C++ was disabled. This confuses some
1492 versions of "make", apparently. Using an intermediate variable solves
1493 this. (Same for CPP_LOBJ.)
1495 3. $(LINK_FOR_BUILD) now includes $(CFLAGS_FOR_BUILD) and $(LINK)
1496 (non-Windows) now includes $(CFLAGS) because these flags are sometimes
1497 necessary on certain architectures.
1499 4. Added a setting of -export-symbols-regex to the link command to remove
1500 those symbols that are exported in the C sense, but actually are local
1501 within the library, and not documented. Their names all begin with
1502 "_pcre_". This is not a perfect job, because (a) we have to except some
1503 symbols that pcretest ("illegally") uses, and (b) the facility isn't always
1504 available (and never for static libraries). I have made a note to try to
1505 find a way round (a) in the future.
1508 Version 6.2 01-Aug-05
1509 ---------------------
1511 1. There was no test for integer overflow of quantifier values. A construction
1512 such as {1111111111111111} would give undefined results. What is worse, if
1513 a minimum quantifier for a parenthesized subpattern overflowed and became
1514 negative, the calculation of the memory size went wrong. This could have
1515 led to memory overwriting.
1517 2. Building PCRE using VPATH was broken. Hopefully it is now fixed.
1519 3. Added "b" to the 2nd argument of fopen() in dftables.c, for non-Unix-like
1520 operating environments where this matters.
1522 4. Applied Giuseppe Maxia's patch to add additional features for controlling
1523 PCRE options from within the C++ wrapper.
1525 5. Named capturing subpatterns were not being correctly counted when a pattern
1526 was compiled. This caused two problems: (a) If there were more than 100
1527 such subpatterns, the calculation of the memory needed for the whole
1528 compiled pattern went wrong, leading to an overflow error. (b) Numerical
1529 back references of the form \12, where the number was greater than 9, were
1530 not recognized as back references, even though there were sufficient
1531 previous subpatterns.
1533 6. Two minor patches to pcrecpp.cc in order to allow it to compile on older
1534 versions of gcc, e.g. 2.95.4.
1537 Version 6.1 21-Jun-05
1538 ---------------------
1540 1. There was one reference to the variable "posix" in pcretest.c that was not
1541 surrounded by "#if !defined NOPOSIX".
1543 2. Make it possible to compile pcretest without DFA support, UTF8 support, or
1544 the cross-check on the old pcre_info() function, for the benefit of the
1545 cut-down version of PCRE that is currently imported into Exim.
1547 3. A (silly) pattern starting with (?i)(?-i) caused an internal space
1548 allocation error. I've done the easy fix, which wastes 2 bytes for sensible
1549 patterns that start (?i) but I don't think that matters. The use of (?i) is
1550 just an example; this all applies to the other options as well.
1552 4. Since libtool seems to echo the compile commands it is issuing, the output
1553 from "make" can be reduced a bit by putting "@" in front of each libtool
1554 compile command.
1556 5. Patch from the folks at Google for configure.in to be a bit more thorough
1557 in checking for a suitable C++ installation before trying to compile the
1558 C++ stuff. This should fix a reported problem when a compiler was present,
1559 but no suitable headers.
1561 6. The man pages all had just "PCRE" as their title. I have changed them to
1562 be the relevant file name. I have also arranged that these names are
1563 retained in the file doc/pcre.txt, which is a concatenation in text format
1564 of all the man pages except the little individual ones for each function.
1566 7. The NON-UNIX-USE file had not been updated for the different set of source
1567 files that come with release 6. I also added a few comments about the C++
1568 wrapper.
1571 Version 6.0 07-Jun-05
1572 ---------------------
1574 1. Some minor internal re-organization to help with my DFA experiments.
1576 2. Some missing #ifdef SUPPORT_UCP conditionals in pcretest and printint that
1577 didn't matter for the library itself when fully configured, but did matter
1578 when compiling without UCP support, or within Exim, where the ucp files are
1579 not imported.
1581 3. Refactoring of the library code to split up the various functions into
1582 different source modules. The addition of the new DFA matching code (see
1583 below) to a single monolithic source would have made it really too
1584 unwieldy, quite apart from causing all the code to be include in a
1585 statically linked application, when only some functions are used. This is
1586 relevant even without the DFA addition now that patterns can be compiled in
1587 one application and matched in another.
1589 The downside of splitting up is that there have to be some external
1590 functions and data tables that are used internally in different modules of
1591 the library but which are not part of the API. These have all had their
1592 names changed to start with "_pcre_" so that they are unlikely to clash
1593 with other external names.
1595 4. Added an alternate matching function, pcre_dfa_exec(), which matches using
1596 a different (DFA) algorithm. Although it is slower than the original
1597 function, it does have some advantages for certain types of matching
1598 problem.
1600 5. Upgrades to pcretest in order to test the features of pcre_dfa_exec(),
1601 including restarting after a partial match.
1603 6. A patch for pcregrep that defines INVALID_FILE_ATTRIBUTES if it is not
1604 defined when compiling for Windows was sent to me. I have put it into the
1605 code, though I have no means of testing or verifying it.
1607 7. Added the pcre_refcount() auxiliary function.
1609 8. Added the PCRE_FIRSTLINE option. This constrains an unanchored pattern to
1610 match before or at the first newline in the subject string. In pcretest,
1611 the /f option on a pattern can be used to set this.
1613 9. A repeated \w when used in UTF-8 mode with characters greater than 256
1614 would behave wrongly. This has been present in PCRE since release 4.0.
1616 10. A number of changes to the pcregrep command:
1618 (a) Refactored how -x works; insert ^(...)$ instead of setting
1619 PCRE_ANCHORED and checking the length, in preparation for adding
1620 something similar for -w.
1622 (b) Added the -w (match as a word) option.
1624 (c) Refactored the way lines are read and buffered so as to have more
1625 than one at a time available.
1627 (d) Implemented a pcregrep test script.
1629 (e) Added the -M (multiline match) option. This allows patterns to match
1630 over several lines of the subject. The buffering ensures that at least
1631 8K, or the rest of the document (whichever is the shorter) is available
1632 for matching (and similarly the previous 8K for lookbehind assertions).
1634 (f) Changed the --help output so that it now says
1636 -w, --word-regex(p)
1638 instead of two lines, one with "regex" and the other with "regexp"
1639 because that confused at least one person since the short forms are the
1640 same. (This required a bit of code, as the output is generated
1641 automatically from a table. It wasn't just a text change.)
1643 (g) -- can be used to terminate pcregrep options if the next thing isn't an
1644 option but starts with a hyphen. Could be a pattern or a path name
1645 starting with a hyphen, for instance.
1647 (h) "-" can be given as a file name to represent stdin.
1649 (i) When file names are being printed, "(standard input)" is used for
1650 the standard input, for compatibility with GNU grep. Previously
1651 "<stdin>" was used.
1653 (j) The option --label=xxx can be used to supply a name to be used for
1654 stdin when file names are being printed. There is no short form.
1656 (k) Re-factored the options decoding logic because we are going to add
1657 two more options that take data. Such options can now be given in four
1658 different ways, e.g. "-fname", "-f name", "--file=name", "--file name".
1660 (l) Added the -A, -B, and -C options for requesting that lines of context
1661 around matches be printed.
1663 (m) Added the -L option to print the names of files that do not contain
1664 any matching lines, that is, the complement of -l.
1666 (n) The return code is 2 if any file cannot be opened, but pcregrep does
1667 continue to scan other files.
1669 (o) The -s option was incorrectly implemented. For compatibility with other
1670 greps, it now suppresses the error message for a non-existent or non-
1671 accessible file (but not the return code). There is a new option called
1672 -q that suppresses the output of matching lines, which was what -s was
1673 previously doing.
1675 (p) Added --include and --exclude options to specify files for inclusion
1676 and exclusion when recursing.
1678 11. The Makefile was not using the Autoconf-supported LDFLAGS macro properly.
1679 Hopefully, it now does.
1681 12. Missing cast in pcre_study().
1683 13. Added an "uninstall" target to the makefile.
1685 14. Replaced "extern" in the function prototypes in Makefile.in with
1686 "PCRE_DATA_SCOPE", which defaults to 'extern' or 'extern "C"' in the Unix
1687 world, but is set differently for Windows.
1689 15. Added a second compiling function called pcre_compile2(). The only
1690 difference is that it has an extra argument, which is a pointer to an
1691 integer error code. When there is a compile-time failure, this is set
1692 non-zero, in addition to the error test pointer being set to point to an
1693 error message. The new argument may be NULL if no error number is required
1694 (but then you may as well call pcre_compile(), which is now just a
1695 wrapper). This facility is provided because some applications need a
1696 numeric error indication, but it has also enabled me to tidy up the way
1697 compile-time errors are handled in the POSIX wrapper.
1699 16. Added VPATH=.libs to the makefile; this should help when building with one
1700 prefix path and installing with another. (Or so I'm told by someone who
1701 knows more about this stuff than I do.)
1703 17. Added a new option, REG_DOTALL, to the POSIX function regcomp(). This
1704 passes PCRE_DOTALL to the pcre_compile() function, making the "." character
1705 match everything, including newlines. This is not POSIX-compatible, but
1706 somebody wanted the feature. From pcretest it can be activated by using
1707 both the P and the s flags.
1709 18. AC_PROG_LIBTOOL appeared twice in Makefile.in. Removed one.
1711 19. libpcre.pc was being incorrectly installed as executable.
1713 20. A couple of places in pcretest check for end-of-line by looking for '\n';
1714 it now also looks for '\r' so that it will work unmodified on Windows.
1716 21. Added Google's contributed C++ wrapper to the distribution.
1718 22. Added some untidy missing memory free() calls in pcretest, to keep
1719 Electric Fence happy when testing.
1723 Version 5.0 13-Sep-04
1724 ---------------------
1726 1. Internal change: literal characters are no longer packed up into items
1727 containing multiple characters in a single byte-string. Each character
1728 is now matched using a separate opcode. However, there may be more than one
1729 byte in the character in UTF-8 mode.
1731 2. The pcre_callout_block structure has two new fields: pattern_position and
1732 next_item_length. These contain the offset in the pattern to the next match
1733 item, and its length, respectively.
1735 3. The PCRE_AUTO_CALLOUT option for pcre_compile() requests the automatic
1736 insertion of callouts before each pattern item. Added the /C option to
1737 pcretest to make use of this.
1739 4. On the advice of a Windows user, the lines
1741 #if defined(_WIN32) || defined(WIN32)
1742 _setmode( _fileno( stdout ), 0x8000 );
1743 #endif /* defined(_WIN32) || defined(WIN32) */
1745 have been added to the source of pcretest. This apparently does useful
1746 magic in relation to line terminators.
1748 5. Changed "r" and "w" in the calls to fopen() in pcretest to "rb" and "wb"
1749 for the benefit of those environments where the "b" makes a difference.
1751 6. The icc compiler has the same options as gcc, but "configure" doesn't seem
1752 to know about it. I have put a hack into configure.in that adds in code
1753 to set GCC=yes if CC=icc. This seems to end up at a point in the
1754 generated configure script that is early enough to affect the setting of
1755 compiler options, which is what is needed, but I have no means of testing
1756 whether it really works. (The user who reported this had patched the
1757 generated configure script, which of course I cannot do.)
1759 LATER: After change 22 below (new libtool files), the configure script
1760 seems to know about icc (and also ecc). Therefore, I have commented out
1761 this hack in configure.in.
1763 7. Added support for pkg-config (2 patches were sent in).
1765 8. Negated POSIX character classes that used a combination of internal tables
1766 were completely broken. These were [[:^alpha:]], [[:^alnum:]], and
1767 [[:^ascii]]. Typically, they would match almost any characters. The other
1768 POSIX classes were not broken in this way.
1770 9. Matching the pattern "\b.*?" against "ab cd", starting at offset 1, failed
1771 to find the match, as PCRE was deluded into thinking that the match had to
1772 start at the start point or following a newline. The same bug applied to
1773 patterns with negative forward assertions or any backward assertions
1774 preceding ".*" at the start, unless the pattern required a fixed first
1775 character. This was a failing pattern: "(?!.bcd).*". The bug is now fixed.
1777 10. In UTF-8 mode, when moving forwards in the subject after a failed match
1778 starting at the last subject character, bytes beyond the end of the subject
1779 string were read.
1781 11. Renamed the variable "class" as "classbits" to make life easier for C++
1782 users. (Previously there was a macro definition, but it apparently wasn't
1783 enough.)
1785 12. Added the new field "tables" to the extra data so that tables can be passed
1786 in at exec time, or the internal tables can be re-selected. This allows
1787 a compiled regex to be saved and re-used at a later time by a different
1788 program that might have everything at different addresses.
1790 13. Modified the pcre-config script so that, when run on Solaris, it shows a
1791 -R library as well as a -L library.
1793 14. The debugging options of pcretest (-d on the command line or D on a
1794 pattern) showed incorrect output for anything following an extended class
1795 that contained multibyte characters and which was followed by a quantifier.
1797 15. Added optional support for general category Unicode character properties
1798 via the \p, \P, and \X escapes. Unicode property support implies UTF-8
1799 support. It adds about 90K to the size of the library. The meanings of the
1800 inbuilt class escapes such as \d and \s have NOT been changed.
1802 16. Updated pcredemo.c to include calls to free() to release the memory for the
1803 compiled pattern.
1805 17. The generated file chartables.c was being created in the source directory
1806 instead of in the building directory. This caused the build to fail if the
1807 source directory was different from the building directory, and was
1808 read-only.
1810 18. Added some sample Win commands from Mark Tetrode into the NON-UNIX-USE
1811 file. No doubt somebody will tell me if they don't make sense... Also added
1812 Dan Mooney's comments about building on OpenVMS.
1814 19. Added support for partial matching via the PCRE_PARTIAL option for
1815 pcre_exec() and the \P data escape in pcretest.
1817 20. Extended pcretest with 3 new pattern features:
1819 (i) A pattern option of the form ">rest-of-line" causes pcretest to
1820 write the compiled pattern to the file whose name is "rest-of-line".
1821 This is a straight binary dump of the data, with the saved pointer to
1822 the character tables forced to be NULL. The study data, if any, is
1823 written too. After writing, pcretest reads a new pattern.
1825 (ii) If, instead of a pattern, "<rest-of-line" is given, pcretest reads a
1826 compiled pattern from the given file. There must not be any
1827 occurrences of "<" in the file name (pretty unlikely); if there are,
1828 pcretest will instead treat the initial "<" as a pattern delimiter.
1829 After reading in the pattern, pcretest goes on to read data lines as
1830 usual.
1832 (iii) The F pattern option causes pcretest to flip the bytes in the 32-bit
1833 and 16-bit fields in a compiled pattern, to simulate a pattern that
1834 was compiled on a host of opposite endianness.
1836 21. The pcre-exec() function can now cope with patterns that were compiled on
1837 hosts of opposite endianness, with this restriction:
1839 As for any compiled expression that is saved and used later, the tables
1840 pointer field cannot be preserved; the extra_data field in the arguments
1841 to pcre_exec() should be used to pass in a tables address if a value
1842 other than the default internal tables were used at compile time.
1844 22. Calling pcre_exec() with a negative value of the "ovecsize" parameter is
1845 now diagnosed as an error. Previously, most of the time, a negative number
1846 would have been treated as zero, but if in addition "ovector" was passed as
1847 NULL, a crash could occur.
1849 23. Updated the files ltmain.sh, config.sub, config.guess, and aclocal.m4 with
1850 new versions from the libtool 1.5 distribution (the last one is a copy of
1851 a file called libtool.m4). This seems to have fixed the need to patch
1852 "configure" to support Darwin 1.3 (which I used to do). However, I still
1853 had to patch ltmain.sh to ensure that ${SED} is set (it isn't on my
1854 workstation).
1856 24. Changed the PCRE licence to be the more standard "BSD" licence.
1859 Version 4.5 01-Dec-03
1860 ---------------------
1862 1. There has been some re-arrangement of the code for the match() function so
1863 that it can be compiled in a version that does not call itself recursively.
1864 Instead, it keeps those local variables that need separate instances for
1865 each "recursion" in a frame on the heap, and gets/frees frames whenever it
1866 needs to "recurse". Keeping track of where control must go is done by means
1867 of setjmp/longjmp. The whole thing is implemented by a set of macros that
1868 hide most of the details from the main code, and operates only if
1869 NO_RECURSE is defined while compiling pcre.c. If PCRE is built using the
1870 "configure" mechanism, "--disable-stack-for-recursion" turns on this way of
1871 operating.
1873 To make it easier for callers to provide specially tailored get/free
1874 functions for this usage, two new functions, pcre_stack_malloc, and
1875 pcre_stack_free, are used. They are always called in strict stacking order,
1876 and the size of block requested is always the same.
1878 The PCRE_CONFIG_STACKRECURSE info parameter can be used to find out whether
1879 PCRE has been compiled to use the stack or the heap for recursion. The
1880 -C option of pcretest uses this to show which version is compiled.
1882 A new data escape \S, is added to pcretest; it causes the amounts of store
1883 obtained and freed by both kinds of malloc/free at match time to be added
1884 to the output.
1886 2. Changed the locale test to use "fr_FR" instead of "fr" because that's
1887 what's available on my current Linux desktop machine.
1889 3. When matching a UTF-8 string, the test for a valid string at the start has
1890 been extended. If start_offset is not zero, PCRE now checks that it points
1891 to a byte that is the start of a UTF-8 character. If not, it returns
1892 PCRE_ERROR_BADUTF8_OFFSET (-11). Note: the whole string is still checked;
1893 this is necessary because there may be backward assertions in the pattern.
1894 When matching the same subject several times, it may save resources to use
1895 PCRE_NO_UTF8_CHECK on all but the first call if the string is long.
1897 4. The code for checking the validity of UTF-8 strings has been tightened so
1898 that it rejects (a) strings containing 0xfe or 0xff bytes and (b) strings
1899 containing "overlong sequences".
1901 5. Fixed a bug (appearing twice) that I could not find any way of exploiting!
1902 I had written "if ((digitab[*p++] && chtab_digit) == 0)" where the "&&"
1903 should have been "&", but it just so happened that all the cases this let
1904 through by mistake were picked up later in the function.
1906 6. I had used a variable called "isblank" - this is a C99 function, causing
1907 some compilers to warn. To avoid this, I renamed it (as "blankclass").
1909 7. Cosmetic: (a) only output another newline at the end of pcretest if it is
1910 prompting; (b) run "./pcretest /dev/null" at the start of the test script
1911 so the version is shown; (c) stop "make test" echoing "./RunTest".
1913 8. Added patches from David Burgess to enable PCRE to run on EBCDIC systems.
1915 9. The prototype for memmove() for systems that don't have it was using
1916 size_t, but the inclusion of the header that defines size_t was later. I've
1917 moved the #includes for the C headers earlier to avoid this.
1919 10. Added some adjustments to the code to make it easier to compiler on certain
1920 special systems:
1922 (a) Some "const" qualifiers were missing.
1923 (b) Added the macro EXPORT before all exported functions; by default this
1924 is defined to be empty.
1925 (c) Changed the dftables auxiliary program (that builds chartables.c) so
1926 that it reads its output file name as an argument instead of writing
1927 to the standard output and assuming this can be redirected.
1929 11. In UTF-8 mode, if a recursive reference (e.g. (?1)) followed a character
1930 class containing characters with values greater than 255, PCRE compilation
1931 went into a loop.
1933 12. A recursive reference to a subpattern that was within another subpattern
1934 that had a minimum quantifier of zero caused PCRE to crash. For example,
1935 (x(y(?2))z)? provoked this bug with a subject that got as far as the
1936 recursion. If the recursively-called subpattern itself had a zero repeat,
1937 that was OK.
1939 13. In pcretest, the buffer for reading a data line was set at 30K, but the
1940 buffer into which it was copied (for escape processing) was still set at
1941 1024, so long lines caused crashes.
1943 14. A pattern such as /[ab]{1,3}+/ failed to compile, giving the error
1944 "internal error: code overflow...". This applied to any character class
1945 that was followed by a possessive quantifier.
1947 15. Modified the Makefile to add libpcre.la as a prerequisite for
1948 libpcreposix.la because I was told this is needed for a parallel build to
1949 work.
1951 16. If a pattern that contained .* following optional items at the start was
1952 studied, the wrong optimizing data was generated, leading to matching
1953 errors. For example, studying /[ab]*.*c/ concluded, erroneously, that any
1954 matching string must start with a or b or c. The correct conclusion for
1955 this pattern is that a match can start with any character.
1958 Version 4.4 13-Aug-03
1959 ---------------------
1961 1. In UTF-8 mode, a character class containing characters with values between
1962 127 and 255 was not handled correctly if the compiled pattern was studied.
1963 In fixing this, I have also improved the studying algorithm for such
1964 classes (slightly).
1966 2. Three internal functions had redundant arguments passed to them. Removal
1967 might give a very teeny performance improvement.
1969 3. Documentation bug: the value of the capture_top field in a callout is *one
1970 more than* the number of the hightest numbered captured substring.
1972 4. The Makefile linked pcretest and pcregrep with -lpcre, which could result
1973 in incorrectly linking with a previously installed version. They now link
1974 explicitly with libpcre.la.
1976 5. configure.in no longer needs to recognize Cygwin specially.
1978 6. A problem in pcre.in for Windows platforms is fixed.
1980 7. If a pattern was successfully studied, and the -d (or /D) flag was given to
1981 pcretest, it used to include the size of the study block as part of its
1982 output. Unfortunately, the structure contains a field that has a different
1983 size on different hardware architectures. This meant that the tests that
1984 showed this size failed. As the block is currently always of a fixed size,
1985 this information isn't actually particularly useful in pcretest output, so
1986 I have just removed it.
1988 8. Three pre-processor statements accidentally did not start in column 1.
1989 Sadly, there are *still* compilers around that complain, even though
1990 standard C has not required this for well over a decade. Sigh.
1992 9. In pcretest, the code for checking callouts passed small integers in the
1993 callout_data field, which is a void * field. However, some picky compilers
1994 complained about the casts involved for this on 64-bit systems. Now
1995 pcretest passes the address of the small integer instead, which should get
1996 rid of the warnings.
1998 10. By default, when in UTF-8 mode, PCRE now checks for valid UTF-8 strings at
1999 both compile and run time, and gives an error if an invalid UTF-8 sequence
2000 is found. There is a option for disabling this check in cases where the
2001 string is known to be correct and/or the maximum performance is wanted.
2003 11. In response to a bug report, I changed one line in Makefile.in from
2005 -Wl,--out-implib,.libs/lib@WIN_PREFIX@pcreposix.dll.a \
2006 to
2007 -Wl,--out-implib,.libs/@WIN_PREFIX@libpcreposix.dll.a \
2009 to look similar to other lines, but I have no way of telling whether this
2010 is the right thing to do, as I do not use Windows. No doubt I'll get told
2011 if it's wrong...
2014 Version 4.3 21-May-03
2015 ---------------------
2017 1. Two instances of @WIN_PREFIX@ omitted from the Windows targets in the
2018 Makefile.
2020 2. Some refactoring to improve the quality of the code:
2022 (i) The utf8_table... variables are now declared "const".
2024 (ii) The code for \cx, which used the "case flipping" table to upper case
2025 lower case letters, now just substracts 32. This is ASCII-specific,
2026 but the whole concept of \cx is ASCII-specific, so it seems
2027 reasonable.
2029 (iii) PCRE was using its character types table to recognize decimal and
2030 hexadecimal digits in the pattern. This is silly, because it handles
2031 only 0-9, a-f, and A-F, but the character types table is locale-
2032 specific, which means strange things might happen. A private
2033 table is now used for this - though it costs 256 bytes, a table is
2034 much faster than multiple explicit tests. Of course, the standard
2035 character types table is still used for matching digits in subject
2036 strings against \d.
2038 (iv) Strictly, the identifier ESC_t is reserved by POSIX (all identifiers
2039 ending in _t are). So I've renamed it as ESC_tee.
2041 3. The first argument for regexec() in the POSIX wrapper should have been
2042 defined as "const".
2044 4. Changed pcretest to use malloc() for its buffers so that they can be
2045 Electric Fenced for debugging.
2047 5. There were several places in the code where, in UTF-8 mode, PCRE would try
2048 to read one or more bytes before the start of the subject string. Often this
2049 had no effect on PCRE's behaviour, but in some circumstances it could
2050 provoke a segmentation fault.
2052 6. A lookbehind at the start of a pattern in UTF-8 mode could also cause PCRE
2053 to try to read one or more bytes before the start of the subject string.
2055 7. A lookbehind in a pattern matched in non-UTF-8 mode on a PCRE compiled with
2056 UTF-8 support could misbehave in various ways if the subject string
2057 contained bytes with the 0x80 bit set and the 0x40 bit unset in a lookbehind
2058 area. (PCRE was not checking for the UTF-8 mode flag, and trying to move
2059 back over UTF-8 characters.)
2062 Version 4.2 14-Apr-03
2063 ---------------------
2065 1. Typo "#if SUPPORT_UTF8" instead of "#ifdef SUPPORT_UTF8" fixed.
2067 2. Changes to the building process, supplied by Ronald Landheer-Cieslak
2068 [ON_WINDOWS]: new variable, "#" on non-Windows platforms
2069 [NOT_ON_WINDOWS]: new variable, "#" on Windows platforms
2070 [WIN_PREFIX]: new variable, "cyg" for Cygwin
2071 * Makefile.in: use autoconf substitution for OBJEXT, EXEEXT, BUILD_OBJEXT
2073 Note: automatic setting of the BUILD variables is not yet working
2074 set CPPFLAGS and BUILD_CPPFLAGS (but don't use yet) - should be used at
2075 compile-time but not at link-time
2076 [LINK]: use for linking executables only
2077 make different versions for Windows and non-Windows
2078 [LINKLIB]: new variable, copy of UNIX-style LINK, used for linking
2079 libraries
2080 [LINK_FOR_BUILD]: new variable
2081 [OBJEXT]: use throughout
2082 [EXEEXT]: use throughout
2083 <winshared>: new target
2084 <wininstall>: new target
2085 <dftables.o>: use native compiler
2086 <dftables>: use native linker
2087 <install>: handle Windows platform correctly
2088 <clean>: ditto
2089 <check>: ditto
2090 copy DLL to top builddir before testing
2092 As part of these changes, -no-undefined was removed again. This was reported
2093 to give trouble on HP-UX 11.0, so getting rid of it seems like a good idea
2094 in any case.
2096 3. Some tidies to get rid of compiler warnings:
2098 . In the match_data structure, match_limit was an unsigned long int, whereas
2099 match_call_count was an int. I've made them both unsigned long ints.
2101 . In pcretest the fact that a const uschar * doesn't automatically cast to
2102 a void * provoked a warning.
2104 . Turning on some more compiler warnings threw up some "shadow" variables
2105 and a few more missing casts.
2107 4. If PCRE was complied with UTF-8 support, but called without the PCRE_UTF8
2108 option, a class that contained a single character with a value between 128
2109 and 255 (e.g. /[\xFF]/) caused PCRE to crash.
2111 5. If PCRE was compiled with UTF-8 support, but called without the PCRE_UTF8
2112 option, a class that contained several characters, but with at least one
2113 whose value was between 128 and 255 caused PCRE to crash.
2116 Version 4.1 12-Mar-03
2117 ---------------------
2119 1. Compiling with gcc -pedantic found a couple of places where casts were
2120 needed, and a string in dftables.c that was longer than standard compilers are
2121 required to support.
2123 2. Compiling with Sun's compiler found a few more places where the code could
2124 be tidied up in order to avoid warnings.
2126 3. The variables for cross-compiling were called HOST_CC and HOST_CFLAGS; the
2127 first of these names is deprecated in the latest Autoconf in favour of the name
2128 CC_FOR_BUILD, because "host" is typically used to mean the system on which the
2129 compiled code will be run. I can't find a reference for HOST_CFLAGS, but by
2130 analogy I have changed it to CFLAGS_FOR_BUILD.
2132 4. Added -no-undefined to the linking command in the Makefile, because this is
2133 apparently helpful for Windows. To make it work, also added "-L. -lpcre" to the
2134 linking step for the pcreposix library.
2136 5. PCRE was failing to diagnose the case of two named groups with the same
2137 name.
2139 6. A problem with one of PCRE's optimizations was discovered. PCRE remembers a
2140 literal character that is needed in the subject for a match, and scans along to
2141 ensure that it is present before embarking on the full matching process. This
2142 saves time in cases of nested unlimited repeats that are never going to match.
2143 Problem: the scan can take a lot of time if the subject is very long (e.g.
2144 megabytes), thus penalizing straightforward matches. It is now done only if the
2145 amount of subject to be scanned is less than 1000 bytes.
2147 7. A lesser problem with the same optimization is that it was recording the
2148 first character of an anchored pattern as "needed", thus provoking a search
2149 right along the subject, even when the first match of the pattern was going to
2150 fail. The "needed" character is now not set for anchored patterns, unless it
2151 follows something in the pattern that is of non-fixed length. Thus, it still
2152 fulfils its original purpose of finding quick non-matches in cases of nested
2153 unlimited repeats, but isn't used for simple anchored patterns such as /^abc/.
2156 Version 4.0 17-Feb-03
2157 ---------------------
2159 1. If a comment in an extended regex that started immediately after a meta-item
2160 extended to the end of string, PCRE compiled incorrect data. This could lead to
2161 all kinds of weird effects. Example: /#/ was bad; /()#/ was bad; /a#/ was not.
2163 2. Moved to autoconf 2.53 and libtool 1.4.2.
2165 3. Perl 5.8 no longer needs "use utf8" for doing UTF-8 things. Consequently,
2166 the special perltest8 script is no longer needed - all the tests can be run
2167 from a single perltest script.
2169 4. From 5.004, Perl has not included the VT character (0x0b) in the set defined
2170 by \s. It has now been removed in PCRE. This means it isn't recognized as
2171 whitespace in /x regexes too, which is the same as Perl. Note that the POSIX
2172 class [:space:] *does* include VT, thereby creating a mess.
2174 5. Added the class [:blank:] (a GNU extension from Perl 5.8) to match only
2175 space and tab.
2177 6. Perl 5.005 was a long time ago. It's time to amalgamate the tests that use
2178 its new features into the main test script, reducing the number of scripts.
2180 7. Perl 5.8 has changed the meaning of patterns like /a(?i)b/. Earlier versions
2181 were backward compatible, and made the (?i) apply to the whole pattern, as if
2182 /i were given. Now it behaves more logically, and applies the option setting
2183 only to what follows. PCRE has been changed to follow suit. However, if it
2184 finds options settings right at the start of the pattern, it extracts them into
2185 the global options, as before. Thus, they show up in the info data.
2187 8. Added support for the \Q...\E escape sequence. Characters in between are
2188 treated as literals. This is slightly different from Perl in that $ and @ are
2189 also handled as literals inside the quotes. In Perl, they will cause variable
2190 interpolation. Note the following examples:
2192 Pattern PCRE matches Perl matches
2194 \Qabc$xyz\E abc$xyz abc followed by the contents of $xyz
2195 \Qabc\$xyz\E abc\$xyz abc\$xyz
2196 \Qabc\E\$\Qxyz\E abc$xyz abc$xyz
2198 For compatibility with Perl, \Q...\E sequences are recognized inside character
2199 classes as well as outside them.
2201 9. Re-organized 3 code statements in pcretest to avoid "overflow in
2202 floating-point constant arithmetic" warnings from a Microsoft compiler. Added a
2203 (size_t) cast to one statement in pcretest and one in pcreposix to avoid
2204 signed/unsigned warnings.
2206 10. SunOS4 doesn't have strtoul(). This was used only for unpicking the -o
2207 option for pcretest, so I've replaced it by a simple function that does just
2208 that job.
2210 11. pcregrep was ending with code 0 instead of 2 for the commands "pcregrep" or
2211 "pcregrep -".
2213 12. Added "possessive quantifiers" ?+, *+, ++, and {,}+ which come from Sun's
2214 Java package. This provides some syntactic sugar for simple cases of what my
2215 documentation calls "once-only subpatterns". A pattern such as x*+ is the same
2216 as (?>x*). In other words, if what is inside (?>...) is just a single repeated
2217 item, you can use this simplified notation. Note that only makes sense with
2218 greedy quantifiers. Consequently, the use of the possessive quantifier forces
2219 greediness, whatever the setting of the PCRE_UNGREEDY option.
2221 13. A change of greediness default within a pattern was not taking effect at
2222 the current level for patterns like /(b+(?U)a+)/. It did apply to parenthesized
2223 subpatterns that followed. Patterns like /b+(?U)a+/ worked because the option
2224 was abstracted outside.
2226 14. PCRE now supports the \G assertion. It is true when the current matching
2227 position is at the start point of the match. This differs from \A when the
2228 starting offset is non-zero. Used with the /g option of pcretest (or similar
2229 code), it works in the same way as it does for Perl's /g option. If all
2230 alternatives of a regex begin with \G, the expression is anchored to the start
2231 match position, and the "anchored" flag is set in the compiled expression.
2233 15. Some bugs concerning the handling of certain option changes within patterns
2234 have been fixed. These applied to options other than (?ims). For example,
2235 "a(?x: b c )d" did not match "XabcdY" but did match "Xa b c dY". It should have
2236 been the other way round. Some of this was related to change 7 above.
2238 16. PCRE now gives errors for /[.x.]/ and /[=x=]/ as unsupported POSIX
2239 features, as Perl does. Previously, PCRE gave the warnings only for /[[.x.]]/
2240 and /[[=x=]]/. PCRE now also gives an error for /[:name:]/ because it supports
2241 POSIX classes only within a class (e.g. /[[:alpha:]]/).
2243 17. Added support for Perl's \C escape. This matches one byte, even in UTF8
2244 mode. Unlike ".", it always matches newline, whatever the setting of
2245 PCRE_DOTALL. However, PCRE does not permit \C to appear in lookbehind
2246 assertions. Perl allows it, but it doesn't (in general) work because it can't
2247 calculate the length of the lookbehind. At least, that's the case for Perl
2248 5.8.0 - I've been told they are going to document that it doesn't work in
2249 future.
2251 18. Added an error diagnosis for escapes that PCRE does not support: these are
2252 \L, \l, \N, \P, \p, \U, \u, and \X.
2254 19. Although correctly diagnosing a missing ']' in a character class, PCRE was
2255 reading past the end of the pattern in cases such as /[abcd/.
2257 20. PCRE was getting more memory than necessary for patterns with classes that
2258 contained both POSIX named classes and other characters, e.g. /[[:space:]abc/.
2260 21. Added some code, conditional on #ifdef VPCOMPAT, to make life easier for
2261 compiling PCRE for use with Virtual Pascal.
2263 22. Small fix to the Makefile to make it work properly if the build is done
2264 outside the source tree.
2266 23. Added a new extension: a condition to go with recursion. If a conditional
2267 subpattern starts with (?(R) the "true" branch is used if recursion has
2268 happened, whereas the "false" branch is used only at the top level.
2270 24. When there was a very long string of literal characters (over 255 bytes
2271 without UTF support, over 250 bytes with UTF support), the computation of how
2272 much memory was required could be incorrect, leading to segfaults or other
2273 strange effects.
2275 25. PCRE was incorrectly assuming anchoring (either to start of subject or to
2276 start of line for a non-DOTALL pattern) when a pattern started with (.*) and
2277 there was a subsequent back reference to those brackets. This meant that, for
2278 example, /(.*)\d+\1/ failed to match "abc123bc". Unfortunately, it isn't
2279 possible to check for precisely this case. All we can do is abandon the
2280 optimization if .* occurs inside capturing brackets when there are any back
2281 references whatsoever. (See below for a better fix that came later.)
2283 26. The handling of the optimization for finding the first character of a
2284 non-anchored pattern, and for finding a character that is required later in the
2285 match were failing in some cases. This didn't break the matching; it just
2286 failed to optimize when it could. The way this is done has been re-implemented.
2288 27. Fixed typo in error message for invalid (?R item (it said "(?p").
2290 28. Added a new feature that provides some of the functionality that Perl
2291 provides with (?{...}). The facility is termed a "callout". The way it is done
2292 in PCRE is for the caller to provide an optional function, by setting
2293 pcre_callout to its entry point. Like pcre_malloc and pcre_free, this is a
2294 global variable. By default it is unset, which disables all calling out. To get
2295 the function called, the regex must include (?C) at appropriate points. This
2296 is, in fact, equivalent to (?C0), and any number <= 255 may be given with (?C).
2297 This provides a means of identifying different callout points. When PCRE
2298 reaches such a point in the regex, if pcre_callout has been set, the external
2299 function is called. It is provided with data in a structure called
2300 pcre_callout_block, which is defined in pcre.h. If the function returns 0,
2301 matching continues; if it returns a non-zero value, the match at the current
2302 point fails. However, backtracking will occur if possible. [This was changed
2303 later and other features added - see item 49 below.]
2305 29. pcretest is upgraded to test the callout functionality. It provides a
2306 callout function that displays information. By default, it shows the start of
2307 the match and the current position in the text. There are some new data escapes
2308 to vary what happens:
2310 \C+ in addition, show current contents of captured substrings
2311 \C- do not supply a callout function
2312 \C!n return 1 when callout number n is reached
2313 \C!n!m return 1 when callout number n is reached for the mth time
2315 30. If pcregrep was called with the -l option and just a single file name, it
2316 output "<stdin>" if a match was found, instead of the file name.
2318 31. Improve the efficiency of the POSIX API to PCRE. If the number of capturing
2319 slots is less than POSIX_MALLOC_THRESHOLD, use a block on the stack to pass to
2320 pcre_exec(). This saves a malloc/free per call. The default value of
2321 POSIX_MALLOC_THRESHOLD is 10; it can be changed by --with-posix-malloc-threshold
2322 when configuring.
2324 32. The default maximum size of a compiled pattern is 64K. There have been a
2325 few cases of people hitting this limit. The code now uses macros to handle the
2326 storing of links as offsets within the compiled pattern. It defaults to 2-byte
2327 links, but this can be changed to 3 or 4 bytes by --with-link-size when
2328 configuring. Tests 2 and 5 work only with 2-byte links because they output
2329 debugging information about compiled patterns.
2331 33. Internal code re-arrangements:
2333 (a) Moved the debugging function for printing out a compiled regex into
2334 its own source file (printint.c) and used #include to pull it into
2335 pcretest.c and, when DEBUG is defined, into pcre.c, instead of having two
2336 separate copies.
2338 (b) Defined the list of op-code names for debugging as a macro in
2339 internal.h so that it is next to the definition of the opcodes.
2341 (c) Defined a table of op-code lengths for simpler skipping along compiled
2342 code. This is again a macro in internal.h so that it is next to the
2343 definition of the opcodes.
2345 34. Added support for recursive calls to individual subpatterns, along the
2346 lines of Robin Houston's patch (but implemented somewhat differently).
2348 35. Further mods to the Makefile to help Win32. Also, added code to pcregrep to
2349 allow it to read and process whole directories in Win32. This code was
2350 contributed by Lionel Fourquaux; it has not been tested by me.
2352 36. Added support for named subpatterns. The Python syntax (?P<name>...) is
2353 used to name a group. Names consist of alphanumerics and underscores, and must
2354 be unique. Back references use the syntax (?P=name) and recursive calls use
2355 (?P>name) which is a PCRE extension to the Python extension. Groups still have
2356 numbers. The function pcre_fullinfo() can be used after compilation to extract
2357 a name/number map. There are three relevant calls:
2359 PCRE_INFO_NAMEENTRYSIZE yields the size of each entry in the map
2360 PCRE_INFO_NAMECOUNT yields the number of entries
2361 PCRE_INFO_NAMETABLE yields a pointer to the map.
2363 The map is a vector of fixed-size entries. The size of each entry depends on
2364 the length of the longest name used. The first two bytes of each entry are the
2365 group number, most significant byte first. There follows the corresponding
2366 name, zero terminated. The names are in alphabetical order.
2368 37. Make the maximum literal string in the compiled code 250 for the non-UTF-8
2369 case instead of 255. Making it the same both with and without UTF-8 support
2370 means that the same test output works with both.
2372 38. There was a case of malloc(0) in the POSIX testing code in pcretest. Avoid
2373 calling malloc() with a zero argument.
2375 39. Change 25 above had to resort to a heavy-handed test for the .* anchoring
2376 optimization. I've improved things by keeping a bitmap of backreferences with
2377 numbers 1-31 so that if .* occurs inside capturing brackets that are not in
2378 fact referenced, the optimization can be applied. It is unlikely that a
2379 relevant occurrence of .* (i.e. one which might indicate anchoring or forcing
2380 the match to follow \n) will appear inside brackets with a number greater than
2381 31, but if it does, any back reference > 31 suppresses the optimization.
2383 40. Added a new compile-time option PCRE_NO_AUTO_CAPTURE. This has the effect
2384 of disabling numbered capturing parentheses. Any opening parenthesis that is
2385 not followed by ? behaves as if it were followed by ?: but named parentheses
2386 can still be used for capturing (and they will acquire numbers in the usual
2387 way).
2389 41. Redesigned the return codes from the match() function into yes/no/error so
2390 that errors can be passed back from deep inside the nested calls. A malloc
2391 failure while inside a recursive subpattern call now causes the
2392 PCRE_ERROR_NOMEMORY return instead of quietly going wrong.
2394 42. It is now possible to set a limit on the number of times the match()
2395 function is called in a call to pcre_exec(). This facility makes it possible to
2396 limit the amount of recursion and backtracking, though not in a directly
2397 obvious way, because the match() function is used in a number of different
2398 circumstances. The count starts from zero for each position in the subject
2399 string (for non-anchored patterns). The default limit is, for compatibility, a
2400 large number, namely 10 000 000. You can change this in two ways:
2402 (a) When configuring PCRE before making, you can use --with-match-limit=n
2403 to set a default value for the compiled library.
2405 (b) For each call to pcre_exec(), you can pass a pcre_extra block in which
2406 a different value is set. See 45 below.
2408 If the limit is exceeded, pcre_exec() returns PCRE_ERROR_MATCHLIMIT.
2410 43. Added a new function pcre_config(int, void *) to enable run-time extraction
2411 of things that can be changed at compile time. The first argument specifies
2412 what is wanted and the second points to where the information is to be placed.
2413 The current list of available information is:
2417 The output is an integer that is set to one if UTF-8 support is available;
2418 otherwise it is set to zero.
2422 The output is an integer that it set to the value of the code that is used for
2423 newline. It is either LF (10) or CR (13).
2427 The output is an integer that contains the number of bytes used for internal
2428 linkage in compiled expressions. The value is 2, 3, or 4. See item 32 above.
2432 The output is an integer that contains the threshold above which the POSIX
2433 interface uses malloc() for output vectors. See item 31 above.
2437 The output is an unsigned integer that contains the default limit of the number
2438 of match() calls in a pcre_exec() execution. See 42 above.
2440 44. pcretest has been upgraded by the addition of the -C option. This causes it
2441 to extract all the available output from the new pcre_config() function, and to
2442 output it. The program then exits immediately.
2444 45. A need has arisen to pass over additional data with calls to pcre_exec() in
2445 order to support additional features. One way would have been to define
2446 pcre_exec2() (for example) with extra arguments, but this would not have been
2447 extensible, and would also have required all calls to the original function to
2448 be mapped to the new one. Instead, I have chosen to extend the mechanism that
2449 is used for passing in "extra" data from pcre_study().
2451 The pcre_extra structure is now exposed and defined in pcre.h. It currently
2452 contains the following fields:
2454 flags a bitmap indicating which of the following fields are set
2455 study_data opaque data from pcre_study()
2456 match_limit a way of specifying a limit on match() calls for a specific
2457 call to pcre_exec()
2458 callout_data data for callouts (see 49 below)
2460 The flag bits are also defined in pcre.h, and are
2466 The pcre_study() function now returns one of these new pcre_extra blocks, with
2467 the actual study data pointed to by the study_data field, and the
2468 PCRE_EXTRA_STUDY_DATA flag set. This can be passed directly to pcre_exec() as
2469 before. That is, this change is entirely upwards-compatible and requires no
2470 change to existing code.
2472 If you want to pass in additional data to pcre_exec(), you can either place it
2473 in a pcre_extra block provided by pcre_study(), or create your own pcre_extra
2474 block.
2476 46. pcretest has been extended to test the PCRE_EXTRA_MATCH_LIMIT feature. If a
2477 data string contains the escape sequence \M, pcretest calls pcre_exec() several
2478 times with different match limits, until it finds the minimum value needed for
2479 pcre_exec() to complete. The value is then output. This can be instructive; for
2480 most simple matches the number is quite small, but for pathological cases it
2481 gets very large very quickly.
2483 47. There's a new option for pcre_fullinfo() called PCRE_INFO_STUDYSIZE. It
2484 returns the size of the data block pointed to by the study_data field in a
2485 pcre_extra block, that is, the value that was passed as the argument to
2486 pcre_malloc() when PCRE was getting memory in which to place the information
2487 created by pcre_study(). The fourth argument should point to a size_t variable.
2488 pcretest has been extended so that this information is shown after a successful
2489 pcre_study() call when information about the compiled regex is being displayed.
2491 48. Cosmetic change to Makefile: there's no need to have / after $(DESTDIR)
2492 because what follows is always an absolute path. (Later: it turns out that this
2493 is more than cosmetic for MinGW, because it doesn't like empty path
2494 components.)
2496 49. Some changes have been made to the callout feature (see 28 above):
2498 (i) A callout function now has three choices for what it returns:
2500 0 => success, carry on matching
2501 > 0 => failure at this point, but backtrack if possible
2502 < 0 => serious error, return this value from pcre_exec()
2504 Negative values should normally be chosen from the set of PCRE_ERROR_xxx
2505 values. In particular, returning PCRE_ERROR_NOMATCH forces a standard
2506 "match failed" error. The error number PCRE_ERROR_CALLOUT is reserved for
2507 use by callout functions. It will never be used by PCRE itself.
2509 (ii) The pcre_extra structure (see 45 above) has a void * field called
2510 callout_data, with corresponding flag bit PCRE_EXTRA_CALLOUT_DATA. The
2511 pcre_callout_block structure has a field of the same name. The contents of
2512 the field passed in the pcre_extra structure are passed to the callout
2513 function in the corresponding field in the callout block. This makes it
2514 easier to use the same callout-containing regex from multiple threads. For
2515 testing, the pcretest program has a new data escape
2517 \C*n pass the number n (may be negative) as callout_data
2519 If the callout function in pcretest receives a non-zero value as
2520 callout_data, it returns that value.
2522 50. Makefile wasn't handling CFLAGS properly when compiling dftables. Also,
2523 there were some redundant $(CFLAGS) in commands that are now specified as
2524 $(LINK), which already includes $(CFLAGS).
2526 51. Extensions to UTF-8 support are listed below. These all apply when (a) PCRE
2527 has been compiled with UTF-8 support *and* pcre_compile() has been compiled
2528 with the PCRE_UTF8 flag. Patterns that are compiled without that flag assume
2529 one-byte characters throughout. Note that case-insensitive matching applies
2530 only to characters whose values are less than 256. PCRE doesn't support the
2531 notion of cases for higher-valued characters.
2533 (i) A character class whose characters are all within 0-255 is handled as
2534 a bit map, and the map is inverted for negative classes. Previously, a
2535 character > 255 always failed to match such a class; however it should
2536 match if the class was a negative one (e.g. [^ab]). This has been fixed.
2538 (ii) A negated character class with a single character < 255 is coded as
2539 "not this character" (OP_NOT). This wasn't working properly when the test
2540 character was multibyte, either singly or repeated.
2542 (iii) Repeats of multibyte characters are now handled correctly in UTF-8
2543 mode, for example: \x{100}{2,3}.
2545 (iv) The character escapes \b, \B, \d, \D, \s, \S, \w, and \W (either
2546 singly or repeated) now correctly test multibyte characters. However,
2547 PCRE doesn't recognize any characters with values greater than 255 as
2548 digits, spaces, or word characters. Such characters always match \D, \S,
2549 and \W, and never match \d, \s, or \w.
2551 (v) Classes may now contain characters and character ranges with values
2552 greater than 255. For example: [ab\x{100}-\x{400}].
2554 (vi) pcregrep now has a --utf-8 option (synonym -u) which makes it call
2555 PCRE in UTF-8 mode.
2557 52. The info request value PCRE_INFO_FIRSTCHAR has been renamed
2558 PCRE_INFO_FIRSTBYTE because it is a byte value. However, the old name is
2559 retained for backwards compatibility. (Note that LASTLITERAL is also a byte
2560 value.)
2562 53. The single man page has become too large. I have therefore split it up into
2563 a number of separate man pages. These also give rise to individual HTML pages;
2564 these are now put in a separate directory, and there is an index.html page that
2565 lists them all. Some hyperlinking between the pages has been installed.
2567 54. Added convenience functions for handling named capturing parentheses.
2569 55. Unknown escapes inside character classes (e.g. [\M]) and escapes that
2570 aren't interpreted therein (e.g. [\C]) are literals in Perl. This is now also
2571 true in PCRE, except when the PCRE_EXTENDED option is set, in which case they
2572 are faulted.
2574 56. Introduced HOST_CC and HOST_CFLAGS which can be set in the environment when
2575 calling configure. These values are used when compiling the dftables.c program
2576 which is run to generate the source of the default character tables. They
2577 default to the values of CC and CFLAGS. If you are cross-compiling PCRE,
2578 you will need to set these values.
2580 57. Updated the building process for Windows DLL, as provided by Fred Cox.
2583 Version 3.9 02-Jan-02
2584 ---------------------
2586 1. A bit of extraneous text had somehow crept into the pcregrep documentation.
2588 2. If --disable-static was given, the building process failed when trying to
2589 build pcretest and pcregrep. (For some reason it was using libtool to compile
2590 them, which is not right, as they aren't part of the library.)
2593 Version 3.8 18-Dec-01
2594 ---------------------
2596 1. The experimental UTF-8 code was completely screwed up. It was packing the
2597 bytes in the wrong order. How dumb can you get?
2600 Version 3.7 29-Oct-01
2601 ---------------------
2603 1. In updating pcretest to check change 1 of version 3.6, I screwed up.
2604 This caused pcretest, when used on the test data, to segfault. Unfortunately,
2605 this didn't happen under Solaris 8, where I normally test things.
2607 2. The Makefile had to be changed to make it work on BSD systems, where 'make'
2608 doesn't seem to recognize that ./xxx and xxx are the same file. (This entry
2609 isn't in ChangeLog distributed with 3.7 because I forgot when I hastily made
2610 this fix an hour or so after the initial 3.7 release.)
2613 Version 3.6 23-Oct-01
2614 ---------------------
2616 1. Crashed with /(sens|respons)e and \1ibility/ and "sense and sensibility" if
2617 offsets passed as NULL with zero offset count.
2619 2. The config.guess and config.sub files had not been updated when I moved to
2620 the latest autoconf.
2623 Version 3.5 15-Aug-01
2624 ---------------------
2626 1. Added some missing #if !defined NOPOSIX conditionals in pcretest.c that
2627 had been forgotten.
2629 2. By using declared but undefined structures, we can avoid using "void"
2630 definitions in pcre.h while keeping the internal definitions of the structures
2631 private.
2633 3. The distribution is now built using autoconf 2.50 and libtool 1.4. From a
2634 user point of view, this means that both static and shared libraries are built
2635 by default, but this can be individually controlled. More of the work of
2636 handling this static/shared cases is now inside libtool instead of PCRE's make
2637 file.
2639 4. The pcretest utility is now installed along with pcregrep because it is
2640 useful for users (to test regexs) and by doing this, it automatically gets
2641 relinked by libtool. The documentation has been turned into a man page, so
2642 there are now .1, .txt, and .html versions in /doc.
2644 5. Upgrades to pcregrep:
2645 (i) Added long-form option names like gnu grep.
2646 (ii) Added --help to list all options with an explanatory phrase.
2647 (iii) Added -r, --recursive to recurse into sub-directories.
2648 (iv) Added -f, --file to read patterns from a file.
2650 6. pcre_exec() was referring to its "code" argument before testing that
2651 argument for NULL (and giving an error if it was NULL).
2653 7. Upgraded Makefile.in to allow for compiling in a different directory from
2654 the source directory.
2656 8. Tiny buglet in pcretest: when pcre_fullinfo() was called to retrieve the
2657 options bits, the pointer it was passed was to an int instead of to an unsigned
2658 long int. This mattered only on 64-bit systems.
2660 9. Fixed typo (3.4/1) in pcre.h again. Sigh. I had changed pcre.h (which is
2661 generated) instead of pcre.in, which it its source. Also made the same change
2662 in several of the .c files.
2664 10. A new release of gcc defines printf() as a macro, which broke pcretest
2665 because it had an ifdef in the middle of a string argument for printf(). Fixed
2666 by using separate calls to printf().
2668 11. Added --enable-newline-is-cr and --enable-newline-is-lf to the configure
2669 script, to force use of CR or LF instead of \n in the source. On non-Unix
2670 systems, the value can be set in config.h.
2672 12. The limit of 200 on non-capturing parentheses is a _nesting_ limit, not an
2673 absolute limit. Changed the text of the error message to make this clear, and
2674 likewise updated the man page.
2676 13. The limit of 99 on the number of capturing subpatterns has been removed.
2677 The new limit is 65535, which I hope will not be a "real" limit.
2680 Version 3.4 22-Aug-00
2681 ---------------------
2683 1. Fixed typo in pcre.h: unsigned const char * changed to const unsigned char *.
2685 2. Diagnose condition (?(0) as an error instead of crashing on matching.
2688 Version 3.3 01-Aug-00
2689 ---------------------
2691 1. If an octal character was given, but the value was greater than \377, it
2692 was not getting masked to the least significant bits, as documented. This could
2693 lead to crashes in some systems.
2695 2. Perl 5.6 (if not earlier versions) accepts classes like [a-\d] and treats
2696 the hyphen as a literal. PCRE used to give an error; it now behaves like Perl.
2698 3. Added the functions pcre_free_substring() and pcre_free_substring_list().
2699 These just pass their arguments on to (pcre_free)(), but they are provided
2700 because some uses of PCRE bind it to non-C systems that can call its functions,
2701 but cannot call free() or pcre_free() directly.
2703 4. Add "make test" as a synonym for "make check". Corrected some comments in
2704 the Makefile.
2706 5. Add $(DESTDIR)/ in front of all the paths in the "install" target in the
2707 Makefile.
2709 6. Changed the name of pgrep to pcregrep, because Solaris has introduced a
2710 command called pgrep for grepping around the active processes.
2712 7. Added the beginnings of support for UTF-8 character strings.
2714 8. Arranged for the Makefile to pass over the settings of CC, CFLAGS, and
2715 RANLIB to ./ltconfig so that they are used by libtool. I think these are all
2716 the relevant ones. (AR is not passed because ./ltconfig does its own figuring
2717 out for the ar command.)
2720 Version 3.2 12-May-00
2721 ---------------------
2723 This is purely a bug fixing release.
2725 1. If the pattern /((Z)+|A)*/ was matched agained ZABCDEFG it matched Z instead
2726 of ZA. This was just one example of several cases that could provoke this bug,
2727 which was introduced by change 9 of version 2.00. The code for breaking
2728 infinite loops after an iteration that matches an empty string was't working
2729 correctly.
2731 2. The pcretest program was not imitating Perl correctly for the pattern /a*/g
2732 when matched against abbab (for example). After matching an empty string, it
2733 wasn't forcing anchoring when setting PCRE_NOTEMPTY for the next attempt; this
2734 caused it to match further down the string than it should.
2736 3. The code contained an inclusion of sys/types.h. It isn't clear why this
2737 was there because it doesn't seem to be needed, and it causes trouble on some
2738 systems, as it is not a Standard C header. It has been removed.
2740 4. Made 4 silly changes to the source to avoid stupid compiler warnings that
2741 were reported on the Macintosh. The changes were from
2743 while ((c = *(++ptr)) != 0 && c != '\n');
2744 to
2745 while ((c = *(++ptr)) != 0 && c != '\n') ;
2747 Totally extraordinary, but if that's what it takes...
2749 5. PCRE is being used in one environment where neither memmove() nor bcopy() is
2750 available. Added HAVE_BCOPY and an autoconf test for it; if neither
2751 HAVE_MEMMOVE nor HAVE_BCOPY is set, use a built-in emulation function which
2752 assumes the way PCRE uses memmove() (always moving upwards).
2754 6. PCRE is being used in one environment where strchr() is not available. There
2755 was only one use in pcre.c, and writing it out to avoid strchr() probably gives
2756 faster code anyway.
2759 Version 3.1 09-Feb-00
2760 ---------------------
2762 The only change in this release is the fixing of some bugs in Makefile.in for
2763 the "install" target:
2765 (1) It was failing to install pcreposix.h.
2767 (2) It was overwriting the pcre.3 man page with the pcreposix.3 man page.
2770 Version 3.0 01-Feb-00
2771 ---------------------
2773 1. Add support for the /+ modifier to perltest (to output $` like it does in
2774 pcretest).
2776 2. Add support for the /g modifier to perltest.
2778 3. Fix pcretest so that it behaves even more like Perl for /g when the pattern
2779 matches null strings.
2781 4. Fix perltest so that it doesn't do unwanted things when fed an empty
2782 pattern. Perl treats empty patterns specially - it reuses the most recent
2783 pattern, which is not what we want. Replace // by /(?#)/ in order to avoid this
2784 effect.
2786 5. The POSIX interface was broken in that it was just handing over the POSIX
2787 captured string vector to pcre_exec(), but (since release 2.00) PCRE has
2788 required a bigger vector, with some working space on the end. This means that
2789 the POSIX wrapper now has to get and free some memory, and copy the results.
2791 6. Added some simple autoconf support, placing the test data and the
2792 documentation in separate directories, re-organizing some of the
2793 information files, and making it build pcre-config (a GNU standard). Also added
2794 libtool support for building PCRE as a shared library, which is now the
2795 default.
2797 7. Got rid of the leading zero in the definition of PCRE_MINOR because 08 and
2798 09 are not valid octal constants. Single digits will be used for minor values
2799 less than 10.
2801 8. Defined REG_EXTENDED and REG_NOSUB as zero in the POSIX header, so that
2802 existing programs that set these in the POSIX interface can use PCRE without
2803 modification.
2805 9. Added a new function, pcre_fullinfo() with an extensible interface. It can
2806 return all that pcre_info() returns, plus additional data. The pcre_info()
2807 function is retained for compatibility, but is considered to be obsolete.
2809 10. Added experimental recursion feature (?R) to handle one common case that
2810 Perl 5.6 will be able to do with (?p{...}).
2812 11. Added support for POSIX character classes like [:alpha:], which Perl is
2813 adopting.
2816 Version 2.08 31-Aug-99
2817 ----------------------
2819 1. When startoffset was not zero and the pattern began with ".*", PCRE was not
2820 trying to match at the startoffset position, but instead was moving forward to
2821 the next newline as if a previous match had failed.
2823 2. pcretest was not making use of PCRE_NOTEMPTY when repeating for /g and /G,
2824 and could get into a loop if a null string was matched other than at the start
2825 of the subject.
2827 3. Added definitions of PCRE_MAJOR and PCRE_MINOR to pcre.h so the version can
2828 be distinguished at compile time, and for completeness also added PCRE_DATE.
2830 5. Added Paul Sokolovsky's minor changes to make it easy to compile a Win32 DLL
2831 in GnuWin32 environments.
2834 Version 2.07 29-Jul-99
2835 ----------------------
2837 1. The documentation is now supplied in plain text form and HTML as well as in
2838 the form of man page sources.
2840 2. C++ compilers don't like assigning (void *) values to other pointer types.
2841 In particular this affects malloc(). Although there is no problem in Standard
2842 C, I've put in casts to keep C++ compilers happy.
2844 3. Typo on pcretest.c; a cast of (unsigned char *) in the POSIX regexec() call
2845 should be (const char *).
2847 4. If NOPOSIX is defined, pcretest.c compiles without POSIX support. This may
2848 be useful for non-Unix systems who don't want to bother with the POSIX stuff.
2849 However, I haven't made this a standard facility. The documentation doesn't
2850 mention it, and the Makefile doesn't support it.
2852 5. The Makefile now contains an "install" target, with editable destinations at
2853 the top of the file. The pcretest program is not installed.
2855 6. pgrep -V now gives the PCRE version number and date.
2857 7. Fixed bug: a zero repetition after a literal string (e.g. /abcde{0}/) was
2858 causing the entire string to be ignored, instead of just the last character.
2860 8. If a pattern like /"([^\\"]+|\\.)*"/ is applied in the normal way to a
2861 non-matching string, it can take a very, very long time, even for strings of
2862 quite modest length, because of the nested recursion. PCRE now does better in
2863 some of these cases. It does this by remembering the last required literal
2864 character in the pattern, and pre-searching the subject to ensure it is present
2865 before running the real match. In other words, it applies a heuristic to detect
2866 some types of certain failure quickly, and in the above example, if presented
2867 with a string that has no trailing " it gives "no match" very quickly.
2869 9. A new runtime option PCRE_NOTEMPTY causes null string matches to be ignored;
2870 other alternatives are tried instead.
2873 Version 2.06 09-Jun-99
2874 ----------------------
2876 1. Change pcretest's output for amount of store used to show just the code
2877 space, because the remainder (the data block) varies in size between 32-bit and
2878 64-bit systems.
2880 2. Added an extra argument to pcre_exec() to supply an offset in the subject to
2881 start matching at. This allows lookbehinds to work when searching for multiple
2882 occurrences in a string.
2884 3. Added additional options to pcretest for testing multiple occurrences:
2886 /+ outputs the rest of the string that follows a match
2887 /g loops for multiple occurrences, using the new startoffset argument
2888 /G loops for multiple occurrences by passing an incremented pointer
2890 4. PCRE wasn't doing the "first character" optimization for patterns starting
2891 with \b or \B, though it was doing it for other lookbehind assertions. That is,
2892 it wasn't noticing that a match for a pattern such as /\bxyz/ has to start with
2893 the letter 'x'. On long subject strings, this gives a significant speed-up.
2896 Version 2.05 21-Apr-99
2897 ----------------------
2899 1. Changed the type of magic_number from int to long int so that it works
2900 properly on 16-bit systems.
2902 2. Fixed a bug which caused patterns starting with .* not to work correctly
2903 when the subject string contained newline characters. PCRE was assuming
2904 anchoring for such patterns in all cases, which is not correct because .* will
2905 not pass a newline unless PCRE_DOTALL is set. It now assumes anchoring only if
2906 DOTALL is set at top level; otherwise it knows that patterns starting with .*
2907 must be retried after every newline in the subject.
2910 Version 2.04 18-Feb-99
2911 ----------------------
2913 1. For parenthesized subpatterns with repeats whose minimum was zero, the
2914 computation of the store needed to hold the pattern was incorrect (too large).
2915 If such patterns were nested a few deep, this could multiply and become a real
2916 problem.
2918 2. Added /M option to pcretest to show the memory requirement of a specific
2919 pattern. Made -m a synonym of -s (which does this globally) for compatibility.
2921 3. Subpatterns of the form (regex){n,m} (i.e. limited maximum) were being
2922 compiled in such a way that the backtracking after subsequent failure was
2923 pessimal. Something like (a){0,3} was compiled as (a)?(a)?(a)? instead of
2924 ((a)((a)(a)?)?)? with disastrous performance if the maximum was of any size.
2927 Version 2.03 02-Feb-99
2928 ----------------------
2930 1. Fixed typo and small mistake in man page.
2932 2. Added 4th condition (GPL supersedes if conflict) and created separate
2933 LICENCE file containing the conditions.
2935 3. Updated pcretest so that patterns such as /abc\/def/ work like they do in
2936 Perl, that is the internal \ allows the delimiter to be included in the
2937 pattern. Locked out the use of \ as a delimiter. If \ immediately follows
2938 the final delimiter, add \ to the end of the pattern (to test the error).
2940 4. Added the convenience functions for extracting substrings after a successful
2941 match. Updated pcretest to make it able to test these functions.
2944 Version 2.02 14-Jan-99
2945 ----------------------
2947 1. Initialized the working variables associated with each extraction so that
2948 their saving and restoring doesn't refer to uninitialized store.
2950 2. Put dummy code into study.c in order to trick the optimizer of the IBM C
2951 compiler for OS/2 into generating correct code. Apparently IBM isn't going to
2952 fix the problem.
2954 3. Pcretest: the timing code wasn't using LOOPREPEAT for timing execution
2955 calls, and wasn't printing the correct value for compiling calls. Increased the
2956 default value of LOOPREPEAT, and the number of significant figures in the
2957 times.
2959 4. Changed "/bin/rm" in the Makefile to "-rm" so it works on Windows NT.
2961 5. Renamed "deftables" as "dftables" to get it down to 8 characters, to avoid
2962 a building problem on Windows NT with a FAT file system.
2965 Version 2.01 21-Oct-98
2966 ----------------------
2968 1. Changed the API for pcre_compile() to allow for the provision of a pointer
2969 to character tables built by pcre_maketables() in the current locale. If NULL
2970 is passed, the default tables are used.
2973 Version 2.00 24-Sep-98
2974 ----------------------
2976 1. Since the (>?) facility is in Perl 5.005, don't require PCRE_EXTRA to enable
2977 it any more.
2979 2. Allow quantification of (?>) groups, and make it work correctly.
2981 3. The first character computation wasn't working for (?>) groups.
2983 4. Correct the implementation of \Z (it is permitted to match on the \n at the
2984 end of the subject) and add 5.005's \z, which really does match only at the
2985 very end of the subject.
2987 5. Remove the \X "cut" facility; Perl doesn't have it, and (?> is neater.
2989 6. Remove the ability to specify CASELESS, MULTILINE, DOTALL, and
2990 DOLLAR_END_ONLY at runtime, to make it possible to implement the Perl 5.005
2991 localized options. All options to pcre_study() were also removed.
2993 7. Add other new features from 5.005:
2995 $(?<= positive lookbehind
2996 $(?<! negative lookbehind
2997 (?imsx-imsx) added the unsetting capability
2998 such a setting is global if at outer level; local otherwise
2999 (?imsx-imsx:) non-capturing groups with option setting
3000 (?(cond)re|re) conditional pattern matching
3002 A backreference to itself in a repeated group matches the previous
3003 captured string.
3005 8. General tidying up of studying (both automatic and via "study")
3006 consequential on the addition of new assertions.
3008 9. As in 5.005, unlimited repeated groups that could match an empty substring
3009 are no longer faulted at compile time. Instead, the loop is forcibly broken at
3010 runtime if any iteration does actually match an empty substring.
3012 10. Include the RunTest script in the distribution.
3014 11. Added tests from the Perl 5.005_02 distribution. This showed up a few
3015 discrepancies, some of which were old and were also with respect to 5.004. They
3016 have now been fixed.
3019 Version 1.09 28-Apr-98
3020 ----------------------
3022 1. A negated single character class followed by a quantifier with a minimum
3023 value of one (e.g. [^x]{1,6} ) was not compiled correctly. This could lead to
3024 program crashes, or just wrong answers. This did not apply to negated classes
3025 containing more than one character, or to minima other than one.
3028 Version 1.08 27-Mar-98
3029 ----------------------
3031 1. Add PCRE_UNGREEDY to invert the greediness of quantifiers.
3033 2. Add (?U) and (?X) to set PCRE_UNGREEDY and PCRE_EXTRA respectively. The
3034 latter must appear before anything that relies on it in the pattern.
3037 Version 1.07 16-Feb-98
3038 ----------------------
3040 1. A pattern such as /((a)*)*/ was not being diagnosed as in error (unlimited
3041 repeat of a potentially empty string).
3044 Version 1.06 23-Jan-98
3045 ----------------------
3047 1. Added Markus Oberhumer's little patches for C++.
3049 2. Literal strings longer than 255 characters were broken.
3052 Version 1.05 23-Dec-97
3053 ----------------------
3055 1. Negated character classes containing more than one character were failing if
3056 PCRE_CASELESS was set at run time.
3059 Version 1.04 19-Dec-97
3060 ----------------------
3062 1. Corrected the man page, where some "const" qualifiers had been omitted.
3064 2. Made debugging output print "{0,xxx}" instead of just "{,xxx}" to agree with
3065 input syntax.
3067 3. Fixed memory leak which occurred when a regex with back references was
3068 matched with an offsets vector that wasn't big enough. The temporary memory
3069 that is used in this case wasn't being freed if the match failed.
3071 4. Tidied pcretest to ensure it frees memory that it gets.
3073 5. Temporary memory was being obtained in the case where the passed offsets
3074 vector was exactly big enough.
3076 6. Corrected definition of offsetof() from change 5 below.
3078 7. I had screwed up change 6 below and broken the rules for the use of
3079 setjmp(). Now fixed.
3082 Version 1.03 18-Dec-97
3083 ----------------------
3085 1. A erroneous regex with a missing opening parenthesis was correctly
3086 diagnosed, but PCRE attempted to access brastack[-1], which could cause crashes
3087 on some systems.
3089 2. Replaced offsetof(real_pcre, code) by offsetof(real_pcre, code[0]) because
3090 it was reported that one broken compiler failed on the former because "code" is
3091 also an independent variable.
3093 3. The erroneous regex a[]b caused an array overrun reference.
3095 4. A regex ending with a one-character negative class (e.g. /[^k]$/) did not
3096 fail on data ending with that character. (It was going on too far, and checking
3097 the next character, typically a binary zero.) This was specific to the
3098 optimized code for single-character negative classes.
3100 5. Added a contributed patch from the TIN world which does the following:
3102 + Add an undef for memmove, in case the the system defines a macro for it.
3104 + Add a definition of offsetof(), in case there isn't one. (I don't know
3105 the reason behind this - offsetof() is part of the ANSI standard - but
3106 it does no harm).
3108 + Reduce the ifdef's in pcre.c using macro DPRINTF, thereby eliminating
3109 most of the places where whitespace preceded '#'. I have given up and
3110 allowed the remaining 2 cases to be at the margin.
3112 + Rename some variables in pcre to eliminate shadowing. This seems very
3113 pedantic, but does no harm, of course.
3115 6. Moved the call to setjmp() into its own function, to get rid of warnings
3116 from gcc -Wall, and avoided calling it at all unless PCRE_EXTRA is used.
3118 7. Constructs such as \d{8,} were compiling into the equivalent of
3119 \d{8}\d{0,65527} instead of \d{8}\d* which didn't make much difference to the
3120 outcome, but in this particular case used more store than had been allocated,
3121 which caused the bug to be discovered because it threw up an internal error.
3123 8. The debugging code in both pcre and pcretest for outputting the compiled
3124 form of a regex was going wrong in the case of back references followed by
3125 curly-bracketed repeats.
3128 Version 1.02 12-Dec-97
3129 ----------------------
3131 1. Typos in pcre.3 and comments in the source fixed.
3133 2. Applied a contributed patch to get rid of places where it used to remove
3134 'const' from variables, and fixed some signed/unsigned and uninitialized
3135 variable warnings.
3137 3. Added the "runtest" target to Makefile.
3139 4. Set default compiler flag to -O2 rather than just -O.
3142 Version 1.01 19-Nov-97
3143 ----------------------
3145 1. PCRE was failing to diagnose unlimited repeat of empty string for patterns
3146 like /([ab]*)*/, that is, for classes with more than one character in them.
3148 2. Likewise, it wasn't diagnosing patterns with "once-only" subpatterns, such
3149 as /((?>a*))*/ (a PCRE_EXTRA facility).
3152 Version 1.00 18-Nov-97
3153 ----------------------
3155 1. Added compile-time macros to support systems such as SunOS4 which don't have
3156 memmove() or strerror() but have other things that can be used instead.
3158 2. Arranged that "make clean" removes the executables.
3161 Version 0.99 27-Oct-97
3162 ----------------------
3164 1. Fixed bug in code for optimizing classes with only one character. It was
3165 initializing a 32-byte map regardless, which could cause it to run off the end
3166 of the memory it had got.
3168 2. Added, conditional on PCRE_EXTRA, the proposed (?>REGEX) construction.
3171 Version 0.98 22-Oct-97
3172 ----------------------
3174 1. Fixed bug in code for handling temporary memory usage when there are more
3175 back references than supplied space in the ovector. This could cause segfaults.
3178 Version 0.97 21-Oct-97
3179 ----------------------
3181 1. Added the \X "cut" facility, conditional on PCRE_EXTRA.
3183 2. Optimized negated single characters not to use a bit map.
3185 3. Brought error texts together as macro definitions; clarified some of them;
3186 fixed one that was wrong - it said "range out of order" when it meant "invalid
3187 escape sequence".
3189 4. Changed some char * arguments to const char *.
3191 5. Added PCRE_NOTBOL and PCRE_NOTEOL (from POSIX).
3193 6. Added the POSIX-style API wrapper in pcreposix.a and testing facilities in
3194 pcretest.
3197 Version 0.96 16-Oct-97
3198 ----------------------
3200 1. Added a simple "pgrep" utility to the distribution.
3202 2. Fixed an incompatibility with Perl: "{" is now treated as a normal character
3203 unless it appears in one of the precise forms "{ddd}", "{ddd,}", or "{ddd,ddd}"
3204 where "ddd" means "one or more decimal digits".
3206 3. Fixed serious bug. If a pattern had a back reference, but the call to
3207 pcre_exec() didn't supply a large enough ovector to record the related
3208 identifying subpattern, the match always failed. PCRE now remembers the number
3209 of the largest back reference, and gets some temporary memory in which to save
3210 the offsets during matching if necessary, in order to ensure that
3211 backreferences always work.
3213 4. Increased the compatibility with Perl in a number of ways:
3215 (a) . no longer matches \n by default; an option PCRE_DOTALL is provided
3216 to request this handling. The option can be set at compile or exec time.
3218 (b) $ matches before a terminating newline by default; an option
3219 PCRE_DOLLAR_ENDONLY is provided to override this (but not in multiline
3220 mode). The option can be set at compile or exec time.
3222 (c) The handling of \ followed by a digit other than 0 is now supposed to be
3223 the same as Perl's. If the decimal number it represents is less than 10
3224 or there aren't that many previous left capturing parentheses, an octal
3225 escape is read. Inside a character class, it's always an octal escape,
3226 even if it is a single digit.
3228 (d) An escaped but undefined alphabetic character is taken as a literal,
3229 unless PCRE_EXTRA is set. Currently this just reserves the remaining
3230 escapes.
3232 (e) {0} is now permitted. (The previous item is removed from the compiled
3233 pattern).
3235 5. Changed all the names of code files so that the basic parts are no longer
3236 than 10 characters, and abolished the teeny "globals.c" file.
3238 6. Changed the handling of character classes; they are now done with a 32-byte
3239 bit map always.
3241 7. Added the -d and /D options to pcretest to make it possible to look at the
3242 internals of compilation without having to recompile pcre.
3245 Version 0.95 23-Sep-97
3246 ----------------------
3248 1. Fixed bug in pre-pass concerning escaped "normal" characters such as \x5c or
3249 \x20 at the start of a run of normal characters. These were being treated as
3250 real characters, instead of the source characters being re-checked.
3253 Version 0.94 18-Sep-97
3254 ----------------------
3256 1. The functions are now thread-safe, with the caveat that the global variables
3257 containing pointers to malloc() and free() or alternative functions are the
3258 same for all threads.
3260 2. Get pcre_study() to generate a bitmap of initial characters for non-
3261 anchored patterns when this is possible, and use it if passed to pcre_exec().
3264 Version 0.93 15-Sep-97
3265 ----------------------
3267 1. /(b)|(:+)/ was computing an incorrect first character.
3269 2. Add pcre_study() to the API and the passing of pcre_extra to pcre_exec(),
3270 but not actually doing anything yet.
3272 3. Treat "-" characters in classes that cannot be part of ranges as literals,
3273 as Perl does (e.g. [-az] or [az-]).
3275 4. Set the anchored flag if a branch starts with .* or .*? because that tests
3276 all possible positions.
3278 5. Split up into different modules to avoid including unneeded functions in a
3279 compiled binary. However, compile and exec are still in one module. The "study"
3280 function is split off.
3282 6. The character tables are now in a separate module whose source is generated
3283 by an auxiliary program - but can then be edited by hand if required. There are
3284 now no calls to isalnum(), isspace(), isdigit(), isxdigit(), tolower() or
3285 toupper() in the code.
3287 7. Turn the malloc/free funtions variables into pcre_malloc and pcre_free and
3288 make them global. Abolish the function for setting them, as the caller can now
3289 set them directly.
3292 Version 0.92 11-Sep-97
3293 ----------------------
3295 1. A repeat with a fixed maximum and a minimum of 1 for an ordinary character
3296 (e.g. /a{1,3}/) was broken (I mis-optimized it).
3298 2. Caseless matching was not working in character classes if the characters in
3299 the pattern were in upper case.
3301 3. Make ranges like [W-c] work in the same way as Perl for caseless matching.
3303 4. Make PCRE_ANCHORED public and accept as a compile option.
3305 5. Add an options word to pcre_exec() and accept PCRE_ANCHORED and
3306 PCRE_CASELESS at run time. Add escapes \A and \I to pcretest to cause it to
3307 pass them.
3309 6. Give an error if bad option bits passed at compile or run time.
3311 7. Add PCRE_MULTILINE at compile and exec time, and (?m) as well. Add \M to
3312 pcretest to cause it to pass that flag.
3314 8. Add pcre_info(), to get the number of identifying subpatterns, the stored
3315 options, and the first character, if set.
3317 9. Recognize C+ or C{n,m} where n >= 1 as providing a fixed starting character.
3320 Version 0.91 10-Sep-97
3321 ----------------------
3323 1. PCRE was failing to diagnose unlimited repeats of subpatterns that could
3324 match the empty string as in /(a*)*/. It was looping and ultimately crashing.
3326 2. PCRE was looping on encountering an indefinitely repeated back reference to
3327 a subpattern that had matched an empty string, e.g. /(a|)\1*/. It now does what
3328 Perl does - treats the match as successful.
3330 ****


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

  ViewVC Help
Powered by ViewVC 1.1.5