ViewVC logotype

Contents of /code/trunk/ChangeLog

Parent Directory Parent Directory | Revision Log Revision Log

Revision 335 - (show annotations)
Sat Apr 12 14:36:14 2008 UTC (13 years, 6 months ago) by ph10
File size: 153509 byte(s)
Do not discard subpatterns with {0} quantifiers, as they may be called as 
1 ChangeLog for PCRE
2 ------------------
4 Version 7.7 05-Mar-08
5 ---------------------
7 1. Applied Craig's patch to sort out a long long problem: "If we can't convert
8 a string to a long long, pretend we don't even have a long long." This is
9 done by checking for the strtoq, strtoll, and _strtoi64 functions.
11 2. Applied Craig's patch to pcrecpp.cc to restore ABI compatibility with
12 pre-7.6 versions, which defined a global no_arg variable instead of putting
13 it in the RE class. (See also #8 below.)
15 3. Remove a line of dead code, identified by coverity and reported by Nuno
16 Lopes.
18 4. Fixed two related pcregrep bugs involving -r with --include or --exclude:
20 (1) The include/exclude patterns were being applied to the whole pathnames
21 of files, instead of just to the final components.
23 (2) If there was more than one level of directory, the subdirectories were
24 skipped unless they satisfied the include/exclude conditions. This is
25 inconsistent with GNU grep (and could even be seen as contrary to the
26 pcregrep specification - which I improved to make it absolutely clear).
27 The action now is always to scan all levels of directory, and just
28 apply the include/exclude patterns to regular files.
30 5. Added the --include_dir and --exclude_dir patterns to pcregrep, and used
31 --exclude_dir in the tests to avoid scanning .svn directories.
33 6. Applied Craig's patch to the QuoteMeta function so that it escapes the
34 NUL character as backslash + 0 rather than backslash + NUL, because PCRE
35 doesn't support NULs in patterns.
37 7. Added some missing "const"s to declarations of static tables in
38 pcre_compile.c and pcre_dfa_exec.c.
40 8. Applied Craig's patch to pcrecpp.cc to fix a problem in OS X that was
41 caused by fix #2 above. (Subsequently also a second patch to fix the
42 first patch. And a third patch - this was a messy problem.)
44 9. Applied Craig's patch to remove the use of push_back().
46 10. Applied Alan Lehotsky's patch to add REG_STARTEND support to the POSIX
47 matching function regexec().
49 11. Added support for the Oniguruma syntax \g<name>, \g<n>, \g'name', \g'n',
50 which, however, unlike Perl's \g{...}, are subroutine calls, not back
51 references. PCRE supports relative numbers with this syntax (I don't think
52 Oniguruma does).
54 12. Previously, a group with a zero repeat such as (...){0} was completely
55 omitted from the compiled regex. However, this means that if the group
56 was called as a subroutine from elsewhere in the pattern, things went wrong
57 (an internal error was given). Such groups are now left in the compiled
58 pattern, with a new opcode that causes them to be skipped at execution
59 time.
62 Version 7.6 28-Jan-08
63 ---------------------
65 1. A character class containing a very large number of characters with
66 codepoints greater than 255 (in UTF-8 mode, of course) caused a buffer
67 overflow.
69 2. Patch to cut out the "long long" test in pcrecpp_unittest when
70 HAVE_LONG_LONG is not defined.
72 3. Applied Christian Ehrlicher's patch to update the CMake build files to
73 bring them up to date and include new features. This patch includes:
75 - Fixed PH's badly added libz and libbz2 support.
76 - Fixed a problem with static linking.
77 - Added pcredemo. [But later removed - see 7 below.]
78 - Fixed dftables problem and added an option.
79 - Added a number of HAVE_XXX tests, including HAVE_WINDOWS_H and
81 - Added readline support for pcretest.
82 - Added an listing of the option settings after cmake has run.
84 4. A user submitted a patch to Makefile that makes it easy to create
85 "pcre.dll" under mingw when using Configure/Make. I added stuff to
86 Makefile.am that cause it to include this special target, without
87 affecting anything else. Note that the same mingw target plus all
88 the other distribution libraries and programs are now supported
89 when configuring with CMake (see 6 below) instead of with
90 Configure/Make.
92 5. Applied Craig's patch that moves no_arg into the RE class in the C++ code.
93 This is an attempt to solve the reported problem "pcrecpp::no_arg is not
94 exported in the Windows port". It has not yet been confirmed that the patch
95 solves the problem, but it does no harm.
97 6. Applied Sheri's patch to CMakeLists.txt to add NON_STANDARD_LIB_PREFIX and
98 NON_STANDARD_LIB_SUFFIX for dll names built with mingw when configured
99 with CMake, and also correct the comment about stack recursion.
101 7. Remove the automatic building of pcredemo from the ./configure system and
102 from CMakeLists.txt. The whole idea of pcredemo.c is that it is an example
103 of a program that users should build themselves after PCRE is installed, so
104 building it automatically is not really right. What is more, it gave
105 trouble in some build environments.
107 8. Further tidies to CMakeLists.txt from Sheri and Christian.
110 Version 7.5 10-Jan-08
111 ---------------------
113 1. Applied a patch from Craig: "This patch makes it possible to 'ignore'
114 values in parens when parsing an RE using the C++ wrapper."
116 2. Negative specials like \S did not work in character classes in UTF-8 mode.
117 Characters greater than 255 were excluded from the class instead of being
118 included.
120 3. The same bug as (2) above applied to negated POSIX classes such as
121 [:^space:].
123 4. PCRECPP_STATIC was referenced in pcrecpp_internal.h, but nowhere was it
124 defined or documented. It seems to have been a typo for PCRE_STATIC, so
125 I have changed it.
127 5. The construct (?&) was not diagnosed as a syntax error (it referenced the
128 first named subpattern) and a construct such as (?&a) would reference the
129 first named subpattern whose name started with "a" (in other words, the
130 length check was missing). Both these problems are fixed. "Subpattern name
131 expected" is now given for (?&) (a zero-length name), and this patch also
132 makes it give the same error for \k'' (previously it complained that that
133 was a reference to a non-existent subpattern).
135 6. The erroneous patterns (?+-a) and (?-+a) give different error messages;
136 this is right because (?- can be followed by option settings as well as by
137 digits. I have, however, made the messages clearer.
139 7. Patterns such as (?(1)a|b) (a pattern that contains fewer subpatterns
140 than the number used in the conditional) now cause a compile-time error.
141 This is actually not compatible with Perl, which accepts such patterns, but
142 treats the conditional as always being FALSE (as PCRE used to), but it
143 seems to me that giving a diagnostic is better.
145 8. Change "alphameric" to the more common word "alphanumeric" in comments
146 and messages.
148 9. Fix two occurrences of "backslash" in comments that should have been
149 "backspace".
151 10. Remove two redundant lines of code that can never be obeyed (their function
152 was moved elsewhere).
154 11. The program that makes PCRE's Unicode character property table had a bug
155 which caused it to generate incorrect table entries for sequences of
156 characters that have the same character type, but are in different scripts.
157 It amalgamated them into a single range, with the script of the first of
158 them. In other words, some characters were in the wrong script. There were
159 thirteen such cases, affecting characters in the following ranges:
161 U+002b0 - U+002c1
162 U+0060c - U+0060d
163 U+0061e - U+00612
164 U+0064b - U+0065e
165 U+0074d - U+0076d
166 U+01800 - U+01805
167 U+01d00 - U+01d77
168 U+01d9b - U+01dbf
169 U+0200b - U+0200f
170 U+030fc - U+030fe
171 U+03260 - U+0327f
172 U+0fb46 - U+0fbb1
173 U+10450 - U+1049d
175 12. The -o option (show only the matching part of a line) for pcregrep was not
176 compatible with GNU grep in that, if there was more than one match in a
177 line, it showed only the first of them. It now behaves in the same way as
178 GNU grep.
180 13. If the -o and -v options were combined for pcregrep, it printed a blank
181 line for every non-matching line. GNU grep prints nothing, and pcregrep now
182 does the same. The return code can be used to tell if there were any
183 non-matching lines.
185 14. Added --file-offsets and --line-offsets to pcregrep.
187 15. The pattern (?=something)(?R) was not being diagnosed as a potentially
188 infinitely looping recursion. The bug was that positive lookaheads were not
189 being skipped when checking for a possible empty match (negative lookaheads
190 and both kinds of lookbehind were skipped).
192 16. Fixed two typos in the Windows-only code in pcregrep.c, and moved the
193 inclusion of <windows.h> to before rather than after the definition of
194 INVALID_FILE_ATTRIBUTES (patch from David Byron).
196 17. Specifying a possessive quantifier with a specific limit for a Unicode
197 character property caused pcre_compile() to compile bad code, which led at
198 runtime to PCRE_ERROR_INTERNAL (-14). Examples of patterns that caused this
199 are: /\p{Zl}{2,3}+/8 and /\p{Cc}{2}+/8. It was the possessive "+" that
200 caused the error; without that there was no problem.
202 18. Added --enable-pcregrep-libz and --enable-pcregrep-libbz2.
204 19. Added --enable-pcretest-libreadline.
206 20. In pcrecpp.cc, the variable 'count' was incremented twice in
207 RE::GlobalReplace(). As a result, the number of replacements returned was
208 double what it should be. I removed one of the increments, but Craig sent a
209 later patch that removed the other one (the right fix) and added unit tests
210 that check the return values (which was not done before).
212 21. Several CMake things:
214 (1) Arranged that, when cmake is used on Unix, the libraries end up with
215 the names libpcre and libpcreposix, not just pcre and pcreposix.
217 (2) The above change means that pcretest and pcregrep are now correctly
218 linked with the newly-built libraries, not previously installed ones.
222 22. In UTF-8 mode, with newline set to "any", a pattern such as .*a.*=.b.*
223 crashed when matching a string such as a\x{2029}b (note that \x{2029} is a
224 UTF-8 newline character). The key issue is that the pattern starts .*;
225 this means that the match must be either at the beginning, or after a
226 newline. The bug was in the code for advancing after a failed match and
227 checking that the new position followed a newline. It was not taking
228 account of UTF-8 characters correctly.
230 23. PCRE was behaving differently from Perl in the way it recognized POSIX
231 character classes. PCRE was not treating the sequence [:...:] as a
232 character class unless the ... were all letters. Perl, however, seems to
233 allow any characters between [: and :], though of course it rejects as
234 unknown any "names" that contain non-letters, because all the known class
235 names consist only of letters. Thus, Perl gives an error for [[:1234:]],
236 for example, whereas PCRE did not - it did not recognize a POSIX character
237 class. This seemed a bit dangerous, so the code has been changed to be
238 closer to Perl. The behaviour is not identical to Perl, because PCRE will
239 diagnose an unknown class for, for example, [[:l\ower:]] where Perl will
240 treat it as [[:lower:]]. However, PCRE does now give "unknown" errors where
241 Perl does, and where it didn't before.
243 24. Rewrite so as to remove the single use of %n from pcregrep because in some
244 Windows environments %n is disabled by default.
247 Version 7.4 21-Sep-07
248 ---------------------
250 1. Change 7.3/28 was implemented for classes by looking at the bitmap. This
251 means that a class such as [\s] counted as "explicit reference to CR or
252 LF". That isn't really right - the whole point of the change was to try to
253 help when there was an actual mention of one of the two characters. So now
254 the change happens only if \r or \n (or a literal CR or LF) character is
255 encountered.
257 2. The 32-bit options word was also used for 6 internal flags, but the numbers
258 of both had grown to the point where there were only 3 bits left.
259 Fortunately, there was spare space in the data structure, and so I have
260 moved the internal flags into a new 16-bit field to free up more option
261 bits.
263 3. The appearance of (?J) at the start of a pattern set the DUPNAMES option,
264 but did not set the internal JCHANGED flag - either of these is enough to
265 control the way the "get" function works - but the PCRE_INFO_JCHANGED
266 facility is supposed to tell if (?J) was ever used, so now (?J) at the
267 start sets both bits.
269 4. Added options (at build time, compile time, exec time) to change \R from
270 matching any Unicode line ending sequence to just matching CR, LF, or CRLF.
272 5. doc/pcresyntax.html was missing from the distribution.
274 6. Put back the definition of PCRE_ERROR_NULLWSLIMIT, for backward
275 compatibility, even though it is no longer used.
277 7. Added macro for snprintf to pcrecpp_unittest.cc and also for strtoll and
278 strtoull to pcrecpp.cc to select the available functions in WIN32 when the
279 windows.h file is present (where different names are used). [This was
280 reversed later after testing - see 16 below.]
282 8. Changed all #include <config.h> to #include "config.h". There were also
283 some further <pcre.h> cases that I changed to "pcre.h".
285 9. When pcregrep was used with the --colour option, it missed the line ending
286 sequence off the lines that it output.
288 10. It was pointed out to me that arrays of string pointers cause lots of
289 relocations when a shared library is dynamically loaded. A technique of
290 using a single long string with a table of offsets can drastically reduce
291 these. I have refactored PCRE in four places to do this. The result is
292 dramatic:
294 Originally: 290
295 After changing UCP table: 187
296 After changing error message table: 43
297 After changing table of "verbs" 36
298 After changing table of Posix names 22
300 Thanks to the folks working on Gregex for glib for this insight.
302 11. --disable-stack-for-recursion caused compiling to fail unless -enable-
303 unicode-properties was also set.
305 12. Updated the tests so that they work when \R is defaulted to ANYCRLF.
307 13. Added checks for ANY and ANYCRLF to pcrecpp.cc where it previously
308 checked only for CRLF.
310 14. Added casts to pcretest.c to avoid compiler warnings.
312 15. Added Craig's patch to various pcrecpp modules to avoid compiler warnings.
314 16. Added Craig's patch to remove the WINDOWS_H tests, that were not working,
315 and instead check for _strtoi64 explicitly, and avoid the use of snprintf()
316 entirely. This removes changes made in 7 above.
318 17. The CMake files have been updated, and there is now more information about
319 building with CMake in the NON-UNIX-USE document.
322 Version 7.3 28-Aug-07
323 ---------------------
325 1. In the rejigging of the build system that eventually resulted in 7.1, the
326 line "#include <pcre.h>" was included in pcre_internal.h. The use of angle
327 brackets there is not right, since it causes compilers to look for an
328 installed pcre.h, not the version that is in the source that is being
329 compiled (which of course may be different). I have changed it back to:
331 #include "pcre.h"
333 I have a vague recollection that the change was concerned with compiling in
334 different directories, but in the new build system, that is taken care of
335 by the VPATH setting the Makefile.
337 2. The pattern .*$ when run in not-DOTALL UTF-8 mode with newline=any failed
338 when the subject happened to end in the byte 0x85 (e.g. if the last
339 character was \x{1ec5}). *Character* 0x85 is one of the "any" newline
340 characters but of course it shouldn't be taken as a newline when it is part
341 of another character. The bug was that, for an unlimited repeat of . in
342 not-DOTALL UTF-8 mode, PCRE was advancing by bytes rather than by
343 characters when looking for a newline.
345 3. A small performance improvement in the DOTALL UTF-8 mode .* case.
347 4. Debugging: adjusted the names of opcodes for different kinds of parentheses
348 in debug output.
350 5. Arrange to use "%I64d" instead of "%lld" and "%I64u" instead of "%llu" for
351 long printing in the pcrecpp unittest when running under MinGW.
353 6. ESC_K was left out of the EBCDIC table.
355 7. Change 7.0/38 introduced a new limit on the number of nested non-capturing
356 parentheses; I made it 1000, which seemed large enough. Unfortunately, the
357 limit also applies to "virtual nesting" when a pattern is recursive, and in
358 this case 1000 isn't so big. I have been able to remove this limit at the
359 expense of backing off one optimization in certain circumstances. Normally,
360 when pcre_exec() would call its internal match() function recursively and
361 immediately return the result unconditionally, it uses a "tail recursion"
362 feature to save stack. However, when a subpattern that can match an empty
363 string has an unlimited repetition quantifier, it no longer makes this
364 optimization. That gives it a stack frame in which to save the data for
365 checking that an empty string has been matched. Previously this was taken
366 from the 1000-entry workspace that had been reserved. So now there is no
367 explicit limit, but more stack is used.
369 8. Applied Daniel's patches to solve problems with the import/export magic
370 syntax that is required for Windows, and which was going wrong for the
371 pcreposix and pcrecpp parts of the library. These were overlooked when this
372 problem was solved for the main library.
374 9. There were some crude static tests to avoid integer overflow when computing
375 the size of patterns that contain repeated groups with explicit upper
376 limits. As the maximum quantifier is 65535, the maximum group length was
377 set at 30,000 so that the product of these two numbers did not overflow a
378 32-bit integer. However, it turns out that people want to use groups that
379 are longer than 30,000 bytes (though not repeat them that many times).
380 Change 7.0/17 (the refactoring of the way the pattern size is computed) has
381 made it possible to implement the integer overflow checks in a much more
382 dynamic way, which I have now done. The artificial limitation on group
383 length has been removed - we now have only the limit on the total length of
384 the compiled pattern, which depends on the LINK_SIZE setting.
386 10. Fixed a bug in the documentation for get/copy named substring when
387 duplicate names are permitted. If none of the named substrings are set, the
388 functions return PCRE_ERROR_NOSUBSTRING (7); the doc said they returned an
389 empty string.
391 11. Because Perl interprets \Q...\E at a high level, and ignores orphan \E
392 instances, patterns such as [\Q\E] or [\E] or even [^\E] cause an error,
393 because the ] is interpreted as the first data character and the
394 terminating ] is not found. PCRE has been made compatible with Perl in this
395 regard. Previously, it interpreted [\Q\E] as an empty class, and [\E] could
396 cause memory overwriting.
398 10. Like Perl, PCRE automatically breaks an unlimited repeat after an empty
399 string has been matched (to stop an infinite loop). It was not recognizing
400 a conditional subpattern that could match an empty string if that
401 subpattern was within another subpattern. For example, it looped when
402 trying to match (((?(1)X|))*) but it was OK with ((?(1)X|)*) where the
403 condition was not nested. This bug has been fixed.
405 12. A pattern like \X?\d or \P{L}?\d in non-UTF-8 mode could cause a backtrack
406 past the start of the subject in the presence of bytes with the top bit
407 set, for example "\x8aBCD".
409 13. Added Perl 5.10 experimental backtracking controls (*FAIL), (*F), (*PRUNE),
410 (*SKIP), (*THEN), (*COMMIT), and (*ACCEPT).
412 14. Optimized (?!) to (*FAIL).
414 15. Updated the test for a valid UTF-8 string to conform to the later RFC 3629.
415 This restricts code points to be within the range 0 to 0x10FFFF, excluding
416 the "low surrogate" sequence 0xD800 to 0xDFFF. Previously, PCRE allowed the
417 full range 0 to 0x7FFFFFFF, as defined by RFC 2279. Internally, it still
418 does: it's just the validity check that is more restrictive.
420 16. Inserted checks for integer overflows during escape sequence (backslash)
421 processing, and also fixed erroneous offset values for syntax errors during
422 backslash processing.
424 17. Fixed another case of looking too far back in non-UTF-8 mode (cf 12 above)
425 for patterns like [\PPP\x8a]{1,}\x80 with the subject "A\x80".
427 18. An unterminated class in a pattern like (?1)\c[ with a "forward reference"
428 caused an overrun.
430 19. A pattern like (?:[\PPa*]*){8,} which had an "extended class" (one with
431 something other than just ASCII characters) inside a group that had an
432 unlimited repeat caused a loop at compile time (while checking to see
433 whether the group could match an empty string).
435 20. Debugging a pattern containing \p or \P could cause a crash. For example,
436 [\P{Any}] did so. (Error in the code for printing property names.)
438 21. An orphan \E inside a character class could cause a crash.
440 22. A repeated capturing bracket such as (A)? could cause a wild memory
441 reference during compilation.
443 23. There are several functions in pcre_compile() that scan along a compiled
444 expression for various reasons (e.g. to see if it's fixed length for look
445 behind). There were bugs in these functions when a repeated \p or \P was
446 present in the pattern. These operators have additional parameters compared
447 with \d, etc, and these were not being taken into account when moving along
448 the compiled data. Specifically:
450 (a) A item such as \p{Yi}{3} in a lookbehind was not treated as fixed
451 length.
453 (b) An item such as \pL+ within a repeated group could cause crashes or
454 loops.
456 (c) A pattern such as \p{Yi}+(\P{Yi}+)(?1) could give an incorrect
457 "reference to non-existent subpattern" error.
459 (d) A pattern like (\P{Yi}{2}\277)? could loop at compile time.
461 24. A repeated \S or \W in UTF-8 mode could give wrong answers when multibyte
462 characters were involved (for example /\S{2}/8g with "A\x{a3}BC").
464 25. Using pcregrep in multiline, inverted mode (-Mv) caused it to loop.
466 26. Patterns such as [\P{Yi}A] which include \p or \P and just one other
467 character were causing crashes (broken optimization).
469 27. Patterns such as (\P{Yi}*\277)* (group with possible zero repeat containing
470 \p or \P) caused a compile-time loop.
472 28. More problems have arisen in unanchored patterns when CRLF is a valid line
473 break. For example, the unstudied pattern [\r\n]A does not match the string
474 "\r\nA" because change 7.0/46 below moves the current point on by two
475 characters after failing to match at the start. However, the pattern \nA
476 *does* match, because it doesn't start till \n, and if [\r\n]A is studied,
477 the same is true. There doesn't seem any very clean way out of this, but
478 what I have chosen to do makes the common cases work: PCRE now takes note
479 of whether there can be an explicit match for \r or \n anywhere in the
480 pattern, and if so, 7.0/46 no longer applies. As part of this change,
481 there's a new PCRE_INFO_HASCRORLF option for finding out whether a compiled
482 pattern has explicit CR or LF references.
484 29. Added (*CR) etc for changing newline setting at start of pattern.
487 Version 7.2 19-Jun-07
488 ---------------------
490 1. If the fr_FR locale cannot be found for test 3, try the "french" locale,
491 which is apparently normally available under Windows.
493 2. Re-jig the pcregrep tests with different newline settings in an attempt
494 to make them independent of the local environment's newline setting.
496 3. Add code to configure.ac to remove -g from the CFLAGS default settings.
498 4. Some of the "internals" tests were previously cut out when the link size
499 was not 2, because the output contained actual offsets. The recent new
500 "Z" feature of pcretest means that these can be cut out, making the tests
501 usable with all link sizes.
503 5. Implemented Stan Switzer's goto replacement for longjmp() when not using
504 stack recursion. This gives a massive performance boost under BSD, but just
505 a small improvement under Linux. However, it saves one field in the frame
506 in all cases.
508 6. Added more features from the forthcoming Perl 5.10:
510 (a) (?-n) (where n is a string of digits) is a relative subroutine or
511 recursion call. It refers to the nth most recently opened parentheses.
513 (b) (?+n) is also a relative subroutine call; it refers to the nth next
514 to be opened parentheses.
516 (c) Conditions that refer to capturing parentheses can be specified
517 relatively, for example, (?(-2)... or (?(+3)...
519 (d) \K resets the start of the current match so that everything before
520 is not part of it.
522 (e) \k{name} is synonymous with \k<name> and \k'name' (.NET compatible).
524 (f) \g{name} is another synonym - part of Perl 5.10's unification of
525 reference syntax.
527 (g) (?| introduces a group in which the numbering of parentheses in each
528 alternative starts with the same number.
530 (h) \h, \H, \v, and \V match horizontal and vertical whitespace.
532 7. Added two new calls to pcre_fullinfo(): PCRE_INFO_OKPARTIAL and
535 8. A pattern such as (.*(.)?)* caused pcre_exec() to fail by either not
536 terminating or by crashing. Diagnosed by Viktor Griph; it was in the code
537 for detecting groups that can match an empty string.
539 9. A pattern with a very large number of alternatives (more than several
540 hundred) was running out of internal workspace during the pre-compile
541 phase, where pcre_compile() figures out how much memory will be needed. A
542 bit of new cunning has reduced the workspace needed for groups with
543 alternatives. The 1000-alternative test pattern now uses 12 bytes of
544 workspace instead of running out of the 4096 that are available.
546 10. Inserted some missing (unsigned int) casts to get rid of compiler warnings.
548 11. Applied patch from Google to remove an optimization that didn't quite work.
549 The report of the bug said:
551 pcrecpp::RE("a*").FullMatch("aaa") matches, while
552 pcrecpp::RE("a*?").FullMatch("aaa") does not, and
553 pcrecpp::RE("a*?\\z").FullMatch("aaa") does again.
555 12. If \p or \P was used in non-UTF-8 mode on a character greater than 127
556 it matched the wrong number of bytes.
559 Version 7.1 24-Apr-07
560 ---------------------
562 1. Applied Bob Rossi and Daniel G's patches to convert the build system to one
563 that is more "standard", making use of automake and other Autotools. There
564 is some re-arrangement of the files and adjustment of comments consequent
565 on this.
567 2. Part of the patch fixed a problem with the pcregrep tests. The test of -r
568 for recursive directory scanning broke on some systems because the files
569 are not scanned in any specific order and on different systems the order
570 was different. A call to "sort" has been inserted into RunGrepTest for the
571 approprate test as a short-term fix. In the longer term there may be an
572 alternative.
574 3. I had an email from Eric Raymond about problems translating some of PCRE's
575 man pages to HTML (despite the fact that I distribute HTML pages, some
576 people do their own conversions for various reasons). The problems
577 concerned the use of low-level troff macros .br and .in. I have therefore
578 removed all such uses from the man pages (some were redundant, some could
579 be replaced by .nf/.fi pairs). The 132html script that I use to generate
580 HTML has been updated to handle .nf/.fi and to complain if it encounters
581 .br or .in.
583 4. Updated comments in configure.ac that get placed in config.h.in and also
584 arranged for config.h to be included in the distribution, with the name
585 config.h.generic, for the benefit of those who have to compile without
586 Autotools (compare pcre.h, which is now distributed as pcre.h.generic).
588 5. Updated the support (such as it is) for Virtual Pascal, thanks to Stefan
589 Weber: (1) pcre_internal.h was missing some function renames; (2) updated
590 makevp.bat for the current PCRE, using the additional files
591 makevp_c.txt, makevp_l.txt, and pcregexp.pas.
593 6. A Windows user reported a minor discrepancy with test 2, which turned out
594 to be caused by a trailing space on an input line that had got lost in his
595 copy. The trailing space was an accident, so I've just removed it.
597 7. Add -Wl,-R... flags in pcre-config.in for *BSD* systems, as I'm told
598 that is needed.
600 8. Mark ucp_table (in ucptable.h) and ucp_gentype (in pcre_ucp_searchfuncs.c)
601 as "const" (a) because they are and (b) because it helps the PHP
602 maintainers who have recently made a script to detect big data structures
603 in the php code that should be moved to the .rodata section. I remembered
604 to update Builducptable as well, so it won't revert if ucptable.h is ever
605 re-created.
607 9. Added some extra #ifdef SUPPORT_UTF8 conditionals into pcretest.c,
608 pcre_printint.src, pcre_compile.c, pcre_study.c, and pcre_tables.c, in
609 order to be able to cut out the UTF-8 tables in the latter when UTF-8
610 support is not required. This saves 1.5-2K of code, which is important in
611 some applications.
613 Later: more #ifdefs are needed in pcre_ord2utf8.c and pcre_valid_utf8.c
614 so as not to refer to the tables, even though these functions will never be
615 called when UTF-8 support is disabled. Otherwise there are problems with a
616 shared library.
618 10. Fixed two bugs in the emulated memmove() function in pcre_internal.h:
620 (a) It was defining its arguments as char * instead of void *.
622 (b) It was assuming that all moves were upwards in memory; this was true
623 a long time ago when I wrote it, but is no longer the case.
625 The emulated memove() is provided for those environments that have neither
626 memmove() nor bcopy(). I didn't think anyone used it these days, but that
627 is clearly not the case, as these two bugs were recently reported.
629 11. The script PrepareRelease is now distributed: it calls 132html, CleanTxt,
630 and Detrail to create the HTML documentation, the .txt form of the man
631 pages, and it removes trailing spaces from listed files. It also creates
632 pcre.h.generic and config.h.generic from pcre.h and config.h. In the latter
633 case, it wraps all the #defines with #ifndefs. This script should be run
634 before "make dist".
636 12. Fixed two fairly obscure bugs concerned with quantified caseless matching
637 with Unicode property support.
639 (a) For a maximizing quantifier, if the two different cases of the
640 character were of different lengths in their UTF-8 codings (there are
641 some cases like this - I found 11), and the matching function had to
642 back up over a mixture of the two cases, it incorrectly assumed they
643 were both the same length.
645 (b) When PCRE was configured to use the heap rather than the stack for
646 recursion during matching, it was not correctly preserving the data for
647 the other case of a UTF-8 character when checking ahead for a match
648 while processing a minimizing repeat. If the check also involved
649 matching a wide character, but failed, corruption could cause an
650 erroneous result when trying to check for a repeat of the original
651 character.
653 13. Some tidying changes to the testing mechanism:
655 (a) The RunTest script now detects the internal link size and whether there
656 is UTF-8 and UCP support by running ./pcretest -C instead of relying on
657 values substituted by "configure". (The RunGrepTest script already did
658 this for UTF-8.) The configure.ac script no longer substitutes the
659 relevant variables.
661 (b) The debugging options /B and /D in pcretest show the compiled bytecode
662 with length and offset values. This means that the output is different
663 for different internal link sizes. Test 2 is skipped for link sizes
664 other than 2 because of this, bypassing the problem. Unfortunately,
665 there was also a test in test 3 (the locale tests) that used /B and
666 failed for link sizes other than 2. Rather than cut the whole test out,
667 I have added a new /Z option to pcretest that replaces the length and
668 offset values with spaces. This is now used to make test 3 independent
669 of link size. (Test 2 will be tidied up later.)
671 14. If erroroffset was passed as NULL to pcre_compile, it provoked a
672 segmentation fault instead of returning the appropriate error message.
674 15. In multiline mode when the newline sequence was set to "any", the pattern
675 ^$ would give a match between the \r and \n of a subject such as "A\r\nB".
676 This doesn't seem right; it now treats the CRLF combination as the line
677 ending, and so does not match in that case. It's only a pattern such as ^$
678 that would hit this one: something like ^ABC$ would have failed after \r
679 and then tried again after \r\n.
681 16. Changed the comparison command for RunGrepTest from "diff -u" to "diff -ub"
682 in an attempt to make files that differ only in their line terminators
683 compare equal. This works on Linux.
685 17. Under certain error circumstances pcregrep might try to free random memory
686 as it exited. This is now fixed, thanks to valgrind.
688 19. In pcretest, if the pattern /(?m)^$/g<any> was matched against the string
689 "abc\r\n\r\n", it found an unwanted second match after the second \r. This
690 was because its rules for how to advance for /g after matching an empty
691 string at the end of a line did not allow for this case. They now check for
692 it specially.
694 20. pcretest is supposed to handle patterns and data of any length, by
695 extending its buffers when necessary. It was getting this wrong when the
696 buffer for a data line had to be extended.
698 21. Added PCRE_NEWLINE_ANYCRLF which is like ANY, but matches only CR, LF, or
699 CRLF as a newline sequence.
701 22. Code for handling Unicode properties in pcre_dfa_exec() wasn't being cut
702 out by #ifdef SUPPORT_UCP. This did no harm, as it could never be used, but
703 I have nevertheless tidied it up.
705 23. Added some casts to kill warnings from HP-UX ia64 compiler.
707 24. Added a man page for pcre-config.
710 Version 7.0 19-Dec-06
711 ---------------------
713 1. Fixed a signed/unsigned compiler warning in pcre_compile.c, shown up by
714 moving to gcc 4.1.1.
716 2. The -S option for pcretest uses setrlimit(); I had omitted to #include
717 sys/time.h, which is documented as needed for this function. It doesn't
718 seem to matter on Linux, but it showed up on some releases of OS X.
720 3. It seems that there are systems where bytes whose values are greater than
721 127 match isprint() in the "C" locale. The "C" locale should be the
722 default when a C program starts up. In most systems, only ASCII printing
723 characters match isprint(). This difference caused the output from pcretest
724 to vary, making some of the tests fail. I have changed pcretest so that:
726 (a) When it is outputting text in the compiled version of a pattern, bytes
727 other than 32-126 are always shown as hex escapes.
729 (b) When it is outputting text that is a matched part of a subject string,
730 it does the same, unless a different locale has been set for the match
731 (using the /L modifier). In this case, it uses isprint() to decide.
733 4. Fixed a major bug that caused incorrect computation of the amount of memory
734 required for a compiled pattern when options that changed within the
735 pattern affected the logic of the preliminary scan that determines the
736 length. The relevant options are -x, and -i in UTF-8 mode. The result was
737 that the computed length was too small. The symptoms of this bug were
738 either the PCRE error "internal error: code overflow" from pcre_compile(),
739 or a glibc crash with a message such as "pcretest: free(): invalid next
740 size (fast)". Examples of patterns that provoked this bug (shown in
741 pcretest format) are:
743 /(?-x: )/x
744 /(?x)(?-x: \s*#\s*)/
745 /((?i)[\x{c0}])/8
746 /(?i:[\x{c0}])/8
748 HOWEVER: Change 17 below makes this fix obsolete as the memory computation
749 is now done differently.
751 5. Applied patches from Google to: (a) add a QuoteMeta function to the C++
752 wrapper classes; (b) implement a new function in the C++ scanner that is
753 more efficient than the old way of doing things because it avoids levels of
754 recursion in the regex matching; (c) add a paragraph to the documentation
755 for the FullMatch() function.
757 6. The escape sequence \n was being treated as whatever was defined as
758 "newline". Not only was this contrary to the documentation, which states
759 that \n is character 10 (hex 0A), but it also went horribly wrong when
760 "newline" was defined as CRLF. This has been fixed.
762 7. In pcre_dfa_exec.c the value of an unsigned integer (the variable called c)
763 was being set to -1 for the "end of line" case (supposedly a value that no
764 character can have). Though this value is never used (the check for end of
765 line is "zero bytes in current character"), it caused compiler complaints.
766 I've changed it to 0xffffffff.
768 8. In pcre_version.c, the version string was being built by a sequence of
769 C macros that, in the event of PCRE_PRERELEASE being defined as an empty
770 string (as it is for production releases) called a macro with an empty
771 argument. The C standard says the result of this is undefined. The gcc
772 compiler treats it as an empty string (which was what was wanted) but it is
773 reported that Visual C gives an error. The source has been hacked around to
774 avoid this problem.
776 9. On the advice of a Windows user, included <io.h> and <fcntl.h> in Windows
777 builds of pcretest, and changed the call to _setmode() to use _O_BINARY
778 instead of 0x8000. Made all the #ifdefs test both _WIN32 and WIN32 (not all
779 of them did).
781 10. Originally, pcretest opened its input and output without "b"; then I was
782 told that "b" was needed in some environments, so it was added for release
783 5.0 to both the input and output. (It makes no difference on Unix-like
784 systems.) Later I was told that it is wrong for the input on Windows. I've
785 now abstracted the modes into two macros, to make it easier to fiddle with
786 them, and removed "b" from the input mode under Windows.
788 11. Added pkgconfig support for the C++ wrapper library, libpcrecpp.
790 12. Added -help and --help to pcretest as an official way of being reminded
791 of the options.
793 13. Removed some redundant semicolons after macro calls in pcrecpparg.h.in
794 and pcrecpp.cc because they annoy compilers at high warning levels.
796 14. A bit of tidying/refactoring in pcre_exec.c in the main bumpalong loop.
798 15. Fixed an occurrence of == in configure.ac that should have been = (shell
799 scripts are not C programs :-) and which was not noticed because it works
800 on Linux.
802 16. pcretest is supposed to handle any length of pattern and data line (as one
803 line or as a continued sequence of lines) by extending its input buffer if
804 necessary. This feature was broken for very long pattern lines, leading to
805 a string of junk being passed to pcre_compile() if the pattern was longer
806 than about 50K.
808 17. I have done a major re-factoring of the way pcre_compile() computes the
809 amount of memory needed for a compiled pattern. Previously, there was code
810 that made a preliminary scan of the pattern in order to do this. That was
811 OK when PCRE was new, but as the facilities have expanded, it has become
812 harder and harder to keep it in step with the real compile phase, and there
813 have been a number of bugs (see for example, 4 above). I have now found a
814 cunning way of running the real compile function in a "fake" mode that
815 enables it to compute how much memory it would need, while actually only
816 ever using a few hundred bytes of working memory and without too many
817 tests of the mode. This should make future maintenance and development
818 easier. A side effect of this work is that the limit of 200 on the nesting
819 depth of parentheses has been removed (though this was never a serious
820 limitation, I suspect). However, there is a downside: pcre_compile() now
821 runs more slowly than before (30% or more, depending on the pattern). I
822 hope this isn't a big issue. There is no effect on runtime performance.
824 18. Fixed a minor bug in pcretest: if a pattern line was not terminated by a
825 newline (only possible for the last line of a file) and it was a
826 pattern that set a locale (followed by /Lsomething), pcretest crashed.
828 19. Added additional timing features to pcretest. (1) The -tm option now times
829 matching only, not compiling. (2) Both -t and -tm can be followed, as a
830 separate command line item, by a number that specifies the number of
831 repeats to use when timing. The default is 50000; this gives better
832 precision, but takes uncomfortably long for very large patterns.
834 20. Extended pcre_study() to be more clever in cases where a branch of a
835 subpattern has no definite first character. For example, (a*|b*)[cd] would
836 previously give no result from pcre_study(). Now it recognizes that the
837 first character must be a, b, c, or d.
839 21. There was an incorrect error "recursive call could loop indefinitely" if
840 a subpattern (or the entire pattern) that was being tested for matching an
841 empty string contained only one non-empty item after a nested subpattern.
842 For example, the pattern (?>\x{100}*)\d(?R) provoked this error
843 incorrectly, because the \d was being skipped in the check.
845 22. The pcretest program now has a new pattern option /B and a command line
846 option -b, which is equivalent to adding /B to every pattern. This causes
847 it to show the compiled bytecode, without the additional information that
848 -d shows. The effect of -d is now the same as -b with -i (and similarly, /D
849 is the same as /B/I).
851 23. A new optimization is now able automatically to treat some sequences such
852 as a*b as a*+b. More specifically, if something simple (such as a character
853 or a simple class like \d) has an unlimited quantifier, and is followed by
854 something that cannot possibly match the quantified thing, the quantifier
855 is automatically "possessified".
857 24. A recursive reference to a subpattern whose number was greater than 39
858 went wrong under certain circumstances in UTF-8 mode. This bug could also
859 have affected the operation of pcre_study().
861 25. Realized that a little bit of performance could be had by replacing
862 (c & 0xc0) == 0xc0 with c >= 0xc0 when processing UTF-8 characters.
864 26. Timing data from pcretest is now shown to 4 decimal places instead of 3.
866 27. Possessive quantifiers such as a++ were previously implemented by turning
867 them into atomic groups such as ($>a+). Now they have their own opcodes,
868 which improves performance. This includes the automatically created ones
869 from 23 above.
871 28. A pattern such as (?=(\w+))\1: which simulates an atomic group using a
872 lookahead was broken if it was not anchored. PCRE was mistakenly expecting
873 the first matched character to be a colon. This applied both to named and
874 numbered groups.
876 29. The ucpinternal.h header file was missing its idempotency #ifdef.
878 30. I was sent a "project" file called libpcre.a.dev which I understand makes
879 building PCRE on Windows easier, so I have included it in the distribution.
881 31. There is now a check in pcretest against a ridiculously large number being
882 returned by pcre_exec() or pcre_dfa_exec(). If this happens in a /g or /G
883 loop, the loop is abandoned.
885 32. Forward references to subpatterns in conditions such as (?(2)...) where
886 subpattern 2 is defined later cause pcre_compile() to search forwards in
887 the pattern for the relevant set of parentheses. This search went wrong
888 when there were unescaped parentheses in a character class, parentheses
889 escaped with \Q...\E, or parentheses in a #-comment in /x mode.
891 33. "Subroutine" calls and backreferences were previously restricted to
892 referencing subpatterns earlier in the regex. This restriction has now
893 been removed.
895 34. Added a number of extra features that are going to be in Perl 5.10. On the
896 whole, these are just syntactic alternatives for features that PCRE had
897 previously implemented using the Python syntax or my own invention. The
898 other formats are all retained for compatibility.
900 (a) Named groups can now be defined as (?<name>...) or (?'name'...) as well
901 as (?P<name>...). The new forms, as well as being in Perl 5.10, are
902 also .NET compatible.
904 (b) A recursion or subroutine call to a named group can now be defined as
905 (?&name) as well as (?P>name).
907 (c) A backreference to a named group can now be defined as \k<name> or
908 \k'name' as well as (?P=name). The new forms, as well as being in Perl
909 5.10, are also .NET compatible.
911 (d) A conditional reference to a named group can now use the syntax
912 (?(<name>) or (?('name') as well as (?(name).
914 (e) A "conditional group" of the form (?(DEFINE)...) can be used to define
915 groups (named and numbered) that are never evaluated inline, but can be
916 called as "subroutines" from elsewhere. In effect, the DEFINE condition
917 is always false. There may be only one alternative in such a group.
919 (f) A test for recursion can be given as (?(R1).. or (?(R&name)... as well
920 as the simple (?(R). The condition is true only if the most recent
921 recursion is that of the given number or name. It does not search out
922 through the entire recursion stack.
924 (g) The escape \gN or \g{N} has been added, where N is a positive or
925 negative number, specifying an absolute or relative reference.
927 35. Tidied to get rid of some further signed/unsigned compiler warnings and
928 some "unreachable code" warnings.
930 36. Updated the Unicode property tables to Unicode version 5.0.0. Amongst other
931 things, this adds five new scripts.
933 37. Perl ignores orphaned \E escapes completely. PCRE now does the same.
934 There were also incompatibilities regarding the handling of \Q..\E inside
935 character classes, for example with patterns like [\Qa\E-\Qz\E] where the
936 hyphen was adjacent to \Q or \E. I hope I've cleared all this up now.
938 38. Like Perl, PCRE detects when an indefinitely repeated parenthesized group
939 matches an empty string, and forcibly breaks the loop. There were bugs in
940 this code in non-simple cases. For a pattern such as ^(a()*)* matched
941 against aaaa the result was just "a" rather than "aaaa", for example. Two
942 separate and independent bugs (that affected different cases) have been
943 fixed.
945 39. Refactored the code to abolish the use of different opcodes for small
946 capturing bracket numbers. This is a tidy that I avoided doing when I
947 removed the limit on the number of capturing brackets for 3.5 back in 2001.
948 The new approach is not only tidier, it makes it possible to reduce the
949 memory needed to fix the previous bug (38).
951 40. Implemented PCRE_NEWLINE_ANY to recognize any of the Unicode newline
952 sequences (http://unicode.org/unicode/reports/tr18/) as "newline" when
953 processing dot, circumflex, or dollar metacharacters, or #-comments in /x
954 mode.
956 41. Add \R to match any Unicode newline sequence, as suggested in the Unicode
957 report.
959 42. Applied patch, originally from Ari Pollak, modified by Google, to allow
960 copy construction and assignment in the C++ wrapper.
962 43. Updated pcregrep to support "--newline=any". In the process, I fixed a
963 couple of bugs that could have given wrong results in the "--newline=crlf"
964 case.
966 44. Added a number of casts and did some reorganization of signed/unsigned int
967 variables following suggestions from Dair Grant. Also renamed the variable
968 "this" as "item" because it is a C++ keyword.
970 45. Arranged for dftables to add
972 #include "pcre_internal.h"
974 to pcre_chartables.c because without it, gcc 4.x may remove the array
975 definition from the final binary if PCRE is built into a static library and
976 dead code stripping is activated.
978 46. For an unanchored pattern, if a match attempt fails at the start of a
979 newline sequence, and the newline setting is CRLF or ANY, and the next two
980 characters are CRLF, advance by two characters instead of one.
983 Version 6.7 04-Jul-06
984 ---------------------
986 1. In order to handle tests when input lines are enormously long, pcretest has
987 been re-factored so that it automatically extends its buffers when
988 necessary. The code is crude, but this _is_ just a test program. The
989 default size has been increased from 32K to 50K.
991 2. The code in pcre_study() was using the value of the re argument before
992 testing it for NULL. (Of course, in any sensible call of the function, it
993 won't be NULL.)
995 3. The memmove() emulation function in pcre_internal.h, which is used on
996 systems that lack both memmove() and bcopy() - that is, hardly ever -
997 was missing a "static" storage class specifier.
999 4. When UTF-8 mode was not set, PCRE looped when compiling certain patterns
1000 containing an extended class (one that cannot be represented by a bitmap
1001 because it contains high-valued characters or Unicode property items, e.g.
1002 [\pZ]). Almost always one would set UTF-8 mode when processing such a
1003 pattern, but PCRE should not loop if you do not (it no longer does).
1004 [Detail: two cases were found: (a) a repeated subpattern containing an
1005 extended class; (b) a recursive reference to a subpattern that followed a
1006 previous extended class. It wasn't skipping over the extended class
1007 correctly when UTF-8 mode was not set.]
1009 5. A negated single-character class was not being recognized as fixed-length
1010 in lookbehind assertions such as (?<=[^f]), leading to an incorrect
1011 compile error "lookbehind assertion is not fixed length".
1013 6. The RunPerlTest auxiliary script was showing an unexpected difference
1014 between PCRE and Perl for UTF-8 tests. It turns out that it is hard to
1015 write a Perl script that can interpret lines of an input file either as
1016 byte characters or as UTF-8, which is what "perltest" was being required to
1017 do for the non-UTF-8 and UTF-8 tests, respectively. Essentially what you
1018 can't do is switch easily at run time between having the "use utf8;" pragma
1019 or not. In the end, I fudged it by using the RunPerlTest script to insert
1020 "use utf8;" explicitly for the UTF-8 tests.
1022 7. In multiline (/m) mode, PCRE was matching ^ after a terminating newline at
1023 the end of the subject string, contrary to the documentation and to what
1024 Perl does. This was true of both matching functions. Now it matches only at
1025 the start of the subject and immediately after *internal* newlines.
1027 8. A call of pcre_fullinfo() from pcretest to get the option bits was passing
1028 a pointer to an int instead of a pointer to an unsigned long int. This
1029 caused problems on 64-bit systems.
1031 9. Applied a patch from the folks at Google to pcrecpp.cc, to fix "another
1032 instance of the 'standard' template library not being so standard".
1034 10. There was no check on the number of named subpatterns nor the maximum
1035 length of a subpattern name. The product of these values is used to compute
1036 the size of the memory block for a compiled pattern. By supplying a very
1037 long subpattern name and a large number of named subpatterns, the size
1038 computation could be caused to overflow. This is now prevented by limiting
1039 the length of names to 32 characters, and the number of named subpatterns
1040 to 10,000.
1042 11. Subpatterns that are repeated with specific counts have to be replicated in
1043 the compiled pattern. The size of memory for this was computed from the
1044 length of the subpattern and the repeat count. The latter is limited to
1045 65535, but there was no limit on the former, meaning that integer overflow
1046 could in principle occur. The compiled length of a repeated subpattern is
1047 now limited to 30,000 bytes in order to prevent this.
1049 12. Added the optional facility to have named substrings with the same name.
1051 13. Added the ability to use a named substring as a condition, using the
1052 Python syntax: (?(name)yes|no). This overloads (?(R)... and names that
1053 are numbers (not recommended). Forward references are permitted.
1055 14. Added forward references in named backreferences (if you see what I mean).
1057 15. In UTF-8 mode, with the PCRE_DOTALL option set, a quantified dot in the
1058 pattern could run off the end of the subject. For example, the pattern
1059 "(?s)(.{1,5})"8 did this with the subject "ab".
1061 16. If PCRE_DOTALL or PCRE_MULTILINE were set, pcre_dfa_exec() behaved as if
1062 PCRE_CASELESS was set when matching characters that were quantified with ?
1063 or *.
1065 17. A character class other than a single negated character that had a minimum
1066 but no maximum quantifier - for example [ab]{6,} - was not handled
1067 correctly by pce_dfa_exec(). It would match only one character.
1069 18. A valid (though odd) pattern that looked like a POSIX character
1070 class but used an invalid character after [ (for example [[,abc,]]) caused
1071 pcre_compile() to give the error "Failed: internal error: code overflow" or
1072 in some cases to crash with a glibc free() error. This could even happen if
1073 the pattern terminated after [[ but there just happened to be a sequence of
1074 letters, a binary zero, and a closing ] in the memory that followed.
1076 19. Perl's treatment of octal escapes in the range \400 to \777 has changed
1077 over the years. Originally (before any Unicode support), just the bottom 8
1078 bits were taken. Thus, for example, \500 really meant \100. Nowadays the
1079 output from "man perlunicode" includes this:
1081 The regular expression compiler produces polymorphic opcodes. That
1082 is, the pattern adapts to the data and automatically switches to
1083 the Unicode character scheme when presented with Unicode data--or
1084 instead uses a traditional byte scheme when presented with byte
1085 data.
1087 Sadly, a wide octal escape does not cause a switch, and in a string with
1088 no other multibyte characters, these octal escapes are treated as before.
1089 Thus, in Perl, the pattern /\500/ actually matches \100 but the pattern
1090 /\500|\x{1ff}/ matches \500 or \777 because the whole thing is treated as a
1091 Unicode string.
1093 I have not perpetrated such confusion in PCRE. Up till now, it took just
1094 the bottom 8 bits, as in old Perl. I have now made octal escapes with
1095 values greater than \377 illegal in non-UTF-8 mode. In UTF-8 mode they
1096 translate to the appropriate multibyte character.
1098 29. Applied some refactoring to reduce the number of warnings from Microsoft
1099 and Borland compilers. This has included removing the fudge introduced
1100 seven years ago for the OS/2 compiler (see 2.02/2 below) because it caused
1101 a warning about an unused variable.
1103 21. PCRE has not included VT (character 0x0b) in the set of whitespace
1104 characters since release 4.0, because Perl (from release 5.004) does not.
1105 [Or at least, is documented not to: some releases seem to be in conflict
1106 with the documentation.] However, when a pattern was studied with
1107 pcre_study() and all its branches started with \s, PCRE still included VT
1108 as a possible starting character. Of course, this did no harm; it just
1109 caused an unnecessary match attempt.
1111 22. Removed a now-redundant internal flag bit that recorded the fact that case
1112 dependency changed within the pattern. This was once needed for "required
1113 byte" processing, but is no longer used. This recovers a now-scarce options
1114 bit. Also moved the least significant internal flag bit to the most-
1115 significant bit of the word, which was not previously used (hangover from
1116 the days when it was an int rather than a uint) to free up another bit for
1117 the future.
1119 23. Added support for CRLF line endings as well as CR and LF. As well as the
1120 default being selectable at build time, it can now be changed at runtime
1121 via the PCRE_NEWLINE_xxx flags. There are now options for pcregrep to
1122 specify that it is scanning data with non-default line endings.
1124 24. Changed the definition of CXXLINK to make it agree with the definition of
1125 LINK in the Makefile, by replacing LDFLAGS to CXXFLAGS.
1127 25. Applied Ian Taylor's patches to avoid using another stack frame for tail
1128 recursions. This makes a big different to stack usage for some patterns.
1130 26. If a subpattern containing a named recursion or subroutine reference such
1131 as (?P>B) was quantified, for example (xxx(?P>B)){3}, the calculation of
1132 the space required for the compiled pattern went wrong and gave too small a
1133 value. Depending on the environment, this could lead to "Failed: internal
1134 error: code overflow at offset 49" or "glibc detected double free or
1135 corruption" errors.
1137 27. Applied patches from Google (a) to support the new newline modes and (b) to
1138 advance over multibyte UTF-8 characters in GlobalReplace.
1140 28. Change free() to pcre_free() in pcredemo.c. Apparently this makes a
1141 difference for some implementation of PCRE in some Windows version.
1143 29. Added some extra testing facilities to pcretest:
1145 \q<number> in a data line sets the "match limit" value
1146 \Q<number> in a data line sets the "match recursion limt" value
1147 -S <number> sets the stack size, where <number> is in megabytes
1149 The -S option isn't available for Windows.
1152 Version 6.6 06-Feb-06
1153 ---------------------
1155 1. Change 16(a) for 6.5 broke things, because PCRE_DATA_SCOPE was not defined
1156 in pcreposix.h. I have copied the definition from pcre.h.
1158 2. Change 25 for 6.5 broke compilation in a build directory out-of-tree
1159 because pcre.h is no longer a built file.
1161 3. Added Jeff Friedl's additional debugging patches to pcregrep. These are
1162 not normally included in the compiled code.
1165 Version 6.5 01-Feb-06
1166 ---------------------
1168 1. When using the partial match feature with pcre_dfa_exec(), it was not
1169 anchoring the second and subsequent partial matches at the new starting
1170 point. This could lead to incorrect results. For example, with the pattern
1171 /1234/, partially matching against "123" and then "a4" gave a match.
1173 2. Changes to pcregrep:
1175 (a) All non-match returns from pcre_exec() were being treated as failures
1176 to match the line. Now, unless the error is PCRE_ERROR_NOMATCH, an
1177 error message is output. Some extra information is given for the
1179 probably the only errors that are likely to be caused by users (by
1180 specifying a regex that has nested indefinite repeats, for instance).
1181 If there are more than 20 of these errors, pcregrep is abandoned.
1183 (b) A binary zero was treated as data while matching, but terminated the
1184 output line if it was written out. This has been fixed: binary zeroes
1185 are now no different to any other data bytes.
1187 (c) Whichever of the LC_ALL or LC_CTYPE environment variables is set is
1188 used to set a locale for matching. The --locale=xxxx long option has
1189 been added (no short equivalent) to specify a locale explicitly on the
1190 pcregrep command, overriding the environment variables.
1192 (d) When -B was used with -n, some line numbers in the output were one less
1193 than they should have been.
1195 (e) Added the -o (--only-matching) option.
1197 (f) If -A or -C was used with -c (count only), some lines of context were
1198 accidentally printed for the final match.
1200 (g) Added the -H (--with-filename) option.
1202 (h) The combination of options -rh failed to suppress file names for files
1203 that were found from directory arguments.
1205 (i) Added the -D (--devices) and -d (--directories) options.
1207 (j) Added the -F (--fixed-strings) option.
1209 (k) Allow "-" to be used as a file name for -f as well as for a data file.
1211 (l) Added the --colo(u)r option.
1213 (m) Added Jeffrey Friedl's -S testing option, but within #ifdefs so that it
1214 is not present by default.
1216 3. A nasty bug was discovered in the handling of recursive patterns, that is,
1217 items such as (?R) or (?1), when the recursion could match a number of
1218 alternatives. If it matched one of the alternatives, but subsequently,
1219 outside the recursion, there was a failure, the code tried to back up into
1220 the recursion. However, because of the way PCRE is implemented, this is not
1221 possible, and the result was an incorrect result from the match.
1223 In order to prevent this happening, the specification of recursion has
1224 been changed so that all such subpatterns are automatically treated as
1225 atomic groups. Thus, for example, (?R) is treated as if it were (?>(?R)).
1227 4. I had overlooked the fact that, in some locales, there are characters for
1228 which isalpha() is true but neither isupper() nor islower() are true. In
1229 the fr_FR locale, for instance, the \xAA and \xBA characters (ordmasculine
1230 and ordfeminine) are like this. This affected the treatment of \w and \W
1231 when they appeared in character classes, but not when they appeared outside
1232 a character class. The bit map for "word" characters is now created
1233 separately from the results of isalnum() instead of just taking it from the
1234 upper, lower, and digit maps. (Plus the underscore character, of course.)
1236 5. The above bug also affected the handling of POSIX character classes such as
1237 [[:alpha:]] and [[:alnum:]]. These do not have their own bit maps in PCRE's
1238 permanent tables. Instead, the bit maps for such a class were previously
1239 created as the appropriate unions of the upper, lower, and digit bitmaps.
1240 Now they are created by subtraction from the [[:word:]] class, which has
1241 its own bitmap.
1243 6. The [[:blank:]] character class matches horizontal, but not vertical space.
1244 It is created by subtracting the vertical space characters (\x09, \x0a,
1245 \x0b, \x0c) from the [[:space:]] bitmap. Previously, however, the
1246 subtraction was done in the overall bitmap for a character class, meaning
1247 that a class such as [\x0c[:blank:]] was incorrect because \x0c would not
1248 be recognized. This bug has been fixed.
1250 7. Patches from the folks at Google:
1252 (a) pcrecpp.cc: "to handle a corner case that may or may not happen in
1253 real life, but is still worth protecting against".
1255 (b) pcrecpp.cc: "corrects a bug when negative radixes are used with
1256 regular expressions".
1258 (c) pcre_scanner.cc: avoid use of std::count() because not all systems
1259 have it.
1261 (d) Split off pcrecpparg.h from pcrecpp.h and had the former built by
1262 "configure" and the latter not, in order to fix a problem somebody had
1263 with compiling the Arg class on HP-UX.
1265 (e) Improve the error-handling of the C++ wrapper a little bit.
1267 (f) New tests for checking recursion limiting.
1269 8. The pcre_memmove() function, which is used only if the environment does not
1270 have a standard memmove() function (and is therefore rarely compiled),
1271 contained two bugs: (a) use of int instead of size_t, and (b) it was not
1272 returning a result (though PCRE never actually uses the result).
1274 9. In the POSIX regexec() interface, if nmatch is specified as a ridiculously
1275 large number - greater than INT_MAX/(3*sizeof(int)) - REG_ESPACE is
1276 returned instead of calling malloc() with an overflowing number that would
1277 most likely cause subsequent chaos.
1279 10. The debugging option of pcretest was not showing the NO_AUTO_CAPTURE flag.
1281 11. The POSIX flag REG_NOSUB is now supported. When a pattern that was compiled
1282 with this option is matched, the nmatch and pmatch options of regexec() are
1283 ignored.
1285 12. Added REG_UTF8 to the POSIX interface. This is not defined by POSIX, but is
1286 provided in case anyone wants to the the POSIX interface with UTF-8
1287 strings.
1289 13. Added CXXLDFLAGS to the Makefile parameters to provide settings only on the
1290 C++ linking (needed for some HP-UX environments).
1292 14. Avoid compiler warnings in get_ucpname() when compiled without UCP support
1293 (unused parameter) and in the pcre_printint() function (omitted "default"
1294 switch label when the default is to do nothing).
1296 15. Added some code to make it possible, when PCRE is compiled as a C++
1297 library, to replace subject pointers for pcre_exec() with a smart pointer
1298 class, thus making it possible to process discontinuous strings.
1300 16. The two macros PCRE_EXPORT and PCRE_DATA_SCOPE are confusing, and perform
1301 much the same function. They were added by different people who were trying
1302 to make PCRE easy to compile on non-Unix systems. It has been suggested
1303 that PCRE_EXPORT be abolished now that there is more automatic apparatus
1304 for compiling on Windows systems. I have therefore replaced it with
1305 PCRE_DATA_SCOPE. This is set automatically for Windows; if not set it
1306 defaults to "extern" for C or "extern C" for C++, which works fine on
1307 Unix-like systems. It is now possible to override the value of PCRE_DATA_
1308 SCOPE with something explicit in config.h. In addition:
1310 (a) pcreposix.h still had just "extern" instead of either of these macros;
1311 I have replaced it with PCRE_DATA_SCOPE.
1313 (b) Functions such as _pcre_xclass(), which are internal to the library,
1314 but external in the C sense, all had PCRE_EXPORT in their definitions.
1315 This is apparently wrong for the Windows case, so I have removed it.
1316 (It makes no difference on Unix-like systems.)
1318 17. Added a new limit, MATCH_LIMIT_RECURSION, which limits the depth of nesting
1319 of recursive calls to match(). This is different to MATCH_LIMIT because
1320 that limits the total number of calls to match(), not all of which increase
1321 the depth of recursion. Limiting the recursion depth limits the amount of
1322 stack (or heap if NO_RECURSE is set) that is used. The default can be set
1323 when PCRE is compiled, and changed at run time. A patch from Google adds
1324 this functionality to the C++ interface.
1326 18. Changes to the handling of Unicode character properties:
1328 (a) Updated the table to Unicode 4.1.0.
1330 (b) Recognize characters that are not in the table as "Cn" (undefined).
1332 (c) I revised the way the table is implemented to a much improved format
1333 which includes recognition of ranges. It now supports the ranges that
1334 are defined in UnicodeData.txt, and it also amalgamates other
1335 characters into ranges. This has reduced the number of entries in the
1336 table from around 16,000 to around 3,000, thus reducing its size
1337 considerably. I realized I did not need to use a tree structure after
1338 all - a binary chop search is just as efficient. Having reduced the
1339 number of entries, I extended their size from 6 bytes to 8 bytes to
1340 allow for more data.
1342 (d) Added support for Unicode script names via properties such as \p{Han}.
1344 19. In UTF-8 mode, a backslash followed by a non-Ascii character was not
1345 matching that character.
1347 20. When matching a repeated Unicode property with a minimum greater than zero,
1348 (for example \pL{2,}), PCRE could look past the end of the subject if it
1349 reached it while seeking the minimum number of characters. This could
1350 happen only if some of the characters were more than one byte long, because
1351 there is a check for at least the minimum number of bytes.
1353 21. Refactored the implementation of \p and \P so as to be more general, to
1354 allow for more different types of property in future. This has changed the
1355 compiled form incompatibly. Anybody with saved compiled patterns that use
1356 \p or \P will have to recompile them.
1358 22. Added "Any" and "L&" to the supported property types.
1360 23. Recognize \x{...} as a code point specifier, even when not in UTF-8 mode,
1361 but give a compile time error if the value is greater than 0xff.
1363 24. The man pages for pcrepartial, pcreprecompile, and pcre_compile2 were
1364 accidentally not being installed or uninstalled.
1366 25. The pcre.h file was built from pcre.h.in, but the only changes that were
1367 made were to insert the current release number. This seemed silly, because
1368 it made things harder for people building PCRE on systems that don't run
1369 "configure". I have turned pcre.h into a distributed file, no longer built
1370 by "configure", with the version identification directly included. There is
1371 no longer a pcre.h.in file.
1373 However, this change necessitated a change to the pcre-config script as
1374 well. It is built from pcre-config.in, and one of the substitutions was the
1375 release number. I have updated configure.ac so that ./configure now finds
1376 the release number by grepping pcre.h.
1378 26. Added the ability to run the tests under valgrind.
1381 Version 6.4 05-Sep-05
1382 ---------------------
1384 1. Change 6.0/10/(l) to pcregrep introduced a bug that caused separator lines
1385 "--" to be printed when multiple files were scanned, even when none of the
1386 -A, -B, or -C options were used. This is not compatible with Gnu grep, so I
1387 consider it to be a bug, and have restored the previous behaviour.
1389 2. A couple of code tidies to get rid of compiler warnings.
1391 3. The pcretest program used to cheat by referring to symbols in the library
1392 whose names begin with _pcre_. These are internal symbols that are not
1393 really supposed to be visible externally, and in some environments it is
1394 possible to suppress them. The cheating is now confined to including
1395 certain files from the library's source, which is a bit cleaner.
1397 4. Renamed pcre.in as pcre.h.in to go with pcrecpp.h.in; it also makes the
1398 file's purpose clearer.
1400 5. Reorganized pcre_ucp_findchar().
1403 Version 6.3 15-Aug-05
1404 ---------------------
1406 1. The file libpcre.pc.in did not have general read permission in the tarball.
1408 2. There were some problems when building without C++ support:
1410 (a) If C++ support was not built, "make install" and "make test" still
1411 tried to test it.
1413 (b) There were problems when the value of CXX was explicitly set. Some
1414 changes have been made to try to fix these, and ...
1416 (c) --disable-cpp can now be used to explicitly disable C++ support.
1418 (d) The use of @CPP_OBJ@ directly caused a blank line preceded by a
1419 backslash in a target when C++ was disabled. This confuses some
1420 versions of "make", apparently. Using an intermediate variable solves
1421 this. (Same for CPP_LOBJ.)
1423 3. $(LINK_FOR_BUILD) now includes $(CFLAGS_FOR_BUILD) and $(LINK)
1424 (non-Windows) now includes $(CFLAGS) because these flags are sometimes
1425 necessary on certain architectures.
1427 4. Added a setting of -export-symbols-regex to the link command to remove
1428 those symbols that are exported in the C sense, but actually are local
1429 within the library, and not documented. Their names all begin with
1430 "_pcre_". This is not a perfect job, because (a) we have to except some
1431 symbols that pcretest ("illegally") uses, and (b) the facility isn't always
1432 available (and never for static libraries). I have made a note to try to
1433 find a way round (a) in the future.
1436 Version 6.2 01-Aug-05
1437 ---------------------
1439 1. There was no test for integer overflow of quantifier values. A construction
1440 such as {1111111111111111} would give undefined results. What is worse, if
1441 a minimum quantifier for a parenthesized subpattern overflowed and became
1442 negative, the calculation of the memory size went wrong. This could have
1443 led to memory overwriting.
1445 2. Building PCRE using VPATH was broken. Hopefully it is now fixed.
1447 3. Added "b" to the 2nd argument of fopen() in dftables.c, for non-Unix-like
1448 operating environments where this matters.
1450 4. Applied Giuseppe Maxia's patch to add additional features for controlling
1451 PCRE options from within the C++ wrapper.
1453 5. Named capturing subpatterns were not being correctly counted when a pattern
1454 was compiled. This caused two problems: (a) If there were more than 100
1455 such subpatterns, the calculation of the memory needed for the whole
1456 compiled pattern went wrong, leading to an overflow error. (b) Numerical
1457 back references of the form \12, where the number was greater than 9, were
1458 not recognized as back references, even though there were sufficient
1459 previous subpatterns.
1461 6. Two minor patches to pcrecpp.cc in order to allow it to compile on older
1462 versions of gcc, e.g. 2.95.4.
1465 Version 6.1 21-Jun-05
1466 ---------------------
1468 1. There was one reference to the variable "posix" in pcretest.c that was not
1469 surrounded by "#if !defined NOPOSIX".
1471 2. Make it possible to compile pcretest without DFA support, UTF8 support, or
1472 the cross-check on the old pcre_info() function, for the benefit of the
1473 cut-down version of PCRE that is currently imported into Exim.
1475 3. A (silly) pattern starting with (?i)(?-i) caused an internal space
1476 allocation error. I've done the easy fix, which wastes 2 bytes for sensible
1477 patterns that start (?i) but I don't think that matters. The use of (?i) is
1478 just an example; this all applies to the other options as well.
1480 4. Since libtool seems to echo the compile commands it is issuing, the output
1481 from "make" can be reduced a bit by putting "@" in front of each libtool
1482 compile command.
1484 5. Patch from the folks at Google for configure.in to be a bit more thorough
1485 in checking for a suitable C++ installation before trying to compile the
1486 C++ stuff. This should fix a reported problem when a compiler was present,
1487 but no suitable headers.
1489 6. The man pages all had just "PCRE" as their title. I have changed them to
1490 be the relevant file name. I have also arranged that these names are
1491 retained in the file doc/pcre.txt, which is a concatenation in text format
1492 of all the man pages except the little individual ones for each function.
1494 7. The NON-UNIX-USE file had not been updated for the different set of source
1495 files that come with release 6. I also added a few comments about the C++
1496 wrapper.
1499 Version 6.0 07-Jun-05
1500 ---------------------
1502 1. Some minor internal re-organization to help with my DFA experiments.
1504 2. Some missing #ifdef SUPPORT_UCP conditionals in pcretest and printint that
1505 didn't matter for the library itself when fully configured, but did matter
1506 when compiling without UCP support, or within Exim, where the ucp files are
1507 not imported.
1509 3. Refactoring of the library code to split up the various functions into
1510 different source modules. The addition of the new DFA matching code (see
1511 below) to a single monolithic source would have made it really too
1512 unwieldy, quite apart from causing all the code to be include in a
1513 statically linked application, when only some functions are used. This is
1514 relevant even without the DFA addition now that patterns can be compiled in
1515 one application and matched in another.
1517 The downside of splitting up is that there have to be some external
1518 functions and data tables that are used internally in different modules of
1519 the library but which are not part of the API. These have all had their
1520 names changed to start with "_pcre_" so that they are unlikely to clash
1521 with other external names.
1523 4. Added an alternate matching function, pcre_dfa_exec(), which matches using
1524 a different (DFA) algorithm. Although it is slower than the original
1525 function, it does have some advantages for certain types of matching
1526 problem.
1528 5. Upgrades to pcretest in order to test the features of pcre_dfa_exec(),
1529 including restarting after a partial match.
1531 6. A patch for pcregrep that defines INVALID_FILE_ATTRIBUTES if it is not
1532 defined when compiling for Windows was sent to me. I have put it into the
1533 code, though I have no means of testing or verifying it.
1535 7. Added the pcre_refcount() auxiliary function.
1537 8. Added the PCRE_FIRSTLINE option. This constrains an unanchored pattern to
1538 match before or at the first newline in the subject string. In pcretest,
1539 the /f option on a pattern can be used to set this.
1541 9. A repeated \w when used in UTF-8 mode with characters greater than 256
1542 would behave wrongly. This has been present in PCRE since release 4.0.
1544 10. A number of changes to the pcregrep command:
1546 (a) Refactored how -x works; insert ^(...)$ instead of setting
1547 PCRE_ANCHORED and checking the length, in preparation for adding
1548 something similar for -w.
1550 (b) Added the -w (match as a word) option.
1552 (c) Refactored the way lines are read and buffered so as to have more
1553 than one at a time available.
1555 (d) Implemented a pcregrep test script.
1557 (e) Added the -M (multiline match) option. This allows patterns to match
1558 over several lines of the subject. The buffering ensures that at least
1559 8K, or the rest of the document (whichever is the shorter) is available
1560 for matching (and similarly the previous 8K for lookbehind assertions).
1562 (f) Changed the --help output so that it now says
1564 -w, --word-regex(p)
1566 instead of two lines, one with "regex" and the other with "regexp"
1567 because that confused at least one person since the short forms are the
1568 same. (This required a bit of code, as the output is generated
1569 automatically from a table. It wasn't just a text change.)
1571 (g) -- can be used to terminate pcregrep options if the next thing isn't an
1572 option but starts with a hyphen. Could be a pattern or a path name
1573 starting with a hyphen, for instance.
1575 (h) "-" can be given as a file name to represent stdin.
1577 (i) When file names are being printed, "(standard input)" is used for
1578 the standard input, for compatibility with GNU grep. Previously
1579 "<stdin>" was used.
1581 (j) The option --label=xxx can be used to supply a name to be used for
1582 stdin when file names are being printed. There is no short form.
1584 (k) Re-factored the options decoding logic because we are going to add
1585 two more options that take data. Such options can now be given in four
1586 different ways, e.g. "-fname", "-f name", "--file=name", "--file name".
1588 (l) Added the -A, -B, and -C options for requesting that lines of context
1589 around matches be printed.
1591 (m) Added the -L option to print the names of files that do not contain
1592 any matching lines, that is, the complement of -l.
1594 (n) The return code is 2 if any file cannot be opened, but pcregrep does
1595 continue to scan other files.
1597 (o) The -s option was incorrectly implemented. For compatibility with other
1598 greps, it now suppresses the error message for a non-existent or non-
1599 accessible file (but not the return code). There is a new option called
1600 -q that suppresses the output of matching lines, which was what -s was
1601 previously doing.
1603 (p) Added --include and --exclude options to specify files for inclusion
1604 and exclusion when recursing.
1606 11. The Makefile was not using the Autoconf-supported LDFLAGS macro properly.
1607 Hopefully, it now does.
1609 12. Missing cast in pcre_study().
1611 13. Added an "uninstall" target to the makefile.
1613 14. Replaced "extern" in the function prototypes in Makefile.in with
1614 "PCRE_DATA_SCOPE", which defaults to 'extern' or 'extern "C"' in the Unix
1615 world, but is set differently for Windows.
1617 15. Added a second compiling function called pcre_compile2(). The only
1618 difference is that it has an extra argument, which is a pointer to an
1619 integer error code. When there is a compile-time failure, this is set
1620 non-zero, in addition to the error test pointer being set to point to an
1621 error message. The new argument may be NULL if no error number is required
1622 (but then you may as well call pcre_compile(), which is now just a
1623 wrapper). This facility is provided because some applications need a
1624 numeric error indication, but it has also enabled me to tidy up the way
1625 compile-time errors are handled in the POSIX wrapper.
1627 16. Added VPATH=.libs to the makefile; this should help when building with one
1628 prefix path and installing with another. (Or so I'm told by someone who
1629 knows more about this stuff than I do.)
1631 17. Added a new option, REG_DOTALL, to the POSIX function regcomp(). This
1632 passes PCRE_DOTALL to the pcre_compile() function, making the "." character
1633 match everything, including newlines. This is not POSIX-compatible, but
1634 somebody wanted the feature. From pcretest it can be activated by using
1635 both the P and the s flags.
1637 18. AC_PROG_LIBTOOL appeared twice in Makefile.in. Removed one.
1639 19. libpcre.pc was being incorrectly installed as executable.
1641 20. A couple of places in pcretest check for end-of-line by looking for '\n';
1642 it now also looks for '\r' so that it will work unmodified on Windows.
1644 21. Added Google's contributed C++ wrapper to the distribution.
1646 22. Added some untidy missing memory free() calls in pcretest, to keep
1647 Electric Fence happy when testing.
1651 Version 5.0 13-Sep-04
1652 ---------------------
1654 1. Internal change: literal characters are no longer packed up into items
1655 containing multiple characters in a single byte-string. Each character
1656 is now matched using a separate opcode. However, there may be more than one
1657 byte in the character in UTF-8 mode.
1659 2. The pcre_callout_block structure has two new fields: pattern_position and
1660 next_item_length. These contain the offset in the pattern to the next match
1661 item, and its length, respectively.
1663 3. The PCRE_AUTO_CALLOUT option for pcre_compile() requests the automatic
1664 insertion of callouts before each pattern item. Added the /C option to
1665 pcretest to make use of this.
1667 4. On the advice of a Windows user, the lines
1669 #if defined(_WIN32) || defined(WIN32)
1670 _setmode( _fileno( stdout ), 0x8000 );
1671 #endif /* defined(_WIN32) || defined(WIN32) */
1673 have been added to the source of pcretest. This apparently does useful
1674 magic in relation to line terminators.
1676 5. Changed "r" and "w" in the calls to fopen() in pcretest to "rb" and "wb"
1677 for the benefit of those environments where the "b" makes a difference.
1679 6. The icc compiler has the same options as gcc, but "configure" doesn't seem
1680 to know about it. I have put a hack into configure.in that adds in code
1681 to set GCC=yes if CC=icc. This seems to end up at a point in the
1682 generated configure script that is early enough to affect the setting of
1683 compiler options, which is what is needed, but I have no means of testing
1684 whether it really works. (The user who reported this had patched the
1685 generated configure script, which of course I cannot do.)
1687 LATER: After change 22 below (new libtool files), the configure script
1688 seems to know about icc (and also ecc). Therefore, I have commented out
1689 this hack in configure.in.
1691 7. Added support for pkg-config (2 patches were sent in).
1693 8. Negated POSIX character classes that used a combination of internal tables
1694 were completely broken. These were [[:^alpha:]], [[:^alnum:]], and
1695 [[:^ascii]]. Typically, they would match almost any characters. The other
1696 POSIX classes were not broken in this way.
1698 9. Matching the pattern "\b.*?" against "ab cd", starting at offset 1, failed
1699 to find the match, as PCRE was deluded into thinking that the match had to
1700 start at the start point or following a newline. The same bug applied to
1701 patterns with negative forward assertions or any backward assertions
1702 preceding ".*" at the start, unless the pattern required a fixed first
1703 character. This was a failing pattern: "(?!.bcd).*". The bug is now fixed.
1705 10. In UTF-8 mode, when moving forwards in the subject after a failed match
1706 starting at the last subject character, bytes beyond the end of the subject
1707 string were read.
1709 11. Renamed the variable "class" as "classbits" to make life easier for C++
1710 users. (Previously there was a macro definition, but it apparently wasn't
1711 enough.)
1713 12. Added the new field "tables" to the extra data so that tables can be passed
1714 in at exec time, or the internal tables can be re-selected. This allows
1715 a compiled regex to be saved and re-used at a later time by a different
1716 program that might have everything at different addresses.
1718 13. Modified the pcre-config script so that, when run on Solaris, it shows a
1719 -R library as well as a -L library.
1721 14. The debugging options of pcretest (-d on the command line or D on a
1722 pattern) showed incorrect output for anything following an extended class
1723 that contained multibyte characters and which was followed by a quantifier.
1725 15. Added optional support for general category Unicode character properties
1726 via the \p, \P, and \X escapes. Unicode property support implies UTF-8
1727 support. It adds about 90K to the size of the library. The meanings of the
1728 inbuilt class escapes such as \d and \s have NOT been changed.
1730 16. Updated pcredemo.c to include calls to free() to release the memory for the
1731 compiled pattern.
1733 17. The generated file chartables.c was being created in the source directory
1734 instead of in the building directory. This caused the build to fail if the
1735 source directory was different from the building directory, and was
1736 read-only.
1738 18. Added some sample Win commands from Mark Tetrode into the NON-UNIX-USE
1739 file. No doubt somebody will tell me if they don't make sense... Also added
1740 Dan Mooney's comments about building on OpenVMS.
1742 19. Added support for partial matching via the PCRE_PARTIAL option for
1743 pcre_exec() and the \P data escape in pcretest.
1745 20. Extended pcretest with 3 new pattern features:
1747 (i) A pattern option of the form ">rest-of-line" causes pcretest to
1748 write the compiled pattern to the file whose name is "rest-of-line".
1749 This is a straight binary dump of the data, with the saved pointer to
1750 the character tables forced to be NULL. The study data, if any, is
1751 written too. After writing, pcretest reads a new pattern.
1753 (ii) If, instead of a pattern, "<rest-of-line" is given, pcretest reads a
1754 compiled pattern from the given file. There must not be any
1755 occurrences of "<" in the file name (pretty unlikely); if there are,
1756 pcretest will instead treat the initial "<" as a pattern delimiter.
1757 After reading in the pattern, pcretest goes on to read data lines as
1758 usual.
1760 (iii) The F pattern option causes pcretest to flip the bytes in the 32-bit
1761 and 16-bit fields in a compiled pattern, to simulate a pattern that
1762 was compiled on a host of opposite endianness.
1764 21. The pcre-exec() function can now cope with patterns that were compiled on
1765 hosts of opposite endianness, with this restriction:
1767 As for any compiled expression that is saved and used later, the tables
1768 pointer field cannot be preserved; the extra_data field in the arguments
1769 to pcre_exec() should be used to pass in a tables address if a value
1770 other than the default internal tables were used at compile time.
1772 22. Calling pcre_exec() with a negative value of the "ovecsize" parameter is
1773 now diagnosed as an error. Previously, most of the time, a negative number
1774 would have been treated as zero, but if in addition "ovector" was passed as
1775 NULL, a crash could occur.
1777 23. Updated the files ltmain.sh, config.sub, config.guess, and aclocal.m4 with
1778 new versions from the libtool 1.5 distribution (the last one is a copy of
1779 a file called libtool.m4). This seems to have fixed the need to patch
1780 "configure" to support Darwin 1.3 (which I used to do). However, I still
1781 had to patch ltmain.sh to ensure that ${SED} is set (it isn't on my
1782 workstation).
1784 24. Changed the PCRE licence to be the more standard "BSD" licence.
1787 Version 4.5 01-Dec-03
1788 ---------------------
1790 1. There has been some re-arrangement of the code for the match() function so
1791 that it can be compiled in a version that does not call itself recursively.
1792 Instead, it keeps those local variables that need separate instances for
1793 each "recursion" in a frame on the heap, and gets/frees frames whenever it
1794 needs to "recurse". Keeping track of where control must go is done by means
1795 of setjmp/longjmp. The whole thing is implemented by a set of macros that
1796 hide most of the details from the main code, and operates only if
1797 NO_RECURSE is defined while compiling pcre.c. If PCRE is built using the
1798 "configure" mechanism, "--disable-stack-for-recursion" turns on this way of
1799 operating.
1801 To make it easier for callers to provide specially tailored get/free
1802 functions for this usage, two new functions, pcre_stack_malloc, and
1803 pcre_stack_free, are used. They are always called in strict stacking order,
1804 and the size of block requested is always the same.
1806 The PCRE_CONFIG_STACKRECURSE info parameter can be used to find out whether
1807 PCRE has been compiled to use the stack or the heap for recursion. The
1808 -C option of pcretest uses this to show which version is compiled.
1810 A new data escape \S, is added to pcretest; it causes the amounts of store
1811 obtained and freed by both kinds of malloc/free at match time to be added
1812 to the output.
1814 2. Changed the locale test to use "fr_FR" instead of "fr" because that's
1815 what's available on my current Linux desktop machine.
1817 3. When matching a UTF-8 string, the test for a valid string at the start has
1818 been extended. If start_offset is not zero, PCRE now checks that it points
1819 to a byte that is the start of a UTF-8 character. If not, it returns
1820 PCRE_ERROR_BADUTF8_OFFSET (-11). Note: the whole string is still checked;
1821 this is necessary because there may be backward assertions in the pattern.
1822 When matching the same subject several times, it may save resources to use
1823 PCRE_NO_UTF8_CHECK on all but the first call if the string is long.
1825 4. The code for checking the validity of UTF-8 strings has been tightened so
1826 that it rejects (a) strings containing 0xfe or 0xff bytes and (b) strings
1827 containing "overlong sequences".
1829 5. Fixed a bug (appearing twice) that I could not find any way of exploiting!
1830 I had written "if ((digitab[*p++] && chtab_digit) == 0)" where the "&&"
1831 should have been "&", but it just so happened that all the cases this let
1832 through by mistake were picked up later in the function.
1834 6. I had used a variable called "isblank" - this is a C99 function, causing
1835 some compilers to warn. To avoid this, I renamed it (as "blankclass").
1837 7. Cosmetic: (a) only output another newline at the end of pcretest if it is
1838 prompting; (b) run "./pcretest /dev/null" at the start of the test script
1839 so the version is shown; (c) stop "make test" echoing "./RunTest".
1841 8. Added patches from David Burgess to enable PCRE to run on EBCDIC systems.
1843 9. The prototype for memmove() for systems that don't have it was using
1844 size_t, but the inclusion of the header that defines size_t was later. I've
1845 moved the #includes for the C headers earlier to avoid this.
1847 10. Added some adjustments to the code to make it easier to compiler on certain
1848 special systems:
1850 (a) Some "const" qualifiers were missing.
1851 (b) Added the macro EXPORT before all exported functions; by default this
1852 is defined to be empty.
1853 (c) Changed the dftables auxiliary program (that builds chartables.c) so
1854 that it reads its output file name as an argument instead of writing
1855 to the standard output and assuming this can be redirected.
1857 11. In UTF-8 mode, if a recursive reference (e.g. (?1)) followed a character
1858 class containing characters with values greater than 255, PCRE compilation
1859 went into a loop.
1861 12. A recursive reference to a subpattern that was within another subpattern
1862 that had a minimum quantifier of zero caused PCRE to crash. For example,
1863 (x(y(?2))z)? provoked this bug with a subject that got as far as the
1864 recursion. If the recursively-called subpattern itself had a zero repeat,
1865 that was OK.
1867 13. In pcretest, the buffer for reading a data line was set at 30K, but the
1868 buffer into which it was copied (for escape processing) was still set at
1869 1024, so long lines caused crashes.
1871 14. A pattern such as /[ab]{1,3}+/ failed to compile, giving the error
1872 "internal error: code overflow...". This applied to any character class
1873 that was followed by a possessive quantifier.
1875 15. Modified the Makefile to add libpcre.la as a prerequisite for
1876 libpcreposix.la because I was told this is needed for a parallel build to
1877 work.
1879 16. If a pattern that contained .* following optional items at the start was
1880 studied, the wrong optimizing data was generated, leading to matching
1881 errors. For example, studying /[ab]*.*c/ concluded, erroneously, that any
1882 matching string must start with a or b or c. The correct conclusion for
1883 this pattern is that a match can start with any character.
1886 Version 4.4 13-Aug-03
1887 ---------------------
1889 1. In UTF-8 mode, a character class containing characters with values between
1890 127 and 255 was not handled correctly if the compiled pattern was studied.
1891 In fixing this, I have also improved the studying algorithm for such
1892 classes (slightly).
1894 2. Three internal functions had redundant arguments passed to them. Removal
1895 might give a very teeny performance improvement.
1897 3. Documentation bug: the value of the capture_top field in a callout is *one
1898 more than* the number of the hightest numbered captured substring.
1900 4. The Makefile linked pcretest and pcregrep with -lpcre, which could result
1901 in incorrectly linking with a previously installed version. They now link
1902 explicitly with libpcre.la.
1904 5. configure.in no longer needs to recognize Cygwin specially.
1906 6. A problem in pcre.in for Windows platforms is fixed.
1908 7. If a pattern was successfully studied, and the -d (or /D) flag was given to
1909 pcretest, it used to include the size of the study block as part of its
1910 output. Unfortunately, the structure contains a field that has a different
1911 size on different hardware architectures. This meant that the tests that
1912 showed this size failed. As the block is currently always of a fixed size,
1913 this information isn't actually particularly useful in pcretest output, so
1914 I have just removed it.
1916 8. Three pre-processor statements accidentally did not start in column 1.
1917 Sadly, there are *still* compilers around that complain, even though
1918 standard C has not required this for well over a decade. Sigh.
1920 9. In pcretest, the code for checking callouts passed small integers in the
1921 callout_data field, which is a void * field. However, some picky compilers
1922 complained about the casts involved for this on 64-bit systems. Now
1923 pcretest passes the address of the small integer instead, which should get
1924 rid of the warnings.
1926 10. By default, when in UTF-8 mode, PCRE now checks for valid UTF-8 strings at
1927 both compile and run time, and gives an error if an invalid UTF-8 sequence
1928 is found. There is a option for disabling this check in cases where the
1929 string is known to be correct and/or the maximum performance is wanted.
1931 11. In response to a bug report, I changed one line in Makefile.in from
1933 -Wl,--out-implib,.libs/lib@WIN_PREFIX@pcreposix.dll.a \
1934 to
1935 -Wl,--out-implib,.libs/@WIN_PREFIX@libpcreposix.dll.a \
1937 to look similar to other lines, but I have no way of telling whether this
1938 is the right thing to do, as I do not use Windows. No doubt I'll get told
1939 if it's wrong...
1942 Version 4.3 21-May-03
1943 ---------------------
1945 1. Two instances of @WIN_PREFIX@ omitted from the Windows targets in the
1946 Makefile.
1948 2. Some refactoring to improve the quality of the code:
1950 (i) The utf8_table... variables are now declared "const".
1952 (ii) The code for \cx, which used the "case flipping" table to upper case
1953 lower case letters, now just substracts 32. This is ASCII-specific,
1954 but the whole concept of \cx is ASCII-specific, so it seems
1955 reasonable.
1957 (iii) PCRE was using its character types table to recognize decimal and
1958 hexadecimal digits in the pattern. This is silly, because it handles
1959 only 0-9, a-f, and A-F, but the character types table is locale-
1960 specific, which means strange things might happen. A private
1961 table is now used for this - though it costs 256 bytes, a table is
1962 much faster than multiple explicit tests. Of course, the standard
1963 character types table is still used for matching digits in subject
1964 strings against \d.
1966 (iv) Strictly, the identifier ESC_t is reserved by POSIX (all identifiers
1967 ending in _t are). So I've renamed it as ESC_tee.
1969 3. The first argument for regexec() in the POSIX wrapper should have been
1970 defined as "const".
1972 4. Changed pcretest to use malloc() for its buffers so that they can be
1973 Electric Fenced for debugging.
1975 5. There were several places in the code where, in UTF-8 mode, PCRE would try
1976 to read one or more bytes before the start of the subject string. Often this
1977 had no effect on PCRE's behaviour, but in some circumstances it could
1978 provoke a segmentation fault.
1980 6. A lookbehind at the start of a pattern in UTF-8 mode could also cause PCRE
1981 to try to read one or more bytes before the start of the subject string.
1983 7. A lookbehind in a pattern matched in non-UTF-8 mode on a PCRE compiled with
1984 UTF-8 support could misbehave in various ways if the subject string
1985 contained bytes with the 0x80 bit set and the 0x40 bit unset in a lookbehind
1986 area. (PCRE was not checking for the UTF-8 mode flag, and trying to move
1987 back over UTF-8 characters.)
1990 Version 4.2 14-Apr-03
1991 ---------------------
1993 1. Typo "#if SUPPORT_UTF8" instead of "#ifdef SUPPORT_UTF8" fixed.
1995 2. Changes to the building process, supplied by Ronald Landheer-Cieslak
1996 [ON_WINDOWS]: new variable, "#" on non-Windows platforms
1997 [NOT_ON_WINDOWS]: new variable, "#" on Windows platforms
1998 [WIN_PREFIX]: new variable, "cyg" for Cygwin
1999 * Makefile.in: use autoconf substitution for OBJEXT, EXEEXT, BUILD_OBJEXT
2001 Note: automatic setting of the BUILD variables is not yet working
2002 set CPPFLAGS and BUILD_CPPFLAGS (but don't use yet) - should be used at
2003 compile-time but not at link-time
2004 [LINK]: use for linking executables only
2005 make different versions for Windows and non-Windows
2006 [LINKLIB]: new variable, copy of UNIX-style LINK, used for linking
2007 libraries
2008 [LINK_FOR_BUILD]: new variable
2009 [OBJEXT]: use throughout
2010 [EXEEXT]: use throughout
2011 <winshared>: new target
2012 <wininstall>: new target
2013 <dftables.o>: use native compiler
2014 <dftables>: use native linker
2015 <install>: handle Windows platform correctly
2016 <clean>: ditto
2017 <check>: ditto
2018 copy DLL to top builddir before testing
2020 As part of these changes, -no-undefined was removed again. This was reported
2021 to give trouble on HP-UX 11.0, so getting rid of it seems like a good idea
2022 in any case.
2024 3. Some tidies to get rid of compiler warnings:
2026 . In the match_data structure, match_limit was an unsigned long int, whereas
2027 match_call_count was an int. I've made them both unsigned long ints.
2029 . In pcretest the fact that a const uschar * doesn't automatically cast to
2030 a void * provoked a warning.
2032 . Turning on some more compiler warnings threw up some "shadow" variables
2033 and a few more missing casts.
2035 4. If PCRE was complied with UTF-8 support, but called without the PCRE_UTF8
2036 option, a class that contained a single character with a value between 128
2037 and 255 (e.g. /[\xFF]/) caused PCRE to crash.
2039 5. If PCRE was compiled with UTF-8 support, but called without the PCRE_UTF8
2040 option, a class that contained several characters, but with at least one
2041 whose value was between 128 and 255 caused PCRE to crash.
2044 Version 4.1 12-Mar-03
2045 ---------------------
2047 1. Compiling with gcc -pedantic found a couple of places where casts were
2048 needed, and a string in dftables.c that was longer than standard compilers are
2049 required to support.
2051 2. Compiling with Sun's compiler found a few more places where the code could
2052 be tidied up in order to avoid warnings.
2054 3. The variables for cross-compiling were called HOST_CC and HOST_CFLAGS; the
2055 first of these names is deprecated in the latest Autoconf in favour of the name
2056 CC_FOR_BUILD, because "host" is typically used to mean the system on which the
2057 compiled code will be run. I can't find a reference for HOST_CFLAGS, but by
2058 analogy I have changed it to CFLAGS_FOR_BUILD.
2060 4. Added -no-undefined to the linking command in the Makefile, because this is
2061 apparently helpful for Windows. To make it work, also added "-L. -lpcre" to the
2062 linking step for the pcreposix library.
2064 5. PCRE was failing to diagnose the case of two named groups with the same
2065 name.
2067 6. A problem with one of PCRE's optimizations was discovered. PCRE remembers a
2068 literal character that is needed in the subject for a match, and scans along to
2069 ensure that it is present before embarking on the full matching process. This
2070 saves time in cases of nested unlimited repeats that are never going to match.
2071 Problem: the scan can take a lot of time if the subject is very long (e.g.
2072 megabytes), thus penalizing straightforward matches. It is now done only if the
2073 amount of subject to be scanned is less than 1000 bytes.
2075 7. A lesser problem with the same optimization is that it was recording the
2076 first character of an anchored pattern as "needed", thus provoking a search
2077 right along the subject, even when the first match of the pattern was going to
2078 fail. The "needed" character is now not set for anchored patterns, unless it
2079 follows something in the pattern that is of non-fixed length. Thus, it still
2080 fulfils its original purpose of finding quick non-matches in cases of nested
2081 unlimited repeats, but isn't used for simple anchored patterns such as /^abc/.
2084 Version 4.0 17-Feb-03
2085 ---------------------
2087 1. If a comment in an extended regex that started immediately after a meta-item
2088 extended to the end of string, PCRE compiled incorrect data. This could lead to
2089 all kinds of weird effects. Example: /#/ was bad; /()#/ was bad; /a#/ was not.
2091 2. Moved to autoconf 2.53 and libtool 1.4.2.
2093 3. Perl 5.8 no longer needs "use utf8" for doing UTF-8 things. Consequently,
2094 the special perltest8 script is no longer needed - all the tests can be run
2095 from a single perltest script.
2097 4. From 5.004, Perl has not included the VT character (0x0b) in the set defined
2098 by \s. It has now been removed in PCRE. This means it isn't recognized as
2099 whitespace in /x regexes too, which is the same as Perl. Note that the POSIX
2100 class [:space:] *does* include VT, thereby creating a mess.
2102 5. Added the class [:blank:] (a GNU extension from Perl 5.8) to match only
2103 space and tab.
2105 6. Perl 5.005 was a long time ago. It's time to amalgamate the tests that use
2106 its new features into the main test script, reducing the number of scripts.
2108 7. Perl 5.8 has changed the meaning of patterns like /a(?i)b/. Earlier versions
2109 were backward compatible, and made the (?i) apply to the whole pattern, as if
2110 /i were given. Now it behaves more logically, and applies the option setting
2111 only to what follows. PCRE has been changed to follow suit. However, if it
2112 finds options settings right at the start of the pattern, it extracts them into
2113 the global options, as before. Thus, they show up in the info data.
2115 8. Added support for the \Q...\E escape sequence. Characters in between are
2116 treated as literals. This is slightly different from Perl in that $ and @ are
2117 also handled as literals inside the quotes. In Perl, they will cause variable
2118 interpolation. Note the following examples:
2120 Pattern PCRE matches Perl matches
2122 \Qabc$xyz\E abc$xyz abc followed by the contents of $xyz
2123 \Qabc\$xyz\E abc\$xyz abc\$xyz
2124 \Qabc\E\$\Qxyz\E abc$xyz abc$xyz
2126 For compatibility with Perl, \Q...\E sequences are recognized inside character
2127 classes as well as outside them.
2129 9. Re-organized 3 code statements in pcretest to avoid "overflow in
2130 floating-point constant arithmetic" warnings from a Microsoft compiler. Added a
2131 (size_t) cast to one statement in pcretest and one in pcreposix to avoid
2132 signed/unsigned warnings.
2134 10. SunOS4 doesn't have strtoul(). This was used only for unpicking the -o
2135 option for pcretest, so I've replaced it by a simple function that does just
2136 that job.
2138 11. pcregrep was ending with code 0 instead of 2 for the commands "pcregrep" or
2139 "pcregrep -".
2141 12. Added "possessive quantifiers" ?+, *+, ++, and {,}+ which come from Sun's
2142 Java package. This provides some syntactic sugar for simple cases of what my
2143 documentation calls "once-only subpatterns". A pattern such as x*+ is the same
2144 as (?>x*). In other words, if what is inside (?>...) is just a single repeated
2145 item, you can use this simplified notation. Note that only makes sense with
2146 greedy quantifiers. Consequently, the use of the possessive quantifier forces
2147 greediness, whatever the setting of the PCRE_UNGREEDY option.
2149 13. A change of greediness default within a pattern was not taking effect at
2150 the current level for patterns like /(b+(?U)a+)/. It did apply to parenthesized
2151 subpatterns that followed. Patterns like /b+(?U)a+/ worked because the option
2152 was abstracted outside.
2154 14. PCRE now supports the \G assertion. It is true when the current matching
2155 position is at the start point of the match. This differs from \A when the
2156 starting offset is non-zero. Used with the /g option of pcretest (or similar
2157 code), it works in the same way as it does for Perl's /g option. If all
2158 alternatives of a regex begin with \G, the expression is anchored to the start
2159 match position, and the "anchored" flag is set in the compiled expression.
2161 15. Some bugs concerning the handling of certain option changes within patterns
2162 have been fixed. These applied to options other than (?ims). For example,
2163 "a(?x: b c )d" did not match "XabcdY" but did match "Xa b c dY". It should have
2164 been the other way round. Some of this was related to change 7 above.
2166 16. PCRE now gives errors for /[.x.]/ and /[=x=]/ as unsupported POSIX
2167 features, as Perl does. Previously, PCRE gave the warnings only for /[[.x.]]/
2168 and /[[=x=]]/. PCRE now also gives an error for /[:name:]/ because it supports
2169 POSIX classes only within a class (e.g. /[[:alpha:]]/).
2171 17. Added support for Perl's \C escape. This matches one byte, even in UTF8
2172 mode. Unlike ".", it always matches newline, whatever the setting of
2173 PCRE_DOTALL. However, PCRE does not permit \C to appear in lookbehind
2174 assertions. Perl allows it, but it doesn't (in general) work because it can't
2175 calculate the length of the lookbehind. At least, that's the case for Perl
2176 5.8.0 - I've been told they are going to document that it doesn't work in
2177 future.
2179 18. Added an error diagnosis for escapes that PCRE does not support: these are
2180 \L, \l, \N, \P, \p, \U, \u, and \X.
2182 19. Although correctly diagnosing a missing ']' in a character class, PCRE was
2183 reading past the end of the pattern in cases such as /[abcd/.
2185 20. PCRE was getting more memory than necessary for patterns with classes that
2186 contained both POSIX named classes and other characters, e.g. /[[:space:]abc/.
2188 21. Added some code, conditional on #ifdef VPCOMPAT, to make life easier for
2189 compiling PCRE for use with Virtual Pascal.
2191 22. Small fix to the Makefile to make it work properly if the build is done
2192 outside the source tree.
2194 23. Added a new extension: a condition to go with recursion. If a conditional
2195 subpattern starts with (?(R) the "true" branch is used if recursion has
2196 happened, whereas the "false" branch is used only at the top level.
2198 24. When there was a very long string of literal characters (over 255 bytes
2199 without UTF support, over 250 bytes with UTF support), the computation of how
2200 much memory was required could be incorrect, leading to segfaults or other
2201 strange effects.
2203 25. PCRE was incorrectly assuming anchoring (either to start of subject or to
2204 start of line for a non-DOTALL pattern) when a pattern started with (.*) and
2205 there was a subsequent back reference to those brackets. This meant that, for
2206 example, /(.*)\d+\1/ failed to match "abc123bc". Unfortunately, it isn't
2207 possible to check for precisely this case. All we can do is abandon the
2208 optimization if .* occurs inside capturing brackets when there are any back
2209 references whatsoever. (See below for a better fix that came later.)
2211 26. The handling of the optimization for finding the first character of a
2212 non-anchored pattern, and for finding a character that is required later in the
2213 match were failing in some cases. This didn't break the matching; it just
2214 failed to optimize when it could. The way this is done has been re-implemented.
2216 27. Fixed typo in error message for invalid (?R item (it said "(?p").
2218 28. Added a new feature that provides some of the functionality that Perl
2219 provides with (?{...}). The facility is termed a "callout". The way it is done
2220 in PCRE is for the caller to provide an optional function, by setting
2221 pcre_callout to its entry point. Like pcre_malloc and pcre_free, this is a
2222 global variable. By default it is unset, which disables all calling out. To get
2223 the function called, the regex must include (?C) at appropriate points. This
2224 is, in fact, equivalent to (?C0), and any number <= 255 may be given with (?C).
2225 This provides a means of identifying different callout points. When PCRE
2226 reaches such a point in the regex, if pcre_callout has been set, the external
2227 function is called. It is provided with data in a structure called
2228 pcre_callout_block, which is defined in pcre.h. If the function returns 0,
2229 matching continues; if it returns a non-zero value, the match at the current
2230 point fails. However, backtracking will occur if possible. [This was changed
2231 later and other features added - see item 49 below.]
2233 29. pcretest is upgraded to test the callout functionality. It provides a
2234 callout function that displays information. By default, it shows the start of
2235 the match and the current position in the text. There are some new data escapes
2236 to vary what happens:
2238 \C+ in addition, show current contents of captured substrings
2239 \C- do not supply a callout function
2240 \C!n return 1 when callout number n is reached
2241 \C!n!m return 1 when callout number n is reached for the mth time
2243 30. If pcregrep was called with the -l option and just a single file name, it
2244 output "<stdin>" if a match was found, instead of the file name.
2246 31. Improve the efficiency of the POSIX API to PCRE. If the number of capturing
2247 slots is less than POSIX_MALLOC_THRESHOLD, use a block on the stack to pass to
2248 pcre_exec(). This saves a malloc/free per call. The default value of
2249 POSIX_MALLOC_THRESHOLD is 10; it can be changed by --with-posix-malloc-threshold
2250 when configuring.
2252 32. The default maximum size of a compiled pattern is 64K. There have been a
2253 few cases of people hitting this limit. The code now uses macros to handle the
2254 storing of links as offsets within the compiled pattern. It defaults to 2-byte
2255 links, but this can be changed to 3 or 4 bytes by --with-link-size when
2256 configuring. Tests 2 and 5 work only with 2-byte links because they output
2257 debugging information about compiled patterns.
2259 33. Internal code re-arrangements:
2261 (a) Moved the debugging function for printing out a compiled regex into
2262 its own source file (printint.c) and used #include to pull it into
2263 pcretest.c and, when DEBUG is defined, into pcre.c, instead of having two
2264 separate copies.
2266 (b) Defined the list of op-code names for debugging as a macro in
2267 internal.h so that it is next to the definition of the opcodes.
2269 (c) Defined a table of op-code lengths for simpler skipping along compiled
2270 code. This is again a macro in internal.h so that it is next to the
2271 definition of the opcodes.
2273 34. Added support for recursive calls to individual subpatterns, along the
2274 lines of Robin Houston's patch (but implemented somewhat differently).
2276 35. Further mods to the Makefile to help Win32. Also, added code to pcregrep to
2277 allow it to read and process whole directories in Win32. This code was
2278 contributed by Lionel Fourquaux; it has not been tested by me.
2280 36. Added support for named subpatterns. The Python syntax (?P<name>...) is
2281 used to name a group. Names consist of alphanumerics and underscores, and must
2282 be unique. Back references use the syntax (?P=name) and recursive calls use
2283 (?P>name) which is a PCRE extension to the Python extension. Groups still have
2284 numbers. The function pcre_fullinfo() can be used after compilation to extract
2285 a name/number map. There are three relevant calls:
2287 PCRE_INFO_NAMEENTRYSIZE yields the size of each entry in the map
2288 PCRE_INFO_NAMECOUNT yields the number of entries
2289 PCRE_INFO_NAMETABLE yields a pointer to the map.
2291 The map is a vector of fixed-size entries. The size of each entry depends on
2292 the length of the longest name used. The first two bytes of each entry are the
2293 group number, most significant byte first. There follows the corresponding
2294 name, zero terminated. The names are in alphabetical order.
2296 37. Make the maximum literal string in the compiled code 250 for the non-UTF-8
2297 case instead of 255. Making it the same both with and without UTF-8 support
2298 means that the same test output works with both.
2300 38. There was a case of malloc(0) in the POSIX testing code in pcretest. Avoid
2301 calling malloc() with a zero argument.
2303 39. Change 25 above had to resort to a heavy-handed test for the .* anchoring
2304 optimization. I've improved things by keeping a bitmap of backreferences with
2305 numbers 1-31 so that if .* occurs inside capturing brackets that are not in
2306 fact referenced, the optimization can be applied. It is unlikely that a
2307 relevant occurrence of .* (i.e. one which might indicate anchoring or forcing
2308 the match to follow \n) will appear inside brackets with a number greater than
2309 31, but if it does, any back reference > 31 suppresses the optimization.
2311 40. Added a new compile-time option PCRE_NO_AUTO_CAPTURE. This has the effect
2312 of disabling numbered capturing parentheses. Any opening parenthesis that is
2313 not followed by ? behaves as if it were followed by ?: but named parentheses
2314 can still be used for capturing (and they will acquire numbers in the usual
2315 way).
2317 41. Redesigned the return codes from the match() function into yes/no/error so
2318 that errors can be passed back from deep inside the nested calls. A malloc
2319 failure while inside a recursive subpattern call now causes the
2320 PCRE_ERROR_NOMEMORY return instead of quietly going wrong.
2322 42. It is now possible to set a limit on the number of times the match()
2323 function is called in a call to pcre_exec(). This facility makes it possible to
2324 limit the amount of recursion and backtracking, though not in a directly
2325 obvious way, because the match() function is used in a number of different
2326 circumstances. The count starts from zero for each position in the subject
2327 string (for non-anchored patterns). The default limit is, for compatibility, a
2328 large number, namely 10 000 000. You can change this in two ways:
2330 (a) When configuring PCRE before making, you can use --with-match-limit=n
2331 to set a default value for the compiled library.
2333 (b) For each call to pcre_exec(), you can pass a pcre_extra block in which
2334 a different value is set. See 45 below.
2336 If the limit is exceeded, pcre_exec() returns PCRE_ERROR_MATCHLIMIT.
2338 43. Added a new function pcre_config(int, void *) to enable run-time extraction
2339 of things that can be changed at compile time. The first argument specifies
2340 what is wanted and the second points to where the information is to be placed.
2341 The current list of available information is:
2345 The output is an integer that is set to one if UTF-8 support is available;
2346 otherwise it is set to zero.
2350 The output is an integer that it set to the value of the code that is used for
2351 newline. It is either LF (10) or CR (13).
2355 The output is an integer that contains the number of bytes used for internal
2356 linkage in compiled expressions. The value is 2, 3, or 4. See item 32 above.
2360 The output is an integer that contains the threshold above which the POSIX
2361 interface uses malloc() for output vectors. See item 31 above.
2365 The output is an unsigned integer that contains the default limit of the number
2366 of match() calls in a pcre_exec() execution. See 42 above.
2368 44. pcretest has been upgraded by the addition of the -C option. This causes it
2369 to extract all the available output from the new pcre_config() function, and to
2370 output it. The program then exits immediately.
2372 45. A need has arisen to pass over additional data with calls to pcre_exec() in
2373 order to support additional features. One way would have been to define
2374 pcre_exec2() (for example) with extra arguments, but this would not have been
2375 extensible, and would also have required all calls to the original function to
2376 be mapped to the new one. Instead, I have chosen to extend the mechanism that
2377 is used for passing in "extra" data from pcre_study().
2379 The pcre_extra structure is now exposed and defined in pcre.h. It currently
2380 contains the following fields:
2382 flags a bitmap indicating which of the following fields are set
2383 study_data opaque data from pcre_study()
2384 match_limit a way of specifying a limit on match() calls for a specific
2385 call to pcre_exec()
2386 callout_data data for callouts (see 49 below)
2388 The flag bits are also defined in pcre.h, and are
2394 The pcre_study() function now returns one of these new pcre_extra blocks, with
2395 the actual study data pointed to by the study_data field, and the
2396 PCRE_EXTRA_STUDY_DATA flag set. This can be passed directly to pcre_exec() as
2397 before. That is, this change is entirely upwards-compatible and requires no
2398 change to existing code.
2400 If you want to pass in additional data to pcre_exec(), you can either place it
2401 in a pcre_extra block provided by pcre_study(), or create your own pcre_extra
2402 block.
2404 46. pcretest has been extended to test the PCRE_EXTRA_MATCH_LIMIT feature. If a
2405 data string contains the escape sequence \M, pcretest calls pcre_exec() several
2406 times with different match limits, until it finds the minimum value needed for
2407 pcre_exec() to complete. The value is then output. This can be instructive; for
2408 most simple matches the number is quite small, but for pathological cases it
2409 gets very large very quickly.
2411 47. There's a new option for pcre_fullinfo() called PCRE_INFO_STUDYSIZE. It
2412 returns the size of the data block pointed to by the study_data field in a
2413 pcre_extra block, that is, the value that was passed as the argument to
2414 pcre_malloc() when PCRE was getting memory in which to place the information
2415 created by pcre_study(). The fourth argument should point to a size_t variable.
2416 pcretest has been extended so that this information is shown after a successful
2417 pcre_study() call when information about the compiled regex is being displayed.
2419 48. Cosmetic change to Makefile: there's no need to have / after $(DESTDIR)
2420 because what follows is always an absolute path. (Later: it turns out that this
2421 is more than cosmetic for MinGW, because it doesn't like empty path
2422 components.)
2424 49. Some changes have been made to the callout feature (see 28 above):
2426 (i) A callout function now has three choices for what it returns:
2428 0 => success, carry on matching
2429 > 0 => failure at this point, but backtrack if possible
2430 < 0 => serious error, return this value from pcre_exec()
2432 Negative values should normally be chosen from the set of PCRE_ERROR_xxx
2433 values. In particular, returning PCRE_ERROR_NOMATCH forces a standard
2434 "match failed" error. The error number PCRE_ERROR_CALLOUT is reserved for
2435 use by callout functions. It will never be used by PCRE itself.
2437 (ii) The pcre_extra structure (see 45 above) has a void * field called
2438 callout_data, with corresponding flag bit PCRE_EXTRA_CALLOUT_DATA. The
2439 pcre_callout_block structure has a field of the same name. The contents of
2440 the field passed in the pcre_extra structure are passed to the callout
2441 function in the corresponding field in the callout block. This makes it
2442 easier to use the same callout-containing regex from multiple threads. For
2443 testing, the pcretest program has a new data escape
2445 \C*n pass the number n (may be negative) as callout_data
2447 If the callout function in pcretest receives a non-zero value as
2448 callout_data, it returns that value.
2450 50. Makefile wasn't handling CFLAGS properly when compiling dftables. Also,
2451 there were some redundant $(CFLAGS) in commands that are now specified as
2452 $(LINK), which already includes $(CFLAGS).
2454 51. Extensions to UTF-8 support are listed below. These all apply when (a) PCRE
2455 has been compiled with UTF-8 support *and* pcre_compile() has been compiled
2456 with the PCRE_UTF8 flag. Patterns that are compiled without that flag assume
2457 one-byte characters throughout. Note that case-insensitive matching applies
2458 only to characters whose values are less than 256. PCRE doesn't support the
2459 notion of cases for higher-valued characters.
2461 (i) A character class whose characters are all within 0-255 is handled as
2462 a bit map, and the map is inverted for negative classes. Previously, a
2463 character > 255 always failed to match such a class; however it should
2464 match if the class was a negative one (e.g. [^ab]). This has been fixed.
2466 (ii) A negated character class with a single character < 255 is coded as
2467 "not this character" (OP_NOT). This wasn't working properly when the test
2468 character was multibyte, either singly or repeated.
2470 (iii) Repeats of multibyte characters are now handled correctly in UTF-8
2471 mode, for example: \x{100}{2,3}.
2473 (iv) The character escapes \b, \B, \d, \D, \s, \S, \w, and \W (either
2474 singly or repeated) now correctly test multibyte characters. However,
2475 PCRE doesn't recognize any characters with values greater than 255 as
2476 digits, spaces, or word characters. Such characters always match \D, \S,
2477 and \W, and never match \d, \s, or \w.
2479 (v) Classes may now contain characters and character ranges with values
2480 greater than 255. For example: [ab\x{100}-\x{400}].
2482 (vi) pcregrep now has a --utf-8 option (synonym -u) which makes it call
2483 PCRE in UTF-8 mode.
2485 52. The info request value PCRE_INFO_FIRSTCHAR has been renamed
2486 PCRE_INFO_FIRSTBYTE because it is a byte value. However, the old name is
2487 retained for backwards compatibility. (Note that LASTLITERAL is also a byte
2488 value.)
2490 53. The single man page has become too large. I have therefore split it up into
2491 a number of separate man pages. These also give rise to individual HTML pages;
2492 these are now put in a separate directory, and there is an index.html page that
2493 lists them all. Some hyperlinking between the pages has been installed.
2495 54. Added convenience functions for handling named capturing parentheses.
2497 55. Unknown escapes inside character classes (e.g. [\M]) and escapes that
2498 aren't interpreted therein (e.g. [\C]) are literals in Perl. This is now also
2499 true in PCRE, except when the PCRE_EXTENDED option is set, in which case they
2500 are faulted.
2502 56. Introduced HOST_CC and HOST_CFLAGS which can be set in the environment when
2503 calling configure. These values are used when compiling the dftables.c program
2504 which is run to generate the source of the default character tables. They
2505 default to the values of CC and CFLAGS. If you are cross-compiling PCRE,
2506 you will need to set these values.
2508 57. Updated the building process for Windows DLL, as provided by Fred Cox.
2511 Version 3.9 02-Jan-02
2512 ---------------------
2514 1. A bit of extraneous text had somehow crept into the pcregrep documentation.
2516 2. If --disable-static was given, the building process failed when trying to
2517 build pcretest and pcregrep. (For some reason it was using libtool to compile
2518 them, which is not right, as they aren't part of the library.)
2521 Version 3.8 18-Dec-01
2522 ---------------------
2524 1. The experimental UTF-8 code was completely screwed up. It was packing the
2525 bytes in the wrong order. How dumb can you get?
2528 Version 3.7 29-Oct-01
2529 ---------------------
2531 1. In updating pcretest to check change 1 of version 3.6, I screwed up.
2532 This caused pcretest, when used on the test data, to segfault. Unfortunately,
2533 this didn't happen under Solaris 8, where I normally test things.
2535 2. The Makefile had to be changed to make it work on BSD systems, where 'make'
2536 doesn't seem to recognize that ./xxx and xxx are the same file. (This entry
2537 isn't in ChangeLog distributed with 3.7 because I forgot when I hastily made
2538 this fix an hour or so after the initial 3.7 release.)
2541 Version 3.6 23-Oct-01
2542 ---------------------
2544 1. Crashed with /(sens|respons)e and \1ibility/ and "sense and sensibility" if
2545 offsets passed as NULL with zero offset count.
2547 2. The config.guess and config.sub files had not been updated when I moved to
2548 the latest autoconf.
2551 Version 3.5 15-Aug-01
2552 ---------------------
2554 1. Added some missing #if !defined NOPOSIX conditionals in pcretest.c that
2555 had been forgotten.
2557 2. By using declared but undefined structures, we can avoid using "void"
2558 definitions in pcre.h while keeping the internal definitions of the structures
2559 private.
2561 3. The distribution is now built using autoconf 2.50 and libtool 1.4. From a
2562 user point of view, this means that both static and shared libraries are built
2563 by default, but this can be individually controlled. More of the work of
2564 handling this static/shared cases is now inside libtool instead of PCRE's make
2565 file.
2567 4. The pcretest utility is now installed along with pcregrep because it is
2568 useful for users (to test regexs) and by doing this, it automatically gets
2569 relinked by libtool. The documentation has been turned into a man page, so
2570 there are now .1, .txt, and .html versions in /doc.
2572 5. Upgrades to pcregrep:
2573 (i) Added long-form option names like gnu grep.
2574 (ii) Added --help to list all options with an explanatory phrase.
2575 (iii) Added -r, --recursive to recurse into sub-directories.
2576 (iv) Added -f, --file to read patterns from a file.
2578 6. pcre_exec() was referring to its "code" argument before testing that
2579 argument for NULL (and giving an error if it was NULL).
2581 7. Upgraded Makefile.in to allow for compiling in a different directory from
2582 the source directory.
2584 8. Tiny buglet in pcretest: when pcre_fullinfo() was called to retrieve the
2585 options bits, the pointer it was passed was to an int instead of to an unsigned
2586 long int. This mattered only on 64-bit systems.
2588 9. Fixed typo (3.4/1) in pcre.h again. Sigh. I had changed pcre.h (which is
2589 generated) instead of pcre.in, which it its source. Also made the same change
2590 in several of the .c files.
2592 10. A new release of gcc defines printf() as a macro, which broke pcretest
2593 because it had an ifdef in the middle of a string argument for printf(). Fixed
2594 by using separate calls to printf().
2596 11. Added --enable-newline-is-cr and --enable-newline-is-lf to the configure
2597 script, to force use of CR or LF instead of \n in the source. On non-Unix
2598 systems, the value can be set in config.h.
2600 12. The limit of 200 on non-capturing parentheses is a _nesting_ limit, not an
2601 absolute limit. Changed the text of the error message to make this clear, and
2602 likewise updated the man page.
2604 13. The limit of 99 on the number of capturing subpatterns has been removed.
2605 The new limit is 65535, which I hope will not be a "real" limit.
2608 Version 3.4 22-Aug-00
2609 ---------------------
2611 1. Fixed typo in pcre.h: unsigned const char * changed to const unsigned char *.
2613 2. Diagnose condition (?(0) as an error instead of crashing on matching.
2616 Version 3.3 01-Aug-00
2617 ---------------------
2619 1. If an octal character was given, but the value was greater than \377, it
2620 was not getting masked to the least significant bits, as documented. This could
2621 lead to crashes in some systems.
2623 2. Perl 5.6 (if not earlier versions) accepts classes like [a-\d] and treats
2624 the hyphen as a literal. PCRE used to give an error; it now behaves like Perl.
2626 3. Added the functions pcre_free_substring() and pcre_free_substring_list().
2627 These just pass their arguments on to (pcre_free)(), but they are provided
2628 because some uses of PCRE bind it to non-C systems that can call its functions,
2629 but cannot call free() or pcre_free() directly.
2631 4. Add "make test" as a synonym for "make check". Corrected some comments in
2632 the Makefile.
2634 5. Add $(DESTDIR)/ in front of all the paths in the "install" target in the
2635 Makefile.
2637 6. Changed the name of pgrep to pcregrep, because Solaris has introduced a
2638 command called pgrep for grepping around the active processes.
2640 7. Added the beginnings of support for UTF-8 character strings.
2642 8. Arranged for the Makefile to pass over the settings of CC, CFLAGS, and
2643 RANLIB to ./ltconfig so that they are used by libtool. I think these are all
2644 the relevant ones. (AR is not passed because ./ltconfig does its own figuring
2645 out for the ar command.)
2648 Version 3.2 12-May-00
2649 ---------------------
2651 This is purely a bug fixing release.
2653 1. If the pattern /((Z)+|A)*/ was matched agained ZABCDEFG it matched Z instead
2654 of ZA. This was just one example of several cases that could provoke this bug,
2655 which was introduced by change 9 of version 2.00. The code for breaking
2656 infinite loops after an iteration that matches an empty string was't working
2657 correctly.
2659 2. The pcretest program was not imitating Perl correctly for the pattern /a*/g
2660 when matched against abbab (for example). After matching an empty string, it
2661 wasn't forcing anchoring when setting PCRE_NOTEMPTY for the next attempt; this
2662 caused it to match further down the string than it should.
2664 3. The code contained an inclusion of sys/types.h. It isn't clear why this
2665 was there because it doesn't seem to be needed, and it causes trouble on some
2666 systems, as it is not a Standard C header. It has been removed.
2668 4. Made 4 silly changes to the source to avoid stupid compiler warnings that
2669 were reported on the Macintosh. The changes were from
2671 while ((c = *(++ptr)) != 0 && c != '\n');
2672 to
2673 while ((c = *(++ptr)) != 0 && c != '\n') ;
2675 Totally extraordinary, but if that's what it takes...
2677 5. PCRE is being used in one environment where neither memmove() nor bcopy() is
2678 available. Added HAVE_BCOPY and an autoconf test for it; if neither
2679 HAVE_MEMMOVE nor HAVE_BCOPY is set, use a built-in emulation function which
2680 assumes the way PCRE uses memmove() (always moving upwards).
2682 6. PCRE is being used in one environment where strchr() is not available. There
2683 was only one use in pcre.c, and writing it out to avoid strchr() probably gives
2684 faster code anyway.
2687 Version 3.1 09-Feb-00
2688 ---------------------
2690 The only change in this release is the fixing of some bugs in Makefile.in for
2691 the "install" target:
2693 (1) It was failing to install pcreposix.h.
2695 (2) It was overwriting the pcre.3 man page with the pcreposix.3 man page.
2698 Version 3.0 01-Feb-00
2699 ---------------------
2701 1. Add support for the /+ modifier to perltest (to output $` like it does in
2702 pcretest).
2704 2. Add support for the /g modifier to perltest.
2706 3. Fix pcretest so that it behaves even more like Perl for /g when the pattern
2707 matches null strings.
2709 4. Fix perltest so that it doesn't do unwanted things when fed an empty
2710 pattern. Perl treats empty patterns specially - it reuses the most recent
2711 pattern, which is not what we want. Replace // by /(?#)/ in order to avoid this
2712 effect.
2714 5. The POSIX interface was broken in that it was just handing over the POSIX
2715 captured string vector to pcre_exec(), but (since release 2.00) PCRE has
2716 required a bigger vector, with some working space on the end. This means that
2717 the POSIX wrapper now has to get and free some memory, and copy the results.
2719 6. Added some simple autoconf support, placing the test data and the
2720 documentation in separate directories, re-organizing some of the
2721 information files, and making it build pcre-config (a GNU standard). Also added
2722 libtool support for building PCRE as a shared library, which is now the
2723 default.
2725 7. Got rid of the leading zero in the definition of PCRE_MINOR because 08 and
2726 09 are not valid octal constants. Single digits will be used for minor values
2727 less than 10.
2729 8. Defined REG_EXTENDED and REG_NOSUB as zero in the POSIX header, so that
2730 existing programs that set these in the POSIX interface can use PCRE without
2731 modification.
2733 9. Added a new function, pcre_fullinfo() with an extensible interface. It can
2734 return all that pcre_info() returns, plus additional data. The pcre_info()
2735 function is retained for compatibility, but is considered to be obsolete.
2737 10. Added experimental recursion feature (?R) to handle one common case that
2738 Perl 5.6 will be able to do with (?p{...}).
2740 11. Added support for POSIX character classes like [:alpha:], which Perl is
2741 adopting.
2744 Version 2.08 31-Aug-99
2745 ----------------------
2747 1. When startoffset was not zero and the pattern began with ".*", PCRE was not
2748 trying to match at the startoffset position, but instead was moving forward to
2749 the next newline as if a previous match had failed.
2751 2. pcretest was not making use of PCRE_NOTEMPTY when repeating for /g and /G,
2752 and could get into a loop if a null string was matched other than at the start
2753 of the subject.
2755 3. Added definitions of PCRE_MAJOR and PCRE_MINOR to pcre.h so the version can
2756 be distinguished at compile time, and for completeness also added PCRE_DATE.
2758 5. Added Paul Sokolovsky's minor changes to make it easy to compile a Win32 DLL
2759 in GnuWin32 environments.
2762 Version 2.07 29-Jul-99
2763 ----------------------
2765 1. The documentation is now supplied in plain text form and HTML as well as in
2766 the form of man page sources.
2768 2. C++ compilers don't like assigning (void *) values to other pointer types.
2769 In particular this affects malloc(). Although there is no problem in Standard
2770 C, I've put in casts to keep C++ compilers happy.
2772 3. Typo on pcretest.c; a cast of (unsigned char *) in the POSIX regexec() call
2773 should be (const char *).
2775 4. If NOPOSIX is defined, pcretest.c compiles without POSIX support. This may
2776 be useful for non-Unix systems who don't want to bother with the POSIX stuff.
2777 However, I haven't made this a standard facility. The documentation doesn't
2778 mention it, and the Makefile doesn't support it.
2780 5. The Makefile now contains an "install" target, with editable destinations at
2781 the top of the file. The pcretest program is not installed.
2783 6. pgrep -V now gives the PCRE version number and date.
2785 7. Fixed bug: a zero repetition after a literal string (e.g. /abcde{0}/) was
2786 causing the entire string to be ignored, instead of just the last character.
2788 8. If a pattern like /"([^\\"]+|\\.)*"/ is applied in the normal way to a
2789 non-matching string, it can take a very, very long time, even for strings of
2790 quite modest length, because of the nested recursion. PCRE now does better in
2791 some of these cases. It does this by remembering the last required literal
2792 character in the pattern, and pre-searching the subject to ensure it is present
2793 before running the real match. In other words, it applies a heuristic to detect
2794 some types of certain failure quickly, and in the above example, if presented
2795 with a string that has no trailing " it gives "no match" very quickly.
2797 9. A new runtime option PCRE_NOTEMPTY causes null string matches to be ignored;
2798 other alternatives are tried instead.
2801 Version 2.06 09-Jun-99
2802 ----------------------
2804 1. Change pcretest's output for amount of store used to show just the code
2805 space, because the remainder (the data block) varies in size between 32-bit and
2806 64-bit systems.
2808 2. Added an extra argument to pcre_exec() to supply an offset in the subject to
2809 start matching at. This allows lookbehinds to work when searching for multiple
2810 occurrences in a string.
2812 3. Added additional options to pcretest for testing multiple occurrences:
2814 /+ outputs the rest of the string that follows a match
2815 /g loops for multiple occurrences, using the new startoffset argument
2816 /G loops for multiple occurrences by passing an incremented pointer
2818 4. PCRE wasn't doing the "first character" optimization for patterns starting
2819 with \b or \B, though it was doing it for other lookbehind assertions. That is,
2820 it wasn't noticing that a match for a pattern such as /\bxyz/ has to start with
2821 the letter 'x'. On long subject strings, this gives a significant speed-up.
2824 Version 2.05 21-Apr-99
2825 ----------------------
2827 1. Changed the type of magic_number from int to long int so that it works
2828 properly on 16-bit systems.
2830 2. Fixed a bug which caused patterns starting with .* not to work correctly
2831 when the subject string contained newline characters. PCRE was assuming
2832 anchoring for such patterns in all cases, which is not correct because .* will
2833 not pass a newline unless PCRE_DOTALL is set. It now assumes anchoring only if
2834 DOTALL is set at top level; otherwise it knows that patterns starting with .*
2835 must be retried after every newline in the subject.
2838 Version 2.04 18-Feb-99
2839 ----------------------
2841 1. For parenthesized subpatterns with repeats whose minimum was zero, the
2842 computation of the store needed to hold the pattern was incorrect (too large).
2843 If such patterns were nested a few deep, this could multiply and become a real
2844 problem.
2846 2. Added /M option to pcretest to show the memory requirement of a specific
2847 pattern. Made -m a synonym of -s (which does this globally) for compatibility.
2849 3. Subpatterns of the form (regex){n,m} (i.e. limited maximum) were being
2850 compiled in such a way that the backtracking after subsequent failure was
2851 pessimal. Something like (a){0,3} was compiled as (a)?(a)?(a)? instead of
2852 ((a)((a)(a)?)?)? with disastrous performance if the maximum was of any size.
2855 Version 2.03 02-Feb-99
2856 ----------------------
2858 1. Fixed typo and small mistake in man page.
2860 2. Added 4th condition (GPL supersedes if conflict) and created separate
2861 LICENCE file containing the conditions.
2863 3. Updated pcretest so that patterns such as /abc\/def/ work like they do in
2864 Perl, that is the internal \ allows the delimiter to be included in the
2865 pattern. Locked out the use of \ as a delimiter. If \ immediately follows
2866 the final delimiter, add \ to the end of the pattern (to test the error).
2868 4. Added the convenience functions for extracting substrings after a successful
2869 match. Updated pcretest to make it able to test these functions.
2872 Version 2.02 14-Jan-99
2873 ----------------------
2875 1. Initialized the working variables associated with each extraction so that
2876 their saving and restoring doesn't refer to uninitialized store.
2878 2. Put dummy code into study.c in order to trick the optimizer of the IBM C
2879 compiler for OS/2 into generating correct code. Apparently IBM isn't going to
2880 fix the problem.
2882 3. Pcretest: the timing code wasn't using LOOPREPEAT for timing execution
2883 calls, and wasn't printing the correct value for compiling calls. Increased the
2884 default value of LOOPREPEAT, and the number of significant figures in the
2885 times.
2887 4. Changed "/bin/rm" in the Makefile to "-rm" so it works on Windows NT.
2889 5. Renamed "deftables" as "dftables" to get it down to 8 characters, to avoid
2890 a building problem on Windows NT with a FAT file system.
2893 Version 2.01 21-Oct-98
2894 ----------------------
2896 1. Changed the API for pcre_compile() to allow for the provision of a pointer
2897 to character tables built by pcre_maketables() in the current locale. If NULL
2898 is passed, the default tables are used.
2901 Version 2.00 24-Sep-98
2902 ----------------------
2904 1. Since the (>?) facility is in Perl 5.005, don't require PCRE_EXTRA to enable
2905 it any more.
2907 2. Allow quantification of (?>) groups, and make it work correctly.
2909 3. The first character computation wasn't working for (?>) groups.
2911 4. Correct the implementation of \Z (it is permitted to match on the \n at the
2912 end of the subject) and add 5.005's \z, which really does match only at the
2913 very end of the subject.
2915 5. Remove the \X "cut" facility; Perl doesn't have it, and (?> is neater.
2917 6. Remove the ability to specify CASELESS, MULTILINE, DOTALL, and
2918 DOLLAR_END_ONLY at runtime, to make it possible to implement the Perl 5.005
2919 localized options. All options to pcre_study() were also removed.
2921 7. Add other new features from 5.005:
2923 $(?<= positive lookbehind
2924 $(?<! negative lookbehind
2925 (?imsx-imsx) added the unsetting capability
2926 such a setting is global if at outer level; local otherwise
2927 (?imsx-imsx:) non-capturing groups with option setting
2928 (?(cond)re|re) conditional pattern matching
2930 A backreference to itself in a repeated group matches the previous
2931 captured string.
2933 8. General tidying up of studying (both automatic and via "study")
2934 consequential on the addition of new assertions.
2936 9. As in 5.005, unlimited repeated groups that could match an empty substring
2937 are no longer faulted at compile time. Instead, the loop is forcibly broken at
2938 runtime if any iteration does actually match an empty substring.
2940 10. Include the RunTest script in the distribution.
2942 11. Added tests from the Perl 5.005_02 distribution. This showed up a few
2943 discrepancies, some of which were old and were also with respect to 5.004. They
2944 have now been fixed.
2947 Version 1.09 28-Apr-98
2948 ----------------------
2950 1. A negated single character class followed by a quantifier with a minimum
2951 value of one (e.g. [^x]{1,6} ) was not compiled correctly. This could lead to
2952 program crashes, or just wrong answers. This did not apply to negated classes
2953 containing more than one character, or to minima other than one.
2956 Version 1.08 27-Mar-98
2957 ----------------------
2959 1. Add PCRE_UNGREEDY to invert the greediness of quantifiers.
2961 2. Add (?U) and (?X) to set PCRE_UNGREEDY and PCRE_EXTRA respectively. The
2962 latter must appear before anything that relies on it in the pattern.
2965 Version 1.07 16-Feb-98
2966 ----------------------
2968 1. A pattern such as /((a)*)*/ was not being diagnosed as in error (unlimited
2969 repeat of a potentially empty string).
2972 Version 1.06 23-Jan-98
2973 ----------------------
2975 1. Added Markus Oberhumer's little patches for C++.
2977 2. Literal strings longer than 255 characters were broken.
2980 Version 1.05 23-Dec-97
2981 ----------------------
2983 1. Negated character classes containing more than one character were failing if
2984 PCRE_CASELESS was set at run time.
2987 Version 1.04 19-Dec-97
2988 ----------------------
2990 1. Corrected the man page, where some "const" qualifiers had been omitted.
2992 2. Made debugging output print "{0,xxx}" instead of just "{,xxx}" to agree with
2993 input syntax.
2995 3. Fixed memory leak which occurred when a regex with back references was
2996 matched with an offsets vector that wasn't big enough. The temporary memory
2997 that is used in this case wasn't being freed if the match failed.
2999 4. Tidied pcretest to ensure it frees memory that it gets.
3001 5. Temporary memory was being obtained in the case where the passed offsets
3002 vector was exactly big enough.
3004 6. Corrected definition of offsetof() from change 5 below.
3006 7. I had screwed up change 6 below and broken the rules for the use of
3007 setjmp(). Now fixed.
3010 Version 1.03 18-Dec-97
3011 ----------------------
3013 1. A erroneous regex with a missing opening parenthesis was correctly
3014 diagnosed, but PCRE attempted to access brastack[-1], which could cause crashes
3015 on some systems.
3017 2. Replaced offsetof(real_pcre, code) by offsetof(real_pcre, code[0]) because
3018 it was reported that one broken compiler failed on the former because "code" is
3019 also an independent variable.
3021 3. The erroneous regex a[]b caused an array overrun reference.
3023 4. A regex ending with a one-character negative class (e.g. /[^k]$/) did not
3024 fail on data ending with that character. (It was going on too far, and checking
3025 the next character, typically a binary zero.) This was specific to the
3026 optimized code for single-character negative classes.
3028 5. Added a contributed patch from the TIN world which does the following:
3030 + Add an undef for memmove, in case the the system defines a macro for it.
3032 + Add a definition of offsetof(), in case there isn't one. (I don't know
3033 the reason behind this - offsetof() is part of the ANSI standard - but
3034 it does no harm).
3036 + Reduce the ifdef's in pcre.c using macro DPRINTF, thereby eliminating
3037 most of the places where whitespace preceded '#'. I have given up and
3038 allowed the remaining 2 cases to be at the margin.
3040 + Rename some variables in pcre to eliminate shadowing. This seems very
3041 pedantic, but does no harm, of course.
3043 6. Moved the call to setjmp() into its own function, to get rid of warnings
3044 from gcc -Wall, and avoided calling it at all unless PCRE_EXTRA is used.
3046 7. Constructs such as \d{8,} were compiling into the equivalent of
3047 \d{8}\d{0,65527} instead of \d{8}\d* which didn't make much difference to the
3048 outcome, but in this particular case used more store than had been allocated,
3049 which caused the bug to be discovered because it threw up an internal error.
3051 8. The debugging code in both pcre and pcretest for outputting the compiled
3052 form of a regex was going wrong in the case of back references followed by
3053 curly-bracketed repeats.
3056 Version 1.02 12-Dec-97
3057 ----------------------
3059 1. Typos in pcre.3 and comments in the source fixed.
3061 2. Applied a contributed patch to get rid of places where it used to remove
3062 'const' from variables, and fixed some signed/unsigned and uninitialized
3063 variable warnings.
3065 3. Added the "runtest" target to Makefile.
3067 4. Set default compiler flag to -O2 rather than just -O.
3070 Version 1.01 19-Nov-97
3071 ----------------------
3073 1. PCRE was failing to diagnose unlimited repeat of empty string for patterns
3074 like /([ab]*)*/, that is, for classes with more than one character in them.
3076 2. Likewise, it wasn't diagnosing patterns with "once-only" subpatterns, such
3077 as /((?>a*))*/ (a PCRE_EXTRA facility).
3080 Version 1.00 18-Nov-97
3081 ----------------------
3083 1. Added compile-time macros to support systems such as SunOS4 which don't have
3084 memmove() or strerror() but have other things that can be used instead.
3086 2. Arranged that "make clean" removes the executables.
3089 Version 0.99 27-Oct-97
3090 ----------------------
3092 1. Fixed bug in code for optimizing classes with only one character. It was
3093 initializing a 32-byte map regardless, which could cause it to run off the end
3094 of the memory it had got.
3096 2. Added, conditional on PCRE_EXTRA, the proposed (?>REGEX) construction.
3099 Version 0.98 22-Oct-97
3100 ----------------------
3102 1. Fixed bug in code for handling temporary memory usage when there are more
3103 back references than supplied space in the ovector. This could cause segfaults.
3106 Version 0.97 21-Oct-97
3107 ----------------------
3109 1. Added the \X "cut" facility, conditional on PCRE_EXTRA.
3111 2. Optimized negated single characters not to use a bit map.
3113 3. Brought error texts together as macro definitions; clarified some of them;
3114 fixed one that was wrong - it said "range out of order" when it meant "invalid
3115 escape sequence".
3117 4. Changed some char * arguments to const char *.
3119 5. Added PCRE_NOTBOL and PCRE_NOTEOL (from POSIX).
3121 6. Added the POSIX-style API wrapper in pcreposix.a and testing facilities in
3122 pcretest.
3125 Version 0.96 16-Oct-97
3126 ----------------------
3128 1. Added a simple "pgrep" utility to the distribution.
3130 2. Fixed an incompatibility with Perl: "{" is now treated as a normal character
3131 unless it appears in one of the precise forms "{ddd}", "{ddd,}", or "{ddd,ddd}"
3132 where "ddd" means "one or more decimal digits".
3134 3. Fixed serious bug. If a pattern had a back reference, but the call to
3135 pcre_exec() didn't supply a large enough ovector to record the related
3136 identifying subpattern, the match always failed. PCRE now remembers the number
3137 of the largest back reference, and gets some temporary memory in which to save
3138 the offsets during matching if necessary, in order to ensure that
3139 backreferences always work.
3141 4. Increased the compatibility with Perl in a number of ways:
3143 (a) . no longer matches \n by default; an option PCRE_DOTALL is provided
3144 to request this handling. The option can be set at compile or exec time.
3146 (b) $ matches before a terminating newline by default; an option
3147 PCRE_DOLLAR_ENDONLY is provided to override this (but not in multiline
3148 mode). The option can be set at compile or exec time.
3150 (c) The handling of \ followed by a digit other than 0 is now supposed to be
3151 the same as Perl's. If the decimal number it represents is less than 10
3152 or there aren't that many previous left capturing parentheses, an octal
3153 escape is read. Inside a character class, it's always an octal escape,
3154 even if it is a single digit.
3156 (d) An escaped but undefined alphabetic character is taken as a literal,
3157 unless PCRE_EXTRA is set. Currently this just reserves the remaining
3158 escapes.
3160 (e) {0} is now permitted. (The previous item is removed from the compiled
3161 pattern).
3163 5. Changed all the names of code files so that the basic parts are no longer
3164 than 10 characters, and abolished the teeny "globals.c" file.
3166 6. Changed the handling of character classes; they are now done with a 32-byte
3167 bit map always.
3169 7. Added the -d and /D options to pcretest to make it possible to look at the
3170 internals of compilation without having to recompile pcre.
3173 Version 0.95 23-Sep-97
3174 ----------------------
3176 1. Fixed bug in pre-pass concerning escaped "normal" characters such as \x5c or
3177 \x20 at the start of a run of normal characters. These were being treated as
3178 real characters, instead of the source characters being re-checked.
3181 Version 0.94 18-Sep-97
3182 ----------------------
3184 1. The functions are now thread-safe, with the caveat that the global variables
3185 containing pointers to malloc() and free() or alternative functions are the
3186 same for all threads.
3188 2. Get pcre_study() to generate a bitmap of initial characters for non-
3189 anchored patterns when this is possible, and use it if passed to pcre_exec().
3192 Version 0.93 15-Sep-97
3193 ----------------------
3195 1. /(b)|(:+)/ was computing an incorrect first character.
3197 2. Add pcre_study() to the API and the passing of pcre_extra to pcre_exec(),
3198 but not actually doing anything yet.
3200 3. Treat "-" characters in classes that cannot be part of ranges as literals,
3201 as Perl does (e.g. [-az] or [az-]).
3203 4. Set the anchored flag if a branch starts with .* or .*? because that tests
3204 all possible positions.
3206 5. Split up into different modules to avoid including unneeded functions in a
3207 compiled binary. However, compile and exec are still in one module. The "study"
3208 function is split off.
3210 6. The character tables are now in a separate module whose source is generated
3211 by an auxiliary program - but can then be edited by hand if required. There are
3212 now no calls to isalnum(), isspace(), isdigit(), isxdigit(), tolower() or
3213 toupper() in the code.
3215 7. Turn the malloc/free funtions variables into pcre_malloc and pcre_free and
3216 make them global. Abolish the function for setting them, as the caller can now
3217 set them directly.
3220 Version 0.92 11-Sep-97
3221 ----------------------
3223 1. A repeat with a fixed maximum and a minimum of 1 for an ordinary character
3224 (e.g. /a{1,3}/) was broken (I mis-optimized it).
3226 2. Caseless matching was not working in character classes if the characters in
3227 the pattern were in upper case.
3229 3. Make ranges like [W-c] work in the same way as Perl for caseless matching.
3231 4. Make PCRE_ANCHORED public and accept as a compile option.
3233 5. Add an options word to pcre_exec() and accept PCRE_ANCHORED and
3234 PCRE_CASELESS at run time. Add escapes \A and \I to pcretest to cause it to
3235 pass them.
3237 6. Give an error if bad option bits passed at compile or run time.
3239 7. Add PCRE_MULTILINE at compile and exec time, and (?m) as well. Add \M to
3240 pcretest to cause it to pass that flag.
3242 8. Add pcre_info(), to get the number of identifying subpatterns, the stored
3243 options, and the first character, if set.
3245 9. Recognize C+ or C{n,m} where n >= 1 as providing a fixed starting character.
3248 Version 0.91 10-Sep-97
3249 ----------------------
3251 1. PCRE was failing to diagnose unlimited repeats of subpatterns that could
3252 match the empty string as in /(a*)*/. It was looping and ultimately crashing.
3254 2. PCRE was looping on encountering an indefinitely repeated back reference to
3255 a subpattern that had matched an empty string, e.g. /(a|)\1*/. It now does what
3256 Perl does - treats the match as successful.
3258 ****


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

  ViewVC Help
Powered by ViewVC 1.1.5